Materials Chemistry and Physics 132 (2012) 509–514
Contents lists available at SciVerse ScienceDirect
Materials Chemistry and Physics
journal homepage: [Link]/locate/matchemphys
Morphological effects of reflux condensation on nanocrystalline anatase gel
and thin films
Sofia Javed a,∗ , Mohammad Mujahid a , Mohammad Islam b , Umair Manzoor c
a
School of Chemical and Materials Engineering (SCME), National University of Sciences and Technology (NUST),1 H-12, Islamabad, Pakistan
b
Center of Excellence for Research in Engineering Materials (CEREM), College of Engineering, King Saud University,2 P.O. Box 800, Riyadh 11421, Saudi Arabia
c
Shaikh Mohammad Alamoudi Chair for Water Researches, Department of Agriculture Engineering, College of Food and Agriculture Sciences,
King Saud University,3 Riyadh, Saudi Arabia
a r t i c l e i n f o a b s t r a c t
Article history: Synthesis of anatase phase of titania is of great interest because of its efficiency in photocatalysis to
Received 21 March 2011 produce renewable energy in different ways, one being dye sensitized solar cells (DSSC). In the present
Received in revised form 21 October 2011 work reflux condensation is utilized to obtain stable anatase nanostructures from sol gel synthesis using
Accepted 24 November 2011
titanium tetraisopropoxide (TTIP) precursor. The refluxing time affects the morphology of as synthesized
titania and that of annealed films as well. This variation of morphology is studied and a possible expla-
Keywords:
nation is made. Both XRD and TEM analysis reveal crystallite size to be less than 10 nm. SEM and TEM are
Nanocrystalline titania
used to study morphology which changes from rod like to spherical particles with increasing refluxing
Nanostructures
Oxides
time. TGA shows anatase stability upto 700 ◦ C and BET method reveals surface area to be larger than
Sol–gel growth 120 m2 g−1 . The films possess great potential for high efficiency when employed in devices.
Thin films © 2011 Elsevier B.V. All rights reserved.
1. Introduction in most cases [5,7,8]. Sol–gel method using titanium alkoxides
[4,7,9–16], hydrothermal method [6,17], template-assisted growth
Materials with at least one dimension of the order of few [18], anodic oxidation of titanium [19], solvothermal approach [20],
nanometers, referred to as nanomaterials or nanostructures, offer or a suitable combination of these are some of the techniques
great potential in various fields due to their unique electronic, opti- reported for titania synthesis in nanocrystalline form. The mor-
cal, magnetic, chemical, and physical properties [1,2]. Titania (TiO2 ) phology and physical/chemical properties of TiO2 nanostructures
in the form of nanoparticles, nanorods, or thin films has attracted depend on particular process, precursor type and concentration,
substantial interest due to their unusual optical, electrical and cat- use of capping agents, synthesis temperature, pressure and time [9].
alytic properties [3]. These unique properties offer their strong Titania is an important component of dye-sensitized solar cells
potential for application in photocatalysis, photovoltaic devices, (DSSCs) whose performance is largely dependent on the optimal
and carrier for metallic nanoparticles [4]. Titania minerals naturally properties of the semiconductor layer acting as charge separator.
occur as three crystallographic phases namely anatase, brookite, Production of stable nanostructures of anatase with high sur-
and rutile [5]. Anatase is considered to be more important than face area and porosity is of great interest for researchers to take
rutile because it bears better photocatalytic activity due to rel- DSSCs to higher efficiencies with all other advantages. Commercial
atively higher bandgap (3.2 eV) than rutile [6]. There is a great nanopowder P-25 by Degussa is commonly used for DSSCs but tita-
interest in the development of synthetic methods to obtain TiO2 nia prepared by sol gel methods is found to give better performance
nanoparticles with phase specificity, well-defined morphology, and [21].
high surface area [7]. Generally, it is difficult to produce phase-pure In this paper, we present results from low-temperature synthe-
TiO2 nanoparticles [4] as a mixture of phases has been reported sis of anatase TiO2 using reflux condensation in sol–gel method.
