Powder Technology 291 (2016) 322–327

Contents lists available at ScienceDirect

Powder Technology

journal homepage: www.elsevier.com/locate/powtec

Moisture sorption isotherms and storage study of spray dried tamarind

pulp powder
Khalid Muzaffar ⁎, Pradyuman Kumar
Department of Food Engineering and Technology, Sant Longowal Institute of Engineering and Technology, Longowal 148106, Punjab, India

a r t i c l e i n f o a b s t r a c t

Article history: Moisture adsorption isotherms and storage study of spray dried tamarind pulp powder were evaluated in this
Received 16 September 2015 work. Adsorption isotherms of tamarind pulp powder were determined at four different temperatures (20, 30,
Received in revised form 4 December 2015 40 and 50 °C) using a gravimetric technique. The sorption isotherms were found to be typical type II sigmoid.
Accepted 29 December 2015
The experimental data obtained was fitted to several mathematical models viz. two-parameter (BET, Oswin,
Available online 30 December 2015
Smith, Caurie, and Iglesias and Chirife), and three-parameter (GAB) relationships. A non-linear least square re-
Keywords:
gression analysis was used to evaluate the model constants. The GAB followed by Oswin model best fitted the
Tamarind pulp powder experimental data. Changes in physicochemical properties of tamarind pulp powder were evaluated during stor-
Sorption isotherm age (at 0, 1, 2, 3, 4, 5 and 6 months), using three different packaging materials (low density polyethylene, LDPE;
Storage aluminum laminated polyethylene, ALP and glass). Color parameters, moisture content, titratable acidity, bulk
LDPE density and flowability of the powder varied to different extent during storage, depending on the type of pack-
ALP aging material used. Compared to other packaging materials, powder packed in LDPE showed considerable
Glass changes in physicochemical properties during storage. The magnitude of the change in physicochemical proper-
ties of the powder measured during storage suggests that glass is best for long term storage of tamarind pulp
powder.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction and purely empirical models (e.g. Oswin and Smith models). The
moisture sorption isotherms are unique for every material and
Water sorption isotherms are considered important thermody- must be evaluated experimentally.
namic tools to determine the interaction between water and food Fruit juice and pulp powders are valuable materials in terms of
components. They represent the relationship of the equilibrium transportation, packaging, storage and shelf life, compared with their
moisture content of a food product with the relative humidity of its liquid counter parts. Spray drying is one of the common technique
surrounding environment at a particular temperature and provide used for production of powders from liquid solutions and suspensions.
useful information for food processing operations such as drying, Tamarind pulp in powder form is one of the important tamarind prod-
packaging and storage [1,2]. A moisture sorption isotherm can be uct. However due to high amount of sugars and acids in tamarind
used to predict the amount of water that a material will hold if it is pulp, higher significant product loss occurs during spray drying because
exposed to air at a certain relative humidity. This moisture content of the stickiness of the powder. Hence to overcome the stickiness prob-
is dependent on the temperature and the environmental relative hu- lem, drying aids are added during spray drying of tamarind pulp [4,5].
midity, as well as on the composition of the material [3]. The sorption Besides this, physicochemical properties of spray dried powder are af-
isotherms are commonly presented by mathematical models based fected by the conditions used during powder production and storage.
on empirical and/or theoretical criteria. In the literature, a large In our previous study, effect of processing conditions on physicochemi-
number of isotherm models are available which can be categorized cal properties of spray dried tamarind pulp powder was studied [5],
into various groups; kinetic models based on an absorbed mono- however there is no investigation about the quality changes of tamarind
layer of water (BET model), kinetic models based on a multi-layer pulp powder during storage.
and condensed film (GAB model), semi-empirical (Halsey model) Thus, the objective of the present study was to provide experimental
data for sorption characteristics of spray dried tamarind pulp powder in
⁎ Corresponding author.
order to model the sorption isotherms using selected models and to
E-mail addresses: khalidmuzaffar76@gmail.com (K. Muzaffar), pradyuman@sliet.ac.in evaluate the changes in physicochemical properties of spay dried tama-
(P. Kumar). rind pulp powder during storage.

http://dx.doi.org/10.1016/j.powtec.2015.12.046
0032-5910/© 2015 Elsevier B.V. All rights reserved.
 K. Muzaffar, P. Kumar / Powder Technology 291 (2016) 322–327 323

