Development of innovative bio-based films with keratin

extracted from duck feathers.

Sandra Alvarez

To cite this version:

Sandra Alvarez. Development of innovative bio-based films with keratin extracted from duck feathers..
Chemical Sciences. Université de Pau et des Pays de l’Adour, 2024. English. �NNT : 2024PAUU3075�.
�tel-04741373�

HAL Id: tel-04741373

[Link]
Submitted on 17 Oct 2024

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 THÈSE
UNIVERSITE DE PAU ET DES PAYS DE L’ADOUR
École doctorale des sciences exactes et leurs applications

Présentée et soutenue le 08 février 2024

par Sandra ALVAREZ
pour obtenir le grade de docteur
de l’Université de Pau et des Pays de l’Adour
Spécialité : Chimie des matériaux

DEVELOPMENT OF INNOVATIVE BIOBASED FILMS WITH

KERATIN EXTRACTED FROM DUCK FEATHERS
—
Développement de films biosourcés innovants à base de kératine
extraite de plumes de canards

MEMBRES DU JURY

RAPPORTEURS
• Alejandro RODRIGUEZ Professeur / Université de Cordoue
• Ana Rita CRUZ DUARTE Professeure associée / Université NOVA de Lisbonne

EXAMINATEURS
• Susana FERNANDES Professeure / Université de Pau et des Pays de l’Adour
• Laurent BILLON Professeur / Université de Pau et des Pays de l’Adour
• Oihana GORDOBIL Chercheuse associée / Université du Pays Basque

DIRECTEUR
• Eduardo ROBLES Professeur assistant / Université de Pau et des Pays de l’Adour

INVITES
• Marc DELGADO AGUILAR Professeur associé / Université de Gérone
2
REMERCIEMENTS

Avant d’entamer tout développement, je souhaite exprimer ma reconnaissance à toutes les personnes
que j’ai pu rencontrer lors de ce long travail et qui ont contribué, directement ou indirectement, à cette
thèse.

Mon directeur de thèse, Eduardo Robles, est la première personne à qui je souhaite adresser mes
profonds remerciements pour m’avoir accordé sa confiance et m’avoir accompagné durant ces trois
années. Merci de m’avoir fait grandir sur le plan scientifique et technique comme sur le plan humain.
Merci également pour tes conseils, ton soutien quotidien et le temps passé à partager tes connaissances.

Je tiens ensuite à remercier tous les membres de mon jury de soutenance d’avoir accepté d’évaluer mon
travail. Merci aux deux rapporteurs, Alejandro Rodriguez et Ana Rita Cruz Duarte, pour le temps
consacré à la lecture de mon manuscrit ainsi qu’aux examinateurs Susana Fernandes, Laurent Billon et
Oihana Gordobil d’avoir été présents. Merci à tous pour vos conseils bienveillants et pour les échanges
constructifs.

Je remercie le département des Landes ainsi que le programme E2S de l’Université de Pau et des Pays
de l’Adour pour le financement de ce projet. Je remercie également le laboratoire IPREM, et plus
particulièrement le site Montois, au sein duquel ce travail s’est déroulé. Merci également à l’IUT de
Mont de Marsan et au technopole Agrolandes pour la mise à disposition de locaux et de matériel.

Mes sincères remerciements aux professeurs et aux membres de l’IUT de Mont de Marsan qui m’ont
permis de disposer du matériel nécessaire à mon travail et pour leur aide. Merci à Marianne Houssier
d’avoir suivi mon travail de thèse ainsi que pour son expertise en biologie et pour sa grande sympathie.

Je tiens ensuite à exprimer ma profonde gratitude à l’Université de Gérone et particulièrement à Marc

Delgado-Aguilar de m’avoir permis d’intégrer le groupe de recherche LEPAMP-PRODIS lors d’une
mobilité internationale de plus de trois mois. Je remercie grandement toute l’équipe de recherche pour
leur accueil chaleureux et tous les doctorants et post-doctorants que j’ai pu rencontrer pour leur aide et
leur sympathie. Merci particulièrement à Gabriela et Giulia, mes deux acolytes de voyage avec qui j’ai
pu visiter et découvrir une partie de l’Espagne pour des souvenirs gravés dans ma mémoire.

Je tiens maintenant à remercier mes collègues de laboratoire devenus de véritables amis. Je commencerai
par remercier Nidal, sans qui mon travail de thèse n’aurait pas été le même. Nous avons commencé à
travailler ensemble et nous avons collaboré durant de longs mois sur ce projet de plumes. Merci pour
ton aide, ton sourire et ton accent si unique. Un grand et sincère merci à Elsa, mon binôme de laboratoire
et mon amie depuis maintenant plus de sept ans. La vie de laboratoire a été plus sereine et mes journées
ensoleillées par ta présence. Merci également à Jérémy qui a été une véritable aide dans les moments de
doutes et qui a donné beaucoup de son temps pour me transmettre son savoir. Enfin, merci à tous les
doctorants, post doctorants, ingénieurs, alternants et stagiaires qui ont constitué notre équipe :
Alexandre, Alix, Bastien A., Bastien F., Hugues, Léa, Léo, Luanna, Chloé, Quentin G., Quentin M.,
Raphaël, Nouha, Charlotte, Tim, Arsène, Peguy, Wadah… Je suis très reconnaissante de tous les

3
moments passés en votre présence et je n’oublierai jamais tous ces rires partagés qui ont été pour moi
une bouffée d’oxygène tout au long de ce parcours.

Je terminerai par remercier tous mes proches, mes amis et ma famille d’être à mes côtés et de m’apporter
leur soutien. Je remercie avec tendresse ma maman, mon modèle de force, qui a toujours cru en moi et
qui m’encourage dans chacune de mes décisions. Merci à mon papa, parti avant de savoir que j’allais
entreprendre ce travail de thèse, et qui m’a toujours encouragé à viser le plus haut possible et de se
donner les moyens de réussir tout en respectant mes choix. Je suis certaine qu’il me suivait et me
soutenait depuis là-haut. Enfin, merci à ma moitié, Laurent, qui partage ma vie depuis bientôt cinq ans.
Merci pour ton amour, ta patience lors des moments de doute et pour tes encouragements.

4
RÉSUMÉ
La France fait partie des plus grands producteurs de canards gras à l’échelle mondiale. Cette production
se concentre principalement dans la région du Sud-Ouest où des tonnes de co-produits sont générées
chaque année. Généralement, ces co-produits sont déposés en déchetterie, incinérés ou enterrés, ou
faiblement valorisés sans aucune valeur ajoutée. De plus, ces déchets agricoles peuvent causer des
problèmes environnementaux de par la diffusion de micro-organismes et bactéries pouvant affecter les
sols et l’être humain. Les plumes représentent une grande partie de ces déchets, encore peu valorisées à
ce jour. C’est pourquoi, ce sujet de thèse a pour but de valoriser les plumes de canards afin de diminuer
leur impact environnemental et de participer à une économie circulaire locale en créant un matériau
biosourcé et biodégradable. Premièrement, les propriétés physico-chimiques des plumes ont été
analysées. Une plume est composée de 90% de kératine ce qui en fait une source de polymère naturel
idéale. Puis, la valorisation de ces plumes a débuté par la mise en place d’un protocole d’extraction de
la kératine par un procédé de solvolyse à l’aide de solvants non toxiques. Une optimisation de ce procédé
est réalisée en fonction de divers paramètres afin d’obtenir le meilleur rendement en maintenant le plus
faible impact environnemental possible. Les propriétés physico-chimiques de la kératine extraite sont
ensuite déterminées dans le but d’identifier l’influence de chaque paramètre d’extraction sur la qualité
de celle-ci. Enfin, des films biosourcés sont réalisés à base de nanocellulose avec incorporation de
différents pourcentages de kératine pour évaluer l’influence de la kératine sur leur propriétés mécaniques
et hydrophobes dans l’intention d’une utilisation en packaging. La kératine extraite des plumes de
canards peut donc être valorisée, via un procédé d’extraction limitant l’impact environnemental et peut
également être utilisée dans divers domaines d’application comme la cosmétique ou la médecine.

5
ABSTRACT
France is one of the world's leading producers of fattened ducks. This production is mainly concentrated
in the southwest region, where tons of by-products are generated each year. These by-products are
generally disposed of in landfills, incinerated, buried, or have limited value-added recovery.
Furthermore, this agricultural waste can cause environmental problems by spreading microorganisms
and bacteria that can affect the soil and humans. Feathers represent a large part of this waste, which is
still poorly exploited. For this reason, this work aims to valorize duck feathers to reduce their
environmental impact and participate in a circular economy by creating a biosourced and biodegradable
material. First, the physical and chemical properties of feathers were analyzed. A feather is composed
of 90% keratin, making it an ideal natural polymer source. Then, the valorization of these feathers began
by establishing a protocol for extracting keratin using a solvolysis process with two non-toxic solvents.
The process was optimized according to various parameters to obtain the best yield with the minimal
environmental impact. The physicochemical properties of the extracted keratin were then determined in
order to identify the influence of each extraction parameter on its quality. Finally, bio-based
nanocellulose films with different percentages of keratin were prepared to evaluate the influence of
keratin on their mechanical and hydrophobic properties, with a view to their use in packaging. The
keratin extracted from duck feathers can thus be valorized through an extraction process that limits the
environmental impact and can also be used in various applications such as cosmetics or medicine.

6
CONTEXT
Planet Earth has been experiencing considerable demographic growth for several decades, continuing
to rise in the upcoming years. The agri-food industry produces increasing amounts of food every year
to provide for the global population. This global production, in turn, generates by-products that, in most
cases, must be disposed of due to their low value. Unused by-products are then incinerated or buried in
landfills, which can cause major environmental issues. The second impact of the sharp increase in food
production is marketing these products in packaging, most often made from plastics derived from the
petrochemical industry. While these synthetic food packaging materials are good for preserving food,
their production, uncontrolled use, and end-of-life can cause significant environmental pollution. For
this reason, in recent years, the study of natural and bio-sourced polymers has become increasingly
attractive to scientists working to create new materials with low environmental impact to replace
conventional plastics.

In this context, the idea of studying the valorization of an undervalued by-product has emerged. In this
case, the by-product in context is duck feathers, a by-product of the fattened duck industry, a major
activity in the Southwest of France. Feathers contain around 90% protein, making them an exciting
source of natural polymers. Furthermore, this local valorization would allow participation in the circular
economy of the region by recovering and reusing a local waste product. For this purpose, duck feathers
were collected, and their properties were assessed. The complete characterization of duck feathers is an
essential step in this work and was motivated by the lack of information available in the literature. The
following step was extracting the keratin protein from the duck feathers. This extraction is carried out
by solvolysis using a non-toxic solvent with low environmental impact. After this extraction, bio-
sourced films based on cellulose and keratin were produced. Cellulose is the most abundant natural
polymer on earth, and its use in the packaging sector is well established. The disadvantage of cellulose
is its high sensitivity to water, which can limit its application. For this reason, films combining cellulose
and keratin have been produced, and the influence of keratin on the properties of cellulose films has
been investigated. Hence, this work provides a solid basis for the extraction of keratin by solvolysis and
its potential valorization through the production of bio-based films.

7
8
 CONTENT

RÉSUMÉ ________________________________________________________________________ 5

ABSTRACT _____________________________________________________________________ 6

CONTEXT ______________________________________________________________________ 7

LIST OF FIGURES _______________________________________________________________ 13

LIST OF TABLES _______________________________________________________________ 17

INDEX OF ABBREVIATIONS _____________________________________________________ 18

Chapter I _______________________________________________________________________ 21

I.1. The poultry sector: a source of natural by-products _________________________________ 23

I.1.1. Worldwide poultry sector __________________________________________________ 23

I.1.2. New Aquitaine region and Landes department poultry sector ______________________ 23

I.2. Feathers as a source of keratin _________________________________________________ 24

I.2.1. Feathers _______________________________________________________________ 24

I.2.2. Proteins________________________________________________________________ 25

I.2.3. Keratin ________________________________________________________________ 26

I.3. Feathers processing __________________________________________________________ 29

I.3.1. Feathers cleaning methods _________________________________________________ 29

I.3.2. Keratin extraction methods ________________________________________________ 30

I.4. Keratin valorization through bio-based films ______________________________________ 39

I.4.1. Pure keratin films ________________________________________________________ 39

I.4.2. Plasticized keratin films ___________________________________________________ 39

I.4.3. Bio-based films by combining keratin and biopolymers __________________________ 40

I.5. Cellulose __________________________________________________________________ 42

I.5.1. Native cellulose _________________________________________________________ 42

I.5.2. Nanocellulose ___________________________________________________________ 44

I.5.3. Cellulose extraction ______________________________________________________ 45

I.5.4. Cellulose films and their properties __________________________________________ 47

I.6. Conclusion ________________________________________________________________ 50

Chapter II_______________________________________________________________________ 51

9
 II.1. Motivation ________________________________________________________________ 53

II.2. Experimental methods _______________________________________________________ 53

II.2.1. Raw material ___________________________________________________________ 53

II.2.2. Sorting in different categories______________________________________________ 53

II.2.3. Grinding process ________________________________________________________ 54

II.2.4. Cleaning process ________________________________________________________ 54

II.2.5. Characterization of raw feathers ____________________________________________ 55

II.3. Results and discussion _______________________________________________________ 58

II.3.1. Proximate analysis ______________________________________________________ 58

II.3.2. Crude fat content________________________________________________________ 59

II.3.3. Protein content _________________________________________________________ 60

II.3.4. SDS-PAGE analysis _____________________________________________________ 60

II.3.5. Feather-solvent interaction ________________________________________________ 60

II.3.6. Chemical durability test __________________________________________________ 62

II.3.7. Hydrophobic properties __________________________________________________ 64

II.3.8. Fourier Transform Infrared (FT-IR) spectroscopy ______________________________ 64

II.3.9. Thermogravimetric Analysis (TGA)_________________________________________ 65

II.3.10. Differential Scanning Calorimetry (DSC) ___________________________________ 66

II.3.11. X-ray diffraction (XRD) _________________________________________________ 67

II.3.12. Scanning Electron Microscopy (SEM) ______________________________________ 68

II.3.13. Solid-state 13C CP MAS Nuclear Magnetic Resonance (NMR) spectroscopy ________ 69

II.4. Conclusion ________________________________________________________________ 71

Chapter III ______________________________________________________________________ 73

III.1. Motivation _______________________________________________________________ 75

III.2. Experimental Methods ______________________________________________________ 75

III.2.1. Materials _____________________________________________________________ 75

III.2.2. Solvents preparation ____________________________________________________ 75

III.2.3. Keratin extraction parameters involved ______________________________________ 76

III.2.4. Calculation of regenerated keratin yield _____________________________________ 78

III.3. Results and discussion ______________________________________________________ 78

III.3.1. Solvent preparation _____________________________________________________ 78

10
 III.3.2. Regenerated keratin yield comparison according to solvent and parameters used _____ 81

III.4. Conclusion _______________________________________________________________ 91

Chapter IV ______________________________________________________________________ 93

IV.1. Motivation _______________________________________________________________ 95

IV.2. Experimental methods ______________________________________________________ 95

IV.2.1. Materials _____________________________________________________________ 95

IV.2.2. Characterization of regenerated keratin _____________________________________ 95

IV.3. Results and Discussion ______________________________________________________ 96

IV.3.1. Keratin density ________________________________________________________ 97

IV.3.2. Scanning Electron Microscopy (SEM) ______________________________________ 97

IV.3.3. Solid-state 13C CP MAS Nuclear Magnetic Resonance (NMR) spectroscopy ________ 98

IV.3.4. Influence of the extraction temperature______________________________________ 99

IV.3.5. Influence of the extraction time___________________________________________ 106

IV.3.6. Influence of feathers size________________________________________________ 111

IV.3.7. Influence of F:S ratio___________________________________________________ 111

IV.3.8. Influence of the acid used for keratin regeneration ____________________________ 114

IV.4. Conclusion ______________________________________________________________ 119

Chapter V _____________________________________________________________________ 121

V.1. Motivation _______________________________________________________________ 123

V.2. Experimental methods ______________________________________________________ 123

V.2.1. Materials _____________________________________________________________ 123

V.2.2. TOCNF preparation ____________________________________________________ 124

V.2.3. TOCNF characterization_________________________________________________ 125

V.2.4. MFC films preparation __________________________________________________ 127

V.2.5. MFC film characterization _______________________________________________ 129

V.2.6. TOCNF films preparation ________________________________________________ 131

V.2.7. TOCNF film characterization _____________________________________________ 132

V.3. Results and discussion ______________________________________________________ 133

V.3.1. MFC(A) film properties _________________________________________________ 133

V.3.2. MFC(B) film properties _________________________________________________ 138

V.3.3. Chemical properties of TOCNF ___________________________________________ 142

11
 V.3.4. TONCF film properties _________________________________________________ 143

V.4. Conclusion _______________________________________________________________ 160

GENERAL CONCLUSION _______________________________________________________ 163

PERSPECTIVES ________________________________________________________________ 165

REFERENCES _________________________________________________________________ 167

12
LIST OF FIGURES
Figure 1. Flocks of ducks by department end 2010 [9]. __________________________________________________________ 24
Figure 2. Detailed feather composition [14]. _____________________________________________________________________ 25
Figure 3. Number of publications related to "duck feather*" and "chicken feather*" in the last decade. _____ 26
Figure 4. The 20 amino acids list. _________________________________________________________________________________ 27
Figure 5. Illustration of protein hierarchical structure from OpenStax Concepts of Biology. __________________ 28
Figure 6. The different interactions of protein structures [25]. _________________________________________________ 29
Figure 7. Phase diagram of two components eutectic mixture [59]. ____________________________________________ 34
Figure 8. Examples of halide salts and hydrogen bond donors used for DES [60]. _____________________________ 34
Figure 9. Mechanism of feather dissolution in DES (ChCl:urea) adapted from [63]. ___________________________ 35
Figure 10. Dissolution of keratin in NMMO solvent and regeneration mechanisms [31]. ______________________ 37
Figure 11. Biopolymers classification by source [92]. ___________________________________________________________ 42
Figure 12. Schematic representation of the hierarchical structure of cellulose fibers [95]. ___________________ 44
Figure 13. Mechanism of cellulose TEMPO-mediated oxidation in water at pH=10-11 [110]. ________________ 46
Figure 14. Properties of nanocellulose (BNC, CNC, NFC) compared with conventional synthetic and bio-based
polymers: (i) Tensile strength in function of elongation at break and (ii) water vapor permeability in function
of oxygen permeability. Reproduced from [120]. ________________________________________________________________ 48
Figure 15. Picture of the Mulard duck species (La Dépêche, DDM illustration - Nedir DEBBICHE). ___________ 53
Figure 16. The seven feather categories after sorting. ___________________________________________________________ 54
Figure 17. Moisture content, ash content, volatile matter, and fixed carbon content values in percentage for
each feather’s category and section. ______________________________________________________________________________ 59
Figure 18. FT-IR spectrum of extracted fat from duck feather after Soxhlet extraction. _______________________ 59
Figure 19. Biuret test result: before the addition of Biuret reagent (left) and after the addition of Biuret
reagent (right). ____________________________________________________________________________________________________ 60
Figure 20. SDS-PAGE of feathers, lane (1) Prestained Protein Ladder (10- 250 kDa), (2) feathers, and (3)
bovine albumin standard. _________________________________________________________________________________________ 61
Figure 21. Picture of the feather-solvent interaction test with different solvents. _____________________________ 61
Figure 22. Picture of feather sections after immersion in different chemical conditions. ______________________ 62
Figure 23. Picture of hydrophobicity test on the different feather’s categories and their sections in
comparison with cotton and cellulose pulp. ______________________________________________________________________ 64
Figure 24. Comparison of FT-IR spectra of each category (left) and different sections (right) of duck feathers.
_____________________________________________________________________________________________________________________ 65
Figure [Link] of TGA curves of each category (left) and different sections (right) of duck feathers. 65
Figure 26. Comparison of DSC curves of each category (left) and different sections (right) of duck feathers. 66
Figure 27. Comparison of XRD curves of each category (left) and different sections (right) of duck feathers. 67
Figure 28. SEM images of different sections of duck feathers: barbs, barbules, rachis inside, and calamus
outside. Corresponding magnification: a) x100, b) and d) x500, c) x1500 and e) x 2000. _____________________ 69
Figure 29. Solid-state 13C CP MAS NMR spectrum of random feather category. The spinning band is noted by *.
_____________________________________________________________________________________________________________________ 70
Figure 30. General scheme for keratin extraction by solvolysis. _________________________________________________ 77
Figure 31. FT-IR spectrum of NMMO solvent at 50% concentration and 75% concentration after rotary
evaporator step. ___________________________________________________________________________________________________ 79
Figure 32. Elaboration mechanism of choline chloride:urea (1:2) eutectic solvent modified from [150]. ____ 79

13
Figure 33. Aspect of ChCl:urea (1:2) mixture after a few minutes at RT. Left: mixture without prior drying of
reagents, right: with drying of reagents. _________________________________________________________________________ 80
Figure 34. Heating time required to homogenize DES solvent as a function of temperature (left) and the
quantity of DES prepared (right). ________________________________________________________________________________ 81
Figure 35. FT-IR spectra of ChCl, urea, and the ChCl:urea (1:2) mixture. _______________________________________ 82
Figure 36. Graphs of feathers dissolution time and regenerated keratin yield as a function of heating
temperature. Keratin was extracted with NMMO 50% (left) and NMMO 75% (right) with fixed parameters:
feather size 5 mm and F:S ratio=0.10 (based on 50g solvent). __________________________________________________ 82
Figure 37. Graph of feathers dissolution time and regenerated keratin yield as a function of heating
temperature. Keratin was extracted with DES with fixed parameters: feather size 5 mm and F:S ratio=0.10
(based on 50g solvent). ____________________________________________________________________________________________ 83
Figure 38. Diagram presenting regenerated keratin yield as a function of extraction time. Keratins extracted
with fixed parameters: feather size = 5mm, F:S ratio = 0.10 (based on 50g solvent), temperature = 140 °C for
DES and 110 °C for NMMO. _______________________________________________________________________________________ 85
Figure 39. Diagram presenting regenerated keratin yield as a function of feathers size. Keratins extracted
with fixed parameters: F:S ratio = 0.10 (based on 50 g solvent), temperature = 140 °C during 180 min for DES
and 120 °C during 480 min for NMMO. ___________________________________________________________________________ 86
Figure 40. Diagram presenting regenerated keratin yield as a function of F:S ratio (based on 100 g of
solvent). Keratins extracted with fixed parameters: temperature = 140 °C during 180 min for DES and 120 °C
during 480 min for NMMO.________________________________________________________________________________________ 87
Figure 41. Pictures of the keratin regeneration step by acidification at pH=4. Picture A): keratin obtained
with DES, picture B) keratin obtained with NMMO. _____________________________________________________________ 88
Figure 42. Diagram presenting regenerated keratin yield and acid volume as a function of acid type used for
keratin regeneration step. Keratins extracted with fixed parameters: feathers size = 5 mm, F:S ratio=0.10
(based on 50 g of solvent), temperature = 140 °C for DES and 120 °C for NMMO. _____________________________ 89
Figure 43. SEM images of regenerated keratin extracted with (A) NMMO 75% : magnification ×1,000 and
×10,000 and (B) DES : magnification ×1,000 and ×10,000. _____________________________________________________ 98
Figure 44. NMR spectra of regenerated keratin extracted with NMMO 75% at 110 °C and DES (ChCl:urea) at
140 °C. The spinning band is noted by *. __________________________________________________________________________ 99
Figure 45. Visual aspect of regenerated keratins extracted with NMMO 75% and DES, depending on the
extraction parameter. ____________________________________________________________________________________________ 100
Figure 46. SDS-PAGE pattern of regenerated keratins. Picture (A) for NMMO 75% solvent, lane (1) Prestained
Protein Ladder (10-250 kDa), (2) 110 °C, (3) 160 °C, (4) 6 h, (5) 10 h, (6) H2O, (7) acetic acid, (8) sulfuric
acid, (9) citric acid, (10) bovine albumin standard. Picture (B) for DES solvent, lane (1) Prestained Protein
Ladder (10-250 kDa), (2) 130 °C, (3) 170 °C, (4) 1h30, (5) 4 h, (6) H2O, (7) acetic acid, (8) sulfuric acid, (9)
citric acid, (10) bovine albumin standard. ______________________________________________________________________ 101
Figure 47. FT-IR spectra of regenerated keratin depending on extraction temperature. Extracted with: (A)
NMMO and (B) DES (ChCl:urea)._________________________________________________________________________________ 102
Figure 48. TGA of regenerated keratin depending on extraction temperature. Extracted with: (A) NMMO 75%
and (B) DES (ChCl:urea). _________________________________________________________________________________________ 103
Figure 49. DSC analysis of regenerated keratin extracted with (A) NMMO 75% and (B) DES (ChCl:urea)
depending on the extraction temperature. ______________________________________________________________________ 104
Figure 50. XRD analysis of regenerated keratin depending on extraction temperature. Extracted with: (A)
NMMO 75% and (B) DES (ChCl:urea). ___________________________________________________________________________ 106

14
Figure 51. FT-IR spectra of regenerated keratin depending on extraction time. Extracted with: (A) NMMO
and (B) DES (ChCl:urea). _________________________________________________________________________________________ 108
Figure 52. TGA of regenerated keratin depending on extraction time. Extracted with: (A) NMMO 75% and (B)
DES (ChCl:urea). __________________________________________________________________________________________________ 109
Figure 53. DSC analysis of regenerated keratin depending on extraction time. Extracted with: (A) NMMO 75%
and (B) DES (ChCl:urea).The endothermic peak located at around 60 °C and indicated by * is an artefact due
to the instrument. ________________________________________________________________________________________________ 109
Figure 54. XRD analysis of regenerated keratin depending on extraction time. Extracted with: (A) NMMO
75% and (B) DES (ChCl:urea). ___________________________________________________________________________________ 110
Figure 55. FT-IR spectra of regenerated keratin depending on F:S ratio. Extracted with: (A) NMMO and (B)
DES (ChCl:urea). __________________________________________________________________________________________________ 111
Figure 56. TGA of regenerated keratin depending on F:S ratio. Extracted with: (A) NMMO 75% and (B) DES
(ChCl:urea). _______________________________________________________________________________________________________ 112
Figure 57. DSC analysis of regenerated keratin depending on F:S ratio. Extracted with: (A) NMMO 75% and
(B) DES (ChCl:urea). The endothermic peak located at around 60 °C and indicated by * is an artefact due to
the instrument. ___________________________________________________________________________________________________ 113
Figure 58. XRD analysis of regenerated keratin depending on F:S ratio. Extracted with: (A) NMMO 75% and
(B) DES (ChCl:urea). ______________________________________________________________________________________________ 114
Figure 59. FT-IR spectra of regenerated keratin depending on regeneration step. Extracted with: (A) NMMO
and (B) DES (ChCl:urea). _________________________________________________________________________________________ 115
Figure 60. TGA of regenerated keratin depending on acid used during regeneration step. Extracted with: (A)
NMMO 75% and (B) DES (ChCl:urea). ___________________________________________________________________________ 116
Figure 61. DSC analysis of regenerated keratin depending on acid used during regeneration step. Extracted
with: (A) NMMO 75% and (B) DES (ChCl:urea).The endothermic peak located at around 60 °C and indicated
by * is an artefact due to the instrument.________________________________________________________________________ 116
Figure 62. XRD analysis of regenerated keratin depending on acid used during regeneration step. Extracted
with: (A) NMMO 75% and (B) DES (ChCl:urea). ________________________________________________________________ 117
Figure 63. Solid-state 13C CP MAS NMR spectrum of regenerated keratin with the use of water or acids. Image
(A) correspond to keratin extracted with NMMO and image (B) correspond to keratin extracted with DES.
The spinning band is noted by *. _________________________________________________________________________________ 118
Figure 64. Pictures of regenerated keratin extracted via NMMO solvolysis, humid form (left) and dry form
(right). ____________________________________________________________________________________________________________ 124
Figure 65. Scheme summarizing the manufacturing processes for MFC and TOCNF films, from process P1 to
process P4. ________________________________________________________________________________________________________ 128
Figure 66. Picture of MFC 2% (left) and MFC(A) films (right), a) MFC(A), b) MFC(A)-K5, c) MFC(A)-K10, d)
MFC(A)-K15, and e) MFC(A)-K20. _______________________________________________________________________________ 134
Figure 67. Schematic diagram of the contact angle principle (left), contact angle values of MFC(A) films. _ 135
Figure 68. The stress-strain curves of MFC(A) films (left) and diagram of stress at break (MPa), strain at
break (%) and Young’s modulus (MPa) values for MFC(A) films (right). ______________________________________ 136
Figure 69. TGA and DTG curves of MFC(A) films. _______________________________________________________________ 137
Figure 70. FT-IR spectra of MFC(A) films. _______________________________________________________________________ 137
Figure 71. Pictures of MFC(B) films. A) MFC(B) thick, B) MFC(B) thin, C) MFC(B)-K5, D) MFC(B)-K10, E)
MFC(B)-K15, F) MFC(B)-K20. ____________________________________________________________________________________ 138
Figure 72. SEM images of MFC(B) films: A) is MFC(B) thin, B) is MFC(B)-K5, C) is MFC(B)-K10, D) is MFC(B)-
K15, E) is MFC(B)-K20. ___________________________________________________________________________________________ 139

15
Figure 73. XRD diffractograms of MFC(B) films. ________________________________________________________________ 140
Figure 74. Contact angle of MFC(B) films. _______________________________________________________________________ 141
Figure 75. Diagram of stress at break (MPa) and strain at break (%) for MFC(B) films. _____________________ 141
Figure 76. TGA and DTG curves of MFC(B) films. _______________________________________________________________ 142
Figure 77. Simplified scheme of TOCNF fabrication process. ___________________________________________________ 143
Figure 78. SEM images of TOCNF-5-P film, magnification x1000 (left) and magnification x10000 (right). _ 144
Figure 79. Picture of TOCNF-5-P film (left), TOCNF-5-S film (center), TOCNF-15-S film (right). _____________ 145
Figure 80. WVTR (g/m²*day) of TOCNF-5-P, TOCNF-5-S and TOCNF-15-S films. _____________________________ 147
Figure 81. Diagram of stress at break (MPa) and strain at break (%) of TOCNF-5-P, TOCNF-5-S and TOCNF-
15-S films. _________________________________________________________________________________________________________ 147
Figure 82. TGA and DTG curves of TOCNF films by solvent casting method.___________________________________ 148
Figure 83. Picture of TOCNF films: TOCNF-5-P (left), TOCNF-5-P-DK10 (center), and TOCNF-5-P-HK10
(right). ____________________________________________________________________________________________________________ 149
Figure 84. XRD diffractograms of TOCNF-5-P films depending on the type of keratin incorporated. ________ 150
Figure 85. WVTR (g/m²*day) of TOCNF-5-P films depending on the type of keratin incorporated.__________ 151
Figure 86. Diagram of stress at break (MPa) and strain at break (%) of TOCNF-5-P films depending on the
type of keratin incorporated. ____________________________________________________________________________________ 151
Figure 87. TGA and DTG curves of TEMPO-5-P films depending on the type of keratin incorporated. _______ 152
Figure 88. Picture of TOCNF-5-P films with humid keratin incorporation: A) 0%, B) 5%, C) 10%, D) 20%. _ 152
Figure 89. WVTR (g/m²*day) of TOCNF-5-P films with humid keratin incorporation. _______________________ 153
Figure 90. Diagram of stress at break (MPa) and strain at break (%) values of TOCNF-5-P films with humid
keratin incorporation. ____________________________________________________________________________________________ 154
Figure 91. TGA and DTG curves of TOCNF-5-P films with incorporation of humid keratin. __________________ 154
Figure 92. Pictures of TOCNF-5-P films, A) TOCNF-5-P-20sorbitol, B) TOCNF-5-P-10PEG, C) TOCNF-5-P-
20PEG, D) TOCNF-5-P-40PEG. ___________________________________________________________________________________ 155
Figure 93. WVTR (g/m²*day) of TOCNF-5-P films with addition of plasticizer, PEG400 or sorbitol. _________ 156
Figure 94. Diagram of stress at break (MPa) and strain at break (%) of TOCNF-5-P with addition of
plasticizer, PEG 400 or sorbitol. _________________________________________________________________________________ 156
Figure 95. TGA and DTG curves of TOCNF-5-P films with addition of plasticizer, PEG 400 or sorbitol. ______ 157
Figure 96. Picture of TOCNF-5-P-HK5-20PEG ___________________________________________________________________ 158
Figure 97. TGA and DTG curve of TOCNF-5-P-HK5-20PEG._____________________________________________________ 159

16
LIST OF TABLES
Table 1. Examples of keratin film elaboration in the literature. ________________________________________________ 41
Table 2. Examples of biopolymers films with keratin in the literature. _________________________________________ 43
Table 3. Examples of cellulose films in the literature and their properties. _____________________________________ 49
Table 4. Dimensional characteristics and percentages of the six feather categories after sorting. ___________ 55
Table 5. Observations regarding the physical properties of feather sections after immersion in different
chemical conditions. _______________________________________________________________________________________________ 63
Table 6. Crystallinity index of the different feathers categories and sections calculated from XRD curves. ___ 68
Table 7. Parameters studied in the preparation of the DES. _____________________________________________________ 76
Table 8. Summary of the different parameters studied for keratin extraction as related to the solvent used. 77
Table 9. Examples of various keratin extraction protocols from poultry feathers and their associated yield. 90
Table 10. Bulk density, sintered density and porosity of keratin extrated with DES or NMMO. _______________ 97
Table 11. Crystallinity Index of regenerated keratin depending on solvent and extraction parameters. ____ 107
Table 12. Thickness, density, grammage, and porosity of MFC(A) films. _______________________________________ 134
Table 13. Thickness, density, grammage, porosity, and CI of MFC(B) films. ___________________________________ 139
Table 14. Thickness, density, grammage, porosity, and CI of TOCNF films. ____________________________________ 146

17
INDEX OF ABBREVIATIONS
Chemical compounds

CH3COOH Acetic acid

CED Copper (III) ethylenediamine
ChCl Choline Chloride
CMC Cellulose microcrystal
CNC Cellulose nanocrystal
CNF Cellulose nanofibers
DES Deep eutectic solvents
H2S Hydrogen sulfide
H2SO4 Sulfuric acid
HCl Hydrochlorydric acid
Ils Ionic liquids
Ker Keratin
KOH Potassium hydroxide
MB Methylene blue
MC Methyl cellulose
MFC Microfibrillated cellulose
Na2CO3H2O Sodium carbonate
NaBr Sodium bromide
NaClO Sodium hypochlorite
NaClO2 Sodium chlorite
NaOAc Sodium acetate
NaOH Sodium hydroxide
NH3 Ammonia
NMMO N-methylmorpholine N-oxide
PEG Polyethylene glycol
Pes-Na Sodium polyethylene sulphate
PHA Polyhydroxyalkanoate
PLA Poly(lactic acid)
polyDADMAC Polydiallyldimethylammonium chloride
PVA Polyvinyl alcool
SDS Sodium dodecyl sulfate
SO2 Sulfur dioxide
TEMPO 2,2,6,6-tetramethylpiperidine-1-oxyl
TOCNF Tempo-oxidized cellulose nanofibers

18
Characterization techniques

ATR Attenuated Total Reflectance

13
Solid-state Cross-Polarization Magic Angle Spinning Carbon-13
C CP/MAS NMR
Nuclear Magnetic Resonance
Ac Ash content
CC Carboxyl Content
CD Cationic Demand
CI Cristallinity Index
DP Degree of Polymerisation
DSC Differential Scanning Calorimetry
E Elastic modulus
Fc Fixed carbon content
FT-IR Fourier Transform Infrared
G Grammage
Mc Moisture content
MW Molecular Weight
P Porosity
SDS-PAGE Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis
SEM Scanning Electron Microscopy
TGA Thermogravimetric Analysis
TS Tensile Strenght
Vc Volatile content
WCA Water Contact Angle
WVP Water Vapor Permeability
WVTR Water Vapor Transmission Rate
XRD X-Ray Diffraction
ε Deformation
ρ Density

Other abbreviations

(v:v) Volume:volume
(v:w) Volume:weight
(w:w) Weight:weight
F:S Feathers:solvent
N/A Not available
RT Room temperature

19
20
 Chapter I
Bibliography

21
22
 I.1. The poultry sector: a source of natural by-products
I.1.1. Worldwide poultry sector
Since the beginning of our era, the world's population has been growing at a more or less variable rate.
According to the historical periods, slow or brutal increases could be observed. The real demographic
growth has been observed since 1950 and will reach 7.7 billion people in 2019, according to a United
Nations report [1]. According to the same report, world population projections predict that the
population will reach 8.5 billion people in 2030, 9.7 billion in 2050, and 10.9 billion in 2100. As a result,
the demand for food production and agriculture is constantly increasing.

The poultry industry, an important sector in global meat consumption, is affected by this population
growth and can cause social and economic problems. Various by-products are generated at the end of
the meat production process. These by-products include skin, bones, fat, blood, feathers, guts, and
numerous microorganisms. The primary concern associated with these by-products is their typical
disposal through burial or incineration, often carried out without adequate precautionary measures. For
example, feathers can be stored in landfills for hours, leading to the proliferation of bacteria and the
formation of bad odors, which can cause environmental and social problems [2]. The need for the
reduction of these negative impacts is on the rise.

I.1.2. New Aquitaine region and Landes department poultry sector

On a smaller scale, the Southwest of France, particularly the New Aquitaine Region, is a major
production area for foie gras ducks, as presented in Figure 1. The fattened duck is a duck of the Mulard
species, a cross between a Barbary duck (Cairina moschata) and a Peking duck (Anas platyrhynchos
domestica). This type of duck is selected for its meat and foie gras production. Its production alone
represents 50% of the French farms. The webfoot sector represents a major part of the region's food
production. In 2019, the production of ducks in the region was 42,653 tons, of which 17,430 were in the
Landes, representing 40.8% of the region's production [3]. A Mulard duck weighs about 5 kg and has a
feather content of about 200 g/duck; this is the equivalent of 8,530,600 fattened ducks and 1,700 tons
of feathers in the New Aquitaine region in 2019. More recent data demonstrate the evolution of this
production, which reached 8.9 million ducks in 2023, representing 1,780 tons of feathers [4]. Several
recycling methods can give these wastes a second life and minimize environmental risks. The best
known is reusing small, soft feathers or down for bedding or clothing. However, these types of feathers
represent a small percentage of the total mass of feathers produced. Other feathers can be processed into
flour for fertilization of fields in agriculture [5], used for animal feed production as animal protein [6],
or used as a source for extracting keratin for the pharmaceutical and biomedical sectors [7], and cosmetic
sector [8].

23
 Figure 1. Flocks of ducks by department end 2010 [9].

I.2. Feathers as a source of keratin

I.2.1. Feathers
Feathers are light but strong keratinous appendages that cover the skin of all birds, although there are
different types of feathers depending on the species and the location of the feather on the bird's body.
Feathers can vary in color, size, shape, and whether soft or hard, but they are all waterproof and
lightweight with good flexibility. They are primarily used to protect the skin, but not only. For example,
hard feathers allow the bird to fly and navigate, while soft feathers are used to regulate their temperature
or to float on water. More specifically, a feather is an integumentary appendage consisting of a hard,
tubular main shaft (quill) attached to the barbs. The quill comprises two distinct parts: the calamus, the
empty part of the quill in contact with the bird's skin, and the rachis, the full part of the quill to which
the barbs are attached. On a smaller scale, the latter comprises barbules, which in turn have curved
microhooks that hold the barbules together. These microhooks can maintain the integrity of the quill by
hanging together, allowing a "zipping/unzipping" effect without damage, very close to Velcro behavior
[10] presented in Figure 2. Feathers are created during a cornification process, in which proteins
accumulate in the epidermis to form a protective hard layer. The same principle can give rise to any
form of keratin-based material in vertebrates, such as claws, beaks, horns, and hair [11]. As shown in
Figure 3, most of the literature on feathers deals with chicken commonly produced worldwide for food
consumption and, therefore, is easy to find. As said before, chicken feathers, like general feathers, are
composed of three distinct parts: rachis, barbs, and barbules, creating a hierarchical structure. Indeed,
the first structure, rachis, can reach 17 cm in length, the second structure, barbules, is between 1 to
4.5 cm, and the last structure, barbs, are at microscale with 0.3 - 0.5 mm. Chicken feather is a strong
material But its density is very light: 0.8 g/cm3 lower than cellulose fibers 1.5 g/cm3 or wool 1.3 g/cm3
[12]. The most interesting aspect of the feather remains its chemical composition. It comprises 91%
protein, 8% water, and 1% lipids [13].

24
 Figure 2. Detailed feather composition [14].

Characterization studies have been performed on chicken feathers, unlike duck feathers, for which there
is a lack of information. Tesfaye et al. analyzed the chemical properties of chicken feathers to determine
a possible recovery route. Proximate analysis showed that chicken feathers comprise 12.33% moisture,
0.83% crude lipid, 2.15% crude fiber, 82.36% crude protein, and 1.49% ash. The authors also
demonstrated the durability of the feathers in contact with various solvents: weak and strong acids, weak
and strong bases, and bleaching agents. When immersed in alkaline media, the feathers were damaged
and dissolved rapidly; similarly, when immersed in a strong acid media, the feathers showed good
resistance to weak acids. Bleaching agents should only be used for short periods, as they can damage
and degrade feathers after several hours. The study also confirmed the feathers' excellent resistance to
water with no degradation [15].

Besides their chemical composition, the mechanical properties of feathers can be of great interest when
considering their use. Therefore, the mechanical properties of chicken feathers were investigated by
Zhan and Wool. This article separated a chicken feather barb from the feather to test its tensile strength
and measure its deformation and Young's modulus. For this purpose, the stress-strain curve of one fiber
was analyzed, and the values obtained were 6.93% strain at break, 203 ± 74 MPa for average tensile
strength, and an average tensile modulus of 3.59 ± 1.09 GPa [16].

