The Journal oí Agriculture oí the university oí Puerto Rico

Published by T H E AGRICULTURAL EXPERIMENT STATION

OF T H E

UNIVERSITY OF PUERTO RICO.

Published Quarterly: January, April, July and October of each year.

VOL. XXVII JULY, 1943 NO. 3

R E L A T I O N O F M O I S T U R E C O N T E N T TO Q U A L I T Y O F
VANILLA BEANS

by

FRANCISCA E. ARANA and A. G. KEVORkiAN(i)

Puerto Rico Experiment Station
U. S. Department of Agriculture
Mayagiiez, P. R.

INTRODUCTION

Vanilla beans undergo during curing a series of chemical transfor-

mations which lead to the formation of flavoring compounds. (1, 3).
Curing consists usually (5, 0, pp. 84-117. 7) of an initial killing or wilting
treatment which arrests the natural vegetative processes and hastens chem-
ical changes. After this treatment, the beans are usually subjected to heat-
ing and subsequent sweating in blankets or else continuous heating. The
process is completed by further partial drying at room temperature and
finally aging or conditioning in wooden boxes.
Curing is accompanied by a loss of moisture which starts during
sweating and continues throughout the drying and conditioning periods.
The moisture left in the beans influences their appearance and flexibility.
A relatively high percentage of moisture is desirable and produces higher
returns as the beans are sold by weight. I t is of importance, therefore, that
the curer of vanilla controls moisture losses during curing.
Figure 1 shows some of the visible differences between freshly har-
vested and cured beans of Vanilla fragrans (Salisb.) Ames, the vanilla
of commerce.

(i) Formerly Assistant Plant Pathologist and Physiologist of this Static»

amd now Agriculturist in the Office of Foreign Agricultural Relatioms, assigned
to Nicaragua.
 10(¡ JOUKNAL OF AGKICTTLTUKE

••
j
I

Iff
m
M
Él
Ti
i
•
;¿j¡| d

I
tiltil -
•Mil yjyf
• I

Figure 1. Vanilla beans; one at left, freshly harvested and two at right,
cured. The white marks on the surface of the cured bean at the
... . extreme right are crystals of vanillin, the most widely known of
the aromatic constituents of vanilla.
 , ...., .,. , .MOISTÜIiR. OpH$ÉNT,;'.ÜNT. 'yÁNJLLA,, 107
. . MOISTURE CONTENT OF UNCTTREI) BEANS • •

Maturation ,is indicated in uncured vanilla beans by the development

of a. yellow coloration at the blossom-end. As ripening increases, this colo-
ration spreads along the entire bean and is often accompanied by longitu-
dinal splitting. Fully mature beans finally acquire a chocolate color.
Analyses were made to determine variations in moisture content of
beans of different maturities. Beans 6.5 to 7.5 inches in length, harvested
about 5 days previously, were graded-according to maturity' as whole
beans, entirely green ; whole; blossom-end yellow; split, blossom-end yellow';
and'split, entirely chocolate. >s From each grade of beans three, random
samples were selected. Moisture determinations were made by vacuum
oven-drying at 68° to 70° C. at 100 mm. pressure. {2. p. 336) . ,-
Whole beans, entirely green were found to have an average moisture
content of 81.2 percent, whole beans with blossom-ends yellow 79.2; split
beans with blossom-ends yellow 79.0, and split beans all chocolate 73.5.
The moisture content of split beans varies to a large degree with the ex-
tent and time of splitting. The percentage of moisture in the beans thus
decreased with maturity.