This is the first time that for the synthesis of TiO2 , reflux conden-
sation method was employed in association with sol–gel synthesis
∗ Corresponding author. Tel.: +92 333 556 1974; fax: +92 51 9085 5002. as a viable alternative to autoclave treatment. Acetic acid is used to
E-mail addresses: sofi[Link]@[Link] (S. Javed), mujahids@[Link] modify the precursor titanium tetra-Isopropoxide (TTIP) to have a
(M. Mujahid), [Link]@[Link] (M. Islam), [Link]@[Link] controlled sol gel synthesis [22], then, refluxing for different times
(U. Manzoor). was carried out to obtain nanostructured TiO2 with different char-
1
[Link]
2 acteristics. The powders so obtained were used to make paste for
[Link]
3
[Link] subsequent film deposition using doctor blade method. This study
0254-0584/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/[Link].2011.11.062
510 S. Javed et al. / Materials Chemistry and Physics 132 (2012) 509–514
Fig. 1. X-ray diffraction spectra of annealed powders from reflux times of 24 h (A),
70 h (B), and 120 h (C).
resulted in mesoporous TiO2 films with very high surface area that
can be exploited for more efficient charge transfer in DSSCs. The Fig. 2. TEM image of sol gel reflux synthesized anatase with crystallite sizes around
correlation between refluxing and the resulting nanostructures is 6 nm.
also explored.
(JEOL 2100). The phase and thermal analysis was done using
2. Experimental X-ray diffraction (Stoe D-64295 Darmstadt) and thermogravimet-
ric analysis/Differential thermal analysis (TGA/DTA, Perkin Elmer)
2.1. Synthesis of titania nanoparticles techniques. Surface area measurements of TiO2 films were made
using BET method.
All chemicals and materials obtained from Sigma–Aldrich and
Merck were of analytical grade and were used without further
3. Results and discussion
purification. 12 g (0.2 mol) of glacial acetic acid was added to
56.852 g of Ti (IV) tetraisopropoxide (TTIP) under stirring at room
Reflux condensation with sol gel method seems to be an easy
temperature. After 15 min 290 ml of distilled water was added as
low temperature procedure to obtain stable phase pure anatase
quickly as possible with vigorous stirring. An instant white precip-
nanostructures with huge surface area. The method is adopted
itate resulted. The mixture was stirred for 1 h to ensure complete
from Ito et al. [17] till peptization. Afterwards instead of autoclave,
hydrolysis. Then 4 ml of concentrated HNO3 was added and the
reflux condensation is applied and reported for the first time. As
mixture was heated from room temperature to 80 ◦ C within 40 min
we know, sol–gel processes involve metal alkoxides molecules
and peptized for 75 min. Distilled water was added to make the
that undergo hydrolysis and subsequently react to condense
volume 370 ml. The resulted sol was then subjected to refluxing at
with each other to form metal–oxygen–metal bridging units. The
100 ◦ C for different durations i.e. 24 h, 70 h and 120 h.
modification of titanium tetraisopropoxide with acetic acid results
The gel obtained was exposed to ultrasonics (Cole-Parmer-
in a controlled sol–gel process and it also affects the growth of the
08895-46 ultrasonic bath) for 20 min. The pH was kept in the range
of 1–2. Water was removed form the gel using rotary evaporator
and ethanol was added to make 40 wt% TiO2 gel. The gel was dried at
60 ◦ C and annealed in air at 450 ◦ C for 40 min in Nabertherm Gmbh
N17/HR-400V Muffle furnace. The heating rate was kept 1.8 ◦ C per
minute.
2.2. Thin films of titania for DSSC
TiO2 gel, PEG and acetyl acetone (3:3:1), ethanol absolute, and
glacial acetic acid were ball milled for 5 days. A few micron thick
films over plain glass and transparent conducting oxide (TCO)
coated glass were deposited by doctor blading. The films were
annealed in a muffle furnace at 450 ◦ C for 40 min at a heating rate
of about 1.8 ◦ C per minute.
2.3. Material characterization
The structural and morphological characterization of TiO2 pow-
ders and films was carried out using scanning electron microscope
(SEM) (JEOL JSM6460) and transmission electron microscope (TEM) Fig. 3. SEM image of as prepared sample A (24 h refluxed).
S. Javed et al. / Materials Chemistry and Physics 132 (2012) 509–514 511
Fig. 4. SEM image of as prepared sample B (70 h refluxed) (a) at 400× (b) at 2700×.