I.2.2. Proteins
Proteins are large and complex functional molecules that play a crucial role in the body of a living
organism. They are natural, renewable, and biodegradable polymers from different sources, such as
animals, plants, or bacteria. Each type of protein provides a distinct functionality depending on its
composition and structure. Like any polymer, proteins are composed of a large sequence of monomers

25
 Figure 3. Number of publications related to "duck feather*" and "chicken feather*" in the last decade.

called amino acids. Each amino acid is linked to another by a peptide bond and forms a polypeptide
chain. Proteins are, therefore, differentiated by their different amino acid chains. An amino acid, as its
name indicates, is composed of an amino group (NH2), a carboxylic group (COOH), and a group of
amino acids (-R), which differ depending on the type of protein. There are 20 natural amino acids [17];
therefore, there are as many different types of proteins as assemblies of different amino acids. The list
of the 20 amino acids is shown in Figure 4. The complexity of proteins also comes from their structure
in space. There are 4 levels of structure, the primary, secondary, tertiary, and quaternary structure
represented in Figure 5.

The primary structure corresponds to the protein sequence, i.e. the sequence of amino acids connected
by peptide bonds. An amino acid sequence is specific to each protein. The conformation of the
polypeptide chain defines the secondary structure. The main characteristic is the planarity of this
polypeptide chain that gives birth to several structures, of which the most important are α-helix and
β-sheet structures [17]. The compact molecular structure of polypeptide chains is called tertiary
structure. It corresponds to the overall three-dimensional conformation of the protein. This complex
level of structure gives the protein these biological functions, which differ according to its composition
and structure. The last level of hierarchical structure is the quaternary structure. It is the complex overall
structure of an oligomeric protein composed of several subunits connected by non-covalent bonds.

I.2.3. Keratin
Keratin is a protein known and used as a material by humans for centuries. The origin of the word keratin
comes from the Greek "κερας" which meant "horn" [18]. Keratin was used for various purposes, such
as ornaments, jewelry, clothing, and tools using natural elements from animals like claws, beaks,
hooves, nails, wool, and skins. Initially, keratin serves as a protective barrier against the elements of
nature or as a defense against predators. Today, keratin is a more widely used term to describe a type of
fibrous protein found in various animal species in the epidermal tissue of vertebrates' skin. Keratin and
collagen represent the most important biopolymers in animals [19], present in reptiles, birds, and
mammals. As mentioned above, a feather is approximately 91% keratin protein. Several studies have
26
 Figure 4. The 20 amino acids list.

examined the feather amino acid composition of different animal origins [20]–[23]. According to the
literature, scientists agree that feather keratin consists of between 7% and 9% cysteine. This makes it a
cysteine-rich protein. A large number of hydrophobic amino acids also make up the largest percentage
of this composition, giving keratin its high resistance and hydrophobic properties. The amount of
cysteine in keratin determines its specificity. Cysteine allows the formation of intramolecular and
intermolecular disulfide bridges (-S-S-), thus influencing its physicochemical properties. The higher the
sulfur content, the more resistant the keratin. There are also other types of interactions that maintain the

27
 Figure 5. Illustration of protein hierarchical structure from OpenStax Concepts of Biology.

integrity of keratin. These include Van der Waals reactions, hydrogen, and ionic bonds, as shown in
Figure 6. Keratin can be divided into two categories: soft keratin and hard keratin. Soft keratin comes
from the epidermis and is low in sulfur, while hard keratin comes from mammals, reptiles, or birds and
is found in appendages such as nails, hooves, horns, and feathers and is high in cysteine [24]. Another
classification also distinguishes between keratins with different secondary structures. These different
ordered patterns are classified as α-helix or β-sheet structures and illustrated in Figure 6.

The first consists of a helical polypeptide chain whose structure is maintained by hydrogen bonds
between the (-N-H-) and (-C=O-) groups of its amino acids. This is the most common form found in
mammals. The second structure is common to birds and reptiles and consists of the parallel or
antiparallel assembly of several keratin strands held together by hydrogen bonds between the strands.
Despite their structural differences, α-helix and β-sheet keratins are similar in their nanoscale filament
structure. Both types of keratin appear as a filament-matrix structure. The notable difference between
the two patterns is their molecular size. The α-keratin has a molecular size ranging from 40 to 68 kDa,
while β-keratin is shorter, ranging from 10 to 22 kDa [19].

28
 Figure 6. The different interactions of protein structures [25].

I.3. Feathers processing

I.3.1. Feathers cleaning methods
Because of its chemical structure and structural properties, feather keratin is of great interest to the
scientific research community. However, feathers come from the agri-food industry after the plucking
of poultry raised and killed for meat production. Feathers are, therefore, in direct contact with all the
pathogens, microorganisms, and bacteria in slaughterhouses and the duck. For example, they are in
contact with meat, dirt, insects, blood, and excrement. Therefore, feathers must be treated and cleaned
to ensure they are clean and healthy for use before any keratin extraction. To this end, various cleaning
processes, products, and optimizations have been studied to clean feathers and eliminate pathogens,
bacteria, and microorganisms. In the literature, different cleaning processes exist. Single or multi-stage
cleaning can be found at high or ambient temperatures using water, Soxhlet extraction with ethanol,
oxidants, surfactants or detergents, and manual washing or washing machines.

Pourjavaheri et al. monitored the microbiological, chemical composition, and mechanical properties of
white chicken feathers treated with surfactants, disinfectants, and sanitizers by comparing them to
determine which method is most effective against bacteria that can cause health problems without
degrading keratin. Methods compared are Soxhlet extraction with ethanol (SEEt), oxidation by ozone,
purification by Chlorine Dioxide (ClO2), purification with polyethylene glycol (PEG) such as non-ionic
surfactant, with Sodium Lauryl Sulphate (SLS) such as an anionic surfactant and
CetylTrimethylAmmonium Chloride (CTAC) such as cationic surfactant and finally purification by a
combination method (SLS–ClO2–SEEt). The results demonstrate that several cleaning methods have
advantages and drawbacks to consider depending on the desired use. Indeed, the extraction by ethanol
is effective against salmonella and other kinds of bacteria but also eliminates grease. On the other hand,
this method is not ideal for mechanical properties and especially tensile strength because of its decrease.
The results favor the combination method (SLS-ClO2-SEEt) [26]. Surfactants were also studied by

29
Sharma et al., and they investigated the effect of a first and second wash on the number of bacteria in
chicken feathers. The first wash consisted of rinsing with water at 60 °C, then degreasing with petroleum
ether for 12 h and rinsing with water. A second cleaning is then performed at 40 °C for 3 h using various
aqueous surfactant solutions such as sodium dodecyl sulfate (SDS) (anionic surfactant), polyethylene
glycol (PEG) (non-ionic surfactant), and cetrimonium bromide (CTAB) (cationic surfactant). The
authors conclude that the most effective surfactant is CTAB, as this treatment significantly reduces
bacteria and eliminates salmonella without damaging the structure of the feathers and fibers [27]. In
addition to the type of product being used, temperature is an important parameter to consider. Strnad
and Kreze studied the optimization of cleaning methods concerning poultry feathers regarding the
temperature, the addition of detergent, and the use of ultrasound. For the study, they used pre-washed
feathers at an industrial level that have been rinsed with water on a grid to eliminate big impurities.
Then, different cleaning conditions were reported to be compared. The results are based on measuring
water turbidity and solid matter remaining in the cleaning bath after drying. Experiments show that
adding non-ionic washing agent and a higher temperature can improve the cleaning by increasing the
proportion of solid matter in the cleaning bath but above 90 °C, the temperature has no more influence
on this parameter. Based on the results of turbidity and organic matter removal and regarding energy
consumption, the authors concluded that the best method of cleaning is at 40 °C with addition of washing
agent [28]. Washing methods may also differ when industrial machines such as washing machines are
used. Safaric et al. used an industrial cleaning with water to eliminate the major dirtiness. Then, at the
laboratory scale, poultry feathers are cleaned in a commercial washing machine at 60 °C with non-ionic
detergent after putting feathers in a mesh bag. Photos were taken after cleaning and showed white
feathers that seem clean. After cleaning, feathers required a drying step by a modified vacuum drying
method at a pressure lower than 80 mbar and a temperature between 36 and 42 °C. Finally, the authors
explain that after 29-34 hours, the feather moisture was equal to 10%, and the smell was neutral [29].
Casadesus et al. also used the washing machine by mixing the feathers with hydrogen peroxide solution
(H2O2) for 50 minutes at 35 °C and 3300 ppm. After the cleaning, feathers are oven dried at 60 °C for
24 h [30]. Pre-extraction cleaning using a mixture of hexane and dichloromethane can also be found in
the experiments of Ma et al. [31]. Utami et al. cleaned duck feathers in two consecutive steps. The first
step involved cleaning with water and detergent, then drying in an oven at 50 °C for 24 hours. The
second step is a grinding and a second cleaning with hydrochloric acid and petroleum ether for 24 hours,
followed by rinsing with water [32]. Simpler cleaning methods are also used, such as Ullah et al. who
wash the feathers several times with hot water and soap, then degrease them by Soxhlet extraction with
petroleum ether [33], and Gupta et al. who, conversely, first degrease the feathers with ether for 24
hours, then clean them with soap and water [34].

It is clear, therefore, that cleaning methods vary widely from one study to the next, with the majority
using different steps and products, most of which are toxic, costly, and not environmentally friendly.

I.3.2. Keratin extraction methods

I.3.2.1. History
From a scientific point of view, the first use of keratin began in the 16th century. According to literature,
a Chinese herbalist, Li Shi-Zen, was the first to use keratin in medicine. He used a substance derived

30
from pyrolyzed human hair to aid in the healing process [35]. At that time and in the following years, it
was possible to extract it from various sources such as claws, beaks, hooves, nails, and hair, among
others. The word keratin first appeared in scientific literature in 1850. At that time, this word described
the solid material present in hard tissues such as horns or hooves [36]. In the 20th century, the emergence
of particular attention concerning the extraction of keratin and the study of different processes on
different sources of keratin can be noticed. In fact, keratin is a protein that does not behave like other
proteins from the point of view of dissolution. The generally used protein dissolution processes do not
work on this material. However, in 1905, an extraction of keratin from hooves using a solution
containing lemon was carried out by the American John Hoffmeier [37]. Subsequently, various
extraction processes were established for medical applications, and keratin powders were used in
cosmetics [38][39]. In the 1920s, scientific research deepened the study of keratin, not only in its
extraction but also in the analysis and understanding of its structure and functionality [7]. Biomaterials
emerged as a scientific concern in 1940. The properties of keratin became important for creating
materials, and several keratin classes were developed [36].

I.3.2.2. Denaturation
One of the most important properties of proteins that determine their behavior and functionality is their
solubility. This is the ability of the protein to dissolve in a solvent to form a homogeneous mixture with
it. This dissolution is controlled by the protein-solvent equilibrium and by interactions within and
between proteins. On the one hand, external parameters, such as temperature and pH environment, can
influence the solubility of proteins [17]. Typically, the solubility of proteins increases with increasing
temperature, and protein extractions are performed at high temperatures.

On the other hand, the internal parameters of the protein play a crucial role in its solubility. Indeed,
interactions at different scales of its structure allow the preservation of its structure. The rupture of
disulfide bridges can allow the denaturation of the protein. "Denaturation" means a loss of order in its
super molecular structure without degradation of the initial molecular weights [17]. A denaturing agent
can break hydrogen and hydrophobic bonds in the protein.

Nowadays, the most commonly used extraction methods are based on oxidation [40], reduction [41],
alkaline hydrolysis [42], ionic liquids [43], eutectic solvents [44][45], microbial and enzymatic methods
[46], steam explosion [47], microwave irradiations [25] and sonication.

I.3.2.3. Chemical extraction

Oxidation

Earland et al. oxidized goose feathers after cleaning and degreasing by Soxhlet extraction with alcohol
and ether. The authors point out that the feather strength requires a large excess of peracetic acid solution
(2%) over the feather mass of 500% to complete oxidation [48]. Peracetic acid was also used by
Brown et al., who compared several methods of keratin extraction, including two oxidations. In the first,
wool fibers were oxidized overnight in a 2% peracetic acid solution, heated to 37 °C and mixed at
180 rpm. The oxidized fibers are then filtered and rinsed with water before extracting the keratin with a
Tris-based solution (pH=10.5) twice for 3 hours under the same conditions. The pH of the resulting
solution is then lowered to pH=4 to precipitate the keratin. The second oxidation is carried out with
31
percarbonate. The wool was first hydrated in water overnight and then oxidized with a solution of NaOH
(30% in water) at 3% of the dry wool mass. Percarbonate is then added at 4.5% of the wool's dry weight
and the mixture is stirred for 3 to 4 hours until the wool is completely dissolved. In the same study, a
combination of reduction and oxidation is also compared by mixing the wool with a 0.2 M sodium
thioglycolate solution (pH=12) at room temperature overnight to oxidize the fibers. The mixture was
then reduced to pH=10.5 after dilution with water and HCl (0.5 M) added before centrifugation to obtain
the supernatant. Oxidation is then carried out for 3 hours with 0.25 mL of 3% hydrogen peroxide added
to the supernatant previously adjusted to pH=7 with HCl. Finally, the pH is lowered to pH=4 with HCl
to precipitate the extracted keratin. This comparison shows that the oxidation methods produce keratin
with a homogeneous molecular weight of over 60 kDa, whereas the combined reduction/oxidation
method produces disparate molecular weights between 40 and 70 kDa [40].

Reduction

Keratin was extracted from feathers and wool by Ma et al. [49] and Wang et al. [50] using a reduction
method. In both studies, the extraction principle is similar: feathers or wool are mixed with a solution
containing urea (8 M) and L-cysteine, with the pH controlled at 10.5 using a NaOH solution. For
feathers, the mixture is heated to 70 °C for 12 h, then centrifuged at 10 000 rpm for 20 min to recover
the supernatant, which is then adjusted to pH=4 with HCl and sodium sulfate to obtain solid keratin.
The extraction yield is estimated to be 60%. Wool is heated for 5 h at 75 °C and 72% dissolution is
obtained. The solution is then filtered and dialyzed for 3 days at room temperature. The keratin is then
precipitated by lowering the pH to 4.0 - 4.5 and recovered by centrifugation. In both cases, the final step
is a freeze-drying to obtain keratin powder. However, Wang et al. claim that this extraction method
reduces the disulfide bonds present in the wool by 62%, resulting in a 25 °C reduction in the maximum
decomposition temperature of the regenerated keratin. The same result was observed in the study by
Ma et al., with a slight decrease of 10 °C in the degradation temperature of regenerated keratin. Tonin
et al. used a mixture of urea (8 M), m-bisulfite (0.5 M), and SDS at pH=6.5 adjusted with NaOH to
dissolved wool. Different extraction parameters were studied by varying the amount of SDS (0, 0.3, and
0.6 SDS g/g wool), time (1 h or 5 h), and temperature (65 °C or 100 °C). Each solution was filtered,
dialyzed for 3 days, and then centrifuged to remove any remaining solids. Yields ranged from 22 to 33%
[51].

Alkaline hydrolysis

Alkaline hydrolysis is also a way of extracting keratin, as shown in the study of Sinkiewicz et al., who
demonstrate that the extraction yield is greater with alkaline hydrolysis than with a reduction method.
Keratin was extracted from feathers using a NaOH solution, comparing the influence of its concentration
ranging from 1.0 to 2.5%. The feather:solvent mixture was heated for 75 min at 70 °C. The results show
that extraction yield increases with NaOH concentration, up to around 94% [52]. The alkaline hydrolysis
method can also be coupled with the enzymatic method, as in the study of Branska et al. where alkaline
hydrolysis is a pre-treatment of the feathers. The crushed feathers are mixed with a KOH solution and
heated to 80 °C for 20 h. In the second step, the enzyme is added to the feathers. The second step consists
of adding enzymes after adjusting the pH of the solution to 9. The combined hydrolysis finally takes
place at 50 °C for 24 h. The percentage of feather dissolution is compared with conventional alkaline

32
hydrolysis using 0.3% and 0.6% KOH at 121 °C for 20 minutes. The results show that the dissolution
rate is around 55% for 0.3% KOH, compared with over 80% for 0.6% KOH. Moreover, adding an
enzyme only slightly increases the dissolution rate [53].

Ionic liquids

Discovered in 1914 by the chemist Paul Walden, ionic liquids (ILs) are now widely used in research
and industry. ILs are defined in academic literature as "liquids composed entirely of ions that are fluid
around or below 100 °C" [54]. ILs are used for their physicochemical properties such as transparency,
non-volatility, non-flammability, low vapor pressure, thermal stability, and high solvation capacity [55].
This type of solvent can be used for keratin extraction. Idris et al. extracted keratin from turkey feathers
using various protic ionic liquids based on the cation N,N-dimethylethanolammonium (DMEA), and the
anions formate, acetate, and chloride. In a reflux condenser, the dissolution process was carried out at
100 °C for 7 h. After keratin regeneration, the yield relative to the starting mass was estimated at around
63%, and 99% of ionic liquid was recovered after distillation at 122 °C [56]. 1-buthyl-3-
methylimmidazolium chloride ([BMIM+Cl-) can also be used, as in the study of Ghosh et al., in which
it was demonstrated that extraction temperature impacts keratin yield. Extraction is carried out using
wool at a ratio of 1:6 (w:w), and the temperatures studied range from 120 to 180 °C. Dissolution is
carried out for 30 minutes, after which the keratin is precipitated with water. Yield is calculated on the
basis of the initial wool mass, and values are 57, 35, and 18% for temperatures of 120, 150, and 180 °C,
respectively. The authors conclude that increasing temperature increases the generation of free,
water-soluble amino acids, which are not regenerated and constitute the yield loss [57].

Deep eutectic solvents

The discovery of deep eutectic solvents (DESs) is attributed to A.P. Abbot in 2001, who mixed a metal
oxide with a quaternary ammonium salt to form a liquid at room temperature [58]. A deep eutectic
solvent is generally a mixture of two or three components, including an H bond donor (HBD) compound
and an H bond acceptor (HBA) compound at a precise molar ratio, where the freezing point of the
mixture is considerably lower than the freezing point of its starting compounds (Figure 7) [59]. In fact,
most of them can be used as a liquid at room temperature. Examples of compounds that can be used in
DES synthesis are listed in Figure 8, so various combinations can be generated, and their properties can
be adapted to the specific situation. DESs are considered "green solvents” due to their low or non-
existent toxicity, biodegradability, non-flammability, and non-volatility. DESs are also a good
alternative to ILs, which can be toxic but have very similar properties. This alternative is also useful
from a cost point of view, as DESs are inexpensive and easy to produce, requiring only a few minutes
of heating at around 60 – 80 °C [59]. A choline chloride (ChCl) and urea mixture is one of the most
widely used eutectic solvents. Separately, ChCl has a melting point of 302 °C and urea of 133 °C, well
above the freezing point of the 1:2 molar ratio mixture (ChCl : urea) of 12 °C [60]. Various studies on
wool and feathers have shown that keratin can be efficiently extracted using DES. Wool dissolution
with the solvent ChCl:urea (1:2) was studied by Jiang et al. [61] and Moore et al. [62].

33
 Figure 7. Phase diagram of two components eutectic mixture [59].

In the first case, the solvent was prepared by mixing 1 mol of ChCl with 2 mol of urea, heating at 80 °C
under an inert atmosphere until the mixture was homogenized entirely. Dissolution was then done by
adding 10 mg wool to 1 g solvent heated to 110 °C or 130 °C for 5 h under an inert atmosphere.
Dissolution was then measured and reached around 35 mg/g at 130 °C, whereas there was no dissolution
at 110 °C. In the second article, the solvent was prepared at a 2:1 molar ratio (ChCl:urea) and heated to
50 °C. Then, 1 g of wool was added to 20 g of solvent for dissolution at 170 °C for 30 minutes. The
resulting solutions were filtered, dialyzed, and freeze-dried in both studies for keratin recovery. In
addition to wool, feathers can also be dissolved using DES, as shown in the study of Nuutinen et al.
[45], which compares the efficiency of the solvents NaOAc:urea and ChCl:urea. In this case, 3 g of
crushed feathers are added to 147 g of previously prepared hot solvent at various temperatures from
80 °C to 100 °C. Time is also studied, ranging from 2 h to 24 h of reaction. The final solution is filtered

Figure 8. Examples of halide salts and hydrogen bond donors used for DES [60].

34
through a pressurized filtration unit, and the keratin is regenerated by adding water, filtering, and finally
freeze drying to measure the yield. Given the keratin yield values, it is evident that dissolution time
plays an important role. Indeed, the yield rises from 1% for 2 h of dissolution to 35% for 14 h, the
highest yield obtained at a temperature of 90 °C. After 14 hours of dissolution, the yield decreases
slightly. Temperature also plays a role since by increasing the temperature to 100 °C, the yield rises
sharply to around 45% for 6 h. Regarding the efficiency of DES types, the yield is measured for a 6 h
reaction at 90 °C with around 23% for NaOAc:urea versus 0% for ChCl:urea.

Feathers and wool can, therefore, be dissolved in DESs, and the keratin extracted. However, the reaction
mechanisms of these processes are poorly investigated in the literature. A recent paper by Zhang et al.
[63] explains the chemical reaction between chicken feathers and various DESs and proposes an
interaction [Link] adapted mechanism is presented in Figure 9. This study is based on a
quantum chemical method verified by an experimental test. The first step presented is the swelling of
the protein in contact with the solvent and heat, which allows easier access to the secondary structure of
the protein. DES then acts as a hydrogen bond disruptor due to the presence of the nucleophilic Cl- atom,
which strongly attracts surrounding hydrogens and tends to form strong hydrogen bonds. This is
particularly true for cystins, which have numerous -OH and -NH2 groups that act as hydrogen bond
donors and cause a reorganization of hydrogen bonds. In addition, computational calculations have
shown that the energy of S-S bonds within a cystine is 40% lower than that of C-C and C-H bonds,
making it easier to break disulfide bridges. S-S bonds can, therefore, be easily attacked by H+ ions
present in solution and converted to -SH bonds. As a result, the protein is denatured, its secondary
structure is destroyed, and it becomes unfolded and solubilized in the solvent.

Disulfide bond Hydrogen bond

Figure 9. Mechanism of feather dissolution in DES (ChCl:urea) adapted from [63].

35
 N-Methylmorpholine N-oxide
Scientists and industry are currently using the solvent N-Methylmorpholine N-oxide (NMMO) widely
used to dissolve cellulose in the quest for more environmentally friendly biomass extraction. The
NMMO is a tertiary aliphatic amine N-oxide compound [64] produced by the oxidation of
N-methylmorpholine by aqueous hydrogen peroxide, and it is generally obtained as a 50 wt%
concentrated in aqueous solution [65]. Its chemical formula is C5H11N2O, and its molar mass of
115.2 [Link]-1 is represented in Figure 10. Most often used as an oxidizing organic solvent, it is not
thermally stable; therefore, its decomposition temperature begins above 100-110 °C, and beyond 130-
140 °C, its decomposition is significant. NMMO solvent can be found in different hydrates because it
is a hygroscopic compound. When dehydrated, its melting temperature is 170 °C, whereas when it forms
hydrates, the melting temperature decreases. The monohydrate form C5H11N2 H2O melts from 74-78 °C,
and the dihydrate form C5H11N2 2H2O melts from 38-39 °C [66]. The most important aspect of NMMO
in its use is its highly polar N-O covalent bond with a localized oxygen density on electrons. Due to this
strong polar power, NMMO is hydrophilic and soluble in water. Its hygroscopicity is also explained by
this phenomenon and by its facility to create hydrogen bonds. Moreover, this compound is slightly
alkaline, with a pKb = 9.25 [65].

Nowadays, keratin extraction using NMMO is almost nonexistent in the literature, but the study by
Ma et al. [31] demonstrates the possibility of dissolving chicken feathers in this solvent. NMMO is
commonly purchased as a 50% concentrated solution, but the authors report that feather dissolution is
ineffective at this concentration and must be concentrated to at least 75%. To achieve this, they mixed
15 g of feathers in 300 g of 50% NMMO, heated to 110 °C in a reduced pressure distillation apparatus
to remove excess water until dissolution was observed. Dissolution occurs over 5 hours, after which the
solution is centrifuged, and the keratin is recovered from the supernatant by precipitation at pH = 4.5,
filtration, and drying. The mass of dry regenerated keratin is 3.8 g, i.e. about 25% of the initial feather
mass. Following this study, the authors propose a scheme for the dissolution of keratin in the NMMO
solution, shown in Figure 10. Based on amino acid quantification analysis, the authors determined that
the percentage of cystine in chicken feathers is initially 1.34%, decreases to 0.34% in the filtrate after
feather dissolution, and increases to 4.30% in regenerated keratin. Remember that cystine combines two
cysteines through a disulfide bridge, contributing to protein cohesion and strength. This variation in
cystine content is explained by the possible rupture of intra- and intermolecular disulfide bridges during
the extraction process when the NMMO comes into contact with the feathers, allowing them to dissolve.
As this dissolution is not controlled, it allows the release of water-soluble polypeptides or amino acids
and contributes to the loss of yield as they are not regenerated. The disulfide bridges are then regenerated
during the precipitation and drying steps by an oxidation mechanism. In addition, these results were
confirmed by Raman analysis, which indicated an increase in the cystine content of regenerated keratin
compared to the cystine content of feathers.

36
 Figure 10. Dissolution of keratin in NMMO solvent and regeneration mechanisms [31].

Microbial and enzymatic extraction

Enzymatic and microbial methods for keratin hydrolysis are non-hazardous methods for degrading and
hydrolyzing keratin into peptides. These methods are environmentally friendly and energy-efficient. The
process involves keratinases, which are microbial proteases that degrade keratinous materials [67].
However, the degradation of keratin, known as a recalcitrant and extremely resistant material, requires
the assembly of different keratinolytic enzymes to achieve its degradation [68]. A wide variety of
enzymes, bacteria, and fungi are known to be effective in keratin degradation and are summarized in
various reviews [55], [69]–[71].

Hydrothermal extraction
Steam flash explosion (SFE) is a green and chemical-free hydrothermal method for producing
bio-sourced materials under high-pressure saturated steam via hydrolysis. The material is placed in a
reactor heated to a high temperature for a few minutes, and then the pressure is rapidly reduced to create
a high degree of decompression, leading to an explosion. During exposure to the high heat, steam
penetrates the material's cells, causing them to rupture during the explosion. This process is used as a
pre-treatment, making the material more dissolvable and reducing the subsequent extraction processing
time. This method is widely used in converting lignocellulosic biomass [72]. In the case of duck feathers,

37
a work by Zhang et al. achieved dissolving them in two stages: the SFE process and the alkali extraction.
Optimal parameters were evaluated for the SFE process at a pressure of 1.6 MPa for 1 min with an
explosive decompression of 0.1 s and for the alkaline extraction step at a concentration of 0.4% NaOH
at a temperature of 25 °C for 1 h and a solvent: feather ratio of 20: 1 (v:w). These optimum parameters
lead to an extraction rate of 65.78% and a keratin yield of 42.78% [47].

Microwave irradiations extraction

Microwave-assisted extraction is another "green" method that is widely used today and continues to
develop. The principle is to expose a material to microwaves in a reactor, which heats the material
uniformly and enables faster extraction than conventional extraction methods. The microwave-assisted
process can be chemical-free, chemical-assisted, or with enzymes to aid extraction [73]. Zoccola et al.
studied the chemical-free method on sheep wool. The wool was subjected to microwaves after being
mixed with water. The power applied varied from 150 to 570 W, and the extraction time was set at
60 min, knowing that it took 5 to 7 min to reach the desired temperature, which also varied from 150 to
180 °C. Following this treatment, the slurry obtained was then filtered, and the protein content was
measured on a keratin powder after freeze-drying. The best protein content was obtained at a
combination of 150 °C for 60 min, with a fiber-to-mass ratio of 1 : 5 (w:w), with a value of 74% [74].
Another study on sheep's wool shows microwave-assisted acid hydrolysis. The process involves treating
the wool with citric acid and ascorbic acid at pH=2.3, with a fiber-to-solvent ratio of 1:18 (w:w). This
mixture is then submitted to microwaves and heated to 160 °C for 30 min. The resulting solution, with
a final pH=4.1, consists of two phases: solid and liquid (with suspended particles). The solid part is
separated from the liquid, and the liquid part is centrifuged to recover the precipitated particles. The
total keratin extraction yield is estimated at 85% [75].

Ultrasound extraction
Ultrasonic extraction is a widely used method for biomass processing. As a result, this process is used
in many sectors, including medicine, agri-food, and perfumery [76]. It has also been demonstrated that
ultrasound can obtain nanoparticles and study nanotechnologies [77]. Sonochemistry is founded on the
principle of matter-energy interaction and can be applied to both liquid-liquid and solid-liquid systems.
When a system is subjected to acoustic waves, acoustic cavities will form in the form of bubbles within
the liquid. The collapse of the bubbles causes strong localized heating and high pressure in a very short
time. In his work, K. S. Suslick reports that temperatures can reach 5,000 °C, with heating and cooling
rates over 109 K/s, where pressures can reach 500 atmospheres [78]. In the case of mixtures with solids,
these cavitations can accelerate the solid particles, creating collisions between them and damaging them
by causing fissures or breakage. Azmi et al. [79] studied the extraction of keratin from turkey feathers
using an ionic liquid assisted by ultrasonic technology. The feather solvent mixture was subjected to
ultrasound with a probe transducer at 20 KHz and different powers of 120 W, 200 W, and 280 W. The
results showed that applying ultrasound at 200 W considerably reduced the feather dissolution time from
120 min for conventional solvent extraction without ultrasound to less than 20 min with ultrasound. It
was also shown that the higher the power, the shorter the dissolution time. In terms of chemical and
thermal characteristics, the authors conclude that the protein structure and properties are maintained.
Another example of ultrasound-assisted extraction was carried out by Eslahi et al. [80] to obtain keratin

38
nanoparticles. After optimization of extraction parameters via enzymatic hydrolysis, the samples were
subjected to ultrasound treatment for 15 min at 80% amplitude. The effect of ultrasound can be seen in
the reduction in particle size after ultrasound compared with extraction without ultrasound. The average
particle size decreased from 297 nm to 127 nm. Therefore, it can be said that the ultrasound-assisted
extraction method has advantages in terms of nanomaterial production and process time reduction.

I.4. Keratin valorization through bio-based films

The global community is aware of the major environmental and pollution problems caused by
conventional plastics produced by the petrochemical industry. Synthetic plastics are used in most
everyday items, whatever their sector of use, and their end-of-life management is poorly controlled. In
2019, worldwide plastic production reached 370 million tons, including 58 million tons in Europe [81].
For example, the packaging sector alone represents 39.6% of European plastics consumption. It is the
sector at the top of the list, ahead of the construction (20.4%) and automotive (9.6%) sectors [81].
Following their use, vast quantities of plastics are discarded into nature, contaminating waterways and
soils and potentially destroying flora and fauna.

Moreover, this is the current problem of the depletion of non-renewable resources, a problem that will
only become more severe in the future. For this reason, scientists have long been exploring more
eco-friendly manufacturing techniques, prioritizing materials sourced from nature that are non-toxic,
biodegradable, and, above all, renewable. The aim is to formulate and manufacture materials with a high
percentage of natural sources and biodegradable properties that can compete with synthetic plastics
thanks to their physicochemical properties. Emerging biopolymers such as cellulose, chitin, chitosan,
starch, and keratin, among others, are gaining prominence for their potential use in producing
bio-sourced films.

I.4.1. Pure keratin films

The unique properties of keratin mentioned above (high chemical resistance, insolubility, high cysteine
content, hydrophobicity, biocompatibility), as well as its abundance on earth and its renewability, make
it an interesting and promising biopolymer in various scientific sectors: biomedical [7], composites [82],
animal feed and agriculture [6], and cosmetics [83]. Some examples of the manufacture of pure keratin
films have been reported, but all the authors agree that a pure keratin film is very brittle and cannot be
used for applications or even for mechanical testing [84][85]. Therefore, it is necessary to increase the
processability and flexibility of the film to make it suitable for application. Adding plasticizers or
crosslinkers is one way to achieve this and is the subject of numerous studies.

I.4.2. Plasticized keratin films

A plasticizer serves as an additive, either liquid or solid, to enhance the plasticity of a material. It
achieves this by altering the three-dimensional structure of the polymer by slipping between its chains
and reducing the number of hydrogen bonds between them, making it more malleable and flexible.
Cross-linking agents can also create three-dimensional networks between the polymer's macromolecular
chains, modifying its physicochemical properties. Table 1 enumerates the distinct procedures and
parameters involved in producing keratin films. Each protocol proposes different preparation and

39
manufacturing parameters, such as the source of the keratin, the quantity of keratin incorporated into
the film, the presence or absence of a plasticizer/crosslinker, the quantity of plasticizer added, and the
preparation parameters (mixing, heating temperature, and duration). Moreover, the protocols diverge in
the manufacturing methodology employed, with two widely used methods being favored: solvent
casting and hot pressing.

The solvent-casting method is simple to implement and inexpensive. A solution containing a dissolved
polymer is poured into a receptacle and allowed to dry until complete evaporation, resulting in a dry
film that can be removed from the container. The hot-pressing method involves pressing a polymer
mixture, using a heated press, between two plates at a given pressure and temperature. Under the effect
of heat and pressure, the solvent evaporates, and a dry film is obtained. However, the solvent-casting
method is still the most widely used today. Film thickness varies according to the chosen process, and
it typically exhibits a transparent, flexible, and yellowish due to the presence of keratin.

I.4.3. Bio-based films by combining keratin and biopolymers

In addition to incorporating plasticizers into films, an alternative approach involves fabricating
composite films comprising multiple polymers that combine their physico-chemical characteristics to
enhance the overall film properties. This research primarily concentrates on bio-based polymers, defined
as polymers derived from natural, renewable resources. These bio-based polymers exhibit the potential
for biodegradability and compostability, making them significantly more environmentally sustainable
than conventional plastics.

Bio-based polymers can be classified by their origin and production methods [90], as represented in
Figure 11 :

• Agro-polymers, which are polymers sourced from biomass (polysaccharides, proteins),

• Polymers produced through microbial processes (PHAs),

• Polymers are synthesized from biomass using chemical procedures (PLA).

Beyond their natural source and possibly biodegradability, bio-based polymers have many other
interesting propertiesio-based polymers have many other interesting properties beyond their natural
source and possibly biodegradability. For example, chitosan is biocompatible, non-toxic, and
antibacterial [91], cellulose is also non-toxic, and has good thermomechanical properties but is highly
water sensitive [92], and starch is odorless, tasteless, non-toxic, and transparent [93]. Table 2 presents
examples of composite films that incorporate keratin. In various literature sources, keratin serves ais a
reinforcing agent rather than the film base, primarily due to its inherent fragility. As previously
discussed, pure keratin films exhibit extreme fragility and are unsuitable for practical applications or
testing purposes. Consequently, some research studies opted for a composite approach, utilizing a
combination of combining two or three polymers, sometimes with the inclusion of additional additives.
In terms of visual characteristics, bio-based composites tend to share similarities with keratin films
incorporating plasticizers; they are typically transparent, slightly yellowish and flexible, as observed in
most studies.

40
Table 1. Examples of keratin film elaboration in the literature.

Keratin Keratin Plasticizer/ Plasticizer Fabrication Film

Mechanical properties WVP Film aspect Ref
source quantity crosslinker quantity process thickness

20 wt% to 40 TS= 10.5 to 5.7 MPa, ε= 40.5

65 wt% to 0.20 to 0.40 to 43.6% 3.02 to 4.11 (x1010g) Semi-transparent
Feather GLY wt% on keratin Hot pressing [86]
45 wt% mm ([Link]-1) smooth brown film
weight (εmax at 63.8% for 35% GLY)

7 g keratin for TS= 16.6 to 2.0 MPa, ε= 1.7

0 to 0.09 g/g 0.126 to 5.19x10-8 to 5.44 x Film without GLY too
Feather 100 mL GLY Solvent casting to 31.9%, E= 10.2 to 0.2 [87]
keratin 0.115 mm 10-7 (g m/h m².Pa) brittle
dispersion MPa
15% mixed in a
10% on dried Transparent and
Feather buffer pH 9.5 + GLY Solvent casting N/A TS~ 3.5 MPa, ε~ 128% N/A [49]
mass keratin yellowish
SDS
63 g.L-1 in
TS~ 18 MPa to 5 Mpa, ε~ 8 to Transparent and
Wool isopropanol/ GLY 11 to 28 wt% Solvent casting N/A N/A [88]
280%, E~ 150 to 45 MPa yellowish
water (50:50)
30 mg in 60 mL TS= 0.30 to 0.03 MPa, E= 24 Transparent and
Feather GLY 0, 2, 5, 10 wt% Solvent casting 5 mm N/A [89]
NaOH to 7.5 MPa yellowish

EG(1) or (1): TS~ 17 MPa, ε~ 65%, E~321 MPa (2): EG: transparent and
70 g of ground PEG(2) or 30 g of TS~ 21 MPa, ε~11%, E~ 530 MPa (3): TS~ yellowish, PEG and
Feather Hot pressing 0.8 mm N/A [33]
feathers DET(3) or plasticizer 19 Mpa, ε~ 2%, E~ 1268 MPa (4): TS~ 15 DET: semitransparent
GLY(4) MPa, ε~ 14%, E~ 381 MPa and brown, GLY: dark

For 7.5 mg: Pure keratin film too

EGDE(1) or 0.02 to 0.04
Wool 100 mg in H2O 7.5 to 30 mg Solvent casting (1) TS= 27 MPa, ε= 6%, E= 350 MPa (2) N/A brittle. Other films [85]
GDE(2) mm
TS= 15 MPa, ε= 13%, E= 210 MPa flexible and transparent

TS= 5.1 to 0.5 MPa,

0.10 to 8.10
0 to 0.30 g/g 0.094 to ε= 16.3 to 31.7% (εmax= 52.8% for 0.02
Feather 3.5 mg Sorbitol Solvent casting (x1010) (g/m s N/A [90]
keratin 0.206 mm g sorbitol/g ker),
Pa)
E= 125 to 3 MPa

41
Among these varied choices of bio-based polymers, cellulose seems to be one of the most studied and
widely used. Micro- and nano-sized cellulose products have found various applications and have been
in commercial production for several years. Biocomposite cellulose, such as cellulose nanocristal
(CNC), has numerous industrial applications, with a significant presence in the paper packaging
industry. In addition, cellulose-based molecules, often called microfibrillated cellulose (MFC) and
collectively known as cellulose fibers, offer a sustainable packaging solution due to their biodegradable
nature [91].

Figure 11. Biopolymers classification by source [92].

I.5. Cellulose
I.5.1. Native cellulose
Cellulose is the most abundant renewable organic polymer, constituting 40 – 60% of the mass of
advanced plants [93]. The term “cellulose” appeared for the first time in 1838 in a report of the French
academy by the French chemist Anselme Payen. He also determined its molecular formula: (C6H10O5)n.
Due to its renewability and the trend towards more natural materials, it is increasingly considered and
used as a biocompatible and environmentally friendly material [94]. The fascination for this material
lies in its specific hierarchical structure and crystallinity degree, which differs according to its source.
Most cellulosic compounds contain both crystalline and amorphous domains in different proportions.
This aspect of cellulose plays an important role in its physicochemical properties and reactivity.
Cellulose is a straight carbohydrate polymer chain consisting of glucose-glucose linkage units called
cellobiose; glucose within the chain is alternately rotated at 180 °C. In addition, glucose molecules
contain hydroxyl (-OH) groups. The high donor reactivity of OH groups provides hydrophilicity,
chirality, and degradability and enables the creation of an intrachain and interchain hydrogen bond
network that stabilizes antiparallel chains [94]. The hierarchical structure is represented in Figure 12
[95]. The cellulose is known to be insoluble in water, and the biphasic structure of cellulose complicates

42
Table 2. Examples of biopolymers films with keratin in the literature.

Fabrication Film Mechanical

Biopolymer Polymers quantity Plasticizer WVP Film aspect Ref
process thickness properties

TS= 34 ± 8 MPa, ε= 7
20 CH / 100 Ker 0.01 – 0.02
Chitosan / Ker / Solvent casting ± 2%, E= 176 ± 82 N/A N/A [85]
(mg) mm
MPa

Chitosan / starch / Ker

Ker: 5 to 20 wt% Sorbitol, 1% (v/v) Solvent casting N/A N/A N/A Visible fibers [96]
fibers

TS~ 24 ± 0.9 MPa, ε~

Collagen hydrolysate / GLY, 0.3 g/g of 0.136 ± 0.005 WVP = 3.345 ± 0.171
40CH/40CS/20Ker Solvent casting 71 ± 5.5%, E~ 263 ± Yellowish [97]
Chitosan / Ker CH/CS/Ker weight mm ([Link]/kPa.h.m²)
14 %

For 1wt% Ker: WVP = 7.4 ± 0.6 to

Bacterial PHA / Ker Transparent and
Ker: 0 to 50 wt% / Hot pressing 0.1 mm ε= 3.90%, E~ 865 62 ± 16 (x10-15) [98]
powder yellowish
MPa, (kg.m/s.m².Pa)

PVA / cellulose / Ker Ker: 60% GLY 10% Solvent casting N/A N/A N/A N/A [42]

TS= 28 to 11 MPa ε=
300 / 83.75 to 300 / Smooth surface,
MC / Ker GLY, 25 µL Solvent casting N/A 45 to 94%, E= 544 to N/A [99]
418.75 yellow or brown
149 MPa

TS~ 44 to 28 MPa, ε~
Cellulose / Ker Ker: 0 to 40% / Solvent casting N/A N/A Transparent [100]
7 to 6%

43
its application during chemical processes or its modification. As in any chemical reaction, contact
between the two species is essential to obtain a reaction. The arrangement of the cellulose molecules in
the fiber plays an important role in its reactivity. Chemical reagents easily penetrate the amorphous
zones, whereas the crystalline zones, which are much more ordered, are less accessible and limit the
reaction at their surface [101].

There are three different types of nanocellulose: microfibrillated cellulose (MFC), cellulose nanocrystal
(CNC), and bacterial nanocellulose (BNC). The latter is commonly less used than the first two.

Figure 12. Schematic representation of the hierarchical structure of cellulose fibers [95].

I.5.2. Nanocellulose
Nowadays, our society promotes science and industries to develop new innovative materials, including
more technology, while keeping the environmental aspect in mind. The field of nanotechnology was
then developed to meet this demand and tends to create new composite materials that meet evolving
specifications. In our case, cellulose is an ideal natural material for developing composite materials.
Indeed, nanoparticles can be extracted from their original form, called nanocellulose. Nanocellulose is
the nomenclature given to cellulose particles having at least two dimensions in the nanometer scale
regardless of the source or the extraction method [102] and usually measures less than 100 nm in
diameter for a few micrometers in length [103]. The hierarchical structure of native cellulose extends
over different size scales. A plant cell membrane is composed of macrofibres, which in turn are
composed of microfibrils, which in turn are constituted by nanofibrils. In these nanofibrils, crystalline
and amorphous parts can be found in succession [104]. Cellulose nanomaterials are composed of
cellulose nanocrystal (CNC), cellulose nanofibrils (CNF) cellulose microcrystal (CMC), and cellulose
microfibril (CMF). Three main paths to obtaining these materials are mechanical, chemical, or bacterial.