2. Materials and methods placed in desiccators filled with saturated solution of Magnesium ni-
trate in order to provide a constant relative humidity of 53% during
2.1. Sample preparation the storage period. The desiccators were then stored at 25 ± 1 °C,
representing room temperature. The powder samples were periodically
Tamarind fruit pods were deshelled and soaked in water in the ratio analyzed (at 0, 1, 2, 3, 4, 5 and 6 months) for different physicochemical
1:2.5 under optimum conditions of 33 min soaking time and 39 °C properties (color, moisture content, acidity, bulk density and
soaking temperature [6]. The mixture was then homogenized and flowability).
sieved to separate fiber, rags and seeds from the pulp. The pulp was
passed through three layers of muslin cloth to obtain fine pulp. 2.4.1. Color measurement
The color of the powder samples was determined by using a color
2.2. Spray drying of tamarind pulp spectrophotometer (CM-3600d, Konica Minolta). The results were
expressed in terms of Hunter color values of L*, a*, and b*, where L* de-
A tall type laboratory scale spray dryer (S.M. Scientech, Calcutta, notes lightness/darkness, a* redness/greenness, and b* yellowness/
India) with cocurrent regime (flow of feed spray and drying air in blueness.
same direction) and a two-fluid nozzle (inside diameter of 0.5 mm) at-
omizer was employed for spray drying process. Feed was metered into 2.4.2. Moisture content
the dryer by means of a peristaltic pump. Based on our previous study, The moisture content (%) of the powder samples was determined
the derived optimum conditions for spray drying of tamarind pulp according to AOAC method [9]. About 2 g of the powder sample was
were 25% carrier (soya protein isolate) concentration, 170 °C inlet air taken in a petriplates and dried in a vacuum oven at a temperature of
temperature and 400 ml/h feed flow rate [5] Feed temperature, com- 70 °C until a constant weight was obtained. The samples were analyzed
pressor air pressure and blower speed were kept at 25.0 ± 0.5 °C, in triplicates and the mean was recorded.
0.06 MPa and 2300 rpm, respectively. After the completion of the exper-
iment, the powder was collected from the cyclone and cylindrical parts 2.4.3. Titratable acidity
of dryer chamber by lightly sweeping the chamber wall as proposed by The titratable acidity (%) of the powder sample was determined by
Bhandari et al. [7]. titration with standardized 0.1 N NaOH to the phenolphthalein end
point, according to the method described by Rangana [10]. Titratable
2.3. Determination of sorption isotherms acidity was measured in terms of tartaric acid. The samples were ana-
lyzed in triplicates and the mean value was calculated.
Sorption isotherms were determined by the gravimetric method.
Eight saturated salt solutions were prepared in order to provide differ- 2.4.4. Bulk density
ent relative humidity values. The salt solutions used and the corre- For the determination of loose bulk density a known quantity of
sponding relative humidities at different temperatures are given in powder sample was freely poured into a 10 ml graduated cylinder
Table 1 [8]. Triplicate samples (1 g) of freshly prepared spray dried tam- (readable at 0.1 ml) and the volume occupied was noted and then
arind pulp powder were placed in previously weighed alumunium used to calculate bulk density (weight/volume).
dishes. The samples were then kept in desiccators over the saturated
salt solution of known relative humidity. The desiccators were placed 2.4.5. Powder flowability
in temperature-controlled cabinets maintained at 20, 30, 40 and 50 °C Powder flowability was measured in terms of cohesive index by
(±1 °C) and the samples were allowed to equilibrate until there was using a Powder flow analyzer attached to a texture analyzer (Stable
no distinct weight change (±0.0001 g). A test tube containing thymol Micro Systems, UK). A fixed powder volume of 25 ml was poured into
was placed inside the desiccators with high relative humidity to prevent the cylindrical vessel of the analyzer prior to testing. During the test,
mold growth during storage. The required time period for equilibration the blade of the analyzer took a downward and then upward movement
was about 3–4 weeks. The total time for removal, weighing, and putting for three cycles inside the cylinder, corresponding to three compaction
back the sample in the desiccator was about 30 s. This minimized the and decompaction phases. The force–displacement curve was thus gen-
degree of atmospheric moisture sorption during weighing. The equilib- erated by the system, exhibiting the force exerted on the cylinder bot-
rium moisture content was determined in a vacuum oven, at 70 °C until tom due to blade movement and powder displacement. A cohesion
constant weight was obtained [9]. The measurements were recorded as coefficient (g.mm) was derived by Texture Exponent software (Stable
the mean of triplicates samples. Micro System, UK) through integrating the negative area underneath
the curve during the decompaction cycle. The cohesion index (mm)
2.4. Storage study was defined as the ratio of the negative area under force displacement
curve to the sample weight.
Freshly prepared tamarind pulp powder samples (15 g) were
packed in three different packaging materials (low density polyethyl- 2.5. Sensory evaluation
ene, LDPE; aluminum laminated polyethylene, ALP and glass) and
A panel comprising of seven trained judges did sensory evaluation of
Table 1 tamarind pulp powder after a regular interval of one month storage pe-
Relative humidities of selected saturated salt solutions at different temperatures. riod by using 9-points hedonic scale scorecard.
Salt Relative humidity (%)
2.6. Analysis of sorption data
20 °C 30 °C 40 °C 50 °C