LOSSES I N •W E I G H T DURING CURING

A study was made of vanilla beans subjected to different killing

treatments to determine whether a significant weight loss other than that
due to loss of water occurs during the chemical transformations which
take place during curing.
The beans used were 6.5 to 7.5 inches long and were at different
stages of maturity. The treatments consisted of freezing for 4 and 20-
hour periods, three 10-second immersions at 30-second intervals in water
heated to 80° C, exposure to ethylene gas for different periods at a con-
centration of 1 part to 10,000 parts of air, rubbing with alcohol, exposure
to the sun, and placing in an oven at 65° O. These treatments, which were
applied singly and in combination, -were followed by sweating in an oven
or in the sun, further drying on racks at room temperature to the desired
loss in weight, and conditioning in closed wooden boxes for 3 months. The
loss in weight of the beans during curing was recorded and the amount
of moisture in the final, cured beans was determined. The results class-
ified according to maturities and the killing treatments used, are shown
in table 1.
 Table 1. Loss in weight and moisture content of vanilla beans of
five stages of maturity during curing after various killing
treatments (i)
WHOLE BEANS, BLOSSOM-END GREEN
Loss in
weight
Total loss Moisture
during cur-
in weight content
Killing treatment of cured I ing plus
during
I moisture
curing beans left in cured
beans (')

PERCENT PERCENT PERCENT

Frozen for 4 hours 74.2 26.6 81.1

Frozen for 20 hours 72.7 28.8 80.6
Frozen for 4 h o u r s ; hot water 72.8 26.8 80.1
Hot w a t e r ; frozen for 4 hours 72.5 26.5 79.8
Hot water . 75.6 20.4 80.6
Rubbed with alcohol; hot water 72.2 26.9 79.7
Ethylene gas for 4 14-hour periods 76.0 20.3 80.9
Ethylene gas for 14 hours . 75.1 19.8 SO.0

WHOLE BEANS, BLOSSOM-END YELLOW

Frozen for 4 h o u r s 72.9 27.9 80.5

Frozen for 20 h o u r s 70.2 27.8 79.0
Frozen for 4 h o u r s ; hot water __ 72.3 26.1 79.5
Hot w a t e r ; frozen for 4 h o u r s 72.1 28.5 80.1
Hot water 75.6 20.7 80.7
Rubbed with alcohol; hot water _. 71.4 26.7 79.0
Ethylene gas for 4 14-hours periods 74.9 21.3 80.3
Ethylene gas for 14 h o u r s 74.1 20.1 80.0

S P L I T BEANS, BLOSSOM-END YELLOW

Ethylene gas for 4 14-hour periods 73.4 20.8 78.9

Ethylene gas for 14 hours 73.6 21.1 79.2

SPLIT BEANS, BLOSSOM-END CHOCOLATE

Hot water 76.3

Ethylene gas for 4 14-hour periods 79.1
Ethylene gas for 14 hours 77.6

S P L I T BEANS, ALL CHOCOLATE

Sun 68.8 19.6 74.9

Oven at 65^ C. -- U9.4 19.8 75.5
1
(i) Three- to 4-pound lots of beans were used in each treatment, except in
split beans, all chocolate, in which 2-pound lots were used.
(2) Calculated by adding t h e total loss in weight d u r i n g curing to the water
present in t h e r e m a i n i n g percent weight of cured beans. These figures
represent approximately t h e average moisture content of these types of
uncured beans. T h u s weight loss in curing, other t h a n that due to loss
of water, w a s negligible.
 MOIST-UEE-. C O N T E N T I N VANILLA 109

I t can be observed in table 1 that regardless of the killing treatrnenl

ami the [Link] moisture content of the cured beans, the total loss in weight
during curing plus the water present in the cured product was from 79.1/
to 81.1 percent of the original weight for whole green beans, and from
"¡9.0 to 80.7 percent of the original weight for blossom-end-ye low beans.
These figures represent approximately the average moisture" content of
these types of uncured beans, as stated previously. I n t h e blossom-end
chocolate split beans the values for the- losses in weight during curing
plus the moisture_-of the cured beans" showed greater variation.' However,
this is to be'expected inasmuch as'the extent of splitting and the degree of
maturation of these types cause appreciable variations in the moisture
content. Therefore, it can be concluded that weight loss, other than that
due to loss of water, was negligible.
Loss rsr- WEIGHT DTTRING CONDITIONING
Moisture losses in curing vanilla occur principally during the sweating
and drying treatments. Weight losses also occur during the conditioning or
Table 2. Losses in weight on a fresh basis in vanilla
beans during the conditioning period (i)
During the first During the second
At the end
of drying 3-month period of 3-month period of
conditioning (2) conditioning (.2)