44
Top-down processes obtain the first two types, while the last type is produced by bacteria from glucose
units in a bottom-up process.

Due to its semi-crystalline structure, nanocellulose can be extracted from its original form in two main
steps by various extraction processes. The first step is to remove non-cellulosic plant cell membrane
components such as lignin, hemicellulose, and other compounds by a pretreatment and then extract
nanocellulose. Extraction processes include mechanical treatments: high-pressurized homogenization,
grinding, crushing, and steam explosion; chemical extractions: alkali or acid retting, degumming,
enzymatic hydrolysis, TEMPO-mediated oxidation, or a combination of chemo-mechanical process.
Depending on the nature of the extraction, the crystalline zones differ greatly in size and aspect.
Different micro- or nano-sized fibrils, crystallites, or particles can be obtained [105][103]. The most
interesting aspect in the use of nanoparticles is based on their remarkable properties. Nanocellulose is
ideal for manufacturing nanocomposites and composite materials due to its high mechanical properties,
low density, composite reinforcement capabilities, and biodegradability [106].

I.5.3. Cellulose extraction

I.5.3.1. Viscose and Lyocell processes
One of the most widely used cellulose extraction processes is the viscose process. Cellulose is a
water-insoluble polymer, which makes it a complicated material to process. The viscose procedure is
widely used to obtain textile fibers from plants. This method uses carbon disulphide (CS2) and sulfuric
acid (H2SO4) based on modifying cellulose hydroxyl groups. Its main drawback is the generation of
sulphide pollution due to the non-recoverability of carbon disulphide [107].

The viscose process, known for its drawbacks including the use of a non-recoverable and polluting
solvent, was revamped by opting for NNMO, a non-toxic, recyclable solvent, ideal for environmentally
friendly extraction. Lyocell fiber, produced via direct cellulose dissolution [108], involves mixing
cellulose with an aqueous NMMO solvent. The dissolution occurs physically, facilitated by NMMO's
N-O polar bond and alkalinity, creating hydrogen bonds with cellulose hydroxyl groups. Care is required
as dissolution happens optimally between 90 and 120°C; beyond, NMMO destabilizes. Regenerated
filaments, termed lyocell fibers, are then purified to remove NMMO traces for recovery and recycling,
a process involving purification and evaporation stages. Achieving up to 99% recovery rates [109], this
method holds significant industrial promise due to its efficiency.

I.5.3.2. TEMPO-mediated oxidation

Cellulose nanofibers 2-3 nm wide and a few microns long can be obtained by oxidizing the cellulose.
This oxidation involves using a stable, water-soluble nitroxyl radical, 2,2,6,6-tetramethylpiperidine-1-
oxyl (TEMPO). This oxidation is, therefore, commonly referred to as TEMPO-mediated oxidation. The
principle of this method lies in the use of a combination of TEMPO, NaBr, and NaClO in an aqueous
solution at around pH=10, stabilized by the gradual addition of NaOH, to oxidize cellulose C6 primary
hydroxyls to carboxylate groups [110]. The mechanism of cellulose TEMPO-mediated oxidation is
presented in Figure 13.

45
Several stages must be completed before obtaining CNF. If the raw material is wood, the process steps
are as follows:

- 1 - Pulping of the raw material,

- 2 - Cleaning and obtention of brown fibers due to the lignocellulosic matter contained in the wood,

- 3 - Bleaching of brown fibers,

- 4 - Cleaning and obtention of bleached cellulose fibers,

- 5 - TEMPO oxidation of bleached fibers,

- 6 - Cleaning and obtention of TEMPO-oxidized nanocellulose fibers (TOCNF),

- 7 - Fiber size reduction by mechanical process (blender, high-pressure homogenizer, ultrasonic

homogenizer) and final obtention of transparent TEMPO-oxidized cellulose nanofibers.

Figure 13. Mechanism of cellulose TEMPO-mediated oxidation in water at pH=10-11 [110].

Serra et al. [111] studied the impact of NaClO quantity on nanofiber oxidation by comparing their
carboxylic content, cationic demand, and degree of polymerization. In this study, the nanofibers were
given different names depending on the amount of NaClO added to the reaction in mmol: CNF-2 to
CNF-15. The quantities of the other constituents are fixed at 16mg TEMPO / g dry fiber, and
0.1 g NaBr / g dry fiber in aqueous solution at pH=10. The results show that the properties of the
nanofibers are significantly influenced by the amount of NaClO added. Indeed, the greater the amount
of NaClO, the higher the carboxylic content and cationic demand, while the degree of polymerization
decreases. Due to the principle of reaction mechanism, carboxylic content increases with the amount of
NaClO added (from 40 to 1392 µeq.g/g), generating a rise in cationic demand (from 70 to 2043 µeq g/g)
as the negative charge on the cellulose surface increases, thereby increasing the amount of cationic agent
needed to neutralize the nanocellulose. Concerning the decrease in the degree of polymerization with

46
the addition of NaClO, measured before passage through the high-pressure homogenizer, cellulose
depolymerization is clear (ranging from 1045 to 197 for CNF-2 to CNF-15 respectively). However, the
authors suggested various hypotheses as to this change, with depolymerization during the TEMPO-
mediated oxidation process and slight depolymerization during the DP analysis process. Physical
properties are also important to consider in nanocellulose and paper manufacture. In particular, the water
retention value (g/g) increases from CNF-2 to CNF-15, reflecting the greater wettability of fibers with
higher NaClO addition. Fiber diameter is also influenced, with a tendency for fiber size to decrease with
NaClO addition, from 16.72 nm for CNF-3 to 7.89 nm for CNF-15.

I.5.4. Cellulose films and their properties

For several years, cellulose and its derivatives have been a focal point of research into manufacturing
bio-sourced films. Table 3 shows examples of different types of cellulose film. Cellulose possesses
various properties necessary in the polymer film manufacturing sector, particularly in the packaging
industry. Packaging aims to protect the food it contains while remaining attractive to the consumer. It
must have precise properties, such as good mechanical and thermal properties, and not interact with the
food, i.e. neutral taste and odor. Color is also an important aspect of product marketing. In addition to
its biocompatibility and biodegradability, cellulose exhibits good mechanical and thermal properties and
covers most of the properties required [91]. As previously mentioned and as shown in Table 3, the two
most commonly used film manufacturing processes are hot pressing and solvent casting. These methods
can significantly impact the final properties of the films produced. In fact, based on measured
mechanical properties, tensile strength, elongation, and Young's modulus are generally higher after hot
pressing than after solvent casting or air drying. The mechanical properties of nanocellulose films are
one of the advantages of this type of film, with high tensile strength values reaching, for example,
232 MPa [112] or 245 MPa [113] for 12.6% and 5.2% elongation, respectively. Figure 14 shows the
tensile strength of nanocellulose as a function of elongation at break compared to various synthetic and
bio-based plastics. It can be seen that nanocellulose has one of the highest tensile strengths, rising to the
top of the graph. However, its elongation at break is on the low average compared to other polymers.
Another advantage of nanocellulose films is their resistance to heat and degradation above 280-290 °C,
usually measured by TGA up to 800 °C, which opens up a wide range of applications [112][114]. In
terms of barrier properties, an essential aspect in the choice of field of application, particularly in the
packaging sector, nanocellulose films possess interesting oxygen barrier properties.

The principle is based on blocking the passage of oxygen through the film. Nanocellulose films can
block gases due to the high hydrogen bonding within their molecular structure and the ability to form
high-density films. This density reflects the high cohesion of the film, which prevents the passage of
gases [115]. This property can be seen in Figure 14, where nanocellulose has one of the lowest oxygen
permeabilities compared to other polymers. However, this graph also shows the main disadvantage of
using nanocellulose films. Cellulose is known to be very sensitive to moisture and water. This
hydrophilic property drastically reduces the range of applications for cellulose films. In this case, it can
be seen that nanocellulose has the highest water vapor permeability values, which translates into high
wettability and high-water permeability through nanocellulose films. Another property that measures
the water sensitivity of a material is its contact angle. A material is considered hydrophobic if its contact

47
angle with a drop of water exceeds 90 °. Below this angle, the material shows an affinity for water and
is considered hydrophilic [116]. As can be seen in Table 3, in these cases, the contact angle of cellulose
films is around 40 ° and therefore highly hydrophilic. Another study shows that a higher value of contact
angle can be obtained for a nanocellulose film with a value of 61 °, confirming the hydrophilic aspect
of cellulose [114].

For this reason, cellulose is generally blended with other synthetic or bio-based polymers to be applied
to food packaging. Examples include nanocrystalline cellulose coupled with chitosan, which enhances
the preservation of mangoes [117], carboxymethyl cellulose mixed with polylactide, which improves
the quality of freshly cut tomatoes [118], or cellulose with silver nanoparticles applied to fresh-cut
melons and prevent microbial contamination [119].

Figure 14. Properties of nanocellulose (BNC, CNC, NFC) compared with conventional synthetic and bio-based
polymers: (i) Tensile strength in function of elongation at break and (ii) water vapor permeability in function of
oxygen permeability. Reproduced from [120].

48
Table 3. Examples of cellulose films in the literature and their properties.

Cellulose Fabrication Film thickness Tensile strength Elastic modulus

Density Strain at break (%) WVP or WVTR Contact Angle (°) Ref
type process (µm) (MPa) (GPa)

TOCNF-2 – Laboratory paper 54.7 ± 1.8 to 1.5 ± 0.08 to 11.05 ± 0.5 to 13.0 ±
N/A 1.15 to 1.44 N/A N/A [111]
TOCNF-15 sheet former 145.9 ± 4.3 2.60 ± 0.07 0.5

Vacuum filtration +
TOCNF 22 ± 2 N/A 245 5.2 14.5 N/A N/A [113]
hot pressing

TOCNF Hot pressing 94 ± 1.4 1.48 ± 0.02 232 ± 22 12.6 ± 1.9 4.79 ± 0.14 [112]

13
CNF Hot pressing 58 ± 6.6% ~ 1.01 ~ 100 ~ 3.80 ~ 5.5 ~ 42 [121]
([Link]/m².[Link])

Vacuum filtration+ 15.4 ± 3 to

CNF 120 to 160 N/A 1.7 ± 0.7 to 2.0 ± 0.7 1.7 ± 0.7 to 2.3 ± 0.5 N/A N/A [122]
hot pressing 31.0 ± 3

2.5 ± 0.5 to 10.2 0.1 ± 0.03 to

CNF Solvent casting 120 to 160 N/A 3.3 ± 0.3 to 6.7 ± 1 N/A N/A [122]
±2 0.2 ± 0.04

Vacuum filtration +
MFC 90 ± 3 1.32 ± 0.08 130.2 ± 17.6 6.3 ± 0.5 3.54 ± 0.32 N/A 42.1 ± 2.7 [123]
hot pressing

Vacuum filtration +
MFC 104 ± 4 1.16 ± 0.05 95.3 ± 4.6 6.8 ± 1.6 1.09 ± 0.07 N/A 39.0 ± 3.4 [123]
air drying

MFC Solvent casting 25 1.15 ± 0.03 105 ± 6 4.46 ± 0.48 6.5 ± 0.6 51 ± 4 (g/m²/day) N/A [124]

49
 I.6. Conclusion
In this first chapter, dealing with the literature review, several topics were presented. The context of the
agri-food poultry industry was defined at global level and locally at the scale of the Nouvelle Aquitaine
region for fattened duck production. It was noted that the poultry sector is an important industry in this
region, generating a wide range of co-products, including feathers. However, the literature is mainly
limited to the study of chicken feathers, with very little focus on duck feathers. As a result, the definition
and structure of a feather were discussed, as well as its unique chemical composition with a high protein
content. This protein, clearly identified in the literature as keratin, is thus present in around 90% of
feathers. Keratin also has specific properties such as a high cysteine content, α-helix or β-sheet structure,
biocompatibility, non-toxicity, and chemical resistance to various solvents. Due to these promising
physicochemical properties, feathers have already been the subject of numerous studies reporting on
their cleaning, keratin extraction, and valorization through bio-sourced films. Various protocols exist
for cleaning and keratin extraction, and most have been implemented using toxic chemicals that are
hazardous to humans and the environment. However, the need to improve processes in the direction of
green chemistry has led to more eco-friendly processes involving non-toxic, bio-sourced, or recoverable
and reusable chemicals. Once the keratin has been extracted, the literature provides examples of
keratin-based bio-sourced films blending two or more polymers. We have seen that plasticizers are often
added due to the fragility of pure keratin films.

Another solution discussed was the use of composite films combining several bio-based polymers to
couple their properties and increase film quality. It was also explained that cellulose is an interesting
bio-based polymer for making films, thanks to its many properties, particularly in the packaging sector.
Nanocellulose can be used in various forms: CMC, CMF, CNC, CNF, and CNC. However, cellulose
remains a hydrophilic material, which may limit its use depending on the application. For this reason,
numerous studies have been carried out combining cellulose with other polymers. Mixing extracted
keratin with cellulose is thus an interesting avenue to explore. In addition, the valorization of local waste
while trying to improve and overcome the disadvantages of cellulose films.

50
 Chapter II
Feathers pre-treatment and
characterization

51
52
 II.1. Motivation
As previously mentioned, the central material of this project is duck feathers. However, beyond the
interesting and promising aspect of feathers, they can cause many difficulties in processing. Indeed, as
we have already seen, feathers come from slaughterhouses and are recovered without prior cleaning or
sorting. Feathers are often mixed with organic matter from ducks, such as blood, excrement, skin, bones,
and viscera. Therefore, these recovery and storage conditions are highly favorable to bacterial and
pathogenic agent proliferation, which can lead to health hazards. Therefore, cleaning the feathers as
soon as they are obtained is essential to avoid any risk. Numerous cleaning protocols for using feathers
are described and applied in the literature. It has been found that most of these protocols include various
steps and toxic or environmentally unfriendly products. Moreover, these studies focus on the
characterization and use of chicken feathers and very rarely on the use of duck feathers. Therefore, this
chapter aims to set up a feather-cleaning protocol on a laboratory scale to obtain clean, odor-free feathers
for subsequent use. Then, after cleaning, the properties of duck feathers are analyzed for comparison
with literature data on the properties of chicken feathers.

II.2. Experimental methods

II.2.1. Raw material
The white duck feathers were kindly provided by Plum'Export, a company specialized in the recovery
and distribution of feathers and down (Saint-Sever, France). These feathers originate from a species of
duck called Mulard or Canard Gras. In France, it is used exclusively for the production of foie gras. Its
feathers are mostly white and may be tinted with black (Figure 15). The feathers were provided without
pre-cleaning or sorting.

Figure 15. Picture of the Mulard duck species (La Dépêche, DDM illustration - Nedir DEBBICHE).

II.2.2. Sorting in different categories

The first step was sorting the feathers according to their size and/or morphology to know the different
proportions of feather types on a duck. This allows for anticipating a potential chemical, physical, and
thermal behavior difference in future analyses. After manually sorting 10 kg of feathers, the choice was
made to select 7 different categories (1, 2, 3, 4, 5, 6, 7), represented in Figure 16. Category 1 is simply
the unsorted mixing of feathers. The number 2 is the largest category and represents the strongest
feathers with a large and rigid calamus and a strong rachis with barbs that are difficult to separate.
Category 3 includes feathers with smaller sizes and less rigid components. The calamus and rachis are

53
present, but the barbs are more easily separable and less rigid to the touch. Category 4 contains smaller,
lighter, less solid feathers with a small calamus and rachis. Category 5 includes very light because they
are the birth of new feathers. Neither the calamus nor the rachis are formed yet; only very manageable
barbs form the feather. Category 6 is very close to down, and these feathers are used in the textile
industry because they are soft and flexible without rigid parts. They are extremely light and have volatile
feathers. Category 7 is down, very well known in the textile industry for its softness and fluffiness.

After sorting, 100 feathers from each category were measured and weighed to determine their average
dimensions. Each feather was manually cut to separate its three distinct parts (calamus, rachis, and
barbs). Table 4 shows the height, mass and percentage of each category and each part of a feather.
Results are not given for categories 5 to 7, as the feathers are too thin or lack all three parts.

Figure 16. The seven feather categories after sorting.

II.2.3. Grinding process

In order to facilitate the cleaning of the feathers, it was necessary to grind them. Each category was
crushed independently of the others to avoid mixing them. A custom-made blade mill for low-shear
shredding equipped with a 5 mm sieve was used for the first grinding. Then, after cleaning and if
necessary, a finer grinding operation can be performed using a Resch ZM 200 mill with a 0.2 mm or
0.08 mm sieve.

II.2.4. Cleaning process

Feather cleaning is a two-step process. The first step involves cleaning with soap to remove dust and
organic matter. The second step is disinfection to destroy any bacteria or pathogens. To this end, 5 g of
shredded feathers (5 mm) are immersed in 400 mL of water with 3-4 g of commercially available
ecological soap, heated to 60 °C for 2 hours. The feathers are then rinsed with hot water and filtered.
Then, the disinfection stage involves mixing the washed feathers with a solution of 60% ethanol in water
at 60 °C for 30 minutes. They are then rinsed under hot water, filtered, and air-dried at RT. Temperature,
cleaning time, and solvent were selected based on disinfection recommendations [125]. The resulting
feathers are white, clean, and odorless.
54
Table 4. Dimensional characteristics and percentages of the six feather categories after sorting.

wt% in Whole-feather Whole-feather Calamus Rachis Barbs

Category
random mix height (mm) weight (g) (wt%) (wt%) (wt%)

2 58 229.5 0.74 30 40 30

3 28 186.7 0.35 26 40 34
4 8 118.8 0.16 26 28 46

5 3 N/A N/A N/A N/A N/A

6 0.2 N/A N/A N/A N/A N/A

7 0.2 N/A N/A N/A N/A N/A

Waste 2 N/A N/A N/A N/A N/A

II.2.5. Characterization of raw feathers

II.2.5.1. Proximate analysis
The moisture content of feathers was assessed using a desiccant balance. Initially, a 0.1 g sample was
heated at 105 °C, and the resultant value was directly obtained from the instrument. Subsequently, the
volatile matter was quantified through heating in an inert environment using a calcination furnace
(Carbolite, USA). Clean ground feathers were deposited into ceramic crucibles, sealed with lids, and
heated at 575 °C until a stable mass was reached. During the same process, the ash content was
determined. Following this, samples were heated in ceramic crucibles without lids until a stable mass
was achieved at 575 °C, and the remaining mass was considered as the ash content. All tests were
performed in triplicate.

The fixed carbon content was calculated using Equation 1:

Fc = 100 - (Mc + Vc + Ac) (1)

Here, Fc represents the fixed carbon content, Mc corresponds to moisture, Vc denotes the volatile content,
and Ac is the ash content.

II.2.5.2. Crude fat content

The fat content was determined through Soxhlet extraction using petroleum ether at 100 °C for six hours
within a Soxhlet extraction apparatus. The extracted fat was subsequently subjected to oven-drying and
weighed to calculate its percentage. This analysis was performed on the random category only and in
six repetitions.

II.2.5.3. Protein content

Two different methods of protein detection and quantification have been carried out. The presence of
protein in duck feathers was confirmed using the Biuret assay and quantified through the Kjeldahl
method. This method relies on the complexation of protein peptide bonds under alkaline conditions by
Copper (II) ions contained within the biuret reagent. When proteins are present, the Biuret reagent gives

55
the solution a purple color. Five grams of ground feathers were dissolved in 100 mL of a 1 mol/L NaOH
solution heated to 80 °C for 4 h. Subsequently, the resulting solution underwent filtration and
centrifugation to eliminate all particles and impurities. Next, the Biuret reagent was introduced to 2 mL
of the resultant solution to observe color shift. The Kjeldahl method was employed for protein
quantification by determining the nitrogen content in the dry raw feathers. An automatic digester (Velp
Scientifica DKL8) and an automatic analyzer (Velp Scientifica UDK 159) were used. The nitrogen
content was calculated using a conversion factor of 6.25. This factor is based on the general observation
that most proteins contain approximately 16% nitrogen [126]. The dry matter content was determined
using an infrared moisture balance and expressed as g/100g of the raw sample. The crude protein content
in g/100g of both the crude sample and dry matter was then determined. These analyses were conducted
in triplicate.

II.2.5.4. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis

(SDS-PAGE)
SDS-PAGE analysis, in accordance with Laemmli's method [127], was carried out to determine the
molecular mass of keratin in the feathers. This analysis was performed only with the random category.
Feathers, ground to a size of 5 mm, were employed to facilitate denaturation using Laemmli SDS sample
buffer (4X). 2 mg/mL bovine albumin standard (66.5 KDa) was used to establish a reference point.
Grounded feathers and albumin solution were incubated in 2x Laemmli SDS sample buffer for five
minutes at 95 °C to ensure denaturation. This buffer consisted of 250 mM Tris-HCl (pH 6.8), 8% SDS,
40% glycerol, 8% β-mercaptoethanol, and 0.02% bromophenol blue. For gel preparation, 10 μL of
10-250 kDa Prestained Protein Ladder was loaded into lane 1. In subsequent lanes, 10 µL of the feather
sample and bovine albumin standard were loaded separately in lanes 2 and 3, respectively.
Electrophoresis was performed using Bolt™ 4-12% Bis-Tris Plus Gel and MOPS SDS running buffer.
The gels were run at 180 V until the dye front reached the bottom of the gel. After electrophoresis, the
gels were stained with Coomassie Brilliant Blue R 250 and rinsed with an aqueous solution of 5 % acetic
acid and 20% methanol until a clear background was achieved.

II.2.5.5. Feather–solvent interaction

As previously stated in the bibliographic chapter, keratin is insoluble in most solvents and has
hydrophobic properties. Here, we want to understand the behavior and stability of feathers in various
polar and non-polar solvents. The solvents used included water, ethanol, isopropanol, acetone,
acetonitrile, toluene, diethyl ether, chloroform, hexane, and pentane. Ground feathers were immersed in
each solvent within centrifuge tubes, maintaining a solid-to-solvent ratio of 1:25 (w:v).

II.2.5.6. Chemical durability test

Only category 2 feathers were manually cut with scissors to separate their three distinct parts (calamus,
rachis, and barbs) to assess the chemical durability of duck feathers and their different parts. A range of
chemicals were employed for testing, including: cold water, a strong acid: 1% sulfuric acid solution at
pH = 1; a weak acid: 1% acetic acid solution at pH = 3; a bleaching agent: 4% sodium chlorite solution;
a strong alkali: 1% sodium hydroxide solution at pH = 12-13; and a weak alkali: 0.5% sodium carbonate
solution at pH = 10-11. For each test, 1 g of cut feathers was placed in covered Petri dishes, completely
immersed in the solvent at RT, and then retrieved after 2 hours, 24 hours, and 7 days. The samples were

56
then dried for 24 hours, and any physicochemical alterations were observed and discussed in aspect and
color.

II.2.5.7. Hydrophobicity test

To evaluate the hydrophobic aspect of the feathers, each category of ground feathers and the three
distinct parts of category 2 were immersed in two solvents, an organic solvent, ethyl ether, and an
aqueous solvent, H2O, in the same container. The aqueous phase was stained with bromothymol blue to
distinguish both phases. The comparison is based on reference samples of materials considered to be
hydrophilic: cotton and cellulose pulp. Each sample was mixed with a vortex mixer for 30 seconds and
left to stand overnight.

II.2.5.8. Fourier Transform Infrared (FT-IR) spectroscopy

Chemical composition and functional groups of ground feathers were conducted using FT-IR
spectroscopy in the Attenuated Total Reflectance (ATR) mode, employing a Jasco FT/IR-4700 infrared
spectrometer. For all spectra, a frequency range from 4000 to 400 cm-¹ was examined, with each
spectrum generated by 64 scans at a resolution of 2 cm-¹.

II.2.5.9. Thermogravimetric Analysis (TGA)

Evaluation of thermal resistance for each ground feather category was performed using a TA Q2500
thermogravimetric analyzer. The samples were heated in a nitrogen atmosphere starting at 30 °C and
gradually increasing to 800 °C at a rate of 10 °C per minute.

II.2.5.10. Differential Scanning Calorimetry (DSC)

The melting temperatures of all categories were determined using a TA Q20 differential scanning
calorimeter with continuous nitrogen purge. A heating rate of 10 °C per minute was used over a
temperature range of -50 °C to 250 °C.

II.2.5.11. X-Ray diffraction (XRD)

The determination of crystallinity rate was conducted using a Bruker AXS D8 Advanced diffractometer
instrument. This analysis was performed on each feather category with CuKa radiation (X = 1.542 A)
operating at 30 kV and 15 mA. Data acquisition covered scattering angles ranging from 5° to 70° at a
scan rate of 0.02°/ min. To calculate the crystallinity index (CI), the Origin software and Equation 2 was
utilized:

CI (%) = (Acrist / Atotal) × 100 (2)

In this equation, Acrist represents the area attributed to the crystalline domains, while Atotal corresponds
to the total area under the diffractogram curve.

II.2.5.12. Scanning Electron Microscopy (SEM)

SEM was used to observe the morphology and microstructure of the feathers. The samples were fixed
to specific stubs and subjected to platinum coating via a sputter coater (Q150T, Quorum Technologies,
Kent, UK) to facilitate the examination. Observations were conducted under high vacuum conditions at
2 kV, utilizing a Gemini SEM 300 FESEM (Field Emission Scanning Electron Microscope) (Zeiss,
Germany).

57
 13
II.2.5.13. Solid-state C CP MAS Nuclear Magnetic Resonance (NMR)
spectroscopy
The solid-state 13C CP MAS NRM spectra of random category were recorded on a Bruker Avance II
9.4 T spectrometer at 298 K using a 4 mm rotor with kel-f cap at a spinning rate of 8000 Hz. A total of
3072 repetitions were performed with a contact time of 2.0 ms and a recycle delay of 5 s. The 1D spectra
were processed with Fourier-Transform, and an external reference was used to determine chemical shifts
(glycine, calibrated on the carbonyl signal at 178 ppm).

II.3. Results and discussion

II.3.1. Proximate analysis
The moisture content of a material can affect its appearance, as well as its thermal, chemical and
mechanical properties. It is necessary to characterize this rate and determine whether there is a
significant difference between the different categories of feathers. Figure 17 shows that moisture content
differs slightly between categories, from 7.65% for category 7 to 9.39% for category 5. It can also be
seen that, in terms of feather sections, the calamus is the moister section at 9.91%, compared with 7.52%
and 7.90% for the rachis and barbs. This average value is slightly lower than the value determined for
chicken feathers, which ranges from 8 to 13% [15].

The volatile content of a material results in gaseous constituents that evaporate when the material is
subjected to a certain pressure and temperature. This usually concerns volatile materials such as
methane, hydrogen and carbon monoxide, or may result from material degradation. Under heat, keratin
decomposes, and gases such as NH3, CO2, H2S, SO2, and thiols [128] evaporate due to the high
concentration of sulfur in keratin. The results are shown in Figure 17. Volatile content ranges from
80.66% to 85.97%. We can observe that categories 1, 2, and 3 have the lowest rates of volatile content
(81%) compared to categories 4 to 7 (85%), which is consistent with the volatile content of the calamus,
which is the lowest compared to the other two sections, rachis and barbs. Compared with chicken
feathers, duck feathers have a higher volatile matter content of around 80-85%, compared with 78-82%
for chicken feathers [15].

The ash content refers to the percentage of inorganic, non-combustible material present in a material
after exposure to high-temperature combustion. According to Figure 17, the ash content of duck feathers
is low, less than 1.33% in general, and less than 1% for categories 1, 2, 3, 4, and 7. Regarding the
different feather sections, the calamus and rachis were the lowest, with a similar rate of 0.43%, compared
with 0.87% for barbs. These results are similar to those found in the literature, with less than 1.5% for
chickens [15].

The fixed carbon content measures the non-volatile, solid carbon-containing material in a material after
it has been subjected to high-temperature combustion or pyrolysis in a controlled environment. It
represents the combustible component that remains once volatile matter has been eliminated. According
to Figure 17, the highest percentages are seen in categories 1 to 3, and the lowest in categories 4 to 7,
with overall values ranging from 5% to 10%. Here, a difference appears with the values determined for
chicken feathers, which have higher fixed carbon content values ranging from 17% to 21% [15].

58
Figure 17. Moisture content, ash content, volatile matter, and fixed carbon content values in percentage for each
feather’s category and section.

II.3.2. Crude fat content

The feathers were subjected to Soxhlet extraction to quantify their fat content. FT-IR analysis of the
extract was performed to confirm that the extracted substance was, effectively, fat. Figure 18 shows the
FT-IR spectrum of the extracted fat, which shows distinct bands, including two prominent bands at 2815
and 2848 cm-1, indicative of stretching vibrations associated with methylene (-CH2-) and methyl (-CH3)
groups. These bands are characteristic of a long carbon chain, a common feature in hydrocarbons and
triglycerides. Furthermore, a weaker band at 1738 cm-1 is observed, corresponding to the (C=O) bond
found in ester functions, also present in triglycerides and fat [129]. After confirmation of the extracted
substance as fat, six separate extractions were carried out, and the resulting residues were weighed to
determine the fat content in duck feathers, which was found to be 0.98 ± 0.12% significantly lower than
the literature value of 3%, fat for chicken feathers [15].

Figure 18. FT-IR spectrum of extracted fat from duck feather after Soxhlet extraction.

59
 II.3.3. Protein content
The initial step in protein analysis involved the qualitative confirmation of protein presence in duck
feathers. Initially, the 5 mm crushed feathers were dissolved in NaOH, resulting in a hydrolysate, which
was then treated with the Biuret reagent, leading to a noticeable change in color. Figure 19 illustrates
two test tubes, one containing the hydrolysate without Biuret (on the left) and the other containing the
hydrolysate with the addition of Biuret (on the right). This change from colorless to purple clearly
indicates the complexation of peptide bonds within proteins by copper (II) ions, confirming the presence
of proteins in duck feathers. Furthermore, the crude protein content in duck feathers was quantified using
the Kjeldahl method with a conversion factor of ''N*6.25'', resulting in a value of 82.97 ± 0.98%.

Figure 19. Biuret test result: before the addition of Biuret reagent (left) and after the addition of Biuret reagent
(right).

II.3.4. SDS-PAGE analysis

The molecular weight of the keratin present in the raw feathers was evaluated via SDS-PAGE analysis.
The resulting gel is shown in Figure 20. The gel shows three different lines, line 1 representing the
protein size marker from 10 to 250 KDa, line 2 where the hydrolyzed feather solution has been applied,
and line 3 where an albumin standard with a molecular size of 66.5 KDa has been applied to verify the
validity of the test. This validity is confirmed by a strong coloration around 65 KDa on line 3, the
albumin standard corresponding closely to the size marker. In line 2, there is no evidence of coloration
for feathers before the 15 KDa mark. There is a slight staining between 15 and 10 KDa, but the staining
is closer to 10 KDa. It can be concluded that the keratin present in duck feathers is a protein with a
molecular size close to 10 KDa. This value aligns with the previous results reported in the literature for
β-keratin [19].

II.3.5. Feather-solvent interaction

The influence of solvents on protein folding, stability, and dynamics is a crucial aspect to consider. Early
studies of protein structures have shown that a significant portion, approximately 40-50%, of the
accessible surface area of proteins is made up of non-polar groups [130]. The purpose of this research
was to understand the role of polar and non-polar solvents in the solubility and stability of feather
proteins. This study aimed to evaluate the strength of internal interactions and the nature of the exposed
surface. The image of the test is shown in Figure 21. It is visible that the feathers react differently when
in contact with different solvents. In some tubes, the feathers remain more or less on the surface of the
solvent, interacting mainly with the air, whereas in some solvents, the feathers are completely immersed

60
 Figure 20. SDS-PAGE of feathers, lane (1) Prestained Protein Ladder (10- 250 kDa), (2) feathers, and (3)
bovine albumin standard.

and settle to the bottom of the container. A ranking from weakest to strongest interaction with the solvent
can be classified in this order: water > chloroform > acetonitrile > toluene > isopropanol > ethanol >
diethyl ether > acetone > hexane > pentane. The feather fibers showed the weakest surface interaction
with water and chloroform solvents. This behavior is likely due to the hydrogen bonding capabilities of
these solvents. In contrast, acetonitrile, a polar aprotic solvent with a more pronounced dipole moment
than polar protic alcohols and toluene, showed a different partitioning behavior. In addition, the feather
fractions exhibited greater partitioning in isopropanol than in ethanol, possibly due to the slightly larger
molecular size of isopropanol, which could hinder the sorption process. The inherent hydrophobic nature
of feathers, due to the presence of alkyl groups, made them strongly inclined towards non-polar solvents
such as hexane and pentane, but the interaction of feathers with the solvents tested shows no clear
behavior as a function of the strength of solvent polarity.

Figure 21. Picture of the feather-solvent interaction test with different solvents.

61
It can be concluded that the folding process is a complex interplay of hydrophobic forces, hydrogen
bonding, and related entropic effects. Indeed, the compatibility between materials can be influenced by
various parameters, including hydrophobicity, lipophilicity, density, pH, temperature, polarity,
molecular structure, particle size, and other factors. This test also helps to show that feathers remain
insoluble in most solvents, whether aqueous or organic.

II.3.6. Chemical durability test

The chemical resistance was assessed using various solvents, acids, bases, and bleaches to determine
the feathers' resistance to these chemical environments. Their visual appearance was observed after the
feathers were immersed for various lengths of time (2 hours, 24 hours, and 7 days). A picture of the
results of this test is shown in Figure 22. Several changes were observed, such as fiber swelling,
degradation, and color change.

Figure 22. Picture of feather sections after immersion in different chemical conditions.

Physical transformations and chemical degradation are critical because they can negatively affect natural
fibers and the materials made from them. Considering the visual appearance of the feathers and the
observations in Table 5, we can conclude that they have a low resistance to strong acids, which
significantly degrade the feathers. We can also see that it is not necessary to have basic or acidic
conditions to notice a change, as with water: after a long time (24 h), the color and appearance of the
feathers are modified to the point of causing degradation of the rachis, for example. As far as bleaching
agents are concerned, their bleaching role is only preserved for a short time because the feathers become
fragile and decompose after a few hours. This degradation is also observed in alkaline environments.
Degradation occurs after short exposure times, like barbs, which appear to dissolve and form a viscous
solution after only 2 hours. After 24 hours, all sections are severely degraded.

62
Table 5. Observations regarding the physical properties of feather sections after immersion in different chemical conditions.

Barbs Calamus Rachis

2h 24h 7d 2h 24h 7d 2h 24h 7d

Less Brittle and A little Degradation

Softer and
Cold water No change Slightly grey No change transparent more matte in No change change in with milky
grey
and softer color color dispersion

Whiter in
Fibers are Softer, mass Degradation,
color and a Degradation No change in No change in
H2SO4 Slightly yellow Degradation smaller and loss, and more but no change
little and whiter color, softer color, softer
yellower matte in color of color
degradation

Softer, more
Whiter and a Whiter, softer, Still very
matte, and Hard and more
CH3COOH Slightly yellow Yellower Yellower little and high No change hard, more
intermediate yellow
degradation degradation yellow
degradation

Less brilliant Whiter with a

Presence of More matte Softer and Slightly More rose
NaClO2 Whiter and with rose dispersion of Degradation
rose tones and rose more rose whiter and softer
tones particles

Viscous,
Less yellow Degradation
immediate Highly More matte More viscous Softer and Softer and
NaOH Yellower and high with powder
swelling, and yellow and yellow, and yellower yellower more yellow
degradation dispersion
change to yellow

A little change of Still Still hard and More matte

Na2CO3H2O Grey Greyer, No change Softer Degradation
color transparent more matte and softer

63
 II.3.7. Hydrophobic properties
Hydrophobicity describes molecules or particles that exhibit an aversion to water. The water-repellent
properties of duck feathers are based on a highly organized, hierarchically branched, multi-scaled
structure. This intricate structure provides the feathers with the appropriate surface roughness necessary
for effective water repellency [131]. The surface of duck feathers can be considered heterogeneous,
consisting of solid material and air. In addition, covalent bonds, such as internal disulfide bridges formed
between pairs of cysteine residues, play a critical role. Hydrophobic interactions between non-polar
residues also contribute significantly to protein stability through their dielectric effect [132].
Consequently, the exposure of polar groups is closely related to the number of internal hydrogen bonds,
underscoring the predominance of a hydrophobic driving force. The interpretation of this test is based
on a comparison of the behavior of feathers with two materials known to be hydrophilic: cotton and
cellulose pulp. Figure 23 shows that these two materials blend easily in the blue aqueous phase and settle
to the bottom of the tube. This demonstrates their affinity for water. Conversely, feathers lie between
the organic and aqueous phases and do not mix with the water. This behavior reflects the hydrophobicity
of the material, which seeks to minimize its contact surface with water.

Figure 23. Picture of hydrophobicity test on the different feather’s categories and their sections in comparison
with cotton and cellulose pulp.

II.3.8. Fourier Transform Infrared (FT-IR) spectroscopy

The chemical composition of each sample was investigated using FT-IR analysis, and the resulting
spectra are presented in Figure 24. Notably, these spectra show consistent characteristic bands both
across different categories and between different feather fractions. Furthermore, the results are in line
with those previously reported in the literature for chicken feathers [15], [133], [134]. Therefore, it can
be assumed that duck feathers do not exhibit significant chemical differences compared to chicken
feathers.

During the infrared analysis of the proteins, distinctive bands generated by the general protein backbone
were identified. The spectra show prominent bands associated with amides A, I, II, and III [135]. The
broad peak at 3300-3200 cm-1 corresponds to amide A and is attributed to the stretching vibration of the
N-H bond. The frequency of this band depends on the strength of the hydrogen bond [136]. The most
prominent band, found at 1700-1600 cm-1, is attributed to the absorption of amide I. This band results
from the vibration of the C=O bond within the peptide group, coupled with a slight in-plane N-H bending
motion [135]–[137]. Amide II is represented by a smaller band at 1550-1500 cm-1 compared to amide I.
64
This band is a result of the bending of the N-H bond in the plane and vibrations associated with the C-N
bond [135]. A weaker band is observed at 1300-1200 cm-1, corresponding to amide III. Determining the
exact nature of this band is difficult due to its position and the contributions from various phenomena
and bonds. It is often attributed to the C-N vibration coupled to N-H bending [15], but additional
absorption may result from CH2 wiggling vibrations in this spectral region [138].

Figure 24. Comparison of FT-IR spectra of each category (left) and different sections (right) of duck feathers.

II.3.9. Thermogravimetric Analysis (TGA)

The thermogravimetric analysis determined the thermal stability and degradation temperature of each
feather category and section. The comparative curves are shown in Figure 25. Each curve exhibits a
characteristic two-stage degradation pattern. In the first stage, up to 100 °C, a small mass loss, typically
in the range of 8-9%, is observed up to a temperature of 100 °C. This initial loss is attributed to the
evaporation of hydrogen-bonded water contained in the feather samples, consistent with the previously

Figure [Link] of TGA curves of each category (left) and different sections (right) of duck
feathers.

65
determined moisture content of the feathers. The main degradation of the feathers material occurs in the
second stage, from 230 °C to about 500 °C. During this stage, approximately 70% of the mass is
degraded. This degradation is attributed to structural denaturation, cleavage of peptide bonds between
molecular chains, and degradation of disulfide bridges. In particular, disulfide bridges typically begin
to degrade around 230 °C [139]. There is a notable difference in the curve for Category 7, which
represents down feathers. In this case, the final mass of the sample is significantly lower, about 15%,
compared to the other categories. This finding suggests that down feathers are more susceptible to
temperature-induced degradation. As for the feather sections, they closely mirror the previously
described patterns. They show identical stages of degradation and corresponding mass loss, highlighting
the consistency of thermal stability regardless of the specific component within the same feather
category.

II.3.10. Differential Scanning Calorimetry (DSC)

Differential scanning calorimetry was used to study the thermal transitions of the feathers. The
investigation focused on determining the melting temperature (Tm) across all feather categories and
fractions to identify variations in thermal resistance based on feather morphology, separating hard and
soft parts. The resulting profiles are shown in Figure 26. Each profile generally shows a consistent
pattern characterized by a primary endothermic peak in the 230-245 °C range, indicating a melting point
transition. This peak corresponds to the melting of the protein crystal structure, the degradation of the
α-helix structure, and cystine decomposition [139]. However, upon closer examination, slight
temperature differences are observed between categories 1, 2, 3, and 4 on one side and categories 5, 6,
and 7 on the other.

Figure 26. Comparison of DSC curves of each category (left) and different sections (right) of duck feathers.

For the first four categories, the melting peak is in the range of 230-235 °C, whereas for the last three
categories, the melting temperature is slightly higher at 245 °C. A similar difference in melting
temperature is observed when considering the three feather fractions: Rachis, Barbs, and Calamus.
Specifically, the calamus and rachis exhibit a melting peak at 230-235 °C, while the barbs exhibit a peak
at 240 °C. These results are consistent with the composition of each category. Categories 1, 2, 3, and 4

66
contain a higher proportion of hard parts (rachis and calamus) compared to categories 5, 6, and 7, which
consist mainly of soft feathers, mainly barbs. Furthermore, this variance in Tm may be due to differences
in crystallinity within the different feather fractions, as higher melting temperatures typically correspond
to increased material crystallinity and vice versa [140].