The gel obtained after refluxing at 100 ◦ C consisted of pure
anatase titania as revealed by XRD spectra of the samples A (24 h
refluxed), B (70 h refluxed) and C (120 h refluxed). XRD after anneal-
ing at 450 ◦ C also represents anatase peaks (Fig. 1) i.e. there is
no phase change. Due to the inherent nature of sol gel process-
ing route, certain degree of organic residues maybe retained in the
powders. It is speculated, however, that the annealing treatment
results in burnout of the organic content. Although all major peaks
with intensities beyond a certain threshold are indexed to be due to
anatase phase, a significant level of background noise in the spec-
tra maybe attributed to the nanocrystalline nature of the powder.
From Scherer equation (t = 0.94 /Bcos, where t is crystallite size,
B is the peak width at half maximum and being the diffraction
angle), the average crystallite size was calculated to be <10 nm for
all refluxing times (Table 1).
The findings of crystallite size from XRD data analysis were
found to be in agreement with the size of crystallites observed dur-
ing high-resolution transmission electron microscopy (HR-TEM)
Fig. 5. SEM image of as prepared sample C (120 h refluxed). examination of the powder particles. A representative TEM micro-
graph of one of the samples is shown in Fig. 2 where the crystallites
in powder particles are revealed using lattice imaging. The crys-
resulting product [22,23]. In this work the difference in durations
tallites within individual particles can be noticed to be mainly
of refluxing at 100 ◦ C has resulted in different morphologies of the
spherical in shape, as indicated by the dotted circles and marked
obtained product. Both refluxing time and precursor modification
diameters in Fig. 2. The size, as estimated from crystallites seen
have played their role in determining the product morphology. It
in the TEM image, is in the range of 5–10 nm. Whereas the size
is speculated that the difference in the thermal energy provided
obtained from Scherer equation is based on XRD peak with maxi-
during refluxing for the three kinds of samples has driven the
mum intensity among all peaks present, HR-TEM micrograph offers
hydrolysis reactions between isopropyl and acetate groups at
a visual examination of the crystallites and supports the results
different rates to give the obtained morphologies. Besides, we have
produced from XRD studies.
obtained the anatase phase only.
Fig. 6. Schematic for evolution of spherical morphology with increased refluxing time.
512 S. Javed et al. / Materials Chemistry and Physics 132 (2012) 509–514
Table 1
BET surface area and crystallite sizes (from XRD plots) of samples A–C.
Sr. no. Reflux time (h) Sample ID Crystallite size Crystallite size Surface area
(nm) before (nm) after (m2 g−1 )
annealing annealing
1 24 A 5.4 7.8 160
2 70 B 6.0 9.6 150
3 120 C 5.2 7.3 125
Fig. 7. TG Analysis for (70 h refluxed) sample B.
Refluxing time and acetic acid modification of TTIP has a strong
influence on the morphology and distribution of the structures as
revealed by SEM micrographs. For 24 h refluxing rod like structures
(Fig. 3) are grown which through a mixture of rod like and spher-
ical particles on refluxing for 70 h (Fig. 4a and b) have evolved to
spherical particles on 120 h refluxing (Fig. 5).
This can be explained by having an insight into the chemistry
of the process. We know that sol gel synthesis involve hydrolysis
and condensation of the precursor to form metal oxygen bonds
(Ti–O–Ti) in the final product. Acetate modified TTIP gives rise
to dimers, trimers or chains of various lengths [22] by conden-
sation between acetate and isopropyl ligands making O bridges
between two molecules. It is speculated that for short refluxing
times, growth of rod-like structures occurred due to enhanced
growth along longitudinal directions due to this reaction. Then dur-
ing longer refluxing times the other ligands also get hydrolysed
and condensed. Rods break due to higher thermal energy provision
causing the transition from rod-like to spherical particle morphol-
ogy (Fig. 6).
Thermal stability of the anatase phase was checked up to 700 ◦ C
by testing dried titania pastes in TGA/DTA using air medium. It
was observed that there is no phase change up to this tempera-
ture. Since all samples exhibited similar behaviour during TGA/DTA
studies, the result from only sample B is presented here. The con-
tinuous mass loss from room temperature to 400 ◦ C is attributed
to the loss of added binder and additional water present as impu-
rity in the material (Fig. 7). The graphical form of the data indicates
that at ∼550 ◦ C, the net weight loss was approximately 65% which
is comparable to the initial loading of TiO2 in gel. Studies have
revealed that PEG pyrolysis and decomposition occurs at temper-
atures around 400 ◦ C [24,25]. Since TGA/DTA is carried out under
air, the decomposition of organic binder is believed to take place
at same temperatures during annealing as indicated by TGA. In this
regard, the conditions selected for our study are expected to ensure Fig. 8. SEM images of annealed thin films obtained from sol–reflux–gels A (a), B (b)
almost complete decomposition of PEG during annealing. and C (c).