Lithium chloride 11.3 11.3 11.2 11.1 The equilibrium moisture content of the powders for each tempera-
Potassium acetate 23.1 21.6 20.8 20.4
ture was plotted against the corresponding water activity (relative
Magnesium chloride 33.1 32.4 31.6 30.5
Potassium carbonate 43.2 43.2 40.0 38.5 humidity/100) to produce the sorption isotherms. Six different mathe-
Magnesium nitrate 54.4 51.4 48.4 45.4 matical models presented in Table 2 were used to fit to the experimental
Potassium iodide 69.9 67.9 66.1 64.5 data using regression analysis. The curve fitting and regression analysis
Sodium chloride 75.5 75.1 74.7 74.4 were performed using Statistica. V.8. (Statsoft, India Pvt. Ltd. New
Potassium chloride 85.1 83.6 82.3 81.2
Delhi). The goodness of the fit of each model was evaluated in terms
324 K. Muzaffar, P. Kumar / Powder Technology 291 (2016) 322–327

Table 2 2.7. Statistical analysis

Moisture sorption isotherm models used for experimental data fitting.

Name of the model Mathematical expression The experimental data for physico-chemical properties of the pow-
GAB [26] 0 :C:K:aw
der was expressed as the mean value ± standard deviation. The exper-
M ¼ ð1−K:awMÞð1−K:aw þC:K:aw Þ
M0 :C:aw
imental data was further analyzed using two way ANOVA. The factors
BET [27] M ¼ ½ð1−aw ÞþðC−1Þ: ð1−aw Þ: aw  (and levels) were packaging material (LDPE, ALP and glass) and storage
Oswin [28] M = A . (aw/1 − aw)B
time (0, 1, 2, 3, 4, 5 and 6 months). The ANOVA included the main effects
Caurie [30] M = exp (A +B .aw)
Smith [29] M = A +Blog(1 − aw) of packaging material, storage time and their interaction.
Iglesias and Chirife [31] M = A + B . (aw/1 −aw)
3. Results and discussion

3.1. Sorption isotherm

of minimum standard error of estimate (SE), minimum percent root
mean square error (%RMSE), minimum mean absolute percentage Moisture sorption behavior of spray dried tamarind pulp powder at
error (P), and the maximum adjusted R-square, (R2Adjusted) and were four different temperatures (20, 30, 40, and 50 °C) is described by Fig. 1.
calculated as follows: The sorption isotherms revealed an increase in equilibrium moisture
content with increase in water activity at a constant temperature, a
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u N characteristic feature of amorphous materials rich in hydrophilic com-
uX
SE ¼ t ðY i −Y i 0Þ2 = F ð1Þ ponents. This behavior can be attributed to the hydrophilic nature of
i¼1 carbohydrates and protein present in spray dried tamarind pulp pow-
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi der. At low and intermediate water activities, the so-called multilayer
u N sorption region, equilibrium moisture content increases linearly with
u1 X
%RMSE ¼ t ðY −Y i 0=Y i Þ2  100 ð2Þ water activity, whereas at high water activities, the so-called capillary
N i¼1 i
condensation region, a sharp increase in equilibrium moisture content
was observed [11,12]. This trend has been reported in the literature
100 X
N
for many food materials including raisins, figs, apricots, prunes [13],
P¼ ðY −Y i 0=Y i Þ ð3Þ
N i¼1 i grapes, apples [14], tomato pulp powder [11] and orange juice powder
[15]. Experimental data indicated that the equilibrium moisture content
XN
W i Y i −Y i 0Þ2 ðN−1Þ decreased with increasing temperature, at a constant water activity.
R2Adjusted ¼ 1− XNi¼1 ð4Þ This behavior may be due to the decrease in the total number of active
i¼1
W i Y i −Y″Þ2 ðN−MÞ sites for water binding as a result of physical and/or chemical changes
induced by temperature. At increased temperatures water molecules
Where get activated to higher energy levels and break away from the water-
binding sites of the food, thus decreasing the equilibrium moisture con-
Y experimental value of equilibrium moisture content; tent [16].
Y′ predicted value of equilibrium moisture content; The results of regression analysis for fitting the experimental data to
Y″ mean of experimental data of equilibrium moisture content six different models along with standard error of estimate (SE) on EMC,
N number of observations; percent root mean square error (%RMSE), mean absolute percentage
F degree of freedom of the regression model; error (P), and adjusted R-square (R2Adjusted) are presented in Table 3. In
W weighting applied to each data point, which was set to unity terms of the mean absolute percentage error (P) of the fits, those with
in these analyses; and less than 10% error can be considered acceptable [17] and an R2Adjusted
M number of coefficients in each equation. of N 0.98 may also indicate a good fit. Regarding the BET model only