II.3.11. X-ray diffraction (XRD)

Crystallinity is a measure of the degree of order within a material. Fundamentally, the greater the degree
of crystallinity in a polymer, the more regularly its molecular chains are aligned. Increasing the degree
of crystallinity increases hardness and density, which in turn can affect the physicochemical and thermal
properties of the material. XRD patterns were analyzed to determine the crystalline nature of the various
feather types and their fractions. As shown in Figure 27, all feather categories and feather fractions
exhibit similar diffraction patterns. The appearance of these curves confirms the semi-crystalline
structure of the feathers, as evidenced by the presence of two prominent peaks at approximately 2θ=9 °
and 2θ=19 °. These peaks correspond to an amorphous structure and a random coil + β-sheet structure,
respectively [141]. These patterns suggest the coexistence of amorphous regions together with
crystalline peaks. The intensity of a peak indicates its relative presence within the material. Notably, the
second peak associated with the random coil + β-sheet structure predominates over the weaker peak
attributed to the α-helix structure. This observation implies that the primary structural element in feathers

Figure 27. Comparison of XRD curves of each category (left) and different sections (right) of duck feathers.

is the random coil + β-sheet structure, in agreement with the existing literature [36], [142], [143]. The
crystallinity index, along with the calculated percentages of α-helix and random coil + β-sheet structures,
are detailed in Table 6. The CI ranges from 52.15% to 60.90%, with an average rate of 57.26% observed
for the random category of duck feathers studied in this research. Looking at different parts of a feather
in ascending order, the rachis, calamus, and barbs exhibit crystallinity indices of 53.05%, 54.63%, and
55.21%, respectively. This progression is logically related to the increasing number of barbs in a feather,
as greater crystallinity is observed. These results confirm previous observations from DSC curves where
the melting temperature of barbs was slightly higher than that of rachis and calamus due to their higher
crystallinity rate, which requires more energy to perturb the crystal lattice. In addition, the results

67
emphasize that the random coil + β-sheet structure is the predominant structural element in feathers,
with proportions ranging from 63.35% to 69.15%, while the α-helix structure is present in proportions
ranging from 30.85% to 36.65%.

Table 6. Crystallinity index of the different feathers categories and sections calculated from XRD curves.

random coil + β-sheet

Category/section CI (%) α-helix structure (%)
structure (%)

1 57.26 31.56 68.44

2 53.64 31.69 68.31

3 54.03 31.98 68.02

4 56.14 33.01 66.99

5 60.90 31.22 68.78

6 57.42 36.65 63.35

7 52.15 32.71 67.29

Barbs 55.21 36.55 63.45

Rachis 53.05 30.85 69.15

Calamus 54.63 31.27 68.73

II.3.12. Scanning Electron Microscopy (SEM)

SEM was used to observe the morphological features of various feather sections. Figure 28 shows SEM
images of the barbs, rachis, and calamus of a duck feather at different magnifications. First, at low
magnification, image (a) shows the hierarchical structure of the feather. This structure consists of three
different hierarchical levels: the rachis, the central axis of the feather, to which the barbs are connected
in parallel, forming the second level, and to which the barbules are attached, corresponding to the third
level. These barbules can be seen more clearly in picture (b). They have two different configurations.
Some barbules are broad and flat, while others are thin and end in microhooks. As mentioned in the
previous chapter, the usefulness of these microhooks was found to be essential in maintaining the
integrity of the feather by clinging to each other and allowing a "zipping/unzipping" effect, as mentioned
by Kovalev et al. [10] in their study. This effect makes it possible to cope with external stresses and
rebuild this particular structure without damaging the feather. Regarding the outer surface of the
calamus, picture (c) shows a flat and relatively smooth surface with several visible imperfections,
probably due to external aggression. An examination of a cross-section of the rachis, image (d), reveals
two distinct structural components.

68
 Figure 28. SEM images of different sections of duck feathers: barbs, barbules, rachis inside, and calamus
outside. Corresponding magnification: a) x100, b) and d) x500, c) x1500 and e) x 2000.

The outer tube is dense and impermeable, while the inner tube contains a non-compact, honeycomb-
like foam structure. At a finer scale, in image (e), it is evident that the cell walls within this honeycomb
structure are composed of fibers that create a porous framework. This honeycomb architecture imparts
lightweight characteristics to the feather while giving it exceptional strength [10].

II.3.13. Solid-state 13C CP MAS Nuclear Magnetic Resonance (NMR) spectroscopy

The solid-state nuclear magnetic resonance (NMR) spectrum of the random category was used to
elucidate its chemical composition, as shown in Figure 29. The spectrum shows distinct peaks indicative
of specific chemical entities within the material. The most prominent peak, with the highest intensity,
appears at 174 ppm and can be attributed to the presence of amide carbonyl carbon (C=O) and is
associated with the molecular structure of keratin, encompassing both the random coil and β-sheet
configurations [56]. Another distinct peak appears at 131 ppm, representing aromatic carbons [144]. In
the 50-60 ppm spectral region, signals corresponding to the α-carbon are observed, while a peak at
44 ppm indicates the presence of the β-carbon. These carbon resonances are associated with amino acids
such as leucine and are also related to the cross-linked cysteine residues involved in the formation of the
intermolecular disulfide bridges of keratin [145][146]. In addition, a complex line shape can be seen in
the 20-30 ppm range, resulting from the carbon resonances of alkyl groups in the side chains of the
molecules [147].

69
Figure 29. Solid-state 13C CP MAS NMR spectrum of random feather category. The spinning band is noted by *.

70
 II.4. Conclusion

In this chapter, an extensive analysis of the physical, physicochemical, and morphological properties of
locally collected duck feathers was carried out and interpreted. Prior to this characterization, the first
step was to crush and clean the collected feathers. A feather cleaning protocol was established, using
chemicals and parameters that were as environmentally friendly as possible to obtain white, clean, and
odorless feathers that could be used.

The main objective was to understand the composition of a sample of feathers as supplied by the
manufacturer. The feathers were sorted into different categories and analyzed separately to determine
the specific characteristics of each category. The purpose of this sorting was to determine if all the
feathers shared the same characteristics and could be used in a blend without prior sorting, eliminating
this time-consuming and labor-intensive step. It has been shown that a sample of several kilograms of
feathers will consist mainly of large to medium-sized hard feathers and only a small proportion of soft
feathers. The composition of duck feathers was determined as follows: moisture content (7 – 10%),
volatile matter content (80 – 86%), ash content (0.5 - 1.5%), fixed carbon content (5 – 10%), crude fat
content (0.98 ± 0.12%), molecular weight of 10 kDa and high protein content (82.97 ± 0.98%).
Following this, the hydrophobicity of the feathers was well demonstrated by water interaction tests, and
their resistivity to acidic and basic media was assessed as medium to weak acids and bases and low to
strong acids and bases. The protein composition of the feathers was confirmed by FT-IR analysis,
showing all peaks relating to proteins (Amide A, Amide I, II, and III), and their double α-helix and
β-sheet structure was observed using NMR and XRD analysis. The latter confirmed the majority
presence of a β-sheet structure in each category, with an average calculated in the random category of
31.56% α-helix structure versus 68.44% β-sheet, with a total crystallinity index of 57.26%. Thermal
analysis (TGA and DSC) showed material degradation starting at 230 °C, with protein structure melting
around 230-245 °C, followed by decomposition up to 500 °C. Morphological analysis revealed a 3-level
hierarchical structure of keratin fibers throughout the feather. Two different structures were perceived:
a compact and smooth arrangement on the outside of the calamus, while on the inside of the rachis, the
structure is honeycombed, giving the feather both rigidity and lightness.

All these properties, specific to feathers and keratin, hold great promise for their use in the manufacture
of bio-sourced materials for a variety of applications. But, before new keratin-based materials can be
developed, they have to be extracted from their source. The following chapter describes in detail the
keratin extraction protocol implemented via a comparative study of two green solvents.

71
72
 Chapter III
Optimization of the keratin
extraction process

73
74
 III.1. Motivation
Following the detailed characterization of duck feathers described in the previous chapter, it is clear that
feathers contain a large amount of keratin, making them an ideal source for keratin extraction. This
protein is already widely used in various fields of scientific activity and can be obtained from various
sources. It has been shown in the literature that keratin can be extracted from human hair, wool, animal
hair, or feathers (chicken, goose, duck). Various keratin extraction protocols have been developed and
used, but the major problem with these extraction processes is the use of highly toxic chemicals to
humans and the environment. Keratin is a highly resistant protein, insoluble in water and most solvents.
This is why we find, for example, strong acids or bases, sulfite or sulfate-based compounds, or
2-mercaptoethanol.

This chapter aims to establish a protocol for extracting keratin from duck feathers using a bio-sourced,
renewable, and non-toxic solvent. Two solvents are then compared based on the yield of regenerated
keratin obtained and the properties of this keratin. Each process step has been optimized to minimize
energy and chemical consumption.

III.2. Experimental Methods

III.2.1. Materials
Clean and ground duck feathers, described in the previous chapter (Category 1), were used as a source
of keratin. Two solvents were used to compare extraction processes: NMMO (50 wt% and 75 wt%) and
DES (ChCl:urea). After dissolution, three acids were used for the keratin regeneration step: sulfuric acid
96%, acetic acid 99%, and citric acid in an aqueous solution prepared at a maximum solubility of 59.2%,
i.e. 3 mol L–1.

III.2.2. Solvents preparation

III.2.2.1. NMMO preparation
A 50 wt% NMMO solution in water was purchased as it is commonly supplied. As mentioned in the
literature section, feathers resist dissolution in a 50 wt% NMMO solution due to insufficient
concentration [31]. Keratin extractions were performed using the 50 wt% NMMO solution as supplied
and then compared with extractions using concentrated NMMO at approximately 75 wt% to verify the
accuracy of these results. Solution concentration was achieved using a rotary evaporator set at 60 °C
and gradually increasing the vacuum to remove excess water for several hours, resulting in a NMMO
concentration of approximately 75 wt%. The resulting concentrated solution was light yellow compared
to the colorless, unconcentrated solution. The concentrated solution was then sealed in a glass bottle and
stored in the refrigerator until needed.

III.2.2.2. Deep Eutectic solvent preparation

A deep eutectic solvent was used for the second extraction option. This DES is based on choline chloride
and urea in a 1:2 molar ratio. This solvent requires prior preparation, as it is made from these two
compounds in a solid state, which must be mixed and heated to obtain the liquid solvent. To this end, a
mass of ChCl is mixed with a mass of urea in a round bottom flask, and then the compounds are heated

75
to a certain temperature under stirring. The solvent is formed and the mixture becomes increasingly
liquid. Heating is maintained until the solution is completely homogenized. The resulting solvent is used
directly for extractions or stored in a closed glass container at RT until needed.

DES preparation times were optimized in terms of temperature, pre-drying of solid compounds, and
solvent quantities. Details of the parameters studied are given in Table 7. For the reagent pre-drying
parameter, compounds were stored in an oven set at 50 °C for a minimum of 24 hours before use.

Table 7. Parameters studied in the preparation of the DES.

Reagent pre- Elaboration time

Studied parameters DES masse (g) Temperature (°C)
drying (min)

Yes 50 60 36
Reagent pre-drying
No 50 60 52

No 100 45 90

No 100 50 60

Temperature No 100 60 51

No 100 70 32

No 100 80 18

No 50 60 36

No 100 60 51

DES quantity No 150 60 68

No 200 60 103

No 250 60 127

III.2.3. Keratin extraction parameters involved

During an extraction process, various parameters affect the yield and quality of the extract obtained.
Here, several parameters were varied over various values to determine the optimum protocol, ensuring
the maximum yield of regenerated keratin. The variable parameters were temperature, feather: solvent
ratio (F:S) (w:w), dissolution time, and feather grinding size.

III.2.3.1. Keratin extraction using NMMO solvent

A specified amount of ground feathers was added to a 50 wt% or 75 wt% NMMO solution, with various
feather: solvent ratios investigated. This mixture was heated at different temperatures and for different
durations. The mixture was then filtered to remove any undissolved feather fragments. The resulting
filtrate was then combined with a significant volume of water to initiate precipitation to reach the keratin
isoelectric point, which is pH=4.5, as established by Pillemer et al. [148]. The aqueous solution was
acidified with various acids to reach a pH of approximately 4. The precipitated keratin was then

76
recovered by filtration through filter paper and homogenized with a small amount of water. The next
step was purification by dialysis using regenerated cellulose dialysis tubing (Spectra/Por Standard RC
Tubing, MWCO: 6-8 kD) over 3 days, with water changes 2 to 3 times a day. Finally, the keratin was
freeze-dried and the resulting dry regenerated keratin was weighed to calculate the regenerated keratin
yield and stored in sealed containers at room temperature until further analysis. The overall extraction
scheme is summarized in Figure 30.

Figure 30. General scheme for keratin extraction by solvolysis.

III.2.3.2. Keratin extraction using DES

The extraction steps using DES remain similar to those performed with NMMO to compare the results.
However, the main parameters, the temperature range, the duration, and the F:S ratio were modified to
suit the solvent. Precipitation, purification, and freeze-drying of the regenerated keratin were performed
as described above.

All the extraction parameters studied for the two solvents are summarized in Table 8.

Table 8. Summary of the different parameters studied for keratin extraction as related to the solvent used.

Feather size
Solvent Temperature (°C) Duration (min) F:S ratio
(mm)

NMMO 100 to 160 360 to 600 0.2 or 5 0.02 to 0.2

DES 120 to 170 90 to 240 0.2 or 5 0.02 to 0.2

77
 III.2.3.3. Influence of the acid used for keratin regeneration
A precipitation step is necessary to recover the keratin dissolved in the solvents. Precipitation takes place
after adding acid to lower the pH. The pH values were measured in real-time using a pH electrode and
progressively adding a given acid using a micropipette. Each volume of acid added was recorded. The
tests were performed in triplicates and the results presented are the mean value. The type of acid used
was then compared by measuring the yield of regenerated keratin. A conventional strong acid such as
sulfuric acid was compared with two more natural acids, acetic acid and citric acid.

III.2.4. Calculation of regenerated keratin yield

Regenerated keratin yield is determined by evaluating various extraction parameters to identify the
factors affecting yield. After freeze-drying, the resulting dry regenerated keratin is collected and its mass
is measured. This mass is then compared to the original keratin content, which, as calculated before,
corresponds to 83% of the original feather mass. Notably, in the previous chapter, the protein content of
the feathers was quantified using the Kjeldahl method (N*6.25), revealing a composition of
83 ± 0.98 wt% crude protein in the original sample. Regenerated keratin yield is calculated following
the Equation 3:
𝑚𝑘
Yield (%) = x 100 (3)
0.83 𝑥 𝑚𝑓

Where mk is the mass of regenerated keratin and mf represents the initial mass of feathers.

III.3. Results and discussion

III.3.1. Solvent preparation
III.3.1.1. NMMO concentration
NMMO, with the chemical formula C5H11NO2-H2O, is commonly supplied at 50% in water. However,
this solvent had to be concentrated to approximately 75% to dissolve the feathers effectively. Therefore,
the water in the solvent was evaporated using a rotary evaporator and the 75% NMMO was then stored
until use. Infrared analysis was performed on the 50% solvent and on the 75% solvent after concentration
to verify that the concentration step did not affect the chemical structure of the compound. The FT-IR
spectra are shown in Figure 31. This graph shows several bands, including a broad band at 3300 cm-1,
characteristic of the O-H bond [149], whose intensity decreases for the 75% NMMO compared to the
50% NMMO. This band can be attributed to the water molecules present in the solvent, the proportion
of which logically decreases with the concentration step. Other visible bands include the weak band at
1650 cm-1 representing aqueous compounds, followed by bands at 2950 cm-1 and 1450 cm-1 attributed
to C-H bonds and a narrow band at 1115 cm-1 corresponding to the C-O bond [149]. No chemical change
was observed except for a decrease in the intensity of the bands related to aqueous compounds.

78
 Figure 31. FT-IR spectrum of NMMO solvent at 50% concentration and 75% concentration after rotary
evaporator step.

III.3.1.2. DES preparation optimization

A deep eutectic solvent requires an elaboration step. A DES is based on a mixture of two or more solid
compounds with very high melting point (MP). The mixture is made between choline chloride
(MP= 302 °C) and urea (MP= 133 °C) in a 1:2 molar ratio, as shown in Figure 32. This mixture is then
heated and the solvent can be used immediately or stored in a closed container at RT. The preparation
of this solvent has been optimized according to 3 parameters: prior drying of the reagents, heating
temperature, and amount of DES prepared.

Figure 32. Elaboration mechanism of choline chloride:urea (1:2) eutectic solvent modified from [150].

The presence of water molecules plays an important role in developing a DES. The interaction is based
on the formation of double ionic hydrogen bonds between an H-bond donor compound (urea) and an
H-bond acceptor compound (ChCl). Each of these two compounds is considered hygroscopic and
contains a certain amount of water. For this reason, the moisture content of both compounds was
measured before and after drying. The mass of the sample before drying was measured on a precision
balance, then placed in an oven at 45 °C for 24 hours, and its final mass measured on a precision balance.
Each sample was made in triplicate. The moisture content of ChCl was measured at 2.99 ± 0.95% and
that of urea was measured at 0.83 ± 0.41%. For the total quantity of solvent studied here (50g), this
represents 1 g of water in the solvent, i.e. 2%. The values obtained appear low for hygroscopic

79
compounds, and the analysis parameters should certainly be adapted with a higher temperature or longer
duration to ensure that all the water in the compounds is evaporated.

Figure 33 shows the comparison between two ChCl:urea (1:2) mixtures before heating. It can be seen
that in the right flask, corresponding to the dried mixture, the reaction of the compounds is almost
immediate with the start of melting to form aggregates, whereas in the left flask, where the reagents
have not been dried, no interaction is visible, and the compounds remain in solid state. In addition, after
heating to 60 °C, the time required for complete homogenization of the mixture was evaluated and
compared. With prior drying of the reagents, the time was reduced to 36 min compared to 52 min without
prior drying of the reagents. This phenomenon can be explained by the presence of water, which also
forms hydrogen bonds with the compounds, hindering the direct interaction between choline chloride
and urea and thus implying a longer reaction time.

Figure 33. Aspect of ChCl:urea (1:2) mixture after a few minutes at RT. Left: mixture without prior drying of
reagents, right: with drying of reagents.

Six different heating temperatures ranging from 45 °C to 90 °C were tested. Each sample was heated
with stirring until the mixture was completely homogenized and the results are shown in Figure 34. The
curve shows that the higher the temperature, the shorter the heating time required for homogenization.
In fact, this time increases from 90 min at 45 °C to less than 18 min at 90 °C. The evolution of time
versus temperature follows a quasi-linear trend, with a coefficient of determination of 0.9051. The DES
heating temperature can therefore be modulated according to the operator's needs. In this particular case,
the temperature is set at 60 °C, which allows to obtain a processing time of less than 1 hour while
maintaining an average heating temperature.

Another aspect of this preparation that can be optimized is the amount of solvent prepared. In fact, the
solvent can be used freshly prepared or prepared in advance since it has the property of remaining liquid
at room temperature. The curve in Figure 34 shows that for a 50 g mass of solvent, the duration is
36 min, while for a 100 g mass, the duration is 51 min. Moreover, for a 250 g mass, the time is 127
minutes. A direct linear tendency would suggest that the values for 100 g and 250 g would be around 70
min and 180 min, respectively, so the real duration time implies a direct correlation between mass and
reaction time, which tends to reduce time when increasing the total volume of DES elaborated. This
represents that 250 g of elaborated solvent would save 53 min to produce the same amount of solvent if

80
250 g is done at once rather than doing lots of 50 g. The evolution of time in function of quantity follows
a quasi-linear trend, with a coefficient of determination of 0.9753. It is, therefore, preferable to produce
large quantities to optimize the time required.

Figure 34. Heating time required to homogenize DES solvent as a function of temperature (left) and the quantity
of DES prepared (right).

Following DES elaboration, FT-IR analysis was carried out on the pure compounds and the mixture to
verify the interaction of reagents and the creation of a new compound. FT-IR spectra are visible in
Figure 35. On the ChCl curve, the characteristic band of the O-H bond is visible at 3220 cm-1 due to its
hygroscopic properties. Then several weak bands are visible at 1480 cm-1 corresponding to CH3 groups,
CH2 groups are visible at 1080 cm-1, then the C-C-O bond is attributed to the band located at 955 cm-1,
and finally at 892 cm-1, we find the N-CH3 bond [151]. The characteristic urea bands are located between
3500 and 3100 cm-1 and represent the amides, which are also observed at 1590 cm-1. The weak band at
1675 cm-1 is related to the C=O bond, and the moderately strong band at 1460 cm-1 represents the C-N
bond [152]. Therefore, each of these two compounds has a different IR spectrum with its own
characteristic bands. If we now look at the curve obtained for DES, we find the characteristic bands of
urea in the zones from 3500 to 3100 cm-1 and from 1700 to 1550 cm-1, corresponding to the NH2 groups
of urea. We also find the CH3 band at 1455 cm-1, only present in ChCl [151]. Therefore, we can conclude
that the presence of the characteristic bands of the two compounds indicates the successful interaction
and effective production of DES.

III.3.2. Regenerated keratin yield comparison according to solvent and parameters used
III.3.2.1. Duration and regenerated keratin yield as a function of temperature
Temperature can affect several aspects of a solid-liquid extraction protocol. A temperature change can
impact the solubility of a compound in a solvent. It is generally recognized that as the temperature
increases, the solubility of a compound increases, and this affects the kinetics of the extraction reaction.
The consistency of the solvent can also be changed by increasing the temperature, which reduces the
viscosity of the solvent or mixture. However, an optimum temperature must be explored and established,
as high extraction temperatures can also lead to undesirable degradation of the extracted compounds. In
this study, the temperature was the first parameter investigated. The comparison was made concerning
the yield of regenerated keratin obtained.

81
 Figure 35. FT-IR spectra of ChCl, urea, and the ChCl:urea (1:2) mixture.

NMMO 50%

The bibliographic chapter stated that extraction using the 50% NMMO solvent was tested to extract
keratin from chicken feathers at 110 °C [31]. However, this study was not conclusive, and the authors
stated that it was impossible to dissolve feathers in NMMO concentrated to only 50% and that it was
necessary to concentrate the solvent. Then, feathers were mixed with 50% NMMO and heated at various
temperatures and the time it took to dissolve completely was evaluated. The results are shown in
Figure 36. The first test was performed at 110 °C, but after more than 10 h of reaction, the feather
dissolution remained incomplete. A test was then conducted at 130 °C but feather dissolution remained
extremely slow and even after several hours of heating, dissolution was considered insufficient.

Figure 36. Graphs of feathers dissolution time and regenerated keratin yield as a function of heating
temperature. Keratin was extracted with NMMO 50% (left) and NMMO 75% (right) with fixed parameters:
feather size 5 mm and F:S ratio=0.10 (based on 50g solvent).

82
The minimum temperature determined for effective feather dissolution within a reasonable time was set
at 140 °C, as the time required was then 450 min (7.5 h) for 53% yield. The higher the temperature, the
shorter the dissolution time. The same is also valid for the yield, which decreases with increasing
temperature.

In the context of optimization and the desire to minimize the energy consumption of this extraction
protocol, the concentration of NMMO was tested at a minimum of 75% to reduce the minimum
dissolution temperature.

NMMO 75%

Once the solvent was concentrated to about 75%, temperature tests could again be performed from
100 °C to 160 °C and the results are presented in Figure 36. In this case, the minimum dissolution
temperature was 110 °C for 468 min and a yield of 48%. Below 110 °C, dissolution is not considered
effective because the time required is too long. There is also a slight decrease in time from 110 °C to
120 °C and a sharp decrease when heating to 130 °C or higher. It can be concluded that the minimum
temperature is the optimum temperature, given the maximum yield of 48%. Higher temperatures reduce
the yield rate, which could lead to protein degradation at higher temperatures.

DES solvent

In the case of DES, the minimum temperature studied is 120 °C, but dissolution is not considered
effective, given the time of over 400 min without complete dissolution of the feathers. On the other
hand, Figure 37 shows that the minimum dissolution temperature is 130 °C, with a duration of 385 min,
for a low yield of about 17%, while increasing the temperature by only 10 °C drastically reduces the
heating time and increases the yield. In fact, at a temperature of 140 °C, the time is more than halved to
158 min, with a yield of about 25%. It's also important to note that, in general, any increase in

Figure 37. Graph of feathers dissolution time and regenerated keratin yield as a function of heating
temperature. Keratin was extracted with DES with fixed parameters: feather size 5 mm and F:S ratio=0.10
(based on 50g solvent).

83
temperature leads to a reduction in heating time and yield reduction. A temperature of 140 °C is a good
compromise between short duration and maximum yield.

III.3.2.2. Duration optimization

Once the extraction temperature has been selected for each solvent and the time required for complete
dissolution of the feathers has been evaluated, the second parameter affecting extraction yield is
dissolution time. A longer extraction time may result in better dissolution of the original material but
there is generally an optimization point beyond which a longer time does not necessarily result in a
better yield. However, attention must be paid to the extraction time selected as it can have a negative
effect by contributing to the long-term degradation of compounds. It is necessary to find a balance
between a reasonable extraction time and preserving the extracted compounds. The durations studied
were chosen according to the total dissolution time of the feathers, with lower and higher durations
selected; results are presented in Figure 38.

NMMO

At a temperature of 110 °C, complete dissolution was estimated to take 468 min, so the time scale
studied was chosen to be between 360 and 600 min. In the range of 360 to 480 minutes, the yield value
remains stable at around 48 %, while after 540 min, the yield drops sharply to around 31-34%. Given
the stability of the yield obtained, allowing the extraction to continue for longer than 360 min is
unnecessary. This is equivalent to stop the extraction before the feathers are completely dissolved. This
could be due to the degradation of the extract. In fact, a longer extraction time allows the feathers to
dissolve completely, but given the yield results, this longer time seems to cause a parallel degradation
of the extract, given the stability of the yield and its decrease after 480 min. In conclusion, the optimum
time here is the minimum.

DES

Regarding the DES, the complete dissolution of the feathers took 158 min at a chosen temperature of
140 °C, so a range from 90 minutes to 240 min was investigated. It can be seen that the yield rate
decreases as a function of extraction time. A sharp decrease occurs from 90 min to 120 min, with the
yield dropping from 56% to 33%. The decrease then remains more or less regular for every 30 min
added to the extraction time. This decrease averages about 3.55%. Again, it is unnecessary to wait until
the feathers are completely dissolved to obtain the best yield, since the minimum time studied is the
optimum.

84
 Figure 38. Diagram presenting regenerated keratin yield as a function of extraction time. Keratins extracted
with fixed parameters: feather size = 5mm, F:S ratio = 0.10 (based on 50g solvent), temperature = 140 °C for
DES and 110 °C for NMMO.

III.3.2.3. Influence of feathers size

Extracting keratin from feathers involves grinding the feathers to a more or less uniform size. However,
in an extraction process, the size of the raw material can play a role in the final extraction yield. The
smaller the size of the solid, the greater its contact surface area and the greater the surface area interacting
with the solvent. The finer the solid size, the more accessible the compounds to be extracted and the
higher the yield. This accessibility is also enhanced by fine particles facilitating mixing of the reaction,
as opposed to large particles, which prevent or slow down mixing. For this reason, two feather mill sizes
were studied: 5 mm and 0.2 mm, and the results are shown in Figure 39.

NMMO

The difference in feather size influences the extraction yield of regenerated keratin. Indeed, for a feather
size of 0.2 mm, the yield is higher than that obtained with 5 mm feathers: 29% versus 22%. It is,
therefore, advantageous to add a grinding stage for the feathers to make them more finely to increase
keratin yield. However, grinding to 0.2 mm is an additional step that requires grinding to 5 mm and then
grinding to 0.2 mm. This double-grinding operation is costly in terms of time and energy. Double
grinding is necessary because feathers, being an organic material, are heated by the friction of the
grinder, which causes feather aggregate in the sieve, resulting in feather heating, degradation, and a
strong odor.

DES

The previous observation is confirmed by the results obtained for the DES. Fine grinding also increases
the yield of regenerated keratin in this case, ranging from 28% for a 0.2 mm size to 25% for a 5 mm
size. Again, fine grinding of feathers is more interesting in terms of keratin yield.

85
Figure 39. Diagram presenting regenerated keratin yield as a function of feathers size. Keratins extracted with
fixed parameters: F:S ratio = 0.10 (based on 50 g solvent), temperature = 140 °C during 180 min for DES and
120 °C during 480 min for NMMO.

III.3.2.4. Solvent quantity optimization

The F:S ratio studied here is the ratio between the mass of feathers incorporated in a given amount of
solvent. The mass of feathers is varied while the amount of solvent remains fixed and the yield obtained
is compared as a function of these ratios. Results are presented in Figure 40. It is important to determine
and select the best solid:solvent ratio to optimize the yield, duration, and cost of an extraction process.
In fact, the goal is to determine the maximum mass of feathers that can be dissolved in a given amount
of solvent to limit the amount of solvent used as much as possible, thus limiting chemical consumption
and cost. Consistency is also generally influenced by the solid:solvent ratio and can lead to variations in
yield depending on the viscosity of the mixture.

NMMO

NMMO was subjected to a test on a 100 g solvent base, various amounts of feathers ranging from 2 g
to 20 g were added to the solvent in a 250 ml round bottom flask, and the mixture was stirred under
magnetic stirring. Feathers are a low-density material, making them light but voluminous. Mixing is
efficient for small feather masses of 2 g to 10 g. It can be seen that as the feather mass increases, the
viscosity of the solution increases, making feathers more difficult to dissolve. From 15 g of feathers, the
volume of the flask had to be changed to 500 mL to accommodate the entire amount of feathers. Mixing
this solution becomes difficult, and magnetic stirring becomes almost insufficient. For the 0.20 ratio
test, the 500 mL flask was replaced by a 1 L glass reactor, and a more powerful mechanical stirring
replaced the magnetic stirring to maintain sufficient agitation, given the high consistency of the solution.

Regarding the keratin yield, it can be seen that it increases as the ratio increases. However, at a ratio of
0.15, the yield decreased from 51% to 40% compared to a ratio of 0.10. Despite this decrease, we can
see that the yield increases again to 57% for the maximum ratio of 0.20. This decrease can be attributed

86
to an inappropriate flask volume or insufficient mechanical stirring, which reduces the dissolution of the
feathers and can lead to a decrease in yield. We can conclude that the optimum ratio includes a large
amount of feathers, as in this case, 20 g for 100 g of solvent.

Figure 40. Diagram presenting regenerated keratin yield as a function of F:S ratio (based on 100 g of solvent).
Keratins extracted with fixed parameters: temperature = 140 °C during 180 min for DES and 120 °C during
480 min for NMMO.

DES

The F:S ratio study with DES was conducted in the same manner as with NMMO. Ratios ranging from
0.02 to 0.10 were obtained in a 250 mL round bottom flask, 0.15 in a 500 mL round bottom flask, and
0.20 in a glass reactor with mechanical stirring. The same observation was made regarding the increase
in solution viscosity as the amount of feathers increased. An increase in keratin yield was also observed
as the ratio increased. The minimum yield of 8% is obtained for a minimum ratio of 0.02, compared to
42% for a ratio of 0.15. Here, no yield was obtained at a ratio of 0.20 because the high viscosity of the
reaction prevented the feathers from dissolving, and the reaction was stopped. The maximum yield for
DES was set at 0.15.

III.3.2.5. Influence of acid type during keratin regeneration

Once the feathers have been dissolved in the solvent, the keratin must be precipitated to recover it. This
precipitation step is also known as regeneration. The principle of precipitation is based on the fact that
substances soluble in a solvent become insoluble when environmental conditions such as temperature,
pH, or pressure are changed. As a result, solute particles that were in solution began to aggregate,
forming solid particles. Here, precipitation is initiated by adding distilled water to the filtered solution
of dissolved feathers. Water is known to be an anti-solvent for keratin. In the case of any solvent,

87
precipitation is initiated when water is added to the mixture a precipitate begins to form. However, this
precipitation is weak, so it was necessary to acidify the solution to induce greater precipitation, as it has
been previously reported that the isoelectric point of keratin is around pH=4.5. After acidifying the
solution to a pH of about 4-4.5, significantly more precipitate was observed. This precipitate is
white/beige or even brown, depending on the solvent used, and in most cases, is located at the bottom
of the beaker. The keratin precipitate is shown in Figure 41. In order to optimize the process and respect
the environment, three different acids were tested to compare the yield of regenerated keratin obtained,
the amount of acid added to lower the pH, and the influence of the type of acid on the properties of the
regenerated keratin. A conventional strong acid (sulfuric acid) was compared to two more natural acids
(acetic acid and citric acid) and to water with no added acid. The results are shown in Figure 42.

A) B)

Figure 41. Pictures of the keratin regeneration step by acidification at pH=4. Picture A): keratin obtained with
DES, picture B) keratin obtained with NMMO.

NMMO
The acidity of the solution was assessed in real-time using a pH electrode immersed in the solution.
After adding 400 mL of water to the solvent solution, the initial pH was between 7.20 and 7.35. Figure 42
shows that a yield of 28% is obtained without acidification, only with the addition of water. This clearly
shows that water is an anti-solvent for keratin and causes its regeneration. Overall, however, the yields
are higher when the solution is acidified. Regardless of the acid used, the yield is relatively stable at
about 43-44%. The maximum yield is obtained with citric acid at 44.25%, sulfuric acid at 43.31% and
acetic acid at 42.78%. On the other hand, the difference is in the volume used, with the minimum for
sulfuric acid being at 5 mL versus 24 mL for acetic acid and 40 mL for citric acid. Acetic acid appears
to be a good compromise because of the high toxicity of sulfuric acid, in terms of both yield and volume
of acid poured.

DES
In the case of DES, the initial pH of the solvent solution after the addition of water is between 9.50 and
9.65, which is higher than in the case of NMMO. For the two natural acids, acetic acid and citric acid,
the yields are 30% and 35%, respectively. The best yield is attributed to sulfuric acid, with a yield of
39%. In terms of the volume of acid added, the minimum is for sulfuric acid with 0.10 mL, followed by

88
acetic acid with 0.83 mL, and finally citric acid with 1.57 57 mL. Overall, the amount of acid required
for DES is much lower than for NMMO.

Figure 42. Diagram presenting regenerated keratin yield and acid volume as a function of acid type used for
keratin regeneration step. Keratins extracted with fixed parameters: feathers size = 5 mm, F:S ratio=0.10
(based on 50 g of solvent), temperature = 140 °C for DES and 120 °C for NMMO.

III.3.2.6. Comparison of regenerated keratin yields with literature and

conventional extraction protocols
In Chapter I-Bibliography, it was presented that there are several ways to extract keratin from different
sources that are described in the literature. The most common chemical extraction protocols are
oxidations (i.e., peracetic acid, percarbonate.), reductions (i.e., urea + L-cysteine, urea + m-bisulphite +
SDS), alkaline hydrolysis (i.e., NaOH, KOH) and the use of ionic liquids and DES. Table 9 shows some
keratin chemical extraction protocols based exclusively on chicken or duck feathers. The key
parameters: solvent, temperature, F:S ratio and time are listed together with the corresponding yield of
regenerated keratin. It's essential to remember that the main objective of the keratin extraction protocol
used in this work is to use green solvents with low environmental impact as an alternative to
conventional solvents that are polluting and dangerous to handle. In this respect, NMMO and DES have
a significant advantage. NMMO is a non-toxic, recyclable solvent widely used in industry to dissolve
cellulose. DES (ChCl:urea) also has its advantages as the cost of reagents is very low, it is biobased and
biocompatible, and, therefore, a non-toxic solvent. From the examples given in Table 9, it can be seen
that the dissolution temperature of the feathers is generally below 100 °C but in the two cases proposed
in this work, the optimum temperatures are 110-120 °C and 140 °C for NMMO and DES, respectively.
However, the temperature for extraction with NMMO is in line with the temperature used by
Ma et al. [31], which is 110 °C for a lower yield than that obtained in this work, 25% versus 77%,
89
respectively. This difference in yield can be explained by a longer dissolution time and a 4-fold increase
in the F:S ratio, resulting in a higher yield. The yield for DES (ChCl:urea) is also higher than some
examples of DES used in the literature. The temperature for ChCl:urea is 140 °C compared to 100 °C
for NaOAc:urea and 80 °C for ChCl:oxalic acid, but the extraction time is drastically reduced and
divided by four, with 90 min compared to 360 min for both examples. The yields obtained with NMMO
or DES do not exceed those obtained with sulfur solvents, but overall, the F:S ratios found in the
literature are very low compared to the F:S ratio optimized in this work. It can be seen that the ratios are
generally between 0.025 and 0.05, while in the present work, the optimum ratios are 0.15 and 0.20 for
DES and NMMO, respectively. This increase in the F:S ratio provides significant solvent savings that
can compensate for the high temperature.

In summary, the keratin extraction protocol implemented in this work shows yields ranging from 56%
to 77% for the DES and NMMO solvents, which are within the average of the yields found in the
literature. In fact, these values are higher than some yields obtained with conventional solvents but
remain lower than those obtained, for example, with sulfur solvents. The two solvents studied in this
project offer a good compromise between yield and reduced environmental impact.

Table 9. Examples of various keratin extraction protocols from poultry feathers and their associated yield.

Keratin Time ηregenerated keratin

Solvent T (°C) F:S Ref.
source (mm) (%)

NMMO 75% Duck feathers 120 0.20 480 77 This work

Chicken
NMMO 75% 110 0.05 300 25 [31]
feathers

DES (ChCl:urea) Duck feathers 140 0.10 90 56 This work

Chicken
DES (NaOAc: urea) 100 0.02 360 45 [45]
feathers

DES (ChCl: oxalic Poultry

80 0.05 360 30 [153]
acid) feathers

[Bmim]Cl + Na2SO3 Duck feathers 90 0.05 60 75 [154]

0.4% NaOH + SFE Duck feathers 25 0.05 60 43 [47]

500 mM Sodium Chicken

50 0.025 360 80 [155]
sulfide feathers

Chicken
0.2 M meta-bisulfite 65 0.028 300 88 [156]
feathers

90
 III.4. Conclusion
Once the feathers were characterized and it was determined that duck feathers were an important source
of keratin, the keratin was extracted. During this chapter, an extraction protocol was established to
constantly minimize the environmental impact of such a process. Two non-toxic solvents were compared
for extraction efficiency, and the yield of regenerated keratin was measured. Several parameters, such
as extraction temperature and time, feather:solvent ratio, and feather size, were evaluated and compared.
The use of both solvents requires a prior preparation step. For NMMO, two concentrations were
compared. A 50% concentration is the unmodified product at the time of purchase, and a 75%
concentration is after a preparation step. In the case of a 50% concentration, feather dissolution is
possible at a temperature of 140 °C, whereas after concentrating it to around 75%, the dissolution
temperature can be lowered to 110 °C. This is why the choice of solvent was made in favor of NMMO
75% for further processing. Solvent preparation is a prerequisite for DES.

Optimal preparation parameters were been established with prior drying of the reagents to reduce
homogenization time, with 60 °C being a good compromise between average temperature and correct
time. It is also more advantageous to produce large quantities, saving time and costs. Once the solvents
were prepared, the feather dissolution step was optimized. For the DES solvent, the optimal parameters
were 140 °C, a duration of 90 minutes, a finer initial feather size (0.2 mm), and an optimal F:S ratio of
0.15. For NMMO, the optimum parameters vary with a lower temperature of 110 °C, a duration of 6 h
(longer than DES)), similar feather size (0.2 mm), and a ratio of 0.20. Once the feathers are dissolved,
the keratin regeneration step is optimized by comparing three acids (sulfuric, acetic, and citric) with
water. Since water is an anti-solvent for keratin, it is sufficient to precipitate it. However, it has been
shown that the addition of acid allows the isoelectric point of the protein to be reached at around pH =
4 - 4.5, increasing keratin regeneration and yield. Conventional strong acids are generally used in keratin
extraction protocols to obtain high yields; however, it has been shown in this work that acetic acid can
be used as a good compromise in terms of quantity required and keratin yield to overcome the toxicity
of conventional strong acids and therefore reduce the environmental impact of this extraction protocol.
Generally, the yields obtained in this work are suitable for processes using bio-sourced and non-toxic
and renewable products. These regenerated keratin yields exceed some results found in the literature,
especially in comparison with other studies on NMMO, for example, or other studies using different
DES, but do not exceed the yields obtained with chemicals such as ionic liquids and sulfur solvents.

On the other hand, this work has shown that a high F:S ratio can be used, compared to the literature,
which studies very low ratios, generally in the range of 0.025 to 0.05. This result allows for solvent
savings and reduces the environmental impact and cost of such a process, which can offset the relatively
high temperature above 100 °C, unlike other extraction protocols that use temperatures below 100 °C.
It is also important to note that DES is a very inexpensive and easy-to-use solvent. As noted above in
the example of the NMMO solvent, about 99% of the solvent can be recovered after use with the
necessary equipment, giving it a great economic and environmental advantage over other non-
recoverable solvents that generate waste treatment costs.

Regarding the prospects of this chapter, it would be interesting to further optimize this extraction process
by testing it on a larger production scale, such as a pilot scale. This would certainly lead to even higher

91
keratin yields. What is more, the cost and environmental impact of this process could be further
optimized by establishing a method for recovering and purifying the NMMO after dissolving the feathers
to be able to reuse the solvent and save on solvent costs and quantities. In addition, it would be interesting
to carry out a cost calculation of the entire extraction process in order to study the feasibility of scale-up
for industry.

In this chapter, only the yield of regenerated keratins has been reported, but it is essential to determine
whether the various extraction parameters studied also impact the quality and properties of these
regenerated keratins. For this reason, the following chapter is dedicated to the physicochemical analysis
of the extracted keratins.

92
 Chapter IV
Physico-chemical characterization
of regenerated keratin

93
94
 IV.1. Motivation
The previous chapter optimized the keratin extraction process according to different parameters:
temperature, time, feather size, F:S ratio, and type of acid used. This comparison was made based on
the regenerated keratin yields obtained as a function of each parameter in the case of both solvents.
However, yield should not be the only information considered in the final choice of an extraction
protocol. The quality of the final material obtained is essential to the choice of application, which will
depend on its properties.

High extraction temperatures can cause extract degradation if the extract is heat sensitive. Long
extraction times can lead to the formation of undesirable compounds after prolonged contact with the
solvent; an inappropriate F:S ratio, where excess solvent strongly dissolves the small amount of starting
material; and too fine material size, which increases the contact surface between solid and solvent and
accelerates the extraction. Therefore, this chapter performs a physicochemical analysis of keratins to
evaluate how each parameter affects their properties. In addition, we have compared these properties
with those of raw feathers to determine if the extraction process alters the original properties of the
keratin.

IV.2. Experimental methods

IV.2.1. Materials
The keratins analyzed are those extracted and regenerated using the extraction procedure and parameters
described in Chapter III with NMMO 75% and DES (ChCl:urea). They are in solid, dry form after
freeze-drying and stored in closed plastic containers until analysis.