S. Javed et al. / Materials Chemistry and Physics 132 (2012) 509–514 513
Fig. 9. Schematic for morphology changes in films from as synthesized samples in effect of ball milling.
Thin films produced after 5 days milling of titania gel with binder In general, ball milling resulted in reducing the size of the struc-
(PEG) and acetyl acetone were examined topographically using tures besides acetyl acetone and binder PEG play their roles of
optical microscopy, SEM (before and after annealing) and AFM. The surface modification and binding respectively. The annealing has
films were annealed at 450 ◦ C, a temperature value chosen on the generated nanoporosity in the film structures upon decomposition
basis of TGA data that indicated almost no change in mass beyond of the binder PEG.
that (Fig. 7). Although the milling time can be further optimized, The surface area values using BET method were obtained for
a relatively high PEG content with respect to the TiO2 powder is the annealed paste powders, as listed in Table 1. The surface
beneficial, as it not only acts as a binder for improved mechanical areas can be predicted from Figs. 8 and 9 to be in the order of
stability and pore-forming agent [26], it also improves DSSC per- A > B > C depending upon the morphology of the starting structures
formance through formation of a high quality film with flat, even and changes brought about by surface modification, size reduc-
surface and a higher degree of dye adsorption [27]. tion during milling and annealing (binder removal and sintering
The surface morphologies of the thin films produced via paste effects). This is consistent with the values obtained i.e. 160, 150
coating are presented in Fig. 8. A schematic representation of the and 125 m2 g−1 for films A, B and C respectively. The mesoporosity
effect of milling and the binder over morphology of the films A–C and high roughness of A films justifies relatively higher value than
is given in Fig. 9. These nanostructures obtained after annealing B films which have a compact structure and poorer porosity. The
have interesting morphologies which retain the effect of varia- C films have least value among the three because of the annealing
tion in refluxing time. Film A resulted in mesoporous structures effects which result in compaction of the nanoparticles. This range
consisting of agglomerates with an average size of ∼500 nm. As of surface area values is highly desirable for the use of such powders
evident in Fig. 8a, these spongy structures are well separated in DSSC.
and have pores with typical diameter of the order of <50 nm.
From B, nanostructures for whom the reflux time was 70 h, the 4. Conclusion
mixed rod like and spherical particle morphology of nanostruc-
tures led to a different film surface microstructure after action We have reported a low temperature refluxing process for the
of milling and annealing effects. As shown in Fig. 8b, the film production of nanocrystalline phase pure titania with crystallite
reveals nearly round agglomerates, 200 nm in size. Also, these size of less than 10 nm. The refluxing time and precursor (TTIP)
agglomerates can be seen to have smaller secondary particles modification (with acetic acid) controls the morphology of the
packed together. The average surface roughness of the film is nanostructures and we can obtain rod like or spherical particles
significantly less with reduced degree of porosity. Similarly the or a mixture of both with change in refluxing time. Change in
film produced from ‘C’ gel consists of well distributed nanosized thermal energy and degree of hydrolysis (of acetate groups) dur-
particles with high surface area and good porosity in the film ing refluxing is considered to bring these structural changes. The
(Fig. 8c) in effect of milling of the relatively larger spherical par- structural features of thin films obtained after milling of the nanos-
ticles. Annealing resulted in porosity due to removal of organic tructures in presence of acetyl acetone also depict the effect of
content. Beside that, the nanoparticles seem to be interconnected to different refluxing durations. From rod like structures after 24 h
allow proper conduction of the current, when used as photoanode refluxing, mesoporous particles of around 500 nm have evolved.
in DSSC. And films obtained from nanostructures after 120 h refluxing
514 S. Javed et al. / Materials Chemistry and Physics 132 (2012) 509–514
consist of spherical particles around 50 nm diameter. So refluxing [8] M. Yan, F. Chen, J. Zhang, M. Anpo, J. Phys. Chem. B 109 (2005) 8673–8678.
time has a great influence on the morphology and consequently on [9] M.I. Zaki, A.H. Gamal, Mekhemer, N.E. Fouad, C. Tushar, Jagadale, S.B. Ogale,
Mater. Res. Bull. 45 (2010) 1470–1475.