PERCENT PERCENT PERCENT

75.8 - 0.2 - 0.2
75.3 .2 .1
72.3 .8 - .3
72.2 1.7 .3
72.1 .1 .0
71.2 1.2 .1
70.9 2.0 - .4
70.5 3.1 -.6
70.4 2.3 - .4 . .
70.3 2.5 .4
70.0 1.7 .6
69.8 2.7 . . - .2
' 69.8 1.9 .3
69.4 3.5 • " • • ' .6
69.4 2.0 .3
69.3 1.8 .3
6S.9 2.9 .9
67.8 4.5 • 1.1
66.6 3.4 .9
63.3 6.7 1.0
62.0 5.9 1.3

(1) Three-pound lots of whole beans with blossom-end

either green or yellow were used.
(2) ^ negative loss of weight signifies an actual gain.
110 JOTIKKAL OF AQKICULTUKE

final phase of the processing during which vanillin crystallizes on the sur-
face of the beans and the characteristic vanilla aroma gradually develops.
The moisture losses during the conditioning treatment in 21 3-pound
lots of beans were determined. These samples had lost from 62.0 to 75.8
percent of their weight at the end of the drying period. The beans were
[Link] being placed in the conditioning boxes and at the end of
3 and 6 months. The data are listed in table 2. •
As shown in table 2, there was an appreciable loss in weight in most
of the lots during the first 3 months of conditioning, ranging from a gain
of 0.2 to a loss of 6.7 percent. The loss in weight was negligible, how-
ever, during the last 3 _months not reaching in any case more than 1.3 per-
cent. It is apparent that moisture content approached an aquilibrium
during conditioning and that the loss in weight during this period was
generally greatest in the beans containing' the most moisture.
NOMOGRAPH FOE M O I S T U E E CAICtHCATIONS
\
A nomograph is given in Figure 2 showing the weights to which
100-pound lots of beans of different moisture contents should be reduced
during curing to obtain a predetermined moisture content in the end-
product.
Example: To curelOO pounds of green beans to a desired final mois-
ture of 30 percent, assuming or knowing 80 percent moisture content of fresh
beans, run a straight line from point 80 on left axis to point 30 on right
inclined axis. Eead off on center -axis final weight of beans after curing,
including conditioning. To obtain the weight to which the beans should
be dried before conditioning, the loss during conditioning should be added
to that obtained from the chart. The loss during conditioning- depends
principally upon the moisture content of the beans at the beginning of
conditioning, the number and length of times the beans are examined,
and the amount of wiping to remove molds. However, for approximate
purposes, the following losses may be assumed: For30-40. percent final
moisture in the cured product, 4-7 pounds of 100 pounds fresh and for
20 to 30 percent moisture, 1-3 pounds.

EVAI/CTATION OF CUBED B E A N S W I T H D I F F E R E N T M O I S T U B E CONTENTS

An experiment was conducted to determine the variation in physical

characteristics, phenol value including vanillin (4), and aroma of vanilla
beans cured by different procedures to varying moisture contents. Data
are shown in table 3.
 MOISTTJKE CONTENT I N VANILLA 111

X
70 r -T 6 0
MOISTOU Ufnettl OF ' • MOlSTURS
»[Link] 100 LSS. . . ÍERCSMT
?HESH WESH BEAHS CUtttO
71 BEAHS ATTEft BEAMS
CURING