IV.2.2. Characterization of regenerated keratin

IV.2.2.1. Density
Bulk density and sintered density were assessed. For bulk density, he method used follows the European
Medicines Agency guidelines [157] with a few modifications. Approximately 100 g of the test sample
(m) was gently introduced into a dry graduated cylinder of 10 mL for NMMO and 4 mL for DES
(readable to 2 mL) without compacting and weighed with 0.1 percent accuracy. If necessary, the powder
was carefully leveled without compacting, and the unsettled apparent volume (V0) was read to the
nearest graduated unit. The formula m/V0 calculated the bulk density in g per mL. Sintered density is
calculated by measuring the mass and volume of a sintered sample using a cylindrical mold of 12 mm
diameter and 10 tons of pressure. The mass is determined using a precise balance, and the volume can
be measured either by geometric formulas for regular shapes or by the displacement method for irregular
shapes. The sintered density is then calculated using the formula ρ= m/V, where ρ is the density, m is
the mass, and V is the volume. Both are typically expressed in g/mL or g/cm³ in powder materials
analysis.

IV.2.2.2. Visual aspect

Keratin color and appearance were visually analyzed. Keratins were sorted and arranged according to
each solvent and extraction parameter used to take pictures against a white background and under the
same light conditions.
95
 IV.2.2.3. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-
PAGE)
The molecular mass of regenerated keratin was studied with SDS-PAGE analysis using the same
protocol described in Chapter II.2.5.4.

IV.2.2.4. Fourier Transform Infrared (FT-IR) spectroscopy

The chemical composition and functional groups of the keratins were analyzed by FT-IR spectroscopy
in the ATR mode, as previously described in Chapter II.2.5.8.

IV.2.2.5. Thermogravimetric Analysis (TGA)

Thermal properties and degradation temperature of keratins were analyzed with TGA according to the
parameters described in Chapter II.2.5.9.

IV.2.2.6. Differential Scanning Calorimetry (DSC)

The thermal properties of the keratins were complemented by DSC analysis to determine phase changes.
This analysis was carried out according to the parameters described in Chapter II.2.5.10, modifying the
maximum analysis temperature to 280 °C.

IV.2.2.7. X-ray diffraction (XRD)

The semi-crystalline behavior of keratin was verified by XRD analysis. The crystallinity index (CI) was
calculated using Origin software and the Equation (2). Analysis and calculations were carried out
according to the parameters described in Chapter II.2.5.11.

IV.2.2.8. Scanning Electron Microscopy (SEM)

The morphology of regenerated keratins was observed with SEM. Observations were made with a
Hitachi S-4100 (Japan) FE-SEM scanning electron microscope at 5 kV in high vacuum mode. The
samples were coated with carbon using the evaporation method using an Emitech K950 Turbo
Evaporator (Germany). Quartz PCI software was used for image acquisition.
13
IV.2.2.9. Solid-state C CP/MAS Nuclear Magnetic Resonance (NMR)
spectroscopy
The solid-state 13C CP MAS NMR spectra of regenerated keratins were recorded following the same
parameters of Chapter II.2.5.13.

IV.3. Results and Discussion

Results are presented separately according to temperature, time, feather size, F:S ratio, and type of acid
used for regeneration to determine the influence of each extraction parameter on the quality and
properties of the keratins obtained. For each analysis, a comparison is made between the two solvents
used: NMMO 75% and DES.

96
 IV.3.1. Keratin density
Only two keratin samples were considered for the density measurement. DES keratin extracted at 140 °C
and NMMO keratin extracted at 110 °C were selected based on the previously established optimum
parameters. The main objective in this case is to compare the two solvents. Such information is very
rare in the literature, as keratin density can vary considerably depending on the type of keratin and its
source. However, feathers are known to be light and strong due to their structure and high keratin
content. Barone et al. measured the fiber density of chicken feathers and found it to be 0.89 g/cm3 [158].
This value confirms the feathers' lightness. In the case of keratin, two methods have been used in this
work to measure its density: bulk density and sintered density. The results of these tests are presented
in Table 10. In both cases, DES keratin had a lower density than NMMO.

The porosity of keratin was calculated from its density, considering that the standard density of keratin
is 1.3 g/cm3. The porosity values therefore logically indicate that DES keratin is more porous than
NMMO keratin.

Table 10. Bulk density, sintered density and porosity of keratin extrated with DES or NMMO.

Solvant of extraction Bulk density (g/cm3) Sintered density (g/cm3) Porosity (%)

NMMO 0.17 1.0048 74.57

DES 0.16 0.9694 77.29

IV.3.2. Scanning Electron Microscopy (SEM)

Microscopic images of two keratin samples were taken, one extracted with NMMO 75% solvent at
110 °C and one extracted with DES at 140 °C. SEM images are shown in Figure 43. Images of the same
magnification were chosen to compare surface morphology and keratin size. First, we can see that the
extracted keratins are in the micrometer range and show morphological differences. At x1000
magnification with the NMMO, we can see a sometimes flat, rough structure formed by several layers
of keratin stacked on top of each other, similar to a scale-like structure. Aggregates are attached to this
flat structure. At ×10,000 magnifications, the aggregates show a fibrous structure with interconnected
filaments, holes, and high porosity. For DES, a flat, relatively smooth, and fibrous structure is observed
at ×1,000 magnifications. Filamentous aggregates are also attached and interconnected to this flat
structure. At ×10,000 magnifications, the structure differs from that of NMMO keratin. The arrangement
is compact, and the aggregates are spherical, forming a random microstructure with a sparse pore
structure. In general, the flat areas observed appear to correspond to the structure of the fibers in the
feathers, which may imply that the feathers are not completely dissolved in the solvents during the
dissolution process. It may be that 100% dissolution is not achieved, but it still allows for an efficient
extraction. The keratin that is completely dissolved in the solvent would then be regenerated, forming

97
partially structured arrangements and aggregates, which would explain the double structure observed in
the images.

(A) (B)

Figure 43. SEM images of regenerated keratin extracted with (A) NMMO 75% : magnification
×1,000 and ×10,000 and (B) DES : magnification ×1,000 and ×10,000.

IV.3.3. Solid-state 13C CP MAS Nuclear Magnetic Resonance (NMR) spectroscopy

Solid-state NMR spectra of keratin extracted with NMMO and DES are shown in Figure 44. These
spectra are similar to the NMR spectra of feathers, studied in Chapter II.3.13. The highest peak is related
to the amide carbonyl carbon (C=O), situated at 174 ppm in both cases, while the weak peaks at around
130 ppm are associated with the aromatic carbons, α-carbon and β-carbon, observed in the region of
50-60 ppm and 44 ppm, respectively. Finally, the complex structure is also present in the 20- 30 ppm
region, indicating the presence of alkyl groups at the ends of the chain. A difference in the spectrum of
keratin compared to that of feathers might have been expected due to the disulfide bridge disruption
during the dissolution of feathers. However, as explained in the Chapter I, disulfide bridges must be
broken before feathers can be dissolved in the solvent. This dissociation of disulfide bridges would lead
to a change in the secondary structure of the protein by altering its β-sheet structure. In this case, a
decrease in the intensity of the β-carbon peak at 40 ppm would be expected, which is not the case here.
The β-carbon peaks are present in both spectra, demonstrating that the keratin disulfide bridges are
maintained during the extraction process or regenerated during the precipitation step. In addition,
cleavage of the disulfide bridges would result in the formation of thiol groups (R-SH) with a specific
peak appearing in the 25-29 ppm region [146], which is already dominated by chain end alkyl groups
between 20 and 30 ppm. These peaks are not visible in the current spectra. This NMR analysis
demonstrates that the material extracted by the solvolysis process, whether with NMMO or DES, is
keratin, exhibiting the same chemical structure as feathers.

98
 Figure 44. NMR spectra of regenerated keratin extracted with NMMO 75% at 110 °C and DES (ChCl:urea) at
140 °C. The spinning band is noted by *.

IV.3.4. Influence of the extraction temperature

IV.3.4.1. Visual aspect
Images of all the keratins studied are shown in Figure 45. After the freeze-drying step, the keratin is
recovered in dry powder form. However, differences in texture and color are visible, depending on the
solvent used and the extraction parameters. First, it should be remembered that the duck feathers used
for keratin extraction were white. Once extracted, the keratin takes on a beige to yellowish color for that
extracted with DES and a dark brown color for that extracted with NMMO. Therefore, a modification
occurs when the feathers are dissolved, making the keratin colored.

In the case of DES, this yellowish color is explained by the formation of chromophores when the feathers
are dissolved in the solvent. A chromophore is a group of atoms alternating between conjugated multiple
and single bonds. This alternation of bonds modifies the absorption wavelength of the molecule, thereby
changing its color. The yellowish color comes from the release of chromophores from aromatic amino
acids such as tryptophan, phenylalanine, and tyrosine [159] during dissolution under the action of heat
and solvent. The chemical structure of these amino acids is shown in Figure 4 (Chapter I). In the case
of NMMO, the color comes from the chromophores formed from the amino acids and the solvent itself.
Since NMMO is thermosensitive, chromophores can be formed under the influence of heat. This
coloring effect is also seen when NMMO is applied to cellulosic fibers in the Lyocell process [160]. It
has been reported in the literature that starting from 120 °C, NMMO degradation is observed by the
increase of the extinction at lambda = 400 nm with the increase of the heating temperature. The increase
in absorbance indicates an increase in the concentration of chromophores [161]. This explains the large
difference in color between DES and NMMO keratin.

From the temperature point of view, certain differences in texture and color between keratins can also
be observed. For DES, at a temperature of 130-140 °C, the texture of the keratin is light and cotton-like
with a light color, whereas from 150 °C the texture changes to a more granular and "sandy" texture,
reaching a dark beige at 170 °C. This may reflect a slight keratin degradation at higher temperatures

99
Figure 45. Visual aspect of regenerated keratins extracted with NMMO 75% and DES, depending on the extraction parameter.

100
compared to lower temperatures. In the case of NMMO, the overall appearance of the keratin is heavier
and sandier than that of DES. Color changes are also observed as a function of temperature, particularly
between 110 °C, medium brown, and 150 °C, dark brown. This may also reflect degradation at high
temperatures.

IV.3.4.2. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-

PAGE)
The molecular weight of the extracted keratins was measured by protein separation using SDS-PAGE
analysis. The resulting gel is shown in Figure 46. For this analysis, a size marker was applied in line 1,
the keratins studied were applied in lines 2 to 9, and then a standard protein was applied in line 10 to
check the validity of the assay. Image (A) represents keratins extracted with NMMO, and image (B)
keratins extracted with DES. For this test, the two extreme values of each parameter were compared, as
well as the 4 precipitations.

In the case of the temperature parameter, the samples 110 °C and 160 °C are compared for NMMO
(Figure (A), lines 2 and 3), and the samples 130 °C and 170 °C are compared for DES (Figure (B), lines
2 and 3). In both cases, the blue color, representing the presence of proteins, is aligned with the 10 kDa
marker line. This value is similar to the molecular weight value previously determined for raw feathers
(Figure 20, Chapter II). This indicates that the molecular weight of the keratin monomers remains intact
after extraction and regeneration relative to its source. These results are consistent with those reported
in the literature [46], [162]–[164] for the molar mass of β-keratin, which is reported to be approximately
10 kDa. However, since the size of the keratin is close to the minimum gel limit, it would have been
interesting to carry out this experiment with a higher gel concentration gradient, allowing the analysis
of molecular masses lower than 10 kDa, in order to be able to confirm these results.

(A) (B)

Figure 46. SDS-PAGE pattern of regenerated keratins. Picture (A) for NMMO 75% solvent, lane (1) Prestained
Protein Ladder (10-250 kDa), (2) 110 °C, (3) 160 °C, (4) 6 h, (5) 10 h, (6) H2O, (7) acetic acid, (8) sulfuric acid,
(9) citric acid, (10) bovine albumin standard. Picture (B) for DES solvent, lane (1) Prestained Protein Ladder
(10-250 kDa), (2) 130 °C, (3) 170 °C, (4) 1h30, (5) 4 h, (6) H2O, (7) acetic acid, (8) sulfuric acid, (9) citric acid,
(10) bovine albumin standard.

101
 IV.3.4.3. Fourier Transform Infrared (FT-IR) spectroscopy
The FT-IR spectra of keratins extracted at different temperatures are shown in Figure 47. High extraction
temperatures can lead to protein degradation as these molecules are temperature-sensitive. Therefore,
the chemical composition of the extracted keratins was analyzed. All the protein-specific absorption
bands previously identified in the FT-IR analysis of duck feathers (paragraph II.3.8, Chapter II) are
found in the keratin spectra.

These characteristic bands reflect the presence of peptide bonds (-CONH-). They include the stretching
vibration of the N-H bond at around 3300-3200 cm-1, the vibration of the C=O bond associated with the
bending of the N-H in the plane at around 1700-1600 cm-1, the bending of the N-H associated with the
C-N bond at 1550-1500 cm-1, and finally a band at 1300-1200 cm-1 combining several contributions
from the C-N, N-H, and CH2 bonds. The correspondence of the absorption bands with the feather
spectrum confirms that the extraction process, whether with NMMO or DES, does not affect the peptide
structure of the keratin chains.

(A) (B)

Figure 47. FT-IR spectra of regenerated keratin depending on extraction temperature. Extracted with: (A) NMMO
and (B) DES (ChCl:urea).

IV.3.4.4. Thermogravimetric Analysis (TGA)

The thermal stability of the extracted keratins was analyzed with TGA. The curves obtained are shown
in Figure 48. The general appearance of the curves is similar to that of raw feathers (paragraph II.3.9,
Chapter II). In fact, keratin degradation under the influence of heat also occurs in two main stages. The
first stage is a slight loss of mass, corresponding to the evaporation of hydrogen-bonded water present
in the sample, and the second stage is a severe degradation, previously attributed to denaturation of the

102
 protein structure, disruption of peptide bonds between molecular chains and cleavage of disulfide
bridges. In summary, the degradation of the protein structure of keratin.

(A) (B)

Figure 48. TGA of regenerated keratin depending on extraction temperature. Extracted with: (A) NMMO 75% and (B)
DES (ChCl:urea).

In Figure 48 (A), corresponding to keratins extracted with the NMMO solvent, the loss of mass due to
humidity is between 6% and 8%, with a slight difference in the slope of the curve. In particular, at the
minimum temperatures of 110 °C and 120 °C, the evaporation is slower than at the other temperatures.
This difference is also noticeable at the beginning of the second phase degradation, with different initial
degradation temperatures. In fact, keratins extracted at 130, 140, 150, and 160 °C degrade from 215 °C,
while keratins extracted at 110 and 120 °C degrade from 230 °C, which is similar to raw feathers. Once
800 °C is reached, the amount of remaining material is highest for temperatures of 110 and 120 °C with
19.5% and 18%, respectively, consistent with the raw feather value of 16-20% and less than 16.5% for
higher temperatures. In Figure 48 (B), corresponding to keratins extracted by DES, the moisture content
is between 4.5 and 7%. This is lower than NMMO keratins and raw feathers. In contrast to NMMO, the
moisture content decreases as the extraction temperature increases, e.g., 7% at 130 °C versus 4.5% at
170 °C. There is no difference in the initial degradation temperature, with all keratins starting to degrade
at 220 °C, which is lower than for raw feathers. The derived curve shows that the higher the extraction
temperature, the greater and faster the degradation of the material. Then, at 800 °C, the 130 °C curve
stands out from the others, with a significant residual material value of 16.8%, compared to about 15%
for the other temperatures. These values remain lower than those obtained for raw feathers.

In conclusion, the extraction temperature parameter can influence the thermal degradation
characteristics of keratins. In both cases, the degradation is slower, and the remaining mass of material
is greater at lower extraction temperatures. The degradation temperature is maintained in the case of
NMMO, but slightly reduced by 10 °C in the case of DES compared to raw feathers.

103
 IV.3.4.5. Differential Scanning Calorimetry (DSC)
The thermal properties of keratins were complemented by DSC analysis to observe phase transitions
and compare them with those observed for raw feathers. The thermograms are represented in Figure 49.
The first endothermic peak visible on the curve is between 100 and 150 °C and is associated with
removing bound water from the keratin. TGA analysis has shown that keratin samples contain a certain
amount of water. Numerous hydrogen bonds are present within the protein structure, and water can also
be bound to keratin via hydrogen bonds, which can raise the minimum temperature to remove this water
to above 100 °C [165]. An endothermic peak follows this between 190 °C and 280 °C. In figure 49 (A),
corresponding to NMMO, the 110 °C curve differs from the others due to different heat flow values.
This should not be taken into account as it is due to a malfunction of the instrument. However, if we
consider the shape of the curve, it can be seen that this curve and the 120 °C curve are different from
the other temperatures. Both curves show a large and bimodal endothermic peak at 190-280 °C with an
epaulement at around 210 °C. These two curves are highly similar to those of feather keratin in Figure 26
(paragraph II.3.10, Chapter II), but the shoulder is only visible after extraction. This endothermic peak
is associated with the thermal denaturation of the keratin crystalline structure, which consists of an
α-helix structure. The present shoulder can be explained by an overlap due to degradation of the cysteine
matrix, as we saw earlier that keratin is a cysteine-rich material [166]. Then, when the dissolution
temperature reaches 130 °C or more, we observe a shift of this endothermic peak towards lower
temperature values, around 230-235 °C, with a decrease in peak depth and peak area. The literature
suggests that this shift may be explained by a reduction in the cysteine in the keratin after extraction
compared to the keratin present in feathers. In addition, a loss of peak intensity may also be associated
with transforming the keratin structure from α-helix to random coil or β-sheet arrangement during the
extraction and regeneration [166].

(A) (B)

Figure 49. DSC analysis of regenerated keratin extracted with (A) NMMO 75% and (B) DES (ChCl:urea)
depending on the extraction temperature.

104
Concerning Figure 49 (B), which deals with DES keratins, a difference can be seen depending on the
temperatures studied. Overall, the curves presented have the same shape as the feather DSC curves,
especially in the case of the curve related to 130 °C. A first endothermic peak at 100-150 °C concerns
the elimination of water, followed by a second well-defined melting peak of the crystalline phase. As
the extraction temperature increases, the endothermic peak at about 230 °C loses depth and definition.
The interpretation of this change is similar to that of the previous NMMO keratins.

There is no doubt about the influence of the extraction temperature on the thermal resistance of the
extracted keratin. For NMMO, the two minimum temperatures of 110 °C and 120 °C preserve the
α-helix keratin crystalline structure through a strong crystalline melting peak. This melting occurs at a
higher temperature than that of raw feathers. The hypothesis proposed here is that an extraction
temperature not exceeding 120 °C allows an efficient dissolution of the feathers while maintaining a
sufficient amount of cysteine to regenerate the disulfide bridges or even generate new disulfide bridges,
thereby increasing the melting temperature of the regenerated keratin. Above an extraction temperature
of 130 °C, the structure is less crystalline and less preserved as the temperature rises. This leads to
progressive keratin degradation and possibly loss of cysteine content after extraction.

IV.3.4.6. X-ray diffraction (XRD)

The crystallinity of proteins is essential for their functional behavior and can be modified or damaged
by dissolution and regeneration processes. It has been shown in the literature that feathers have a
semi-crystalline structure consisting of two distinct structures: α-helix and random coil + β-sheet.
Compared to wool, feathers have a higher β-sheet content than α-helix, making them stronger and more
robust [36], [142], [143]. This information was verified in the context of this work in paragraph II.3.11
(Chapter II). From the XRD analyses performed on the raw feathers and the calculation of their CI, it
was found that the CI of a duck feather, in this case, is about 57.26%, with 31.56% α-helix and 68.44%
random coil + β-sheet. Thus, the percentage of β-sheet structure is much higher than that of α-helix.

Figure 50 shows the XRD curves of extracted keratins as a function of extraction temperature. Only the
two extremes were selected for this analysis. Figure 50 (A) shows keratin extracted at 110 °C and 160 °C
for NMMO, and Figure 50 (B) shows keratin extracted at 130 °C and 170 °C for DES. The shape of the
curve, with two broad diffraction peaks at about 2θ = 9° and 19°, demonstrates the semi-crystalline
structure of keratin. This behavior is very similar to that of duck feathers. As in feathers, the peak at
2θ = 9° is assigned to the α-helix structure and the peak at 19° to the random coil + β-sheet structure.
The CI was also calculated, and the values are listed in Table 11.

In the case of NMMO, the CIs calculated for the two temperatures are close, 46.3% for 110 °C versus
45.9% for 160 °C, although this difference appears too minor to be considered significant. However, the
CI of keratin is lower than that of raw feathers, so the NMMO extraction process results in a loss of
crystallinity of about 10%. Slight changes in the α-helix and β-sheet structure rates are observed, with a
decrease in the β-sheet structure rate at higher temperatures.

105
 (A) (B)

Figure 50. XRD analysis of regenerated keratin depending on extraction temperature. Extracted with: (A)
NMMO 75% and (B) DES (ChCl:urea).

In the case of DES, the CI for temperatures of 130 °C and 170 °C is similar to that of feathers. In this
case, there is no loss of crystallinity after extraction. However, there is a significant difference in the
α-helix ratio, which decreases from 45.7% to 31.3% at the higher temperature. This reflects a transition
from α-helix keratin to random coil + β-sheet structure, which is considered to be less robust than the
latter.

IV.3.5. Influence of the extraction time

IV.3.5.1. Visual aspect
In Figure 45, we can see that the extraction time slightly influences the visual appearance of the keratin.
In the case of DES, keratin extracted at 90 min and 120 min has the lightest color, light beige. Then,
after 2.5 h of extraction, the color becomes slightly deeper. The texture is also affected from 1.5 h to
3 h; the texture is cotton-like, while after 3 h, it becomes denser and sandier. For NMMO, the opposite
is observed after 6 h; the texture is heavier and sandier, while after 7 h, the keratin is lighter and more
cotton-like. In this case, the color is almost invisible to the naked eye.

IV.3.5.2. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-

PAGE)
In the case of the duration parameter, the resulting SDS-PAGE gel is presented in Figure 46. Samples
6 h and 10 h (Image (A), lines 4 and 5) were analyzed for NMMO, and samples 1.5 and 4 h (Image (B),
lines 4 and 5) were analyzed for DES. The results and their interpretation are similar to those obtained
in paragraph IV.3.4.2, with keratin masses close to 10 kDa in both cases, but a difference can be seen in
NMMO, with a wider stain at 6 h than at 10 h. The 6 h stain is extended between 10 kDa and 15 kDa,
whereas the 10 h stain is finer and located at 10 kDa. This result shows that the size of the keratin chains
is slightly larger at shorter extraction times. In addition, the proposed reaction mechanism between

106
NMMO solvent and keratin would result in peptide bond cleavage, resulting in a lower molecular
weight, as discussed earlier (Chapter I).

Table 11. Crystallinity Index of regenerated keratin depending on solvent and extraction parameters.

Random coil +
Solvent Parameter Value CI (%) α-helix (%)
β-sheet (%)

110 °C 46.3 29.7 70.3

Temperature
160 °C 45.9 33.5 66.5

6h 50.7 31.6 68.4

Time
10 h 49.9 37.9 62.1

0.02 55.8 27.1 72.9

NMMO F:S ratio
0.20 51.9 35.9 64.1

Acetic 51.8 34.1 65.9

Sulfuric 50.7 35.0 64.9

Acids
Citric 53.4 30.9 69.1

Water 54.1 37.8 62.2

130 °C 56.5 45.7 54.3

Temperature
170 °C 56.4 31.3 68.7

1h30 59.9 42.6 57.4

Time
4h 56.9 29.6 70.4

0.02 55.9 21.3 78.7

DES F:S ratio
0.15 57.1 35.6 64.4

Acetic 53.3 36.2 63.8

Sulfuric 53.9 29.2 70.8

Acids
Citric 49.5 36.1 63.9

Water 50.3 23.6 76.4

107
 IV.3.5.3. Fourier Transform Infrared (FT-IR) spectroscopy
The FT-IR spectra of keratins extracted during different times are shown in Figure 51. The general
appearance of the spectra is again similar to that of the raw feathers. All the characteristic bands of the
peptide bonds named Amide A and Amide I-III are found at the same wavel engths. Moreover, no
additional bands are visible, testifying to the preservation of the chemical structure of keratin after
solvolysis with either NMMO or DES.

(A) (B)

Figure 51. FT-IR spectra of regenerated keratin depending on extraction time. Extracted with: (A) NMMO and
(B) DES (ChCl:urea).

IV.3.5.4. Thermogravimetric Analysis (TGA)

The influence of extraction time on keratin degradation temperature is shown in Figure 52 via the TGA
curves. As with the TGA curves analyzed previously, the behavior of the curves is similar and includes
the two distinct phases of degradation. Curves (A) for NMMO extraction show clear differences. As the
extraction time increases, the percentage of water in the keratin decreases from 4.5 to 7% for 10 h and
6 h, respectively. One curve stands out from the others, namely the 6 h duration, which shows a
temperature of 190 °C at the start of degradation, much lower than 220 °C for the other extraction
durations. However, the slope of the main degradation curve is less sharp for the 6h curve, which means
that degradation in this case is slower than for the other durations. Then, when 800 °C is reached, the
percentage of residual material is highest for the 6 h and 7 h curves, at about 18-19%.

The curves (B) for DES are similar to the moisture content measured between 4 and 6%, but in the
second stage, the degradation temperature of the extracted keratin with the maximum value of 4 h starts
at 190 °C, 40 °C lower than the degradation temperature of the raw feathers, while the degradation starts
at 220 °C for the other durations. At 800 °C, the values of the remaining material do not follow a clear
pattern, ranging from 13 to 16%.

108
In conclusion, the extraction time influences the degradation temperature of the extracted keratins.
Given the results, in the case of NMMO, it is necessary to choose an extraction time greater than 7 h to
maintain a degradation temperature similar to that of feather keratin. Furthermore, in the case of DES,
an extraction time of less than 4 h is required to maintain a degradation temperature similar to that of
feathers.

(A) (B)

Figure 52. TGA of regenerated keratin depending on extraction time. Extracted with: (A) NMMO 75% and (B) DES
(ChCl:urea).

IV.3.5.5. Differential Scanning Calorimetry (DSC)

The DSC curves are shown in Figure 53. The first endothermic pic is located between 100 °C and
150 °C, which is associated with the elimination of water. Then, the second bimodal endothermic peak

(A) (B)

Figure 53. DSC analysis of regenerated keratin depending on extraction time. Extracted with: (A) NMMO 75%
and (B) DES (ChCl:urea).The endothermic peak located at around 60 °C and indicated by * is an artefact due to
the instrument.

109
 is broad, located between 190 °C and 280 °C, and its shape depends on the solvent used. The
interpretation of this peak is identical to that of the previous DSC analysis. For NMMO, the first
minimum is about 210 °C, and the second is 250-265 °C with a greater depth. For DES, the first
minimum is at about 225-230 °C and the second at about 255-260 °C with almost the same depth.
Therefore, given the similarity of the curves for each of the solvents, it can be assumed that the extraction
time is not a parameter influencing the thermal properties of keratin.

IV.3.5.6. X-ray diffraction (XRD)

Figure 54 shows the XRD curves of extracted keratins as a function of extraction time. Only the two
extremes were selected for this analysis. Figure 54 (A) shows keratin extracted during 6 h and 10 h for
NMMO, and Figure 54 (B) shows keratin extracted during 1.5 h and 4 h for DES. In any case, the semi-
crystalline structure of the keratin is preserved, with two broad peaks at about 2θ = 9 ° and 19 °. These
peaks were previously interpreted.

In the case of NMMO, the extraction time does not seem to influence the CI of the regenerated keratin
since the calculated CIs are 50.7% and 49.9% for extractions of 6 h and 10 h, respectively. A difference
can be seen in the composition of the crystalline structure, with the β-sheet rate decreasing with the
longer extraction time. This rate decreases from 68.4% to 62.1% for 6 h and 10 h, respectively.

In the case of DES, a slight decrease in CI is observed, from 59.9% to 56.9% for 1.5 h and 4 h,
respectively. However, there is a significant difference in the curve.

The intensity of the diffraction peak at 2θ = 19 ° is significantly higher than that of the peak at 2θ = 9 °
for a duration of 1.5 h. This means that the β-sheet rate is much higher than that of the α-helix structure
at low durations, with rates of 70.4% versus 29.6%. This composition is very similar to that of feathers.

(A) (B)

Figure 54. XRD analysis of regenerated keratin depending on extraction time. Extracted with: (A) NMMO 75% and
(B) DES (ChCl:urea).

110
 IV.3.6. Influence of feathers size
IV.3.6.1. Visual aspect
As mentioned earlier, the initial size of the dissolved feathers affects the yield of keratin obtained. In
fact, the finer the feather size, the higher the yield. A comparison of the keratin obtained with an initial
feather grinding size of 0.2 mm or 5 mm indicates no change in visual appearance or color. In Figure 45,
it can be seen that for both solvents, NMMO and DES, the two different-sized feathers are similar.
Therefore, this parameter was not considered for further analysis. As previously explained, fine grinding
requires additional specific equipment and additional time to achieve fine grinding, which increases
costs and energy consumption. This step is avoided when setting up such a process.

IV.3.7. Influence of F:S ratio

IV.3.7.1. Visual aspect
According to Figure 45, the difference in aspect created by the F:S ratio is most obvious when using
DES. At the 0.02 ratio, the keratin is grayish, and its texture is very light and brittle under the fingers.
The powdery appearance of other keratins is not observed in this case. With a ratio of 0.05, the keratin
is slightly darker than that obtained with a ratio of 0.10 or 0.15. This may reflect that the 0.02 and 0.05
ratios are insufficient, with the excess solvent possibly causing keratin degradation. In the case of
NMMO, the color change is extremely subtle to the human eye, and the texture becomes sandier at ratios
of 0.10 and above.

IV.3.7.2. Fourier Transform Infrared (FT-IR) spectroscopy

The FT-IR spectra depicted in Figure 55 illustrate the keratins extracted at various F:S ratio. The overall
profile of the spectra closely resembles that of the untreated feathers. Notably, the distinctive bands
corresponding to the peptide bonds, specifically Amide A and Amide I-III, appear at identical

(A) (B)

Figure 55. FT-IR spectra of regenerated keratin depending on F:S ratio. Extracted with: (A) NMMO and (B) DES
(ChCl:urea).

111
wavelengths. No additional bands appeared, but there is a difference in the intensity of the bands located
between 3000 and 2830 cm-1 for both solvents for the 0.02 F:S ratio compared to the other F:S ratios.
This difference is also present in the curve corresponding to the 0.05 ratio for DES. This wavelength
corresponds to stretching the -CH2 and -CH3 groups [15], whose proportion increases when the solvent
is in excess with low ratios.

IV.3.7.3. Thermogravimetric Analysis (TGA)

The influence of the F:S ratio could be demonstrated by TGA analysis. In Figure 56 (A) and (B), the
curve for the 0.02 ratio reveals a significantly lower degradation temperature than the other ratios. For
NMMO, the main degradation starts at 180 °C, with a loss of 12% at 230 °C, and for DES, the
degradation also starts at 180 °C, with a loss of 11% at 220 °C. This early degradation may be due to
keratin damage during the extraction process due to excess solvent. Indeed, as determined in the
previous chapter, the minimal yield associated with this ratio confirms that feather dissolution is highly
effective in breaking many hydrogen bonds and disulfide bridges, which are not regenerated during the
precipitation stage. The keratin structure is thus damaged and weakened, making it less heat-resistant.
This is also confirmed by the amount of residual matter at 800 °C, which is lowest for the 0.02 ratio
with 14% for DES and 13% for NMMO, in contrast to the other ratios with an average of 16% for both
solvents.

(A) (B)

Figure 56. TGA of regenerated keratin depending on F:S ratio. Extracted with: (A) NMMO 75% and (B) DES
(ChCl:urea).

IV.3.7.4. Differential Scanning Calorimetry (DSC)

Figure 57 shows the DSC curves obtained for the different ratios studied. Both curves (A) and (B) show
the endothermic peak at 100-150 °C, indicating the elimination of water from the samples. Moreover,
one curve stands out in both cases: the 0.02 F:S ratio. In the yield calculations and previous analyses, it
has been shown that this ratio is detrimental to the extraction of keratin since the excess solvent degrades

112
it during the extraction. In fact, this hypothesis can be verified in its thermal resistance analysis. As
observed in the TGA analysis, the melting of the crystalline structure occurs before the other ratios. In
graph (A), the melting of the α-helix structure occurs at about 175 °C, and in graph (B) at about 180 °C.
In both cases, the peak is broad, with almost only a single trough, indicating a major loss of α-helix
structure, which would then be hardly visible on the curve and mixed with keratin degradation.

The curves for the other ratios are bimodal in the case of DES and feature a more complex shape in the
case of NMMO. The interpretation of this peak has already been described above. Here, the F:S ratio
affects the thermal properties of the extracted keratin. The loss of thermal properties is confirmed for
the 0.02 ratio.

(A) (B)

Figure 57. DSC analysis of regenerated keratin depending on F:S ratio. Extracted with: (A) NMMO 75% and
(B) DES (ChCl:urea). The endothermic peak located at around 60 °C and indicated by * is an artefact due to
the instrument.

IV.3.7.5. X-Ray diffraction (XRD)

XRD curves of extracted keratins as a function of F:S ratio are presented in Figure 58. Only the two
extreme ratios were selected for this analysis. Figure 58 (A) shows keratin extracted with an F:S ratio
of 0.02 and 0.20 for NMMO and Figure 58 (B) shows keratin extracted with F:S ratio of 0.02 and 0.15
for DES. On both graphs, the semi-crystalline structure of the keratin is preserved with the presence of
two broad peaks at about 2θ = 9 ° and 19 °. These peaks were previously interpreted.

In the case of both NMMO and DES, there is a clear difference between the two ratios studied. Indeed,
the curve corresponding to the F:S ratio of 0.02 shows a diffraction peak intensity at 2θ=19 ° higher
than that of the diffraction peak at 2θ=9 °. As seen from the values in Table 11, this means that the
random coil + β-sheet phase is superior to the α-helix phase. Thus, at low ratios, the β-sheet phase
predominates in regenerated keratin, reaching 72.9% for NMMO and 78.7% for DES. At higher ratios,

113
these values decrease to 64.1% and 64.4%, respectively. This large amount of β-sheet structure explains
the "brittle" and crystallized appearance of the keratin studied in the visual analysis section.

(A) (B)

Figure 58. XRD analysis of regenerated keratin depending on F:S ratio. Extracted with: (A) NMMO 75% and (B)
DES (ChCl:urea).

IV.3.8. Influence of the acid used for keratin regeneration

IV.3.8.1. Visual aspect
The type of acid used for keratin regeneration is the parameter with the greatest influence on the
appearance and color of the keratin, as can be seen in Figure 45. In the case of DES, keratin regenerated
without acid, with the addition of water only, is very different from keratin regenerated with acid. Its
texture is very light and "crunchy" under the fingers, with a unique dark yellow color. The other three
acid-precipitated keratins are similar in texture, but the color is slightly darker in the case of citric acid.

The same observations apply to NMMO. Water-regenerated keratin is very different from the other
three, with the same texture as DES. Sulfuric and acetic acid-precipitated keratins are similar, with the
same color and texture, while citric acid-precipitated keratins are heavy and sandy, with a darker color.

IV.3.8.2. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-

PAGE)
In the case of both solvents, the 4 keratins obtained by acid or acid-free regeneration were compared in
Figure 46. Lines 6 to 9 represent keratins regenerated with acetic acid, sulfuric acid, citric acid, and
water, respectively, in the two images (A) and (B). Again, no change in molecular mass is observed in
all cases; the molecular mass of the keratins is around 10 kDa, in agreement with the previous
interpretation in section IV.3.4.2.

114
 IV.3.8.3. Fourier Transform Infrared (FT-IR) spectroscopy
FT-IR spectra of regenerated keratins with and without acidification are shown in Figure 59. The band
interpretation is identical to previous interpretations, with an apparent similarity to the spectrum of raw
feathers. There is no difference between the different types of acidification. Therefore, this parameter
does not affect the chemical structure of the keratin.

(A) (B)

Figure 59. FT-IR spectra of regenerated keratin depending on regeneration step. Extracted with: (A) NMMO and
(B) DES (ChCl:urea).

IV.3.8.4. Thermogravimetric Analysis (TGA)

Figure 60 shows the TGA curves for keratins regenerated with different acids or without acidification.
In both cases, the behavior of the curves is similar, showing the two phases of degradation. In the case
of DES, the average moisture content is 6%, degradation starts at 215-220 °C, and the amount of residual
material at 800 °C is 16%. These values are similar to previous interpretations. For NMMO, the average
moisture content is 5%, and degradation starts at 225-230 °C, with a slight difference visible for acetic
and sulfuric acids, which seem to produce slightly more degradation than citric acid and water, with an
additional 2% loss at 215 °C. The amount of residual material is between 16 and 18%, similar to previous
interpretations. In this case, the type of acid used does not affect the thermal properties of the keratin.

IV.3.8.5. Differential Scanning Calorimetry (DSC)

Figure 61 shows the DSC curves for keratins regenerated with different acids. In part (A) of this figure,
the curves show the same behavior, with endothermic peaks aline on the same temperatures. There is a
slight endothermic peak for water removal at about 125 °C, followed by two endothermic troughs at
210 °C and 255 °C for NMMO keratin. Then, for DES, the water-related peak is also present at the same
temperature. A difference can be seen in the acetic acid curve, which contains the two endothermic
peaks at 227 and 260 °C, with two broad but defined peaks.

115
 (A) (B)

Figure 60. TGA of regenerated keratin depending on acid used during regeneration step. Extracted with: (A)
NMMO 75% and (B) DES (ChCl:urea).

On the other hand, the degradation trough is less defined for sulfuric acid, acetic acid, and water, with
the curve continuing to decline towards the endothermic axis at 280 °C. Therefore, it can be concluded
that the degradation of keratin precipitated with sulfuric acid, citric acid, or water is more spread over
the temperature range from 240 °C to over 280 °C. However, the type of acid used does not seem to
affect the thermal resistance of keratin phase transitions.

(A) (B)

Figure 61. DSC analysis of regenerated keratin depending on acid used during regeneration step. Extracted with: (A)
NMMO 75% and (B) DES (ChCl:urea).The endothermic peak located at around 60 °C and indicated by * is an artefact
due to the instrument.

116
 IV.3.8.6. X-Ray diffraction (XRD)
XRD graphs of extracted keratins regenerated with different acids are presented in Figure 62. Graphs
(A) and (B), for NMMO and DES, respectively, contain all acids types studied for keratin regeneration:
acetic acid, sulfuric acid, citric acid, and also without acid. Here, the semi-crystalline structure of keratin
is preserved for each type of regeneration, and the two known diffraction peaks are present at the same
angles. These peaks were previously interpreted. Overall, all calculated CIs are close to 50%, and there
is no clear trend for α-helix and random coil + β-sheet structures. For example, the rates of random coil
+ β-sheet from lowest to highest for NMMO are: water < sulfuric acid < acetic acid < citric acid, while
for DES, they are: acetic acid < citric acid < sulfuric acid < water. However, the values obtained are not
very different. Therefore, it can be assumed that the type of acid used does not influence the crystal
structure of regenerated keratin.

(A) (B)

Figure 62. XRD analysis of regenerated keratin depending on acid used during regeneration step. Extracted with: (A)
NMMO 75% and (B) DES (ChCl:urea).

IV.3.8.7. Solid-state 13C CP MAS Nuclear Magnetic Resonance (NMR)

Here, the influence of the acid type used on keratin's chemical structure after regeneration has been
analyzed via NMR. The NMR spectra are shown in Figure 63. These spectra show all the characteristic
bands previously determined in the NMR of feathers (paragraph II.3.13, Chapter II) and section IV.3.3.
All the spectra are similar, demonstrating that despite the lack of acidification with water only, the
regenerated material has the same chemical composition as keratin regenerated by acidification. As
described in the previous chapter, the type of acid used has an influence on yield but this analysis shows
that not on chemical composition.

117
(A) (B)

Figure 63. Solid-state 13C CP MAS NMR spectrum of regenerated keratin with the use of water or acids. Image (A)
correspond to keratin extracted with NMMO and image (B) correspond to keratin extracted with DES. The spinning
band is noted by *.

118
 IV.4. Conclusion
In this chapter, a physicochemical analysis of keratins extracted and regenerated from duck feathers has
been carried out. The main objective of this work was to determine the physicochemical properties of
these keratins to determine the influence of each extraction parameter studied on them. Thus, by
comparing the properties with those of the raw feathers, we were able to observe the changes that take
place during the extraction and regeneration process, and we were also able to compare the influence of
the solvent. The first observation was made to compare the visual aspect of each extracted keratin. A
considerable difference was observed, with a yellowish coloration with DES versus a dark brown
coloration with NMMO. It was explained that this coloration was thanks to the formation of
chromophores due to the presence of cyclic amino acids in the keratin and to the formation of
chromophores during the reaction of NMMO with heat. SEM images revealed that the morphology of
extracted keratin is a mixture of fibrous structure and a random assembly of aggregates that also form
porous arrangements. We have seen that the solvent used affects the morphology of the keratin.
Subsequent chemical structure analysis by NMR and FT-IR showed that the extracted material was
keratin and that its chemical structure was similar to that of raw feathers. The solvent used did not affect
the chemical composition of the keratin.

Furthermore, the band attributed to β-carbon in the NMR spectra indicated that the β-sheet structure is
preserved during the extraction process, reflecting the preservation of keratin disulfide bridges or their
regeneration after dissolution. SDS-PAGE analysis determined the molecular mass of the extracted
keratin to be approximately 10 kDa, which agrees with the literature and the molecular mass of raw
feather keratin. Therefore, no change in the size of the keratin chains is observed after extraction,
regardless of the solvent or parameters studied. Other methods should be employed to have more precise
data for this measurement. Overall, the degradation temperature of feathers is maintained at about
230 °C after extraction with NMMO, while extraction with DES reduces this value by 10 °C. The higher
the extraction temperature, the faster the degradation and the smaller the material remains. The lower
the extraction temperature, the more likely it is that the properties of the material will be preserved.
Concerning the duration parameter, it was concluded that a minimum of 7 h of extraction is necessary
for NMMO to maintain the above degradation temperature and that in the case of DES, it is better to
choose an extraction of less than 4 h to not lower the degradation temperature. It was also shown that in
addition to extremely low yields, the F:S ratio of 0.02 significantly reduces the thermal properties of
keratin.