other properties like surface area, porosity and roughness of the [10] T. Moritz, J. Reiss, K. Diesner, D. Su, A. Chemseddine, J. Phys. Chem. B 101 (1997)
anatase nanostructures. The thin films obtained for 24 h and 120 h 8052–8053.
refluxed samples have got nice properties in terms of surface area, [11] Z. Zhang, C.C. Wang, R. Zakaria, J.Y. Ying, J. Phys. Chem. B 102 (1998)
10871–10878.
morphology and porosity characteristics. These thin films have a [12] H. Uchida, S. Hirao, T. Torimoto, S. Kuwabata, T. Sakata, H. Mori, H. Yoneyama,
potential to be used in photoanodes of DSSCs. Langmuir 11 (1995) 3725–3729.
[13] X. Ding, Z. Qi, Y. He, J. Mater. Sci. Lett. 14 (1995) 21–22.
[14] S. Nishimoto, B. Ohtani, H. Kajiwara, T. Kagiya, J. Chem. Soc.-Faraday Trans. 1
Acknowledgements 81 (1985) 61–68.
[15] I. Moriguchi, H. Maeda, Y. Teraoka, S. Kagawa, Chem. Mater. 9 (1997)
We acknowledge National University of Sciences and Tech- 1050–1057.
[16] C.C. Wang, J.Y. Ying, Chem. Mater. 11 (1999) 3113–3120.
nology, Pakistan for providing funds for this work and School of
[17] S. Ito, T.N. Murakami, P. Comte, P. Liska, C. Grätzel, M.K. Nazeeruddin, M. Grätzel,
Material Science and Engineering, Nanyang Technological Univer- Thin Solid Films 516 (2008) 4613–4619.
sity for providing assistance in performing TEM. [18] S.M. Liu, L.M. Gan, L.H. Liu, W.D. Zhang, H.C. Zeng, Chem. Mater. 14 (2002)
1391–1397.
[19] H. Tsuchiya, J.M. Macak, L. Taveira, E. Balaur, A. Ghicov, K. Sirotna, P. Schmuki,
References Electrochem. Commun. 7 (2005) 576–580.
[20] R.K. Wahi, Y. Liu, J.C. Falkner, V.L. Colvin, J. Colloid Interf. Sci. 302 (2006)
[1] E. Serrano, G. Rus, J. Gartia-Martinez, Renew. Sustain. Energy Rev. 13 (2009) 530–536.
2373–2384. [21] C. Longo, M.D. Paoli, J. Braz. Chem. Soc. 14 (2003) 889–901.
[2] Y. Azizian-Kalandaragh, A. Khodayari, M. Behboudnia, Mater. Sci. Semicond. [22] P.B. Dunbar, J. Mater. Sci. 35 (2000) 367–374.
Process. 12 (2009) 142–145. [23] K.M.S. Khalil, T. Baird, M.I. Zaki, A.A. E1-Samahy, A.M. Awad, Colloids Surfaces
[3] N. Jagtap, M. Bhagwat, P. Awati, V. Ramaswamy, Thermochim. Acta 427 (2005) A: Physiochem. Eng. Aspects 132 (1998) 31–44.
37–41. [24] K.J. Voorhees, S.F. Baugh, D.N. Stevenson, J. Anal. Appl. Pyrolysis 30 (1994)
[4] Y. Ju-Nam, J.R. Lead, Sci. Total Environ. 400 (2008) 396–414. 47–57.
[5] Y. Li, T.J. White, S.H. Lim, J. Solid State Chem. 177 (2004) 1372–1381. [25] N. Grassie, G.A. Perdoma Mendoza, Polym. Deg. Stab. 9 (1984) 155–165.
[6] H. Yin, Y. Wada, T. Kitamura, S. Kambe, S. Murasawa, H. Mori, T. Sakata, S. [26] L. Ma, M. Liu, T. Peng, K. Fan, L. Lu, K. Dai, Mater. Chem. Phys. 118 (2009)
Yanagida, J. Mater. Chem. 11 (2001) 1694–1703. 477–483.
[7] J. Yang, H. Peterlik, M. Lomoschitz, U. Schubert, J. Non-Cryst. Solids 356 (2010) [27] K. Sakai, Y. Hirashita, T. Aihara, A. Fukuyama, T. Ikari, K. Kukita, S. Furukawa,
1217–1227. Thin Solid Films 519 (2011) 5760–5762.