72 -

73

74

75

76 -

77

78

79 -

80

/lOO-x'j
81
--•5 y=ioo
Voo-z/
82 *-

Figure 2. Nomograph relating the three variables: X=moisture content of

fresh vanilla beans in percent fresh basis, Y=weight to which
each 100 lbs. should be reduced during curing, and Z=moisture
content of cured beans in percent cured weight basis. See example
given in the text. .*
 Table,3 Evaluation of vanilla beans cured to different m o i s t u r e contents (1)

Phenol
Killing Final Degrti Mold value Aroma
treatment moisture Chocolate 0/ Vanillin develop- dry e-i
<») content color flexibility crystals ment (s) basis O
PErKOENl PEKOENT I PERCENT
Hot water 26.5 Dark Medium Few 0.0 6.3 Developed, suave
Hot water 30.7 Dark Very None 0.0 6.3 Developed, not suave
Hot water 49.9 Light Very None 0.0 6.4 Crude, slightly fermented, n o t suave tr1
Freezing 26.2 Reddish. Very Some 0.0 6.8 More developed, suave
Freezing 32.6 Reddish Very None 0.0 6.7 Developed, very suave o
Freezing 50.5 Light Very None 0.0 6.9 Developed, slightly fermented, suave
Sun 26.5 Dark Medium None 0.0 6.7 More developed, suave
Sun 30.9 Light Very None 4.0 6.3 Developed, suave
Sun 53.0 Light Very None 7.0 6.9 More developed,slightly fermented, n o t suave
Ethylene 24.2 Dark Little Few 0.0 7.3 More developed, suave
Ethylene 33.6 Light Medium Few 3.0 7.5 More developed, suave O
54.3 Very
Ethylene Light None 3.0 6.9 Developed, slightly fermented, not suave
a1
ir
(i) Beans killed by h o t water and freezing were obtained from the s a m e batch and beans killed by sun and
ethylene from a n o t h e r batch. H
(2) Beans killed by h o t water, freezing a n d ethylene were oven-sweated a n d beans killed in t h e sun were
also sun-sweated. All t h e lots were t h e n dried further at room t e m p e r a t u r e to t h e desired loss in weight
"and conditioned for 3 months. *

(3) Beans examined for mold a t periods v a r y i n g from 4 to 9 days.

 MOISTURE' CONTENT I N VANILLA 113

""' As can be observed in table 3, the moisture content of the' beans

affected the physical, .appearance and aroma of the beans in each of the
different curing treatments but did -not produce any significant effect
in the phenol váliié. .
In all treatments, beans with ¡moisture contents varying from 50 to
54 percent, had a : fermented aroma ¡generally lacking suavity and develop-
ment. Beans containing 24 to 27 percent moisture had a more developed
and suave aroma .but little flexibility while beans with. 31 to 34 percent
- moisture had a Well developed and suave aroma^ and also "possessed a high
degree of flexibility. . '-•
Beans with lower moisture content were generally darker in color
and as expected - their vanillin crystallization was greater except in the
sun-cured beans in which no crystals appeared. The beans killed by freezing
were characterized by great flexibility even when their moisture content
was low and had a reddish color.
No appreciable molding occurred in any of the treatments when the
beans contained 24.2 to 26.5 percent moisture nor in the beans subjected
to hot water or freezing having 30.7 to 50.5 percent moisture. Beans
subjected to the ethylene and sun-killing procedures developed more mold
than did those subjected to hot water or freezing. This may be due to
-the fact that in the former procedures the beans were not as thoroughly
and uniformly killed. The effectiveness of the killing treatment is usually
evidenced by a change in coloration which takes place in the beans during
the sweating period or exposure to heat. The beans killed in hot water and
by freezing became dark brown in color after a shorter period of sweating
in electric ovens at 45°C., than the ethylene and- sun-treated beans. More-
over, freezing and hot water treatment may have a sterilizing effect in
the beans.