Furthermore, changes in crystallinity were observed as a function of the parameters studied. Indeed, the
lowest CIs were calculated with the DES solvent and the highest with the NMMO solvent. Extractions
with NMMO kept the CI value of the feathers as close as possible. In any case, the random coil + β-sheet
structure is still dominant in regenerated keratin. It would be interesting to perform a quantitative
analysis of the amount of cysteine and cystine present in the feathers, during feather dissolution, and in
the regenerated keratin to understand the dissolution process better and to confirm the hypothesis that

119
disulfide bridges are either preserved during dissolution, regenerated during the regeneration stage, or
both.

Keratin is a semi-crystalline material that can assume different morphologies: fibrous or aggregated,
with good thermal and chemical properties and low density. As a natural material in the human body,
its main advantages are biocompatibility and biodegradability. It is a highly interesting natural material
for various applications, including cosmetics, hair care products or creams and treatments;
pharmaceuticals and medical devices; drug delivery systems; agriculture, biofertilizers; and packaging,
bio-based films, and coatings.

120
 Chapter V
Elaboration and characterization of
bio-based films with cellulose and
keratin

121
122
 V.1. Motivation
Faced with the environmental degradation caused, in part, by the production and use of plastics, the
packaging industry was required to change and explore more environmentally sustainable alternatives
to synthetic plastics. Traditionally, packaging has been composed mainly of petroleum-based plastics,
which are known to cause significant environmental problems throughout their life cycle. As a result,
France has imposed restrictions on synthetic plastics and encouraged manufacturers to adopt more
environmentally friendly and cost-effective methods and processes. This shift in research focuses not
only on environmental considerations but also on maintaining performance standards. As a result,
research is focused on innovative materials that meet environmental goals while maintaining
performance expectations. One such solution is using renewable bio-based and natural polymers with a
low environmental impact. Cellulose, a naturally abundant material, has been used in packaging,
particularly in the form of films, but some drawbacks limit its broad application. In particular, its
hygroscopic nature challenges achieving optimal water resistance, a critical requirement for the food
industry. This chapter aims to improve the hydrophobic properties of cellulose by incorporating keratin
into the film formulation. Two types of cellulose have been tested for use in the films: commercially
available microfibrillated cellulose and laboratory-produced tempo-oxidized nanocellulose. This work
was conducted in collaboration with the LEPAMAP-PRODIS laboratory in Girona, Spain, as part of an
international mobility program. Once the films were prepared, a physicochemical and mechanical
analysis was carried out to compare the properties of the different formulations. Different manufacturing
processes, solvent casting, and hot pressing were also studied to optimize the appearance and properties
of the films. Finally, different keratin additions and the presence or absence of a plasticizer were studied.

V.2. Experimental methods

V.2.1. Materials
Commercially-available 2% micro fibrillated cellulose (MFC), TEMPO-oxidized cellulose nanofibers
(TOCNF) (University of Maine, USA), and laboratory-made TOCNF were used as a film-forming base.
The MFC has been used as purchased without modification. TOCNF from UMaine was used as received,
and TOCNF was developed in the LEPAMAP-PRODIS laboratory from commercially dried bleached
Eucalyptus pulp. 2,2,6,6-Tetramethyl-1-piperidinyloxy (TEMPO) reagent 98%, sodium hypochlorite
(NaClO) (9.8% w/v), sodium bromide (NaBr), sodium hydroxide (NaOH) (0.5M) and distilled water
were used for TEMPO-mediated oxidation. Borate buffer 1M pH=8.5, methyl blue 300mg/L, and
hydrochloric acid 0.1M were used to analyze the carboxyl content (CC) of TOCNF. Cationic polymer
polydiallyldimethylammonium chloride (polyDADMAC) and anionic polymer (Pes-Na) were used to
analyze the cationic demand (CD) of TOCNF. Finally, copper (III) ethylenediamine (CED) reagent was
used to calculate the degree of polymerization (DP) of TOCNF. The keratin selected for film formulation
was NMMO-extracted keratin at 110 °C. It was used either as a dry powder after freeze-drying and
stored in a closed container at room temperature or humid before freeze-drying and stored in a closed
container in the refrigerator (cf. Figure 64). Commercial sorbitol and polyethylene glycol (PEG) 400
were used as plasticizers in the film’s preparation and without modification. Two different types of

123
 Figure 64. Pictures of regenerated keratin extracted via NMMO solvolysis, humid form (left) and dry form
(right).

membrane were used for film preparation: nylon membrane filters (Cytiva, WhatmanTM) of 0.45 µm
for MFC films and mixed cellulose esters membrane (Merck) of 0.22 µm for TOCNF films.

V.2.2. TOCNF preparation

The literature mentions that cellulose nanofibers have highly appreciated properties in the papermaking
and packaging sectors due to their low density, biodegradability, biocompatibility, and good mechanical
strength. Various processes for chemical modification of cellulose exist and are being implemented, but
TEMPO-mediated oxidation is one of the most widely used at the research level. It is based on the
oxidation of primary cellulose C6 -OH to -COOH group. This process allows the extraction of
transparent nanocellulose at room temperature, but the production cost remains high due to the high cost
of the TEMPO reagent. However, pretreatment such as TEMPO oxidation reduces the number of times
the cellulose must pass through a high-pressure homogenizer, thereby reducing energy consumption
[111].

TOCNFs were prepared according to the protocol of Serra et al. [111] with slight modifications. For
this experiment, the final dry mass of nanofibers aimed is 25 g. The first step consists of pulping the raw
material, ashless blotting paper, by hydrating the fibers in water under mechanical agitation at
30,000 rpm for a few minutes at RT. Once hydrated, the fibers are filtered and recovered and the
consistency is measured (≈17%). In the second step, TEMPO and NaBr are solubilized in distilled water
under mechanical agitation at RT for approximately 20 to 30 min until complete solubilization. In the
protocol used, the TEMPO required is 16 mg per g of dry fibers, and the NaBr required is 100 mg per g
of dry fibers. Next, a quantity of hydrated pulp (calculated according to its consistency) is added to the
TEMPO + NaBr + H2O mixture, stirring for 10-15 minutes to allow the fibers to untangle. This step is
critical for efficient oxidation. It is important to note that consistency of about 1% fibers in water is
recommended during the reaction, so the amount of water has been calculated according to the target
dry fiber mass. Next, a pH electrode is inserted into the mixture to measure the pH of the reaction in
real-time. Initially, the reaction has a pH of about 6.3. The oxidation reaction begins when NaClO is
added to the mixture. NaClO added was calculated as 5 mmol per g of dry fibers. The amount of NaClO
added will affect the degree of oxidation generated on the cellulose fibers. In this case, 5 mmol of NaClO

124
will produce a nanocellulose called hereafter "TOCNF-5". Maintaining a constant pH of around 10 is
necessary to avoid degradation of the cellulose fibers. This is achieved by adding dropwise NaOH.

The reaction is considered complete when the pH of the solution is constant without adding further
NaOH. Once the oxidation reaction is complete, the solution is filtered, and the cellulose fibers are
rinsed several times with distilled water. At this stage, the fibers obtained are white and oxidized, with
a consistency of about 12%. The fibers are then reduced in size to nanofibers using a high-pressure
homogenizer (PANDA 2000, GEA NIRO-Soavi, Parma, Italy). A suspension with 1% consistency of
the previously oxidized fibers is passed through the homogenizer at different pressures: 3 passes at
300 bar, 3 at 600 bar, and 3 at 900 bar. A gel-like transparent suspension is obtained and stored in a
refrigerator until use. The final consistency after homogenization of TOCNF-5 is approximately 1%.

V.2.3. TOCNF characterization

V.2.3.1. Degree of polymerization (DP)
Intrinsic viscosity was measured following UNE 57-039-92 standard method with the following
modification: instead of 20 °C, the analysis was performed under 25 °C. The concept of capillary
viscosimeter is based on the flow time of a given volume of polymer solution through a thin capillary
and compared with that of a solvent. It turns out that the flow time of either is proportional to viscosity
and inversely proportional to density. This method involves measuring the viscosity of solutions of
cellulose mixed with CED at different concentrations at a temperature of 25 °C. The different solutions
were prepared with 0, 40, 60, 80, 100, and 120 mg of dry TOCNF to which 5 mL of CED and about 5
mL of water were added (the amount of H2O varied according to the mass of the fibers and the fibers
consistency). The solution without fibers is used as a reference for the calculation.

The DP is defined as the molecular weight of a polymer divided by the molecular weight of the repeating
unit:
𝑀𝑊 𝑐𝑒𝑙𝑙𝑢𝑙𝑜𝑠𝑒
DP = 𝑀𝑊 𝑔𝑙𝑢𝑐𝑜𝑠𝑒 𝑢𝑛𝑖𝑡 (4)

Where MW cellulose is the molecular weight of TOCNF added to the solutions, and MW glucose unit
is the molecular weight of the glucose unit (162 g/mol).

The molecular weight of cellulose can then be calculated using Mark-Houwink equation:

Ƞ = K*(MW cellulose)a (5)

Where Ƞ is the intrinsic viscosity (dL/g), MW cellulose is the molecular weight of TOCNF, K and a are
constants (2.28 and 0.76, respectively) [111]. As a result, the calculations were performed using these
different constants, and the DP value was then averaged.

In order to calculate this intrinsic viscosity, further calculation steps are required before plotting the
inherent viscosity and intrinsic viscosity as a function of solution concentration. The equations are as
follows:

125
 𝑡 𝑐𝑠
Ƞrel = 𝑡 𝑠𝑜𝑙𝑣𝑒𝑛𝑡 (6)

𝑡𝑐𝑠−𝑡 𝑠𝑜𝑙𝑣𝑒𝑛𝑡
Ƞsp = 𝑡 𝑠𝑜𝑙𝑣𝑒𝑛𝑡
(7)

𝐿𝑁(Ƞ𝑟𝑒𝑙)
Ƞi = (8)
𝐶

Ƞ𝑠𝑝
Ƞ= 𝐶
(9)

Where Ƞrel is the relative viscosity, tcs is the average of the times measured for the cellulose solution
studied, tsolvent is the time average of the reference solution, Ƞsp is the specific viscosity, Ƞi is the inherent
viscosity, C is the cellulose solution concentration (mg/ml).

V.2.3.2. Carboxyl content (CC)

The carboxyl groups in cellulose are important for its reactivity. The presence of these groups, which
can be oxidized, is responsible for the ion exchange capacity and can retain cationic groups. Methylene
Blue (MB), used as a cationic, water-soluble color indicator, can bind to the carboxyl groups of
cellulose. It has been shown that at a wavelength of 664 nm, the absorbance of MB is proportional to its
concentration [167] and related to the ion exchange capacity of the cellulose. Absorption of samples
was analyzed by a UV-visible absorbance analysis with a Shimadzu UV-Vis spectrophotometer
UV-160A (Shimadzu Europa GmbH, Duisburg, Germany). This method requires the construction of a
calibration curve showing the absorbance of solutions as a function of Methylene Blue concentration
(in ppm). TOCNF samples were prepared by mixing 0.5 g wet TOCNF with 5 mL borate buffer and
5 mL methylene blue. The mixtures were centrifuged at 2300 rpm for 1 hour, then 2 mL supernatant
was collected and diluted in a 25 mL volumetric flask with hydrochloric acid. The absorbance was then
measured at 664 nm, and the value obtained was plotted on the calibration curve. The analysis was
performed in duplicate.

V.2.3.3. Cationic demand (CD)

The cationic demand of TOCNF was carried out by preparing samples containing 0.2 g wet TOCNF
(i.e. approx. 0.002 g dry TOCNF) and 10 mL polyDADMAC. After vigorous mixing, the samples were
centrifuged at 2 300 rpm for 1 h, and the cationic demand was then measured using a particle detector
charge Mütek PCD 04 (BTG Mütek GmbH, Herrsching, Germany). The principle of CD analysis is
based on the interaction of the cationic titrant with the negative charges present in the cellulose. The
cellulose retains a certain quantity of cationic charges while the excess titrant remains free in the
solution. The excess of remaining cationic charges is then measured by dropwise addition of the anionic
polymer titrant (Pes-Na). A 1 mL supernatant sample is placed in the analyzer, and its charge is
measured in real-time as the anionic titrant is added. The experiment ends when the sample charge
stabilizes at zero. All cationic charges have regained their value with the anionic titrant. The experiment
was performed in duplicate. The volume of Pes-Na poured is then noted and used to calculate the CD
using the following Equation 10:

126
 CD = (CpolyDADMAC x VpolyDADMAC) – (Cpes-Na x Vpes-Na) / WTOCNF (10)

Where CpolyDADMAC and VpolyDADMAC are the concentration (0.001 meq/L) and volume (L) of
polyDADMAC, Cpes-Na and Vpes-Na are the concentration (0.001 meq/L) and volume (L) of Pes-Na and
WTOCNF is the weight of TOCNF sample (g).

V.2.4. MFC films preparation

V.2.4.1. MFC(A) films preparation
The first cellulose films tested in this study were based on 2% MFC and named MFC(A). A final weight
of 0.5 g was targeted for each film. Different formulations containing keratin were tested. The
percentages of keratin tested were 0, 5, 10, 15, and 20% (corresponding films are noted MFC(A),
MFC(A)-K5, MFC(A)-K10, MFC(A)-K15 and MFC(A)-K20).

In the first step of the elaboration protocol, the MFC mass is weighed to obtain the desired final mass,
considering its consistency. Here, 25 g of 2% MFC was weighed, and 150-200 mL of distilled water
was added for dilution. For films containing keratin, the weight of cellulose was adjusted, and keratin
was added in dry powder form. Homogenization was performed at 10,000 rpm for 5 minutes using a
high-shear homogenizer (IKA Ultra-turrax® T25 digital). The solution was then membrane-filtered to
obtain a gel-like film. The gel-like film was carefully peeled off the membrane, as it is very fragile, and
placed between two non-stick papers, which were placed between two absorbent papers, placed between
two metal plates, and hot-pressed with a heat press (Carver, hydraulic press, auto series). Pressing
occurred under 2000 lbf (≈ 1 bar) for 60 s, then 3,000 lbf (≈ 1.5 bar) for 60 s, then 7,000 lbf (≈ 3.5 bar)
for 720 s at 105 °C (a total of 14 min). After the cycle, the films remained slightly moist and had to be
reheated directly between the two metal plates for about 1 minute to complete the drying process. A
general scheme of the manufacturing process is shown in Figure 65, and the process described here is
called P1. The resulting films were stored at RT.

V.2.4.2. MFC(B) films preparation

A second film production was performed to refine the thickness of the first films obtained. The target
film mass was now 0.25 g. A "control" film was repeated at 0.5 g for comparison. The same cellulose
is used as in the previous case, but the amount of cellulose differs. A mass of 12.5 g of wet MFC 2% for
the pure MFC film was weighed. When keratin is added, this mass varies accordingly. As in the previous
case, different amounts of keratin were tested, ranging from 0 to 20%, and these films are named
according to the same principle, from MFC(B) to MFC(B)-K20. The solutions are then homogenized
for a longer time, 15 minutes at 10,000 rpm, using a high-shear homogenizer (IKA Ultra-turrax® T25
digital), and the solution is filtered through a nylon membrane. Once the "gel-like" film is formed, it is
left on its filtration membrane, placed between two anti-adherent papers, then between two absorbent
papers, and cold-pressed between two metal plates using a manual press under 1 ton for 5 min. This
pre-pressing step is designed to dewater the film and facilitate its drying during hot pressing. As in the
previous protocol, the film is gently removed from its membrane and placed between two metal plates.
The pressing parameters are as follows: 15 min at 9,000 lbf (≈ 4.5 bar), followed by a change of

127
absorbent paper, and another 15 min at 9,000 lbf (≈4.5 bar) at 105 °C. This second manufacturing
process is summarized in the scheme shown in Figure 65 and named P2. The resulting films were dried
and stored at RT until analysis.

P1

P2

P3

P4

Figure 65. Scheme summarizing the manufacturing processes for MFC and TOCNF films, from process P1 to
process P4.

128
 V.2.5. MFC film characterization
V.2.5.1. MFC(A) characterization
Film thickness

Film thickness was assessed using a micrometer. Each measurement was taken at a different point on
the film, 20 values were recorded, and the final thickness of each film was averaged.

Film density

The mass of each film was measured using a precision balance (± 0.1 mg). The diameter of the films
was measured with a caliper. Estimated volume and density were calculated using Equation 11:
𝑀
𝜌= 𝑉
with 𝑉 = π ∗ r 2 ∗ 𝑒 (11)

Where ρ is the density, M is the mass (g), V is the estimated volume (cm3), r is the radius (cm), and e is
the thickness (cm).

Grammage and porosity

The grammage (G) of MFC(A) films was calculated in g/m². The porosity (P) of each film was
determined from its previously calculated density and the density of each component according to
Equation 12:
𝜌𝑀𝐹𝐶
P=1– (12)
𝜌𝐶+ 𝜌𝐾

Where ρMFC is the film density, ρC is the density of pure cellulose assumed to be 1.5 g/cm3 multiplied
by the fraction contained in the film, and ρK is the density of keratin evaluated previously at 1.3 g/cm3
multiplied by the fraction contained in the film.

Water contact angle (WCA)

Film surface properties were measured using a drop shape analyzer (Krüss DSA 10 Mk2) via the contact
angle sessile drop method using distilled water. Five measurements were performed for each film, and
the contact angle value was averaged. Data were collected using the ADVANCE software.

Mechanical test

Tensile tests of MFC(A) films were performed in a universal testing machine (3R, France). Samples
were cut to obtain 40 mm length and 7 mm width. The tests were carried out with a clamp distance of
20 mm, a preload of 0.5 N, and a 10 mm/min speed. Data were collected using SyntX-Lite software.
The results presented are the mean values of eight trials.

Thermogravimetric analysis (TGA)

Thermal degradation of MFC(A) films was determined by using a TA Q2500 thermogravimetric

analyzer. Films were heated from 30 °C to 800 °C with a 10 °C/min rate under a nitrogen atmosphere.
129
 Differential scanning calorimetry (DSC)

The melting temperature of MFC(A) films was studied from -50 to 250 °C with a heating rate of
10 °C/min using a TA Q20 differential scanning calorimeter and continuous nitrogen purge.

Fourier Transform Infrared (FT-IR) spectroscopy

FT-IR spectroscopy in the Attenuated Total Reflectance (ATR) mode was conducted on MFC(A) films
using a Jasco FT/IR-4700 infrared spectrometer. Each spectrum was generated by 32 scans at a
resolution 4 cm-1 on a range between 4000 and 400 cm-1.

V.2.5.2. MFC(B) characterization

Film thickness

The thickness of each MFC(B) film was measured with a micrometer (Starrett, USA) on five different
points of the film surface, and the value was averaged.

Film density

The mass of each film was measured using a precision balance (± 0.1 mg). The diameter of the films
was measured with a caliper. Estimated volume and density were calculated using the previous
Equation 11.

Grammage and porosity

The grammage of MFC(B) films was determined and averaged over 3 specimens of each composition
studied. Each film was cut with a chosen dimension of 20×10 mm, its mass was weighed on a precision
balance to within ± 0.1 mg, and its grammage (G) was calculated in g/m². The porosity (P) of each
MFC(B) film was determined following the previous Equation 12.

Scanning electron microscopy (SEM)

The scanning electron microscopy technique was employed to examine the morphology and
microstructure of MFC(B) films. For this purpose, samples underwent carbon coating through the
evaporation method using a turbo evaporator device (Emitech K950, Germany). The observations were
conducted in high vacuum mode at 5kV using a FE-SEM Hitachi S-400 (Japan), and image acquisition
was performed using Quartz PCI software.

X-ray diffraction (XRD)

X-ray diffraction (XRD) patterns for MFC(B) films were acquired using a PANanalytical X’Pert PRO
diffractometer to analyze the crystallinity index (CI) of films. The measurements were performed with
CuKa radiation (X = 152.4 pm), and data were recorded utilizing variable slits.

The crystallinity index (CI) calculation was executed using Equation 13.

130
 𝑇𝑜𝑡𝑎𝑙 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑐𝑟𝑦𝑠𝑡𝑎𝑙𝑙𝑖𝑛𝑒 𝑝𝑒𝑎𝑘𝑠
CI (%) = 𝑇𝑜𝑡𝑎𝑙 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑡ℎ𝑒 𝑐𝑢𝑟𝑣𝑒
∗ 100 (13)

Water contact angle (WCA)

The surface properties of the films were assessed using a Drop Shape Analyzer (Krüss Advance)
employing the contact angle sessile drop method with distilled water. Data acquisition was performed
through the ADVANCE software, and the analysis was replicated in triplicate.

Mechanical test

Mechanical tests on MFC(B) films were conducted using a Hounsfield tensile tester with a 2.5 KN
loading cell and pneumatic clamps. Five samples were cut to dimensions of 80 mm length and 15 mm
width for each film type. Before analysis, the samples were conditioned in a climatic chamber for at
least 24 hours at 23 °C and 50% humidity. The tests were executed with a clamp distance of 60 mm, a
preload of 0.01 N, and a speed of 5 mm/min. Data were recorded using the Data Hawk software, and
the presented results represent the mean values obtained from the five trials.

Thermogravimetric analysis (TGA)

Thermogravimetric analysis (TGA) data were acquired using a Mettler Toledo thermogravimetric
analyzer, model TGA-SDTA851. Samples weighing 10 mg were heated at 10 °C/min from room
temperature to 600 ºC under a nitrogen gas flow rate of 40 cm³/min. Two replicates were performed,
and errors were maintained below 1%.

V.2.6. TOCNF films preparation

The TOCNF films were produced during the international mobility at the LEPAMAP-PRODIS research
group. The main objective was to maintain a film-making protocol as identical as possible to the
MFC(B) film-making process to compare the properties of each cellulose and observe the influence of
keratin additions. Therefore, the manufacturing process for TOCNF films follows the same principle as
for MFC(B) films, with a few modifications to adapt to changes in material and equipment. These films
were made using the P3 process, shown in Figure 65, which resembles the P2 process without the cold-
pressing step. TOCNF-5 is mixed with water and various percentages of keratin. For this study, two
types of keratin were tested: dry keratin powder after freeze-drying and humid keratin before freeze-
drying. Films made with a determined percentage of dry keratin are called TOCNF-5-P-DKX, and films
containing an "x" percentage of humid keratin are called TOCNF-5-P-HKX. Some formulations also
feature a plasticizer, sorbitol, or PEG 400, added to the solution. This mixture is then homogenized
using a homogenizer (IKA Ultra-turrax® T25 digital) for 15 minutes at 10,000 rpm. The solution is then
filtered on a cellulose ester membrane. The resulting "gel-like" film is gently peeled from its membrane,
and the paper/cellulose film/metal plate assembly is hot-pressed at 105 °C under 4 bars for 15 min.

A second manufacturing method was also tested for TOCNF films, the P4 process, which involves
solvent casting. In this case, two types of TOCNF were used: TOCNF-5 and TOCNF-15. The two types
of keratin were also tested with a percentage "X", dry and humid, so the films are called, in this case:

131
TOCNF-5-S-DKX / TOCNF-5-S-HKX or TOCNF-15-S-DKX or TOCNF-15-S-HKX. The first mixing
step is identical to P2, but this solution is poured into a 90 mm diameter petri dish and dried by
evaporating the solvent at RT for several days (approximately 7-10 days) until the film is completely
dry. The film is then removed from the petri dish.

V.2.7. TOCNF film characterization

Film thickness

The thickness of TOCNF films was measured with a micrometer (Starrett, USA) on five different points
of the film surface, and the value was averaged.

Film density

The mass of TOCNF film was measured using a precision balance (± 0.1 mg). The diameter of the films
was measured with a caliper. Estimated volume and density were calculated using the previous
equation 11.

Grammage and porosity

The grammage of TOCNF films was determined and averaged over 3 specimens of each composition
studied. Each film was cut with a chosen dimension of 20×10 mm, its mass was weighed on a precision
balance to within ± 0.1 mg, and its grammage (G) was calculated in g/m². The porosity (P) of each film
was determined following the previous Equation 12.

Scanning electron microscopy (SEM)

SEM images of TOCNF films were acquired using the same parameters as for MFC(B) films.

X-ray diffraction (XRD)

XRD spectra of TOCNF films were produced using the same parameters as for MFC(B) films. CI values
were calculated following Equation 13.

Water contact angle (WCA)

Water contact angle values of TOCNF films were recorded as for MFC(B) films.

Water vapor transmission rate (WVTR)

The WVTR of TOCNF films was measured using the cup method according to ASTM E96 standard.
The principle is based on measuring the passage of water vapor through a solid material over a specific
period. TOCNF films were conditioned 48 hours in advance in a climatic chamber at 23 °C and 50%
humidity. A waterproof plastic box was filled with oven-dried silica beads. A cellulose film was laid
down and held between the container and its lid. The center of the lid was cut in a circle. In this way,
the surface area of the film in contact with its environment is controlled by the diameter of the circle.
The box was weighed, and its mass was noted as the initial mass. The box was then stored in a climatic

132
chamber at 23 °C and 50% humidity for 6 h. Every hour, the box was removed from the climatic chamber
to measure its mass with a precision balance, then returned to the climatic chamber until the next
measurement. The WVTR was then calculated using Equation 14:
𝐺
WVTR = 𝑡 ∗ 𝐴 (14)

Where, G is the weight change (g), t = time of exposition (h), A = film test area (m²).

WVTR is calculated for 24 h and expressed in g/m²*day.

Mechanical test

The mechanical properties of TOCNF films, particularly stress at break and strain at break at break,
were determined using the same protocol for MFC(B) films.

Thermogravimetric analysis (TGA)

Thermal degradation of TOCNF films was analyzed following the same parameters as MFC(A) films
except for TOCNF-5-P, TOCNF-5-P-DK10, and TOCNF-5-P-HK10, which were analyzed using the
same parameters as for MFC(B) films.

V.3. Results and discussion

V.3.1. MFC(A) film properties
The first film study was based on hot-pressed MFC using the P1 manufacturing process. The
addition of keratin was then studied, ranging from 5% to 20%.

V.3.1.1. Visual aspect

MFC(A) films were the first to be produced with MFC 2% (Figure 66). They served as the basis for
setting up an efficient film-pressing protocol. In this study, five different formulations were investigated.
The pure cellulose film, MFC(A), is the reference film for comparison with other films made from
cellulose and added dry keratin. The different percentages of added keratin are 5, 10, 15 and 20%. In
Figure 66, a picture of these different films shows their visual appearance after pressing. The first
observation leads to the conclusion that the appearance of the films is non-homogeneous in the presence
of keratin despite the homogenization step. The added keratin is in the form of aggregates dispersed
over the entire film surface. The more keratin is added, the more heterogeneous the surface. This non-
homogeneity can cause mechanical resistance problems by creating points of brittleness where the
keratin grains are located. Film color is also affected by the addition of keratin. Indeed, the keratin added
to the films comes from extraction with NMMO, and we have seen that this keratin has a dark brown
color, giving the films a yellowish coloration that increases with the percentage of keratin added.

Moreover, from the picture, it can be concluded that the pressing parameters are not optimal, as a white
ring around the films may reflect poor pressure or temperature selection. This white coloration should
be avoided in future film production.

133
 A B C

D E

Figure 66. Picture of MFC 2% (left) and MFC(A) films (right), a) MFC(A), b) MFC(A)-K5, c) MFC(A)-K10, d)
MFC(A)-K15, and e) MFC(A)-K20.

V.3.1.2. MFC(A) film thickness, density, grammage, and porosity

Film thickness, density, grammage, and porosity were measured and are reported in Table 12. The film
thickness increased with the addition of keratin compared to the film without keratin. This increase in
thickness can be explained by the presence of the keratin particles previously shown in the film pictures,
which explains their non-homogeneity. Despite the hot pressing, these particles result in an additional
thickness of approximately 30 µm, i.e., 44%. Density is an important indicator of film structure because
it reflects the degree of fiber packing. The more compacted and tightly packed the fibers are, the higher
the density. In this case, it can be observed that the maximum density is obtained with pure cellulose
film, and then a decrease in density is observed after the addition of keratin. The keratin creates free
spaces between the cellulose fibers, preventing them from compacting. The decrease in density is about
0.27 g/cm3, i.e., 24%. Density is directly related to film porosity. The higher the density, the lower the
porosity. Here, the values follow this trend, with a low porosity of 19.9% for the densest film (MFC(A))
and a higher porosity value of 48.2% for the film with the lowest density (MFC(A)-K15). Grammage
values remain similar at around 80 g/cm². However, this value was only obtained on one sample of each
formulation, and it would be preferable to carry out further measurements to obtain a standard deviation
and confirm these values.

Table 12. Thickness, density, grammage, and porosity of MFC(A) films.

Film Thickness (µm) Density (g/cm3) Grammage (g/m²) Porosity (%)

MFC(A) 68 ± 7 1.12 ± 0.12 80.85 19.9 ± 7.8
MFC(A)-K5 96 ± 8 0.86 ± 0.07 82.81 43.3 ± 4.7
MFC(A)-K10 91 ± 8 0.90 ± 0.08 81.93 39.6 ± 5.4
MFC(A)-K15 107 ± 13 0.76 ± 0.09 81.38 48.2 ± 6.3
MFC(A)-K20 97 ± 6 0.86 ± 0.05 82.94 42.7 ± 3.6

134
 V.3.1.3. Water contact angle (WCA)
The water contact angle determines the hydrophobic or hydrophilic nature of a material surface. A drop
of liquid is placed on a surface. If the surface is hydrophilic, the drop will tend to spread out until it
reaches equilibrium. On the other hand, if the surface is hydrophobic, the drop will not spread. The
ability of a surface to be wetted then depends on the value of the contact angle between the liquid and
its surface. It is generally known that if the contact angle is greater than 90 ° (θ>90 °), the surface is
hydrophobic; if the contact angle is less than 90 ° (θ<90 °), the surface is hydrophilic. The contact angle
principle is illustrated in Figure 67 with the results of the CA analysis. The hydrophilic character of
cellulose is well confirmed in this case. The MFC(A) film exhibits a contact angle of 85.5 °, which is
very close to the limiting value of 90 °, but remains below it. However, this value is well above that of
studies in the literature [123], [168], [169], which report an average CA of around 40 ° for MFC films,
whether manufactured by solvent casting or pressing. Contrary to initial hypotheses, the addition of
keratin results in a significant reduction of the contact angle. With only 5% keratin, the CA decreases
by 21% to 67.3 °. The contact angle decreases as the keratin content increases, reaching 57.1 ° for 20%
keratin. As mentioned earlier, keratin, like cellulose, contains numerous hydrogen bonds in its chemical
structure, so it may be possible for keratin's hydrogen bonds to create additional interactions with water,
resulting in a more hydrophilic film surface.

Figure 67. Schematic diagram of the contact angle principle (left), contact angle values of MFC(A) films.

V.3.1.4. Mechanical properties

The MFC(A) films were subjected to tensile tests on a universal machine to evaluate their mechanical
properties. The stress-strain curves and values for stress at break and strain at break are shown in
Figure 68. For MFC(A) pure cellulose film prepared by hot pressing, the mechanical properties are
130 MPa stress at break and 17.8% strain. These values agree with other studies reported in the
literature, particularly that of Nguyen et al., who produced MFC films by the pressing method with an
identical stress at break of 130 MPa and a lower strain at break of 6.3% for a film thickness of 90 µm
and a density of 1.32 g/cm3 [123]. In the case of MFC(A) films, only one manufacturing method was

135
Figure 68. The stress-strain curves of MFC(A) films (left) and diagram of stress at break (MPa), strain at break (%)
and Young’s modulus (MPa) values for MFC(A) films (right).

studied: hot pressing, but if the results of this study are compared with literature values for MFC films
produced by solvent casting, it can be found that the values obtained by pressing are higher than those
obtained by solvent casting, with, for example, a stress at break between 81-105 MPa and a strain at
break between 3.86-4.46% in the work of Kumar et al. [124]. A comparative study between pressing
and solvent casting methods, showing that hot pressing provides superior mechanical properties, is also
available in the literature [123].

Figure 68 shows that the film with the highest mechanical properties is the MFC(A) base film and that
the more keratin is added, the lower the mechanical properties. Indeed, adding 5% keratin drops the
values to 65 MPa for stress at break and 10.4% for strain at break. A decrease is also observed for the
other quantities added. This finding is also presented in the literature with the work of Zhang et al. [170],
who compared the mechanical properties of pure cellulose films with cellulose/keratin films. In their
case, both parameters, stress at break and strain at break, were reduced with the presence of keratin. The
presence of keratin can reduce the interaction between cellulose molecules and the material's cohesion,
thus reducing the mechanical strength.

V.3.1.5. Thermal properties

The thermal degradation of MFC(A) films was analyzed by thermogravimetric analysis. The TGA and
DTG curves are shown in Figure 69. All the curves representing the different percentages of keratin
added show a similar behavior, with thermal decomposition in two phases. The first phase consists of a
slight mass loss of about 5% before reaching 100 °C, due to the elimination of water. A second
degradation phase occurs from about 280 °C to 400-450 °C. This is the main degradation of cellulose,
which reaches a degradation rate of up to 1.33%/°C at 350 °C. These results are consistent with other
studies on the thermal resistance of cellulose in the literature [171]. A slight difference is observed when
keratin is added to the film. In fact, for a formulation containing 15% and 20% keratin, the degradation

136
rate is reduced to 1.28%/°C with a maximum at 355°C and 1.25%/°C at 356 °C. In this case, the addition
of keratin decelerates the thermal degradation of cellulose films and increases the maximum degradation

Figure 69. TGA and DTG curves of MFC(A) films.

temperature by about 5 °C. For all the formulations shown here, the percentage of solid mass remaining
is similar, ranging from 19-20% at 800 °C.

V.3.1.6. Chemical properties

The chemical structure of MFC(A) films was analyzed by FT-IR analysis, and the spectra obtained are
shown in Figure 70. All the characteristic bands of cellulose are present in these spectra. A broad
absorption band at 3310 cm-1 is assigned to the vibrational stretching of hydroxyl groups and includes
both intra- and inter-molecular hydrogen bonds within the MFC chains [123]. The stretching of the C-H
bond of the CH2 is visible at 2880 cm-1, then the band at 1640 cm-1 is associated with the bending of the

Figure 70. FT-IR spectra of MFC(A) films.

137
-OH groups of the residual water present in the cellulose, and finally, the band at 1200 cm-1 corresponds
to the stretching vibration of the C-O-C bond at 1200 cm-1. These results agree with data reported in the
literature [123], [171]–[173].

V.3.2. MFC(B) film properties

This second film production aims to improve the pressing process of the first films presented
above. For this purpose, MFC(B) films were produced using the P2 manufacturing process with
an additional cold pressing step. Different keratin contents were tested, ranging from 5% to
20%.

V.3.2.1. Visual aspect

In this second formation of microfibrillated cellulose films, called MFC(B) films, the issue of white
coloration on the contours of previous films was resolved. Cold pressing to remove the excess water
remaining after filtration and higher hot-pressing enabled the film to dry more effectively, resulting in
a smoother surface with a more homogeneous color and appearance (Figure 71). Different quantities of
keratin powder were then incorporated: 5, 10, 15 and 20%. However, the problem of non-homogeneity
with the addition of keratin remains unchanged; surface particles are still present, and the observations
are identical to those for MFC(A) films.

A B C D E F

Figure 71. Pictures of MFC(B) films. A) MFC(B) thick, B) MFC(B) thin, C) MFC(B)-K5, D) MFC(B)-K10, E)
MFC(B)-K15, F) MFC(B)-K20.

V.3.2.2. MFC(B) film thickness, density, grammage, and porosity

The thickness, density, grammage, and porosity values of the MFC(B) films are shown in Table 13. The
thicknesses of the MFC(B) films are lower than those of the MFC(A) films because the amount of
cellulose used was halved (except for the MFC(B) thick films) to obtain thinner films. The goal was
achieved with a thickness that was also halved. In addition, the pressing parameters were changed, and
a higher pressure was applied. For comparison, the MFC(B) thick film was prepared with the same
amount of cellulose as the MFC(A) film. It can be seen that the latter is much thinner, so the pressing
method influences the film thickness. It can also be seen that the thickness increases with the addition
of keratin. The density of the MFC(B) pure cellulose film is higher than the MFC(A) film, 1.40 g/cm3
compared to 1.12 g/cm3, further confirming that the pressure exerted during film formation influences
its density and thus the packing of the cellulose fibers. The density of the MFC(B) film is close to the

138
density of pure cellulose, assumed to be 1.5 g/cm3, reflecting the tight packing of the fibers in the film.
The porosity of the films is directly related to their density, as a result of which a highly low porosity is
obtained for the MFC(B) thick film at 6.3%. Overall, porosity is higher when adding keratin, as
previously stated. There is a difference between the average grammage of MFC(B) and MFC(A) films,
with values between 30-40 g/m² and 80-83 g/m², respectively, due to the reduced mass of the films for
the same diameter.

Table 13. Thickness, density, grammage, porosity, and CI of MFC(B) films.

Thickness Density Grammage

Film Porosity (%) CI (%)
(µm) (g/cm3) (g/m²)
MFC(B) thin 29 ± 3 1.08 ± 0.04 31.33 ± 1.15 27.9 ± 2.6 64
MFC(B) thick 45 ± 2 1.40 ± 0.04 62.67 ± 1.76 6.3 ± 2.9 69
MFC(B)-K5 46 ± 4 0.84 ± 0.03 37.67 ± 1.26 43.8 ± 1.9 56
MFC(B)-K10 36 ± 6 0.97 ± 0.16 33.83 ± 1.15 34.4 ± 10.9 54
MFC(B)-K15 32 ± 5 1.19 ± 0.10 38.33 ± 3.33 18.5 ± 7.1 57
MFC(B)-K20 36 ± 4 0.86 ± 0.09 31.00 ± 3.28 41.0 ± 6.2 53

V.3.2.3. Morphological properties

SEM images of MFC and keratin-based films are shown in Figure 72. The MFC(B) film is structured
as a layer of entangled fibers that form a flat but not smooth surface. This morphology is also found in

A B C

D E

Figure 72. SEM images of MFC(B) films: A) is MFC(B) thin, B) is MFC(B)-K5, C) is MFC(B)-K10, D) is
MFC(B)-K15, E) is MFC(B)-K20.

139
other film formulations. However, it has been shown that keratin creates a non-homogeneous film
surface due to the aggregation of its molecules. Here, these aggregates are visible as clusters that
roughen the surface. These clusters are crushed due to the pressing protocol to which these films were
subjected. Thus, these SEM images confirm the heterogeneity of the surface of MFC(B) films with
keratin.

V.3.2.4. Crystalline properties

The crystallinity of the microfibrillated cellulose films was evaluated by XRD analysis. The
diffractograms are shown in Figure 73. The crystallinity index of each formulation was calculated, and
the values are shown in Table 13. The general appearance of the diffractograms shows a semi-crystalline
structure for MFC with two diffraction peaks located at 2θ=12-18 ° and 2θ = 22.5 °. The first diffraction
peak is attributed to two lattice planes (1-10) and (110), and the second most intense diffraction peak is
attributed to (200) lattice diffraction, which are characteristic cellulose peaks. These diffraction peaks
correspond to regions where cellulose fibers are ordered and form the crystalline phase, along with
randomly arranged fibers that form the amorphous portion. The CI values of MFC(B) films range from
53 to 69%, confirming other studies on the crystallinity rate of cellulose films obtained by hot pressing.
In this study, the CI is highest for the MFC(B) thick film at 69% and for the MFC(B) thin film at 64%,
while after the addition of keratin, the CI decreases to 53% for the MFC(B)-K20 film. These results
were theoretically expected since adding keratin would prevent the cellulose fibers from arranging
themselves and reduce the orderly structuring of the fibers. A decrease in crystallinity can be a positive
aspect as it reduces the brittleness of the film by reducing its stiffness and increasing its ductility.

Figure 73. XRD diffractograms of MFC(B) films.

V.3.2.5. Water contact angle (WCA)

The contact angle values are shown in Figure 74. The CA values for the two pure MFC(B) films are
highly similar at 69.7 ° and 69.4 °. These are lower than for the MFC(A) films but were not achieved
with the same apparatus or under the same conditions. On the other hand, the decreasing CA tendency

140
with the addition of keratin is similar to that of MFC(A) films, confirming the established hypothesis.
With the addition of 20% keratin, the CA drops to 54.3 °, close to the contact angle of the MFC(A) film
with 20% keratin. The decrease is smaller in this case than with MFC(A) films.

Figure 74. Contact angle of MFC(B) films.

V.3.2.6. Mechanical properties

The mechanical properties of MFC(B) films produced by hot-pressing are shown in Figure 75. It can
be seen that the film thickness affects the tensile strength, with a higher stress-at-break value for the
thick MFC(B) film than for the thin MFC(B) film, at 75 MPa versus 64 MPa. MFC(B) films are, on
average, half the thickness of previous MFC(A) films, which explains the reduction in mechanical
properties. However, as the film thickness increases, its elongation tends to decrease as its stiffness

Figure 75. Diagram of stress at break (MPa) and strain at break (%) for MFC(B) films.

141
increases. Since film elongation is a critical property in the packaging sector, it was decided to continue
the study with "thin" films to compare the addition of keratin. As previously observed with MFC(A)
films, adding keratin reduces the stress-at-break values. However, for a 5% addition, the strain at break
is increased to an average of 2.1%, a 75% increase over the thin MFC(B) film.

V.3.2.7. Thermal properties

Thermogravimetric analysis of MFC(B) films is shown in Figure 76. The two degradation phases are
present, with the first phase reaching 150 °C for 6% moisture removal. This is followed by cellulose
degradation from 280 °C to 400 °C with a high mass loss of 68%. Finally, the degradation continues up
to 600 °C with a final mass loss of 78%. When keratin is added, the same interpretation can be made,
as for MFC(A) films. The DGT curve demonstrates that the more keratin is added, the slower the
degradation. In this case, it is worth adding keratin to formulations to improve their thermal properties.

Figure 76. TGA and DTG curves of MFC(B) films.

V.3.3. Chemical properties of TOCNF

The production of TOCNF from bleached pulp was carried out by TEMPO-mediated oxidation and after
passing through a high-pressure homogenizer. A transparent gel suspension solution was obtained with
a consistency of about 1%, as shown in Figure 77. Prior to film production, the chemical properties of
TOCNF were analyzed. CC, CD, and DP were evaluated to verify proper oxidation of the cellulose
fibers.