Moisture Content of Foreign Beans

Moisture contents were determined of several samples of beans from

Mexico, Comores, Tahiti, Madagascar, and Guadeloupe, which are among
the most important vanilla-producing countries of the world. Eesults are
shown in table 4.
114 JOURNAL OF AGRICULTURE

Table 4. Moisture content of cured vanilla beans from - different countries

Source Moisture content

PERCENT

Mexico 40.3
Mexico (i) 38.3
Comores . 35.4
Tahiti 37.5
Tahiti (i) 33.4
Madagascar 31.3
Guadeloupe (i) 17.8
Guadeloupe 15.2
(i) The moisture contents of these beans were determined
several months after receipt.
As can be seen in table 4, the moisture content was over 30 percent
in the beans of all the countries exeept those of Guadeloupe.

SUMMABY

The moisture content of xincured vanilla beans was found to decrease

with maturity. The-average value obtained for whole beans entirely green
was 81.2 percent and for whole, blossom-end yellow 79.2 percent. The
moisture of split beans varied to a large degree with the extent and time
of splitting.
Losses in weight in vanilla beans, other than those due to loss of
moisture, were shown to be negligible during the curing process, regard-
less of the procedure used.
Although the greatest loss in weight in the beans occurred during
the sweating and drying periods, a significant loss in weight also occurred
during the first 3 months of the conditioning period. Moisture losses during
conditioning were greater for beans of higher moisture content.
A nomograph is included showing the weights to which 100-pound
lots of beans of different moisture contents should be reduced during
curing to obtain a known moisture content in the end-product.
Moisture content was found to affect the appearance and aroma of
the beans but did not produce any effect in the phenol value. Beans cured
by different procedures with a moisture content of over 50 percent had a
fermented aroma generally lacking suavity and development. Beans with
24 to 27 percent moisture had a more developed and suave aroma but
little flexibility while those with 31 to 34 percent had well-developed arid
suave aroma and good flexibility. Beans with similar moisture content, sub-
 MOISTUEE CONTENT TÉ VANILLA Üfr
*
.jected to hot-water or freezing treatments were found to be less liable to
mold infection than sim or ethylene-treated beans.
Commercial lots of vanilla beans from Mexico, Madagascar, Comores,
and Tahiti were found to contain over 30 percent moisture.
Puerto Eican vanilla beans should be cured to a final moisture content
of 30 to 35 percent if good flexibility and development and suavity of
aroma, which are the principal factors affecting the sales value are to be
obtained, judging from the results of our experiments.

A clciiowledgments

Appreciation is expressed to ícLa Cooperativa de Cosecheros de Vai-

nilla de Puerto Pico" and to the Puerto Eico Peconstruction Adminis-
tration for the loan of vanilla beans used in these experiments. The authors
are also indebted to Mario Tomassini for help in the curing and to
Merriam Jones for the construction of the nomograph.
116 JOtJEiSTAL OF Á&EICÜLTÜEE

.' LlTERATUBE CITED

1. Ai'ana, Francisca E. Action of a B-glucosidase in the curing of vanilla.

| Food Res. 8:343-51, Jl '43.
2. Association of Official Agricultural Chemists. Official and Tentative
' Methods of Analysis, p. 336. Washington, D. C. 1940.
3. Balls, A. K. and Francisca E. Arana. The curing of vanilla. Indus, and
Engin. Chem., Indus. Ed. 33: 1073-1075, fig. 1. 1941.
4. and Francisca B. Arana. Determination and significance of
phenols in vanilla extract. Assoc. Off. Agr. Chem. Jour. 24: 507^512,
. fig. 1-2. 1941.
5. , Arthur G. Kevorkian and Francisca E. Arana. Process
for curing vanilla beans. Patent No. 2,274,120, U. S. Patent Office,
Feb. 24, 1942.
6. Chalot, Ch. and U. Bernard. Culture et preparation de la vanille. (Vanilla
culture and curing.) Bibliothéque du Jardín Coloniale. pp. 84-117, illus.
1920. Paris.
7. Kevorkian, A. G. Vanilla-processing studies. Puerto Rico (U. S.) Agr. Exp.
Sta. Rept. 1339: 14-22. 1940.