The carboxyl content of TOCNF prepared by the TEMPO-mediated oxidation method using
NaBr/NaClO/TEMPO at pH 10 was determined by the UV-visible absorption method at 664 nm. The
method involved using methylene blue to measure the amount of COOH groups in the cellulose
quantitatively. In this case, the TOCNFs were prepared using 5 mmol NaClO per gram of cellulose and

142
 Figure 77. Simplified scheme of TOCNF fabrication process.

are therefore called TOCNF-5. It has been shown in the literature that the amount of NaClO added to
the solution is responsible for the amount of cellulose oxidation since the higher the amount of NaClO,
the higher the CC. In the present case, after the analysis, the CC measured for the TOCNF-5 sample is
868 µeq•g/g. This confirms other values in the literature for the same amount of NaClO added. The
analysis of the cationic demand of TOCNF is based on measuring the proportion of negative charges
present in the cellulose.

The CD is closely related to the CC generated by TEMPO-mediated oxidation and the presence of COO-
groups. The CD value for TOCNF-5 was 1335 µeq·g/g, slightly higher than reported in other studies for
the same amount of NaClO [111], [174]. TEMPO-mediated oxidation is used to reduce the size of
cellulose fibers as they pass through the homogenizer more efficiently. COO- groups within the cellulose
cause repulsion between the fibers, resulting in easier separation and size reduction. Therefore,
theoretically, the more cellulose is oxidized, the shorter the fibers and the lower the DP. This theory was
validated and presented by Patiño-Masó et al. [174], who demonstrated that the DP decreases with the
amount of NaClO used. In the present case, the DP measured by the intrinsic viscosity method is 338.
However, this value is lower than those usually found in the literature [111], [174]. The hypothesis
adopted is based on the rapid introduction of NaClO into the cellulose solution and the consequent
increase in pH to around 12, which could have led to fiber degradation. However, this DP value remains
between those obtained for TOCNF-5 and TOCNF-6.

This chemical characterization of the TOCNF-5 obtained in this study shows that the fibers obtained are
well oxidized and that the TEMPO-mediated oxidation was successful. Therefore, the fibers can be used
to form nanocellulose films.

V.3.4. TONCF film properties

Following the preparation of films based on microfibrillated cellulose, another type of cellulose was
used to form films to compare their different properties. In this section, TOCNFs with different degrees
of oxidation were used. A first commercial TOCNF-15 was compared with a TOCNF-5 prepared in the
laboratory during the international mobility at the LEPAMPA-PRODIS research group. Different
film-forming parameters and the method used were studied: solvent casting or hot pressing. It was stated
in Chapter I - bibliography that these two methods provide cellulose films with different properties, and

143
these assertions were studied and confirmed in this work. The use of dry or humid keratin was studied
to overcome the problem of film homogeneity. Different percentages of keratin: 0, 5, 10, and 20% were
incorporated in the formulations. Finally, the incorporation of a plasticizer and its amount were
evaluated to improve the mechanical properties of TOCNF films.

V.3.4.1. Scanning electron microscopy (SEM)

The first analysis used to study TOCNF films is scanning electron spectroscopy to examine the film
surface and verify the presence of nanofibers. These films are made from TEMPO-mediated oxidized
cellulose, which produces nanoscale fibers after treatment in a high-pressure homogenizer. According
to the Handbook of Nanocellulose and Cellulose Nanocomposites, CNFs are generally considered to be
between 20 and 50 nm in width and between 500 and 2000 nm in length [175]. Figure 78 shows two
SEM images of the TOCNF-5-P film at magnifications of x1000 and x10,000. In these images, a cross-
section of the film is presented, showing both the interior and the surface of the film. The structure is
composed of several overlapping layers of fibers. The fibers are visible in the nanometer range in the
enlarged image, confirming that the CNFs have been produced by TEMPO-mediated oxidation.

Figure 78. SEM images of TOCNF-5-P film, magnification x1000 (left) and magnification x10000 (right).

V.3.4.2. Optimum manufacturing process selection

Visual aspect

It was mentioned in Chapter I that pressing and solvent-casting methods produce films with different
properties. Here, the visual appearance of films made by solvent casting and hot pressing of TOCNF-5
and TOCNF-15 was evaluated using the picture in Figure 79. It is important to note that the hot-pressing
method could not be used with TOCNF-15 because the reduced size of the cellulose molecules greatly
increased membrane filtration time. This increased filtration time made the manufacturing protocol
time-consuming and difficult to implement.

In terms of visual appearance, there are several differences. TOCNF-5-P film is similar in appearance
to cellophane, with a whitish color. The surface is wrinkled and cloudy, but the film is transparent on
contact. After solvent casting, the films have a smooth, flat surface with a high brilliance. These films

144
have a more "plasticized" appearance than those obtained by pressing. The picture shows that they are
colorless and highly transparent. Visual appearance is an important criterion in the packaging sector,
and depending on the application, the visual appearance can be tuned.

Figure 79. Picture of TOCNF-5-P film (left), TOCNF-5-S film (center), TOCNF-15-S film (right).

TOCNF film thickness, density, grammage, and porosity

The thickness, density, grammage, and porosity of TOCNF films are shown in Table 14. According to
the results; the solvent-casting method generates higher density films than the pressing method. Film
formation via solvent casting is a slow method of solvent evaporation, and this slow formation may
allow the nanocellulose fibers to arrange themselves in a more orderly and compact manner within the
film, compared with rapid drying via hot pressing. TOCNF-15 was found to produce films with higher
density, heavier grammage, and lower porosity than TOCNF-5, using an identical protocol. TOCNF-15
is reported to have a lower degree of polymerization than TOCNF-5 [111]. In this case, shorter polymer
chains enable better spatial arrangement, resulting in denser film structuring and lower porosity. Despite
its simplicity and low energy consumption, the solvent casting method takes a long time (7 to 10 days
at RT), compared with just a few minutes for hot pressing.

Barrier properties

The barrier properties of a material are critical to its use in food packaging. The purpose of packaging
is to protect and preserve food from its environment. The packaging acts as a barrier to moisture and
oxygen in the atmosphere, thereby extending the shelf life of the product. Poor barrier properties can
lead to accelerated product degradation, making it unsuitable for use in food packaging. In a study, Silva
et al. listed several food products with WVTR limits suitable for preservation [120]. The products listed
range from coffee and milk to cheese, fruit, and vegetables.

In most cases, the required WVTR does not exceed 100 g/m²·day at a test temperature of 23 °C. The
only exception is fruit and vegetables, with an accepted WVTR of 10-4000 g/m²·day. A list of
conventional plastics commonly used is provided with each product. These plastics have very low
WVTR values, making them excellent moisture barrier materials. The WVTR values of some
conventional plastics can be found in the publication by Ortega-Toro et al., which include HDPE, LDPE,
145
PP, PET, EVOH, PS, PLA, among others [176]. Their WVTR values do not exceed 50 g/m²·day and
can even be below 10 g/m²·day for some plastics.

Table 14. Thickness, density, grammage, porosity, and CI of TOCNF films.

Film Thickness (µm) Density (g/cm3) Grammage (g/m²) Porosity (%) CI (%)

TOCNF-5-P 36 ± 5 1.13 ± 0.19 45.01 ± 0.28 14.1 ± 8.8 62

TOCNF-5-P-DK10 47 ± 10 0.84 ± 0.07 45.53 ± 1.63 43.5 ± 3.5 64

TOCNF-5-P-HK5 35 ± 7 1.36 ± 0.12 33.40 ± 0.39 11.4 ± 6.8 -

TOCNF-5-P-HK10 46 ± 4 1.10 ± 0.13 45.43 ± 2.31 20.7 ± 4.4 61

TOCNF-5-P-HK20 35 ± 9 0.99 ± 0.24 33.58 ± 1.41 25.5 ± 13.1 -

TOCNF-5-P-10PEG 30 ± 7 1.25 ± 0.26 36.4 ± 3.07 22.0 ± 17.4 -

TOCNF-5-P-20PEG 25 ± 5 1.22 ± 0.04 30.2 ± 2.22 14.6 ± 3.1 -

TOCNF-5-P-40PEG 18 ± 1 1.46 ± 0.45 27.25 ± 5.66 11.2 ± 6.1 -

TOCNF-5-P-20sorbitol 31 ± 4 1.18 ± 0.09 35.98 ± 4.86 21.5 ± 5.8 -

TOCNF-5-P-HK5-20PEG 36 ± 5 0.75 ± 0.07 25.34 ± 4.29 47.3 ± 4.9 -

TOCNF-5-S 28.9 ± 1 1.33 ± 0.04 38.45 ± 2.08 11.4 ± 2.5 -

TOCNF-15-S 36 ± 3 1.47 ± 0.07 52.48 ± 3.64 7.4 ± 4.9 -

In the present study, TOCNF-5-P, TOCNF-5-S, and TOCNF-15-S films were exposed to an
environment of 23 °C and 50% humidity to measure their WVTR using the cup method. The
measurements were reproduced in triplicate and presented in Figure 80. The WVTR values for
TOCNF-5-P, TOCNF-5-S, and TOCNF-15-S films are 343, 288, and 227 g/m²·day respectively. These
values can be correlated with porosity values, which also decrease in the same order, indicating that
water molecules can more easily pass through the film as porosity increases, increasing the WVTR. In
their current state, these films do not have sufficient barrier properties for use in food packaging or
possibly in certain cases for fruits and vegetables. However, the values obtained are similar to, or even
lower than, some films made from CNF in the literature with values ranging from 686 to 1210 g/m²·day
[177].

146
 Figure 80. WVTR (g/m²*day) of TOCNF-5-P, TOCNF-5-S and TOCNF-15-S films.

Mechanical properties

The mechanical properties of TOCNF films produced by the hot-pressing and solvent-casting processes
are shown in Figure 81. It was previously observed that, in general, the hot-pressing process produces
greater mechanical strength than the solvent-casting process for cellulose films, but this finding was
verified in our study. Two types of cellulose, TOCNF-5 and TOCNF-15, and the manufacturing process
were compared. The results clearly show that the stress-at-break value of TOCNF-5-P films is higher
than that of TOCNF-5-S and TOCNF-15-S. However, the highest elongation is maintained by the
TOCNF-15-S film.

Figure 81. Diagram of stress at break (MPa) and strain at break (%) of TOCNF-5-P, TOCNF-5-S and
TOCNF-15-S films.

Overall, the mechanical strength properties of the TOCNF films are lower than those of the MFC films
previously investigated. The TOCNF films presented in this study have values well below the stress at
147
break and strain at break values found in the literature for CNF and TOCNF films [121], [124], [177].
This difference may be due to a pressing process that is not optimized for this type of cellulose and
would require temperature and/or pressing parameters adjustment to optimize the mechanical properties
of the films.

Thermal properties

Figure 82 shows the TGA and DTG curves for TOCNF-5 and TOCNF-15 films produced by the solvent
casting method, while Figure 87 shows the TGA and DTG curves for TOCNF-5-P film produced by the
hot-pressing method. It can be seen that the TGA curves for TOCNF films are different from the MFC
films. The first degradation before 100 °C is similar to MFC films, with a mass loss of about 7%. The
moisture content of TOCNF cellulose is therefore higher than that of MFC cellulose. Differences can
be seen in the second degradation phase which starts at 220 °C and progresses to 350 °C with 61% mass
loss, and then, to 800 °C with 18% solids remaining in the case of solvent casting. In addition, the DGT
curve for these two films shows two degradation peaks with maxima at 247 °C and 305 °C. This
difference is due to the presence of carboxylic functional groups at the C6 position generated by
TEMPO-mediated oxidation. This oxidation destabilizes the molecules, resulting in degradation at a
lower temperature than that of unoxidized cellulose. In addition, the decarbonation of these functional
groups occurs prior to the degradation of the cellulose chains [178]. As a result, the peak at 247 °C
reaches a degradation rate of 0.5%/°C for TOCNF-15, the most oxidized cellulose, versus 0.43%/°C for
the least oxidized cellulose. In the case of pressing with TOCNF-5-P film in Figure 87, degradation
occurs from about 240 °C to 350 °C with a loss of 60% and a remaining solid mass of 25% at 800 °C.
A shoulder is visible on the DTG curve at 278 °C, presumably due to the presence of carboxyl groups.

Figure 82. TGA and DTG curves of TOCNF films by solvent casting method.

148
In conclusion, films made with TOCNF have a lower thermal resistance than MFC films. Moreover, if
we compare the thermal properties of the solvent casting and pressing methods, the pressing method
confers the best thermal resistance properties, with a degradation onset temperature 20 °C higher for the
same oxidized cellulose.

V.3.4.3. Optimization of film homogenization: dry keratin versus humid keratin.

To overcome the above-mentioned homogenization problems, keratin in humid form was tested prior to
the lyophilization step in order to incorporate a humid phase instead of a dry phase and to improve the
dispersion of the keratin within the film. To this end, 10% keratin was added to a film of TOCNF-5-P.

Visual aspect

Picture of the three formulations with/without keratin and dry/humid are shown in Figure 83. The visual
appearance of these three films differs depending on the type of keratin used. The pure cellulose film,
(TOCNF-5-P), has a cellophane-like appearance, is whitish in color and has a wrinkled surface. When
10% dry keratin is added, the film becomes yellowish in color and appears brittle to the touch. However,
the homogeneity of the film was greatly improved by Ultra-Turrax homogenization using a different
model of apparatus, which allowed better dispersion of the keratin particles. A few keratin particles
remain visible on the surface. TOCNF-5-P-HK10 exhibits the most favorable appearance, with a
significantly smoother and more homogeneous surface. There is also a pronounced yellowish tint due
to the presence of keratin. Here, the addition of humid keratin helps to homogenize the film and improve
its visual appearance after pressing.

Figure 83. Picture of TOCNF films: TOCNF-5-P (left), TOCNF-5-P-DK10 (center), and TOCNF-5-P-
HK10 (right).

TOCNF film thickness, density, grammage and porosity

Comparing the values for density, grammage, thickness and porosity in Table 14, we can see that density
and porosity are affected by the type of keratin used. The addition of dry keratin results in a decrease in
film density to 0.81 g/cm3 and a sharp increase in porosity to 43.5%. When the same percentage of
humid keratin is added, the density is virtually maintained at 1.10 g/cm3 and the porosity is slightly
increased compared to the film without keratin but remains much lower at 20.7% than when dry keratin
is used. In addition, a literature study demonstrated that mechanical properties are closely correlated

149
with film density: the higher the density, the higher the mechanical strength [179]. Accordingly, humid
keratin is preferable.

Crystalline properties (XRD)

Diffractograms of TOCNF-5-P films are shown in Figure 84 to compare their crystallinity according to
the type of keratin used. The CIs have been calculated and are shown in Table 14. The CI values are
62%, 64% and 61% for TOCNF-5-P, TOCNF-5-P-DK10 and TOCNF-5-P-HK10 films, respectively.
Here, the addition of keratin does not affect the CI, so the type of keratin, dry or humid, does not
influence the crystallinity of the TOCNF films.

Figure 84. XRD diffractograms of TOCNF-5-P films depending on the type of keratin incorporated.

Barrier properties

The contact angle of TOCNF-5-P, TOCNF-5-P-DK10 and TOCNF-5-P-HK10 films was measured in
triplicate on each film. However, the results are not shown as they are not interpretable. The contact
angle values increased with analysis time, which is contrary to logic. In fact, over time, the water droplet
crushes on the surface, causing the contact angle to decrease. In the case of TOCNF-5 films, which are
highly hydrophilic, they absorbed the water droplet deposited on their surface almost immediately and
the film swelled. Once the film was swollen, the baseline on the image was altered and the values
recorded by the camera were distorted. However, it is important to point this out to explain the
hydrophilicity of TOCNF films.

The WVTR values for TOCNF-5 films are shown in Figure 85. The values are 343, 358 and
408 g/m²*day for TOCNF-5-P, TOCNF-5-P-DK10 and TOCNF-5-P-HK10 films, respectively. The
results show that the addition of keratin tends to increase the WVTR of the films and even more so with
moist keratin. This was also shown by the contact angle of MFC(A) and MFC(B) films decreasing with
the addition of keratin. Thus, the addition of keratin increases the water sensitivity of the cellulose films,

150
contrary to the intended objective. However, in this case the standard deviations are large and the
difference between the values may be insignificant.

Figure 85. WVTR (g/m²*day) of TOCNF-5-P films depending on the type of keratin incorporated.

Mechanical properties

The mechanical properties of TOCNF-5-P films with 10% dry or humid keratin incorporation are shown
in Figure 86. In line with the results presented previously, the stress at break of TOCNF-5-DK10 film
is lower than that of TOCNF-5-P film without keratin, with a value of 28 MPa versus 45 MPa. However,
this gives a slightly higher elongation with an average value of 0.44% versus 0.31% without keratin. A
significant difference is observed with the addition of humid keratin, which, in contrast to dry keratin,
increases the tensile strength of the film by up to 53 MPa while maintaining an elongation of about
0.35%. Therefore, the addition of humid keratin is preferred to improve the mechanical properties of
TOCNF films.

Figure 86. Diagram of stress at break (MPa) and strain at break (%) of TOCNF-5-P films depending on the
type of keratin incorporated.

151
 Thermal properties

Figure 87 shows the TGA and DTG curves of TOCNF-5-P films with fresh or dry keratin addition. The
interpretation of the curves is identical to that previously made for the TOCNF-5-P film. The
degradation temperatures are similar in the 3 cases studied here, the only difference being seen in the
DTG which shows that the addition of 10% fresh keratin slightly reduces the degradation rate, with a
decrease in the minimum intensity of the DTG curve at 330 °C. It is therefore preferable to use humid
keratin.

Figure 87. TGA and DTG curves of TEMPO-5-P films depending on the type of keratin incorporated.

V.3.4.4. Influence of keratin percentage

Visual aspect

The percentage of keratin added affects the appearance and color of TOCNF-5-P films. Figure 88 shows
picture of different films with humid keratin added ranging from 5 to 20%. The formulation with 5%
keratin provides the most favorable appearance with a smooth surface. This surface is considerably
smoother than when 10% or 20% keratin is added, or without keratin. On the other hand, when 20%

A B C D

Figure 88. Picture of TOCNF-5-P films with humid keratin incorporation: A) 0%, B) 5%, C) 10%, D) 20%.

152
keratin is added, the film turns dark beige and its surface is very wrinkled. Excessive keratin amount
causes faster drying and wrinkling of the film. In this case, the addition of 5% is preferable.

TOCNF film thickness, density, grammage and porosity

The influence of different humid keratin percentages was determined as a function of film thickness,
density, grammage and porosity. The results are shown in Table 14. As previously stated, in general,
the addition of keratin decreases the density and increases the porosity of the films. However, in the
case of the 5% keratin formulation, it turns out that this addition is beneficial. Density is increased from
1.13 to 1.36 g/cm3 and porosity is slightly reduced from 14.1 to 11.4%.

Barrier properties

The WVTR values of films with different percentages of keratin were evaluated and are shown in
Figure 89. As in previous analyses, the addition of keratin increases the WVTR of TONF-5-P films, but
no further trends can be deduced from the results. Values ranged from 408 to 500 g/m²*day in the
presence of keratin.

Figure 89. WVTR (g/m²*day) of TOCNF-5-P films with humid keratin incorporation.

Mechanical properties

Once the addition of humid keratin was shown to improve the mechanical properties of TOCNF films,
different percentages of humid keratin were incorporated into the film formulation, and their mechanical
properties were compared. The results presented in Figure 90 confirm that adding humid keratin
increases the stress and strain at break of TOCNF films with the 5% keratin formulation. Indeed, the
highest values are obtained for the TOCNF-5-P-HK5 film, with a stress at break of 63 MPa and a strain

153
at break of 0.84%. This represents an increase of about 40% and 170% in tensile strength and elongation,
respectively, compared to the TOCNF-5-P film without keratin.

Figure 90. Diagram of stress at break (MPa) and strain at break (%) values of TOCNF-5-P films with humid
keratin incorporation.

Thermal properties

The TGA and DTG curves for TOCNF-5-P and TOCNF-5-P-HK10 films are shown in Figure 87, and
those for TOCNF-5-HK5 and TOCNF-5-P-HK20 films are shown in Figure 91. Each curve shows a
moisture content of about 6.5% at 100 °C, followed by degradation starting at 230 °C and reaching about
63-65% loss at 400 °C. As previously stated, the shoulder at 260-270 °C is attributed to the

Figure 91. TGA and DTG curves of TOCNF-5-P films with incorporation of humid keratin.

154
decarbonisation of the carboxyl functional groups. In this case, the results show that incorporating
keratin modifies the thermal properties developed with TOCNF.

V.3.4.5. Plasticizer selection and quantity influence

Visual aspect

TOCNF-5-P films with added plasticizer have the same overall appearance as pure TOCNF-5-P film.
Figure 92 shows that they are slightly whitish and wrinkled. Whether sorbitol or glycerol is used, the
appearance is similar.

A B C D

Figure 92. Pictures of TOCNF-5-P films, A) TOCNF-5-P-20sorbitol, B) TOCNF-5-P-10PEG, C)

TOCNF-5-P-20PEG, D) TOCNF-5-P-40PEG.

TOCNF film thickness, density, grammage, and porosity

The choice of plasticizer and its concentration was evaluated based on density, thickness, grammage,
and porosity, as shown in Table 14. First, two formulations containing 20% PEG 400 and 20% sorbitol
were compared. Here, sorbitol had only a minor effect on the properties mentioned above, except for an
increase in porosity compared to the film without sorbitol, 21.5% versus 14.1%. On the other hand, the
addition of 20% PEG 400 slightly increases the density of the film but hardly increases the porosity of
the cellulose film. The PEG 400 option is, therefore, preferable.

Two other proportions were then tested: 10% and 40%. There was a direct proportional relation between
the PEG addition and film thickness, grammage, and porosity. On the other hand, the density increases
with the addition of PEG. In this case, the higher the PEG 400 addition, the greater the improvement.

Barrier properties

First, the influence of the plasticizer on the barrier properties of the films was analyzed by measuring
the WVTR of films containing 20% PEG 400 and 20% sorbitol. The measurements were performed on
only one specimen of each film due to time constraints and are shown in Figure 93. Since PEG is known
to be hydrophilic, it is logical to see that the incorporation of PEG 400 strongly increases the WVTR,
with a 94% increase for 20% PEG addition compared to a 26% increase for 20% sorbitol.

Regarding the amount of PEG incorporated, a trend towards higher WVTR was observed with
increasing PEG amount. The values are 531, 665, and 664 g/m²·day for 10%, 20%, and 40%, with a

155
 Figure 93. WVTR (g/m²*day) of TOCNF-5-P films with addition of plasticizer, PEG400 or sorbitol.

value stagnation after 20%. However, it is necessary to repeat the experiment to obtain the standard
deviation and confirm that the results are significant.

Mechanical properties

The influence of the plasticizer incorporated in the formulation of TOCNF films and its quantity on their
mechanical properties was evaluated. The first step was to compare 20% sorbitol with 20% PEG 400.
Figure 94 shows that adding sorbitol increases stress at break and strain at break by about 24% and 81%,
respectively. On the other hand, adding 20% PEG 400 kept the stress-at-break value almost unchanged
but significantly increased the strain-at-break by 194%, reaching a value of 0.91% elongation. For this

Figure 94. Diagram of stress at break (MPa) and strain at break (%) of TOCNF-5-P with addition of plasticizer,
PEG 400 or sorbitol.

156
reason, PEG 400 was selected, and different concentrations were studied with 10 and 40% addition.
Regarding elongation, 20% PEG is still the optimal ratio compared to 10% or 40%.

Thermal properties

The influence of plasticizer incorporation and its amount was analyzed by thermogravimetric analysis,
as shown in Figure 95. The behavior of the curves follows the behavior of the TGA curves of TOCNF
films studied earlier and their interpretation has already been presented. Here, sorbitol is compared to
PEG 400. The degradation initiation temperature is identical for the same percentage of both plasticizers
added. The difference lies at the end of degradation: at 400 °C, the degradation percentage is 62.7% for
20% PEG 400 and 66% for 20% sorbitol; then, at 800 °C, it is 22.7% and 18.4%, respectively. The use
of PEG 400 is, therefore, preferable for this application. A difference is observed regarding the amount
of PEG added with the curve representing 40% PEG, a higher degradation rate and a lower residual
solid mass. The more PEG is added, the lower the thermal properties compared to a pure TOCNF film.
This result was also observed in the literature [180].

Figure 95. TGA and DTG curves of TOCNF-5-P films with addition of plasticizer, PEG 400 or sorbitol.

V.3.4.6. Formulation combining TOCNF - keratin – plasticizer

Visual aspect

A picture of the TOCNFR-5-P-HK5-20PEG film can be seen in Figure 96. The visual appearance of the
film is similar to that of a pure, homogeneous TOCNF-5-P film. It is very lightly colored due to the
presence of 5% humid keratin.

157
 TOCNF film thickness, density, grammage, and porosity

The values in Table 14 show that the combination of the three compounds results in a high-density loss
of 33% and a reduction in grammage to 25.34 g/m² compared to the TOCNF-5-P base film. Conversely,
porosity is highly increased, being 235% higher.

Figure 96. Picture of TOCNF-5-P-HK5-20PEG

Mechanical properties

The three-component TOCNF-keratin-PEG formulation was tested for mechanical properties and the
values obtained are a stress at break = 21.55 ± 9.43 MPa and a strain at break = 1.52 ± 0.34%. These
results are only half of the expected theoretical results. The incorporation of 5% humid keratin alone
had increased both the stress at break and the strain at break, while the addition of 20% PEG 400 alone
had shown preservation of the stress at break and a significant increase in the elongation of the base
film. Theoretically, the composition of the three components should maintain or improve tensile strength
and increase elongation. However, the stress-at-break value here is reduced by about 50% compared to
the TOCNF-5-P base film, but its elongation is increased by about 390%.

These results can be related to the difference in density of TOCNF-5-P and TOCNF-5-P-HK-20PEG
films, as well as to their difference in porosity, visible in Table 14. The values are 1.13 g/cm3 and
0.75 g/cm3 for density and 14% versus 47% for porosity, respectively. The density and porosity of a
material indicate its cohesion: the closer the cellulose fibers are packed together, the higher the density
and the lower the porosity, making the material stronger and conversely. Adding PEG and keratin
reduces the number of nanofibers in the film and can reduce cohesion and interaction between the fibers.

This latest formulation, combining TOCNF-Keratin-PEG, is difficult to use in packaging because it does
not have the minimum required mechanical properties. If its mechanical properties are compared with
those of conventional plastics listed in the literature by Silva et al. [13], the film studied in this work is
of low strength. If referred to Figure 14 (Chapter I), which shows the tensile strength and elongation at
break of various polymers such as PS, PLA, cellophane, PET, HDPE, PVA, and PP, among others, the
elongation values are at least 2% and tend to reach 1000%, for tensile strength values mostly between

158
10 MPa and 200 MPa. Therefore, the mechanical properties of the TOCNF films studied here must be
improved before they can be considered in the food packaging sector.

Thermal properties

The TGA and DTG curves of the TOCNF-5-P-HK5-20PEG formulation are shown in Figure 97 and
compared with the TGA of the TOCNF-5-P film shown in Figure 87. Here, the calculated moisture
content is slightly higher in the TOCNF-5-P-HK5-20PEG formulation at 6.6% versus 6.2% for the pure
TOCNF film. This is probably due to the presence of PEG, known to be hydrophilic. The onset of
degradation starts at the same temperature, but the highest degradation rate occurs at 335 °C for the
blend versus 323 °C for the pure TOCNF-5 film. In this case, the incorporation of 5% keratin and 20%
PEG 400 shifts the main degradation of the film by 12 °C. On the other hand, after degradation, at
400 °C, the mass loss is more important for the blended film, with 66.1%, compared to 61.8% for the
pure cellulose film.

Figure 97. TGA and DTG curve of TOCNF-5-P-HK5-20PEG.

159
 V.4. Conclusion
The main objective of this chapter was to produce a bio-sourced polymer film that could compensate
for the use of conventional synthetic polymers, which are used in large quantities in the food packaging
industry. Cellulose is the most abundant natural polymer on earth and its use in packaging is
well-established. However, the main drawback of cellulose is its hydrophilic nature, which makes it
highly sensitive to water and limits its applications. Therefore, The idea was to incorporate keratin,
previously extracted from duck feathers, to valorize it and study its influence on the properties of
cellulose films. Two types of cellulose were used: microfibrillated cellulose (commercial) and
nanocellulose oxidized by TEMPO mediation carried out in the laboratory. Different parameters were
studied, such as the manufacturing method (solvent casting or hot pressing), the type of keratin added
(dry or humid) and its quantity ranging from 5% to 20%, the use of a plasticizer (sorbitol or PEG 400)
and its quantity. Finally, a formulation combining the three components of cellulose-keratin-plasticizer
was prepared.

First, TOCNFs were prepared and their chemical characterization, including carboxyl content, cationic
demand, and degree of polymerization, confirmed the effective preparation of TOCNF-5. These
nanofibers were then used to prepare TOCNF films according to different protocols. The solvent casting
method produces smooth, transparent films with high density and better barrier properties with a lower
WVTR than films produced by hot pressing but with lower tensile strength. Films obtained by pressing
differ in their visual appearance, with reduced transparency and a more or less crumpled appearance
similar to cellophane. However, pressing increases the degradation temperature of the film by 20 °C.
Despite its ease of use, the solvent casting method is time-consuming and can take several days, whereas
the pressing method can reduce film production to a few minutes. Considering all these results and
considerations, the pressing method was chosen. Homogenization problems observed with MFC films
after pressing led to testing the incorporation of humid keratin compared to dry keratin. Humid keratin
was found to be the optimal choice to increase film homogeneity and to improve visual appearance, to
slightly increase mechanical properties, unlike dry keratin, which decreases them, and to reduce the
thermal degradation rate of the film.

On the other hand, the use of humid keratin increases the WVTR value compared to dry or keratin-free
films. Despite this and all its other advantages, humid keratin is the best option. The percentage of added
keratin was studied in the formulation of MFC(A) and MFC(B) films and influenced some film
properties. As the amount of keratin added increased, the density decreased, the porosity increased, the
crystallinity decreased, the contact angle decreased, reflecting increased sensitivity to water, the
degradation rate slowed, and the overall mechanical properties decreased. However, an exception was
found for adding 5% keratin, which positively influenced the increase in elongation in the case of
MFC(B) films. The percentage of keratin was also studied in the formulation of TOCNF films, and the
5% value was also positive. This addition improves the visual appearance of the film, increases both the
stress at break and the strain at break, and significantly increases the WVTR. Since the elongation values
of the films obtained in this study, especially for TOCNF films, were low, incorporating a plasticizer
was considered. Two widely used plasticizers were compared: sorbitol and PEG 400, and the choice fell
160
on the addition of 20% PEG 400, which increased the film elongation by 81%. Since PEG is known to
be a hydrophilic compound, its incorporation increased the WVTR value. Finally, a formulation of 75%
TOCNF + 20% PEG + 5% humid keratin was prepared. The film density decreased to 0.75 g/cm3 with
a high porosity of 47%. The blend increased the maximum degradation temperature by 12°C compared
to a pure TOCNF film and significantly increased the film elongation by 390%. On the other hand, the
stress at break was reduced.

Finally, this work presented the production of cellulose-based polymer films with keratin incorporation.
The addition of keratin to the formulation has been shown to improve some of the films' properties, and
the results obtained are encouraging for further study. The results presented here prove that keratin is
an interesting natural polymer for the production of bio-based films. As a follow-up to this study, the
prospects are to optimize the hot-pressing process of the films, to test the incorporation of humid keratin
into MFC films, to test other types and amounts of plasticizers, to test keratins extracted using different
extraction processes that may impart different properties to the keratin and may improve its
compatibility with cellulose, and a study of the oxygen barrier property would also be beneficial. Once
the manufacturing process has been optimized and the properties of the films obtained can be exploited,
it will be essential to carry out a life cycle assessment to evaluate the environmental impact of this bio-
sourced material, as well as an analysis of the biodegradability of these films.

161
162
GENERAL CONCLUSION
The literature review in the first chapter explores the global and local context of the agri-food poultry
industry, specifically in the Nouvelle Aquitaine region, focusing on the production of fattened ducks.
While acknowledging the region's importance in poultry, the literature predominantly examines chicken
feathers, leaving a gap in understanding duck feathers. The discussion highlights the feather's definition,
structure, and predominantly keratin composition, comprising approximately 90% of the feather.
Keratin has unique properties, including high cysteine content, α-helix or β-sheet structure,
biocompatibility, non-toxicity, and solvent resistance. Studies have explored the potential of feathers in
bio-based films through cleaning and keratin extraction. Existing protocols often involve hazardous
chemicals, leading to a transition to environmentally friendly processes using non-toxic and reusable
chemicals. Extracted keratin is used in bio-sourced films, frequently blended with other polymers and
plasticizers added to manage brittleness. Composite films of bio-based polymers are considered to
improve film quality. Cellulose, a promising bio-based polymer, particularly in packaging, is discussed,
with nanocellulose variants explored. Despite the hydrophilic nature of cellulose, studies combining it
with other polymers, especially extracted keratin, are proposed to overcome limitations, providing an
interesting avenue for exploration and local waste valorization while eliminating the disadvantages of
cellulose films.

The second chapter comprehensively analyzes locally collected duck feathers, including physical,
physicochemical, and morphological properties. A feather cleaning protocol is established, prioritizing
environmentally friendly chemicals to obtain pristine feathers suitable for various applications. The
objective is to understand the composition of the feathers, categorized by size, to determine if they can
be mixed without prior sorting. The compositional analysis shows moisture (7-10%), volatiles (80-
86%), ash (0.5-1.5%), fixed carbon (5-10%), crude fat (0.98 ± 0.12%), molecular weight of 10 kDa and
high protein content (82.97 ± 0.98%). Water interaction tests demonstrated hydrophobicity and assessed
resistance to acidic and basic media. The protein composition is confirmed by FT-IR analysis with
characteristic peaks, and NMR and XRD analyses show a predominant β-sheet structure. Thermal
analysis indicates material degradation at 230 °C, with protein melting at 230-245 °C. Morphological
analysis reveals a hierarchical keratin fiber structure, with different arrangements in the calamus and
rachis, providing both stiffness and lightness. These unique properties, specific to feathers and keratin,
offer promising prospects for bio-based materials applications.

The third chapter details the extraction of keratin, emphasizing a protocol aimed at minimizing its
environmental impact. Two non-toxic solvents, NMMO and DES, are compared for efficiency. Various
parameters were evaluated, including temperature, time, feather:solvent ratio, and feather size. Optimal
solvent preparation conditions, especially for DES, are established to reduce homogenization time. The
economic and environmental advantages of DES with its recyclable nature are highlighted. Feather
dissolution parameters are optimized for both solvents with NMMO requiring a lower temperature
(110 °C) and longer time (6 hours). A feather:solvent ratio is used, allowing solvent savings and

163
minimizing environmental impact. Keratin regeneration steps are optimized by comparing acids
(sulfuric, acetic, citric) with water, showing that adding acetic acid at pH 4-4.5 improves keratin
regeneration and yield. In particular, acetic acid is a viable alternative to conventional strong acids,
reducing environmental impact. The yields obtained exceed some literature results and demonstrate the
potential of bio-sourced, non-toxic processes.

Chapter IV performs a thorough physicochemical analysis of extracted keratins to understand the

influence of extraction parameters and solvents on their properties. Visual observations revealed a
significant color difference between keratins extracted with DES (yellowish) and NMMO (dark brown)
due to the formation of chromophores. SEM images showed a mixture of fibrous structures and porous
arrangements in the keratin morphology, with the solvent influencing these features. Chemical structure
analysis using NMR and FT-IR confirmed that the extracted keratin retained its similarity to raw feathers
and that the choice of solvent did not alter its chemical composition. The random coil + β-sheet structure,
which is critical to the properties of keratin, was preserved during extraction, as indicated by NMR
spectra. By SDS-PAGE analysis, the molecular mass of the extracted keratin (approximately 10 kDa)
agreed with literature values and raw feather keratin. No significant changes in chain size were observed
after extraction, regardless of the solvent or parameters studied. Thermal degradation analyses showed
that feathers extracted with NMMO maintained a degradation temperature of about 230 °C, while DES
extraction lowered this value by 10 °C. The duration parameter influenced the degradation temperatures,
requiring a minimum of 7 h for NMMO and less than 4 h for DES. The crystallinity index (CI) varied
with the solvent used, with DES yielding the lowest CIs and NMMO maintaining values close to feather.
Keratin's diverse morphologies, excellent thermal and chemical properties, low density,
biocompatibility, and biodegradability position it as a valuable natural material for applications such as
cosmetics, pharmaceuticals, agriculture, and bio-based films and coatings.

Chapter V focuses on developing a bio-based polymer film as an alternative to conventional synthetic

polymers in food packaging. Cellulose, abundant in nature, was chosen, but its hydrophilic nature limits
applications. Keratin extracted from duck feathers is incorporated to study its effect on the properties of
cellulose film and, therefore, overcome its weaknesses mentioned above. Two types of cellulose,
microfibrillated cellulose (MFC) and TEMPO-oxidized nanocellulose (TOCNF), were used. Various
parameters such as manufacturing method, keratin type and amount, plasticizer (sorbitol or PEG 400),
and their amounts are investigated. The study reveals that pressing is the preferred manufacturing
method, and using humid keratin improves film homogeneity, visual appearance, and mechanical
properties while reducing the thermal degradation rate. Moreover, adding 5% keratin allows the
improvement of some film properties. Incorporating 20% PEG 400 as a plasticizer increases the
elongation of the film by 81% and increases the water vapor transmission rate (WVTR). A final
formulation with 75% TOCNF, 20% PEG, and 5% moist keratin shows improved degradation
temperature and film elongation which is highly promising for further study.

164
PERSPECTIVES
The research work presented in this manuscript details the implementation of a valorization process for
by-products from the agri-food industry. This study led to the development of a keratin extraction
process that is environmentally friendlier than conventional extraction processes while maintaining very
reasonable yields. The second part of this work presented a solid basis for a cellulose film production
protocol and encouraging results regarding using keratin to form bio-sourced materials.
Following, there are proposals for perspectives that can be used for further developments:

• Conduct a keratin production cost study for the entire extraction protocol, from duck feather
harvesting to keratin powder production.
• Conduct a life cycle analysis of the extracted keratin to determine the environmental impact of
the extraction process.
• Conduct a feasibility study to scale the extraction process from laboratory to pilot scale.
• Implement solvent recovery to minimize solvent usage and costs.
• Continue to optimize the hot-pressed cellulose film manufacturing process.
• Test the incorporation of a different keratin into cellulose films derived from a different
extraction, which may have different properties and characteristics.
• Once the film formulation has been optimized, a life cycle and biodegradability analysis will
be performed.

165
166
 REFERENCES
[1] U. N. N. (UNDESA) YORK, World population prospects 2019, no. 141. 2019.
[2] T. Tesfaye, B. Sithole, and D. Ramjugernath, “Valorisation of chicken feathers: a review on
recycling and recovery route—current status and future prospects,” Clean Technologies and
Environmental Policy. 2017, doi: 10.1007/s10098-017-1443-9.
[3] N.- Aquitaine, “MÉMENTO 2020 Industries agroalimentaires,” 2020.
[4] E. Nouvelle-aquitaine, “Nouvelle- aquitaine,” 2023.
[5] R. G. Gurav and J. P. Jadhav, “A novel source of biofertilizer from feather biomass for banana
cultivation,” Environ. Sci. Pollut. Res., vol. 20, no. 7, pp. 4532–4539, 2013, doi:
10.1007/s11356-012-1405-z.
[6] H. Chen et al., “Valorization of Livestock Keratin Waste: Application in Agricultural Fields,”
Int. J. Environ. Res. Public Health, vol. 19, no. 11, 2022, doi: 10.3390/ijerph19116681.
[7] J. G. Rouse and M. E. Van Dyke, “A review of keratin-based biomaterials for biomedical
applications,” Materials (Basel)., vol. 3, no. 2, pp. 999–1014, 2010, doi: 10.3390/ma3020999.
[8] A. L. V. Villa et al., “Feather keratin hydrolysates obtained from microbial keratinases: Effect
on hair fiber,” BMC Biotechnol., vol. 13, 2013, doi: 10.1186/1472-6750-13-15.
[9] Agreste, “Filière palmipèdes gras,” Limoges, France, 2017.
[10] A. Kovalev, A. E. Filippov, and S. N. Gorb, “Unzipping bird feathers,” J. R. Soc. Interface, vol.
11, no. 92, 2014, doi: 10.1098/rsif.2013.0988.
[11] L. Alibardi, “Immunolocalization of alpha-keratins and feather beta-proteins in feather cells and
comparison with the general process of cornification in the skin of mammals,” Ann. Anat., vol.
195, no. 2, pp. 189–198, 2013, doi: 10.1016/[Link].2012.08.005.
[12] N. Reddy and Y. Yang, “Structure and properties of chicken feather barbs as natural protein
fibers,” J. Polym. Environ., 2007, doi: 10.1007/s10924-007-0054-7.
[13] G. Bansal and V. Singh, “Review on Chicken Feather Fiber (CFF) a Livestock Waste in
Composite Material Development,” Int. J. Waste Resour., vol. 06, no. 04, 2016, doi:
10.4172/2252-5211.1000254.
[14] T. N. Sullivan, B. Wang, H. D. Espinosa, and M. A. Meyers, “Extreme lightweight structures:
avian feathers and bones,” Materials Today. 2017, doi: 10.1016/[Link].2017.02.004.
[15] T. Tesfaye, B. Sithole, D. Ramjugernath, and V. Chunilall, “Valorisation of chicken feathers:
Characterisation of chemical properties,” Waste Manag., 2017, doi:
10.1016/[Link].2017.06.050.
[16] M. Zhan and R. P. Wool, “Mechanical properties of chicken feather fibers,” Polym. Compos.,
2011, doi: 10.1002/pc.21112.
[17] L. Zhang and M. Zeng, “Proteins as Sources of Materials,” in Monomers, Polymers and
Composites from Renewable Ressources, Elsevier S., M. Belgacem and A. Gandini, Eds. 2008,
pp. 479–493.

167
[18] S. Kokot, “Keratin,” in Encyclopedia of Materials: Science and Technology, 2001, pp. 4363–
4368.
[19] B. Wang, W. Yang, J. McKittrick, and M. A. Meyers, “Keratin: Structure, mechanical properties,
occurrence in biological organisms, and efforts at bioinspiration,” Progress in Materials Science.
2016, doi: 10.1016/[Link].2015.06.001.
[20] K. M. Arai, R. Takahashi, Y. Yokote, and K. Akahane, “The primary structure of feather keratins
from duck (Anas platyrhynchos) and pigeon (Columba livia),” Biochim. Biophys. Acta
(BBA)/Protein Struct. Mol., 1986, doi: 10.1016/0167-4838(86)90182-2.
[21] M. . Papadopoulos, A. R. El Boushy, A. E. Roodbeen, and E. H. Ketelaars, “Effects of processing
time and moisture content on amino acid composition and nitrogen characteristics of feather
meal,” Anim. Feed Sci. Technol., vol. 14, pp. 279–290, 1986.
[22] K. Saravanan and B. Dhurai, “Exploration on amino acid content and morphological structure in
chicken feather fiber,” J. Text. Apparel, Technol. Manag., vol. 7, no. 3, pp. 1–6, 2012.
[23] W. H. Ward, C. H. Binkley, and N. S. Snell, “Amino Acid Composition of Normal Wools, Wool
Fractions, Mohair, Feather, and Feather Fractions,” Text. Res. J., vol. 25, no. 4, pp. 314–325,
1955, doi: 10.1177/004051755502500403.
[24] S. Sharma and A. Gupta, “Sustainable Management of Keratin Waste Biomass: Applications and
Future Perspectives,” Brazilian Arch. Biol. Technol., 2016, doi: 10.1590/1678-4324-
2016150684.
[25] I. S. Rodríguez-Clavel, S. P. Paredes-Carrera, S. O. Flores-Valle, E. J. Paz-García, J. C. Sánchez-
Ochoa, and R. M. Pérez-Gutiérrez, “Effect of Microwave or Ultrasound Irradiation in the
Extraction from Feather Keratin,” J. Chem., vol. 2019, no. Figure 1, 2019, doi:
10.1155/2019/1326063.
[26] F. Pourjavaheri, F. Mohaddes, P. Bramwell, F. Sherkat, and R. A. Shanks, “Purification of avian
biological material to refined keratin fibres,” RSC Adv., vol. 5, no. 86, pp. 69899–69906, 2015,
doi: 10.1039/c5ra08947f.
[27] S. Sharma et al., “Study of different treatment methods on chicken feather biomass,” IIUM Eng.
J., 2017, doi: 10.31436/iiumej.v18i2.806.
[28] S. Strnad and T. Kreze, “Poultry feather wastes - optimization of cleaning procedure for new
products development,” IN-TECH 2011 proceedings, Int. Conf. Innov. Technol., pp. 532–535,
2011.
[29] R. Šafarič et al., “Preparation and characterisation of waste poultry feathers composite
fibreboards,” Materials (Basel)., 2020, doi: 10.3390/ma13214964.
[30] M. Casadesús et al., “Effect of chemical treatments and additives on properties of chicken
feathers thermoplastic biocomposites,” J. Compos. Mater., vol. 52, no. 26, pp. 3637–3653, 2018,
doi: 10.1177/0021998318766652.
[31] B. Ma, Q. Sun, J. Yang, J. Wizi, X. Hou, and Y. Yang, “Degradation and regeneration of feather
keratin in NMMO solution,” Environ. Sci. Pollut. Res., vol. 24, no. 21, pp. 17711–17718, 2017,
doi: 10.1007/s11356-017-9410-x.
[32] U. Baroroh Lili Utami, S. Heru, and B. Cahyono, “Study of Duck Feather Modification using
NaOH to Removal Iron in Acid Mine Drainage (AMD),” E3S Web Conf., vol. 202, pp. 1–11,
2020, doi: 10.1051/e3sconf/202020205002.

168
[33] A. Ullah, T. Vasanthan, D. Bressler, A. L. Elias, and J. Wu, “Bioplastics from Feather Quill,”
Am. Chem. Soc., 2011, doi: 10.1021/bm201112n.
[34] A. Gupta, R. Perumal, R. B. M. Yunus, and N. B. Kamarudin, “Extraction of keratin protein
from chicken feather,” Chemeca Eng. a Better World, 2011.
[35] L. Shizhen, “Ben Cao Gang Mu,” Time Lit. Art Press Chang. Jilin, China, 2005.
[36] S. Feroz, N. Muhammad, J. Ranayake, and G. Dias, “Keratin - Based materials for biomedical
applications,” Bioact. Mater., vol. 5, no. 3, pp. 496–509, 2020, doi:
10.1016/[Link].2020.04.007.
[37] J. Hofmeier, “Horn-lime plastic masses from keratin substances.,” Ger Pat no 184915, 1905.
[38] C. Beyer, “The keratine or horny substance of the hair.,” Ger pat no 22643, 1907.
[39] H. Dale, “Keratin and other coatings for pills,” Pharm J, pp. 129:494–5, 1932.
[40] E. M. Brown, K. Pandya, M. M. Taylor, and C.-K. Liu, “Comparison of Methods for Extraction
of Keratin from Waste Wool,” Agric. Sci., 2016, doi: 10.4236/as.2016.710063.
[41] K. Yamauchi and A. Yamauchi, “Preparation of stable aqueous solution of keratins, and
physicochemical and biochemical properties of films,” Am. Chem. Soc. Polym. Prepr. Div.
Polym. Chem., vol. 39, no. 1, pp. 357–358, 1998.
[42] B. Y. Alashwal, M. Saad Bala, A. Gupta, S. Sharma, and P. Mishra, “Improved properties of
keratin-based bioplastic film blended with microcrystalline cellulose: A comparative analysis,”
J. King Saud Univ. - Sci., 2020, doi: 10.1016/[Link].2019.03.006.
[43] C. D. Tran and T. M. Mututuvari, “Cellulose, Chitosan and Keratin Composite Materials: Facile
and Recyclable Synthesis, Conformation and Properties,” ACS Sustain. Chem. Eng., vol. 4, no.
3, pp. 1850–1861, 2016, doi: 10.1021/acssuschemeng.6b00084.
[44] D. Wang and R. C. Tang, “Dissolution of wool in the choline chloride/oxalic acid deep eutectic
solvent,” Mater. Lett., 2018, doi: 10.1016/[Link].2018.08.056.
[45] E. M. Nuutinen et al., “Green process to regenerate keratin from feathers with an aqueous deep
eutectic solvent,” RSC Adv., 2019, doi: 10.1039/c9ra03305j.
[46] N. Eslahi, F. Dadashian, and N. H. Nejad, “An investigation on keratin extraction from wool and
feather waste by enzymatic hydrolysis,” Prep. Biochem. Biotechnol., vol. 43, no. 7, pp. 624–648,
2013, doi: 10.1080/10826068.2013.763826.
[47] Y. Zhang, W. Zhao, and R. Yang, “Steam Flash Explosion Assisted Dissolution of Keratin from
Feathers,” ACS Sustain. Chem. Eng., 2015, doi: 10.1021/acssuschemeng.5b00310.
[48] C. Earland, P. R. Blakey, and J. G. P. Stell, “Molecular orientation of some keratins,” Nature,
1962, doi: 10.1038/1961287a0.
[49] B. Ma, X. Qiao, X. Hou, and Y. Yang, “Pure keratin membrane and fibers from chicken feather,”
Int. J. Biol. Macromol., 2016, doi: 10.1016/[Link].2016.04.039.
[50] K. Wang, R. Li, J. H. Ma, Y. K. Jian, and J. N. Che, “Extracting keratin from wool by using l-
cysteine,” Green Chem., 2016, doi: 10.1039/c5gc01254f.
[51] C. Tonin, A. Aluigi, C. Vineis, A. Varesano, A. Montarsolo, and F. Ferrero, “Thermal and
structural characterization of poly(ethylene-oxide)/keratin blend films,” J. Therm. Anal.
Calorim., vol. 89, no. 2, pp. 601–608, 2007, doi: 10.1007/s10973-006-7557-7.

169
[52] I. Sinkiewicz, A. Śliwińska, H. Staroszczyk, and I. Kołodziejska, “Alternative Methods of
Preparation of Soluble Keratin from Chicken Feathers,” Waste and Biomass Valorization, 2017,
doi: 10.1007/s12649-016-9678-y.
[53] B. Branska et al., “Chicken feather and wheat straw hydrolysate for direct utilization in
biobutanol production,” Renew. Energy, vol. 145, pp. 1941–1948, 2020, doi:
10.1016/[Link].2019.07.094.
[54] R. D. Rogers and K. R. Seddon, “Ionic Liquids - Solvents of the Future?,” Science, vol. 302, no.
5646. pp. 792–793, 2003, doi: 10.1126/science.1090313.
[55] A. Shavandi, T. H. Silva, A. A. Bekhit, and A. E. D. A. Bekhit, “Keratin: Dissolution, extraction
and biomedical application,” Biomater. Sci., vol. 5, no. 9, pp. 1699–1735, 2017, doi:
10.1039/c7bm00411g.
[56] A. Idris, R. Vijayaraghavan, A. F. Patti, and D. R. Macfarlane, “Distillable protic ionic liquids
for keratin dissolution and recovery,” ACS Sustain. Chem. Eng., vol. 2, no. 7, pp. 1888–1894,
2014, doi: 10.1021/sc500229a.
[57] A. Ghosh, S. Clerens, S. Deb-Choudhury, and J. M. Dyer, “Thermal effects of ionic liquid
dissolution on the structures and properties of regenerated wool keratin,” Polym. Degrad. Stab.,
vol. 108, pp. 108–115, 2014, doi: 10.1016/[Link].2014.06.007.
[58] A. P. Abbott, G. Capper, D. L. Davies, H. L. Munro, R. K. Rasheed, and V. Tambyrajah,
“Preparation of novel, moisture-stable, lewis-acidic ionic liquids containing quaternary
ammonium salts with functional side chains,” Chem. Commun., vol. 1, no. 19, pp. 2010–2011,
2001, doi: 10.1039/b106357j.
[59] E. L. Smith, A. P. Abbott, and K. S. Ryder, “Deep Eutectic Solvents (DESs) and Their
Applications,” Chem. Rev., vol. 114, no. 21, pp. 11060–11082, 2014, doi: 10.1021/cr300162p.
[60] Q. Zhang, K. De Oliveira Vigier, S. Royer, and F. Jérôme, “Deep eutectic solvents: Syntheses,
properties and applications,” Chem. Soc. Rev., 2012, doi: 10.1039/c2cs35178a.
[61] Z. Jiang et al., “Dissolution and regeneration of wool keratin in the deep eutectic solvent of
choline chloride-urea,” Int. J. Biol. Macromol., 2018, doi: 10.1016/[Link].2018.07.161.
[62] K. E. Moore, D. N. Mangos, A. D. Slattery, C. L. Raston, and R. A. Boulos, “Wool
deconstruction using a benign eutectic melt,” RSC Adv., 2016, doi: 10.1039/c5ra26516a.
[63] Y. Zhang, S. Wang, Z. Fang, H. Li, and J. Fang, “Molecular design and experimental study of
deep eutectic solvent extraction of keratin derived from feathers,” Int. J. Biol. Macromol., vol.
241, no. April, p. 124512, 2023, doi: 10.1016/[Link].2023.124512.
[64] T. Rosenau, A. Hofinger, A. Potthast, and P. Kosma, “On the conformation of the cellulose
solvent N-methylmorpholine-N-oxide (NMMO) in solution,” Polymer (Guildf)., vol. 44, no. 20,
pp. 6153–6158, 2003, doi: 10.1016/S0032-3861(03)00663-3.
[65] T. Rosenau, A. Potthast, H. Sixta, and P. Kosma, “The chemistry of side reactions and byproduct
formation in the system NMMO/cellulose (Lyocell process),” Prog. Polym. Sci., vol. 26, no. 9,
pp. 1763–1837, 2001, doi: 10.1016/S0079-6700(01)00023-5.
[66] K. E. Perepelkin, “Lyocell fibres based on direct dissolution of cellulose in N-methylmorpholine
N-oxide: Development and prospects,” Fibre Chem., vol. 39, no. 2, pp. 163–172, 2007, doi:
10.1007/s10692-007-0032-9.
[67] C. Chaitanya Reddy et al., “Valorization of keratin waste biomass and its potential applications,”

170
 J. Water Process Eng., vol. 40, no. January, p. 101707, 2021, doi: 10.1016/[Link].2020.101707.
[68] L. Lange, Y. Huang, and P. K. Busk, “Microbial decomposition of keratin in nature—a new
hypothesis of industrial relevance,” Appl. Microbiol. Biotechnol., vol. 100, no. 5, pp. 2083–2096,
2016, doi: 10.1007/s00253-015-7262-1.
[69] M. A. Hassan, D. Abol-Fotouh, A. M. Omer, T. M. Tamer, and E. Abbas, “Comprehensive
insights into microbial keratinases and their implication in various biotechnological and
industrial sectors: A review,” Int. J. Biol. Macromol., vol. 154, pp. 567–583, 2020, doi:
10.1016/[Link].2020.03.116.
[70] N. E. Nnolim, C. C. Udenigwe, A. I. Okoh, and U. U. Nwodo, “Microbial Keratinase: Next
Generation Green Catalyst and Prospective Applications,” Front. Microbiol., vol. 11, no.
December, 2020, doi: 10.3389/fmicb.2020.580164.
[71] N. Sahni, P. P. Sahota, and U. G. Phutela, “Bacterial keratinases and their prospective
applications: A review,” [Link], vol. 4, no. 6, pp. 768–783, 2015.
[72] Y. Yu et al., “Steam explosion of lignocellulosic biomass for multiple advanced bioenergy
processes: A review,” Renew. Sustain. Energy Rev., vol. 154, no. October 2021, p. 111871, 2022,
doi: 10.1016/[Link].2021.111871.
[73] S. G. Giteru, D. H. Ramsey, Y. Hou, L. Cong, A. Mohan, and A. E. D. A. Bekhit, “Wool keratin
as a novel alternative protein: A comprehensive review of extraction, purification, nutrition,
safety, and food applications,” Compr. Rev. Food Sci. Food Saf., vol. 22, no. 1, pp. 643–687,
2023, doi: 10.1111/1541-4337.13087.
[74] M. Zoccola et al., “Microwave-assisted chemical-free hydrolysis of wool keratin,” Text. Res. J.,
vol. 82, no. 19, pp. 2006–2018, 2012, doi: 10.1177/0040517512452948.
[75] G. J. Dias, T. N. Haththotuwa, D. S. Rowlands, M. Gram, and A. E. D. A. Bekhit, “Wool keratin
– A novel dietary protein source: Nutritional value and toxicological assessment,” Food Chem.,
vol. 383, no. October 2021, p. 132436, 2022, doi: 10.1016/[Link].2022.132436.
[76] M. Vinatoru, “An overview of the ultrasonically assisted extraction of bioactive principles from
herbs,” Ultrason. Sonochem., vol. 8, no. 3, pp. 303–313, 2001, doi: 10.1016/S1350-
4177(01)00071-2.
[77] A. Gedanken, “Using sonochemistry for the fabrication of nanomaterials,” Ultrason. Sonochem.,
vol. 11, no. 2, pp. 47–55, 2004, doi: 10.1016/[Link].2004.01.037.
[78] K. S. Suslick, “Sonochemistry,” vol. 247, no. 1439–1445, 1990, doi:
10.1126/science.247.4949.1439 View.
[79] N. A. Azmi, A. Idris, and N. S. M. Yusof, “Ultrasonic technology for value added products from
feather keratin,” Ultrason. Sonochem., vol. 47, no. March, pp. 99–107, 2018, doi:
10.1016/[Link].2018.04.016.
[80] N. Eslahi, N. Hemmatinejad, and F. Dadashian, “From feather waste to valuable nanoparticles,”
Part. Sci. Technol., vol. 32, no. 3, pp. 242–250, 2014, doi: 10.1080/02726351.2013.851135.
[81] “Plastics – the Facts 2020,” Wemmel, 2020.
[82] J. R. Barone, W. F. Schmidt, and C. F. E. Liebner, “Compounding and molding of polyethylene
composites reinforced with keratin feather fiber,” Compos. Sci. Technol., vol. 65, no. 3–4, pp.
683–692, 2005, doi: 10.1016/[Link].2004.09.030.

171
[83] P. Mokrejs, M. Hutta, J. Pavlackova, P. Egner, and L. Benicek, “The cosmetic and
dermatological potential of keratin hydrolysate,” J. Cosmet. Dermatol., vol. 16, no. 4, pp. e21–
e27, 2017, doi: 10.1111/jocd.12319.
[84] J. R. Barone, W. F. Schmidt, and C. F. E. Liebner, “Thermally processed keratin films,” J. Appl.
Polym. Sci., vol. 97, no. 4, pp. 1644–1651, 2005, doi: 10.1002/app.21901.
[85] T. Tanabe, N. Okitsu, and K. Yamauchi, “Fabrication and characterization of chemically
crosslinked keratin films,” Mater. Sci. Eng. C, 2004, doi: 10.1016/[Link].2003.11.004.
[86] Y. Dou, B. Zhang, M. He, G. Yin, and Y. Cui, “The structure, tensile properties and water
resistance of hydrolyzed feather keratin-based bioplastics,” Chinese J. Chem. Eng., 2016, doi:
10.1016/[Link].2015.11.007.
[87] G. Rocha Plácido Moore, S. Maria Martelli, C. Gandolfo, P. José do Amaral Sobral, and J.
Borges Laurindo, “Influence of the glycerol concentration on some physical properties of feather
keratin films,” Food Hydrocoll., vol. 20, no. 7, pp. 975–982, 2006, doi:
10.1016/[Link].2005.11.001.
[88] B. Fernández-d’Arlas, “Tough and Functional Cross-linked Bioplastics from Sheep Wool
Keratin,” Sci. Rep., 2019, doi: 10.1038/s41598-019-51393-5.
[89] N. Ramakrishnan, S. Sharma, A. Gupta, and B. Y. Alashwal, “Keratin based bioplastic film from
chicken feathers and its characterization,” Int. J. Biol. Macromol., 2018, doi:
10.1016/[Link].2018.01.037.
[90] S. M. Martelli, G. R. P. Moore, and J. B. Laurindo, “Mechanical properties, water vapor
permeability and water affinity of feather keratin films plasticized with sorbitol,” J. Polym.
Environ., vol. 14, no. 3, pp. 215–222, 2006, doi: 10.1007/s10924-006-0017-4.
[91] Y. Liu et al., “A review of cellulose and its derivatives in biopolymer-based for food packaging
application,” Trends Food Sci. Technol., vol. 112, no. April, pp. 532–546, 2021, doi:
10.1016/[Link].2021.04.016.
[92] A. Arif et al., “Bioplastics from waste biomass of marine and poultry industries,” J. Biosci., vol.
48, no. 2, 2023, doi: 10.1007/s12038-023-00332-8.
[93] K. Kamide, “Cellulose and cellulose derivatives: Molecular characterization and its
applications,” Elsevier., Japan, 2005.
[94] D. Klemm, B. Heublein, H. P. Fink, and A. Bohn, “Cellulose: Fascinating biopolymer and
sustainable raw material,” Angew. Chemie - Int. Ed., vol. 44, no. 22, pp. 3358–3393, 2005, doi:
10.1002/anie.200460587.
[95] F. Casanova et al., “Novel micro-and nanocellulose-based delivery systems for liposoluble
compounds,” Nanomaterials, vol. 11, no. 10, pp. 1–38, 2021, doi: 10.3390/nano11102593.
[96] C. G. Flores-Hernández et al., “All green composites from fully renewable biopolymers:
Chitosan-Starch reinforced with Keratin from feathers,” Polymers (Basel)., 2014, doi:
10.3390/polym6030686.
[97] B. Ocak, “Collagen hydrolysate/chitosan/keratin-based ternary films from bio-wastes of the
leather and poultry industries against plastic waste generation,” Biomass Convers. Biorefinery,
no. 0123456789, 2022, doi: 10.1007/s13399-022-03297-0.
[98] P. Pardo-Ibáñez, A. Lopez-Rubio, M. Martínez-Sanz, L. Cabedo, and J. M. Lagaron, “Keratin-
polyhydroxyalkanoate melt-compounded composites with improved barrier properties of interest

172
 in food packaging applications,” J. Appl. Polym. Sci., vol. 131, no. 4, pp. 1–10, 2014, doi:
10.1002/app.39947.
[99] B. M. Liebeck, N. Hidalgo, G. Roth, C. Popescu, and A. Böker, “Synthesis and Characterization
of Methyl Cellulose/Keratin Hydrolysate Composite Membranes,” Polymers (Basel)., vol. 9, no.
3, 2017, doi: 10.3390/polym9030091.
[100] Y. Ma, R. Qi, S. Jia, and Z. Wang, “Preparation and Characterization of Keratin – cellulose
Composite fi lms,” in 2nd Annual International Conference on Advanced Material Engineering,
2016, pp. 38–44.
[101] D. Ciolacu, F. Ciolacu, and V. I. Popa, “Amorphous cellulose - structure and characterization,”
Cellul. Chem. Technol., vol. 45, no. 1–2, pp. 13–21, 2011.
[102] F. Li, E. Mascheroni, and L. Piergiovanni, “The potential of NanCellulose in the Packaging
Field: A Review,” Packag. Technol. Sci., vol. 28, pp. 475–508, 2015, doi: 10.1002/pts.2121.
[103] A. A. B. Omran et al., “Micro-and nanocellulose in polymer composite materials: A review,”
Polymers (Basel)., vol. 13, no. 2, pp. 1–30, 2021, doi: 10.3390/polym13020231.
[104] J. H. Kim et al., “Review of nanocellulose for sustainable future materials,” Int. J. Precis. Eng.
Manuf. - Green Technol., vol. 2, no. 2, pp. 197–213, 2015, doi: 10.1007/s40684-015-0024-9.
[105] P. Phanthong, P. Reubroycharoen, X. Hao, G. Xu, A. Abudula, and G. Guan, “Nanocellulose:
Extraction and application,” Carbon Resour. Convers., vol. 1, no. 1, pp. 32–43, 2018, doi:
10.1016/[Link].2018.05.004.
[106] A. Dufresne, “Nanocellulose: A new ageless bionanomaterial,” Mater. Today, vol. 16, no. 6, pp.
220–227, 2013, doi: 10.1016/[Link].2013.06.004.
[107] A. J. Sayyed, N. A. Deshmukh, and D. V. Pinjari, “A critical review of manufacturing processes
used in regenerated cellulosic fibres: viscose, cellulose acetate, cuprammonium, LiCl/DMAc,
ionic liquids, and NMMO based lyocell,” Cellulose, vol. 26, no. 5, pp. 2913–2940, 2019, doi:
10.1007/s10570-019-02318-y.
[108] E. Borbély, “Lyocell, The New Generation of Regenerated Cellulose,” Acta Polytech.
Hungarica, vol. 5, no. 3, 2008.
[109] X. Jiang, Y. Bai, X. Chen, and W. Liu, “A review on raw materials, commercial production and
properties of lyocell fiber,” J. Bioresour. Bioprod., vol. 5, no. 1, pp. 16–25, 2020, doi:
10.1016/[Link].2020.03.002.
[110] A. Isogai, T. Saito, and H. Fukuzumi, “TEMPO-oxidized cellulose nanofibers,” Nanoscale, vol.
3, no. 1, pp. 71–85, 2011, doi: 10.1039/c0nr00583e.
[111] A. Serra, I. González, H. Oliver-Ortega, Q. Tarrès, M. Delgado-Aguilar, and P. Mutjé,
“Reducing the amount of catalyst in TEMPO-oxidized cellulose nanofibers: Effect on properties
and cost,” Polymers (Basel)., vol. 9, no. 11, 2017, doi: 10.3390/polym9110557.
[112] Y. Qing, R. Sabo, Y. Wu, and Z. Cai, “High-performance cellulose nanofibril composite films,”
BioResources, vol. 7, no. 3, pp. 3064–3075, 2012, doi: 10.15376/biores.7.3.3064-3075.
[113] E. Kaffashsaie et al., “Direct conversion of raw wood to TEMPO-oxidized cellulose nanofibers,”
Carbohydr. Polym., vol. 262, no. March, p. 117938, 2021, doi: 10.1016/[Link].2021.117938.
[114] S. Sharma, X. Zhang, S. S. Nair, A. Ragauskas, J. Zhu, and Y. Deng, “Thermally enhanced high
performance cellulose nano fibril barrier membranes,” RSC Adv., vol. 4, no. 85, pp. 45136–

173
 45142, 2014, doi: 10.1039/c4ra07469f.
[115] M. A. Hubbe et al., “Nanocellulose in Thin Films, Coatings, and Plies for Packaging
Applications: A Review,” BioResources, vol. 12, no. 1, pp. 2143–2233, 2017, doi:
10.15376/biores.12.1.2143-2233.
[116] T. T. Chau, W. J. Bruckard, P. T. L. Koh, and A. V. Nguyen, “A review of factors that affect
contact angle and implications for flotation practice,” Adv. Colloid Interface Sci., vol. 150, no.
2, pp. 106–115, 2009, doi: 10.1016/[Link].2009.07.003.
[117] D. Dey et al., “Physical, antifungal, and biodegradable properties of cellulose nanocrystals and
chitosan nanoparticles for food packaging application,” Mater. Today Proc., vol. 38, pp. 860–
869, 2021, doi: 10.1016/[Link].2020.04.885.
[118] C. Patanè et al., “Nutritional changes during storage in fresh-cut long storage tomato as affected
by biocompostable polylactide and cellulose based packaging,” Lwt, vol. 101, no. November
2018, pp. 618–624, 2019, doi: 10.1016/[Link].2018.11.069.
[119] A. Fernández, P. Picouet, and E. Lloret, “Cellulose-silver nanoparticle hybrid materials to
control spoilage-related microflora in absorbent pads located in trays of fresh-cut melon,” Int. J.
Food Microbiol., vol. 142, no. 1–2, pp. 222–228, 2010, doi: 10.1016/[Link].2010.07.001.
[120] F. A. G. S. Silva, F. Dourado, M. Gama, and F. Poças, “Nanocellulose bio-based composites for
food packaging,” Nanomaterials, vol. 10, no. 10, pp. 1–29, 2020, doi: 10.3390/nano10102041.
[121] I. Hasan, J. Wang, and M. Tajvidi, “Tuning physical, mechanical and barrier properties of
cellulose nanofibril films through film drying techniques coupled with thermal compression,”
Cellulose, vol. 28, no. 18, pp. 11345–11366, 2021, doi: 10.1007/s10570-021-04269-9.
[122] N. Shahi, B. Min, B. Sapkota, and V. K. Rangari, “Eco-friendly cellulose nanofiber extraction
from sugarcane bagasse and film fabrication,” Sustain., vol. 12, no. 15, pp. 1–15, 2020, doi:
10.3390/su12156015.
[123] S. Van Nguyen and B. K. Lee, “Microfibrillated cellulose film with enhanced mechanical and
water-resistant properties by glycerol and hot-pressing treatment,” Cellulose, vol. 28, no. 9, pp.
5693–5705, 2021, doi: 10.1007/s10570-021-03894-8.
[124] V. Kumar et al., “Comparison of nano- and microfibrillated cellulose films,” Cellulose, vol. 21,
no. 5, pp. 3443–3456, 2014, doi: 10.1007/s10570-014-0357-5.
[125] CDC, “Guideline for disinfection and sterilization in healthcare facilities,” Healthc. Infect.
Control Pract. Advis. Comm., no. May, pp. 1–158, 2008.
[126] B. Jiang, R. Tsao, Y. Li, and M. Miao, “Food Safety: Food Analysis Technologies/Techniques,”
Encycl. Agric. Food Syst., vol. 3, pp. 273–288, 2014, doi: 10.1016/B978-0-444-52512-3.00052-
8.
[127] U. K. Laemmli, “Cleavage of structural proteins during the assembly of the head of
bacteriophage T4,” Nature, vol. 227, no. 5259, pp. 680–685, 1970, doi: 10.1038/227680a0.
[128] M. Brebu and I. Spiridon, “Thermal degradation of keratin waste,” J. Anal. Appl. Pyrolysis, 2011,
doi: 10.1016/[Link].2011.03.003.
[129] A. Rohman and Y. B. Che Man, “FTIR spectroscopy combined with chemometrics for analysis
of lard in the mixtures with body fats of lamb, cow, and chicken,” Int. Food Res. J., vol. 17, no.
3, pp. 519–526, 2010.

174
[130] B. Lee and F. M. Richards, “The interpretation of protein structures: Estimation of static
accessibility,” J. Mol. Biol., vol. 55, no. 3, 1971, doi: 10.1016/0022-2836(71)90324-X.
[131] Y. Liu, X. Chen, and J. H. Xin, “Hydrophobic duck feathers and their simulation on textile
substrates for water repellent treatment,” Bioinspiration and Biomimetics, vol. 3, no. 4, 2008,
doi: 10.1088/1748-3182/3/4/046007.
[132] A. J. Cozzone, “Proteins: Fundamental Chemical Properties,” eLS, pp. 1–10, 2002, doi:
10.1038/[Link].0001330.
[133] F. Pourjavaheri-Jad, R. Shanks, M. Czajka, and A. Gupta, “Purification and Characterisation of
Feathers prior to Keratin Extraction,” 2014.
[134] S. Sharma, S., Arun, G., Saufi, S. M., Chik, T., Chua, G. K., Pradeep Kumar, P., ... & Malini,
“Extraction and characterization of keratin from chicken feather waste biomass : a study,” Natl.
Conf. Postgrad. Res., 2016.
[135] J. Kong and S. Yu, “Fourier transform infrared spectroscopic analysis of protein secondary
structures,” Acta Biochim. Biophys. Sin. (Shanghai)., 2007, doi: 10.1111/j.1745-
7270.2007.00320.x.
[136] A. Barth, “Infrared spectroscopy of proteins,” Biochimica et Biophysica Acta - Bioenergetics.
2007, doi: 10.1016/[Link].2007.06.004.
[137] P. I. Haris and F. Severcan, “FTIR spectroscopic characterization of protein structure in aqueous
and non-aqueous media,” J. Mol. Catal. - B Enzym., vol. 7, no. 1–4, pp. 207–221, 1999, doi:
10.1016/S1381-1177(99)00030-2.
[138] M. Jackson and H. H. Mantsch, “The use and misuse of FTIR spectroscopy in the determination
of protein structure,” Crit. Rev. Biochem. Mol. Biol., vol. 30, no. 2, pp. 95–120, 1995, doi:
10.3109/10409239509085140.
[139] E. Menefee and G. Yee, “Thermally-Induced Structural Changes in Wool,” Text. Res. J., vol. 35,
no. 9, pp. 801–812, 1965, doi: 10.1177/004051756503500904.
[140] S. Alahyaribeik and A. Ullah, “Methods of keratin extraction from poultry feathers and their
effects on antioxidant activity of extracted keratin,” Int. J. Biol. Macromol., 2020, doi:
10.1016/[Link].2020.01.144.
[141] M. A. Khosa, J. Wu, and A. Ullah, “Chemical modification, characterization, and application of
chicken feathers as novel biosorbents,” RSC Adv., vol. 3, no. 43, pp. 20800–20810, 2013, doi:
10.1039/c3ra43787f.
[142] T. N. Sullivan, A. Pissarenko, S. A. Herrera, D. Kisailus, V. A. Lubarda, and M. A. Meyers, “A
lightweight, biological structure with tailored stiffness: The feather vane,” Acta Biomater., 2016,
doi: 10.1016/[Link].2016.05.022.
[143] R. K. Donato and A. Mija, “Keratin associations with synthetic, biosynthetic and natural
polymers: An extensive review,” Polymers. 2020, doi: 10.3390/polym12010032.
[144] H. Yoshimizu, H. Mimura, and I. Ando, “13C CP /MAS NMR Study of the Conformation of
Stretched or Heated Low-Sulfur Keratin Protein Films,” Macromolecules, vol. 24, no. 4, pp.
862–866, 1991, doi: 10.1021/ma00004a008.
[145] L. Utiu, D. E. Demco, R. Fechete, M. Möller, and C. Popescu, “Morphology and molecular
dynamics of hard α-keratin based micro-tubes by 1H and 13C solid-state NMR,” Chem. Phys.
Lett., vol. 517, no. 1–3, pp. 86–91, 2011, doi: 10.1016/[Link].2011.10.036.

175
[146] M. J. Duer, N. McDougal, and R. C. Murray, “A solid-state NMR study of the structure and
molecular mobility of α-keratin,” Phys. Chem. Chem. Phys., vol. 5, no. 13, pp. 2894–2899, 2003,
doi: 10.1039/b302506c.
[147] A. Idris, R. Vijayaraghavan, U. A. Rana, D. Fredericks, A. F. Patti, and D. R. MacFarlane,
“Dissolution of feather keratin in ionic liquids,” Green Chem., vol. 15, no. 2, pp. 525–534, 2013,
doi: 10.1039/c2gc36556a.
[148] L. Pillemer, E. Ecker, and J. Wells, “THE SPECIFICITY OF KERATINS,” J Exp Med., vol. 69,
no. 2, pp. 191–7, 1939, doi: 10.1084/jem.69.2.191.
[149] V. Kulichikhin et al., “A role of coagulant in structure formation of fibers and films spun from
cellulose solutions,” Materials (Basel)., vol. 13, no. 16, pp. 1–19, 2020, doi:
10.3390/MA13163495.
[150] X. Zhang, P. Zhu, Q. Li, and H. Xia, “Recent Advances in the Catalytic Conversion of Biomass
to Furfural in Deep Eutectic Solvents,” Front. Chem., vol. 10, no. May, pp. 1–12, 2022, doi:
10.3389/fchem.2022.911674.
[151] C. Du, B. Zhao, X. B. Chen, N. Birbilis, and H. Yang, “Effect of water presence on choline
chloride-2urea ionic liquid and coating platings from the hydrated ionic liquid,” Sci. Rep., vol.
6, no. July, 2016, doi: 10.1038/srep29225.
[152] J. Grdadolnik and Y. Maréchal, “Urea and urea-water solutions - An infrared study,” J. Mol.
Struct., vol. 615, no. 1–3, pp. 177–189, 2002, doi: 10.1016/S0022-2860(02)00214-4.
[153] X. Zhang, Y. Feng, and X. Yang, “Extraction of Keratin from Poultry Feathers with Choline
Chloride-Oxalic Acid Deep Eutectic Solvent,” Fibers Polym., vol. 22, no. 12, pp. 3326–3335,
2021, doi: 10.1007/s12221-021-0255-z.
[154] Y. Ji, J. Chen, J. Lv, Z. Li, L. Xing, and S. Ding, “Extraction of keratin with ionic liquids from
poultry feather,” Sep. Purif. Technol., vol. 132, pp. 577–583, 2014, doi:
10.1016/[Link].2014.05.049.
[155] S. Sharma, A. Gupta, A. Kumar, C. G. Kee, H. Kamyab, and S. M. Saufi, “An efficient
conversion of waste feather keratin into ecofriendly bioplastic film,” Clean Technol. Environ.
Policy, 2018, doi: 10.1007/s10098-018-1498-2.
[156] S. Isarankura Na Ayutthaya, S. Tanpichai, and J. Wootthikanokkhan, “Keratin Extracted from
Chicken Feather Waste: Extraction, Preparation, and Structural Characterization of the Keratin
and Keratin/Biopolymer Films and Electrospuns,” J. Polym. Environ., vol. 23, no. 4, pp. 506–
516, 2015, doi: 10.1007/s10924-015-0725-8.
[157] European Medicines Agency, “Bulk density and tapped density of powder - General Chapter.,”
vol. 44, no. September, 2010.
[158] J. R. Barone and W. F. Schmidt, “Polyethylene reinforced with keratin fibers obtained from
chicken feathers,” Compos. Sci. Technol., vol. 65, no. 2, pp. 173–181, Feb. 2005, doi:
10.1016/[Link].2004.06.011.
[159] G. J. Smith, “Photodegradation of Keratin and Other Structural Proteins,” J. Photochem.
Photobiol. B-Biology, vol. 27, no. 3, pp. 187–198, 1995.
[160] T. Rosenau, A. Potthast, W. Milacher, I. Adorjan, A. Hofinger, and P. Kosma, “Discoloration of
cellulose solutions in N-methylmorpholine-N-oxide (Lyocell). Part 2: Isolation and
identification of chromophores,” Cellulose, vol. 12, no. 2, pp. 197–208, 2005, doi:
10.1007/s10570-004-0210-3.
176
[161] F. Wendler, G. Graneß, and T. Heinze, “Characterization of autocatalytic reactions in modified
cellulose/NMMO solutions by hermal analysis and UV/VIS spectroscopy,” Cellulose, vol. 12,
no. 4, pp. 411–422, 2005, doi: 10.1007/s10570-005-2201-4.
[162] M. Dąbrowska, A. Sommer, I. Sinkiewicz, A. Taraszkiewicz, and H. Staroszczyk, “An optimal
designed experiment for the alkaline hydrolysis of feather keratin,” Environ. Sci. Pollut. Res.,
vol. 29, no. 16, pp. 24145–24154, 2022, doi: 10.1007/s11356-021-17649-2.
[163] A. Mukherjee, Y. H. Kabutare, and P. Ghosh, “Dual crosslinked keratin-alginate fibers formed
via ionic complexation of amide networks with improved toughness for assembling into braids,”
Polym. Test., vol. 81, no. November 2019, p. 106286, 2020, doi:
10.1016/[Link].2019.106286.
[164] S. Sadeghi, F. Dadashian, and N. Eslahi, “Recycling chicken feathers to produce adsorbent
porous keratin-based sponge,” Int. J. Environ. Sci. Technol., vol. 16, no. 2, pp. 1119–1128, 2019,
doi: 10.1007/s13762-018-1669-z.
[165] T. Tesfaye, B. Sithole, D. Ramjugernath, and T. Mokhothu, “Valorisation of chicken feathers:
Characterisation of thermal, mechanical and electrical properties,” Sustain. Chem. Pharm., 2018,
doi: 10.1016/[Link].2018.05.003.
[166] C. Tonin et al., “Study on the conversion of wool keratin by steam explosion,”
Biomacromolecules, vol. 7, no. 12, pp. 3499–3504, 2006, doi: 10.1021/bm060597w.
[167] G. F. Davidson, “6—The acidic properties of cotton cellulose and derived oxycelluloses. Part II.
The absorption of methylene blue,” J. Text. Inst. Trans., vol. 39, no. 3, pp. T65–T86, 1948, doi:
10.1080/19447024808659403.
[168] A. Mayrhofer, S. Kopacic, and W. Bauer, “Extensive Characterization of Alginate, Chitosan and
Microfibrillated Cellulose Cast Films to Assess their Suitability as Barrier Coating for Paper and
Board,” Polymers (Basel)., vol. 15, no. 16, 2023, doi: 10.3390/polym15163336.
[169] G. Rodionova, M. Lenes, Ø. Eriksen, and Ø. Gregersen, “Surface chemical modification of
microfibrillated cellulose: Improvement of barrier properties for packaging applications,”
Cellulose, vol. 18, no. 1, pp. 127–134, 2011, doi: 10.1007/s10570-010-9474-y.
[170] L. Zhang, J. Liu, S. Y. Sui, C. H. Dong, J. T. Xu, and P. Zhu, “Study on the preparation and
characterization of cross-linked film of human hair keratin and cellulose,” Ferroelectrics, vol.
521, no. 1, pp. 77–85, 2017, doi: 10.1080/00150193.2017.1390964.
[171] H. Yang, R. Yan, H. Chen, D. H. Lee, and C. Zheng, “Characteristics of hemicellulose, cellulose
and lignin pyrolysis,” Fuel, vol. 86, no. 12–13, pp. 1781–1788, 2007, doi:
10.1016/[Link].2006.12.013.
[172] J. I. Morán, V. A. Alvarez, V. P. Cyras, and A. Vázquez, “Extraction of cellulose and preparation
of nanocellulose from sisal fibers,” Cellulose, vol. 15, no. 1, pp. 149–159, 2008, doi:
10.1007/s10570-007-9145-9.
[173] C. N. Wu and K. C. Cheng, “Strong, thermal-stable, flexible, and transparent films by self-
assembled TEMPO-oxidized bacterial cellulose nanofibers,” Cellulose, vol. 24, no. 1, pp. 269–
283, 2017, doi: 10.1007/s10570-016-1114-8.
[174] J. Patiño-Masó, F. Serra-Parareda, Q. Tarrés, P. Mutjé, F. X. Espinach, and M. Delgado-Aguilar,
“TEMPO-oxidized cellulose nanofibers: A potential bio-based superabsorbent for diaper
production,” Nanomaterials, vol. 9, no. 9, 2019, doi: 10.3390/nano9091271.
[175] H. Kargarzadeh, M. Ioelovich, I. Ahmad, S. Thomas, and A. Dufresne, “Methods for Extraction
177
 of Nanocellulose from Various Sources,” Handb. Nanocellulose Cellul. Nanocomposites, pp. 1–
49, 2017, doi: 10.1002/9783527689972.ch1.
[176] R. Ortega-Toro, J. Bonilla, P. Talens, and A. Chiralt, “Future of Starch-Based Materials in Food
Packaging,” in Starch-Based Materials in Food Packaging: Processing, Characterization and
Applications, Elsevier Inc., 2017, pp. 257–312.
[177] N. M. Stark, “Opportunities for cellulose nanomaterials in packaging films: A review and future
trends,” J. Renew. Mater., vol. 4, no. 5, pp. 313–326, 2016, doi: 10.7569/JRM.2016.634115.
[178] P. R. Sharma and A. J. Varma, “Thermal stability of cellulose and their nanoparticles: Effect of
incremental increases in carboxyl and aldehyde groups,” Carbohydr. Polym., vol. 114, pp. 339–
343, 2014, doi: 10.1016/[Link].2014.08.032.
[179] E. Amini, I. Hafez, M. Tajvidi, and D. W. Bousfield, “Cellulose and lignocellulose nanofibril
suspensions and films: A comparison,” Carbohydr. Polym., vol. 250, no. August, p. 117011,
2020, doi: 10.1016/[Link].2020.117011.
[180] D. F. Parra, J. Fusaro, F. Gaboardi, and D. S. Rosa, “Influence of poly (ethylene glycol) on the
thermal, mechanical, morphological, physical-chemical and biodegradation properties of poly
(3-hydroxybutyrate),” Polym. Degrad. Stab., vol. 91, no. 9, pp. 1954–1959, 2006, doi:
10.1016/[Link].2006.02.008.

178
179
 ECOLE DOCTORALE :
Sciences Exactes et leurs Applications (ED211)

LABORATOIRE :
Institut des Sciences Analytiques et de Physico-Chimie pour l’Environnement
et les Matériaux (IPREM)