Fig. 1. Adsorption isotherms of spray dried tamarind pulp powder at various temperatures.
 K. Muzaffar, P. Kumar / Powder Technology 291 (2016) 322–327 325

Table 3
Estimated parameters of isotherm models for spray dried tamarind pulp powder at different temperatures.

Goodness of fit parameters Model coefficients/constants

Model Temperature (°C) SE⁎ %RMSE⁎⁎ P (%) Adjusted-R-Square A B C

GAB 20 0.624 3.577 3.338 0.9956 7.7031, Mo 6.5691, C 0.9375, K

30 0.682 4.476 4.221 0.9931 7.2252, Mo 6.4869, C 0.9357, K
40 0.639 4.527 3.939 0.9932 6.3523, Mo 8.1833, C 0.9619, K
50 0.435 1.642 3.451 0.9962 5.6038, Mo 8.3491, C 0.9875, K
BET 20 0.245 3.824 3.620 0.9999 6.7714, Mo 9.0380, C –
30 0.234 3.695 3.477 0.9983 6.1819, Mo 9.9163, C –
40 0.269 5.251 4.514 0.9982 5.6897, Mo 12.3717, C –
50 0.136 2.111 1.854 0.9995 5.1050, Mo 12.2353, C –
Oswin 20 0.714 6.093 5.406 0.9952 12.4952 0.6147 –
30 0.741 6.761 6.110 0.9932 11.6746 0.6120 –
40 0.777 9.350 7.742 0.9913 11.0874 0.6277 –
50 0.699 11.307 8.298 0.9928 10.2142 0.6622 –
Caurie 20 1.171 9.733 7.508 0.987 0.9779 3.0348 –
30 0.931 8.207 6.002 0.9892 0.946 2.9429 –
40 0.952 9.541 6.851 0.9870 0.914 2.9519 –
50 1.685 13.016 10.265 0.9889 0.7786 3.0669 –
Smith 20 1.375 12.163 10.273 0.9826 0.8582 −17.75 –
30 1.128 11.310 9.807 0.9842 1.028 −16.1837 –
40 1.207 12.700 10.959 0.9791 0.9959 −15.4952 –
50 1.191 15.075 11.951 0.9792 0.6030 −14.9805 –
Iglesias and Chirife 20 1.704 22.736 12.995 0.9726 5.5210 5.7694 –
30 1.485 19.450 11.766 0.9726 4.9196 5.5737 –
40 1.072 13.345 8.380 0.9835 4.3505 5.6264 –
50 0.656 5.456 9.525 0.9889 3.5888 5.6588 –
⁎ Standard error of estimate.
⁎⁎ %RMSE = percent root mean squared error, p; mean absolute percentage error.

experimental data with aw b 0.50 was fitted to the BET equation, because primary binding sites on the product surface. The larger the value of C,
above that value the model hypothesis fails and the equation is not able the stronger the bonds between water molecules in the monolayer
to predict sorption behavior accurately. and the binding sites on the surface of the sorbent. The parameter ‘K’
Table 3 shows that GAB model gives the minimum values for SSE, is a correction factor for multilayer molecules relative to the bulk liquid;
RMSE and P and the maximum R2Adjusted when used to predict the mois- when K = 1 the molecules beyond the monolayer have the same char-
ture sorption isotherm of tamarind pulp powder at different tempera- acteristics as pure water. ΔHc = the difference between the heat of
tures. So the GAB best describes the experimental adsorption data sorption for monolayer of water and the heat of sorption for multilayer
through the entire range of water activity. This observation is similar water, and ΔHk = the difference between the latent heat of condensa-
to that obtained by Goula et al. [11] and Sormoli and Langrish [15] tion of pure water and the heat of sorption for multilayer water. ΔHx
who studied sorption isotherms of tomato pulp powder and orange is a constant parameter to express the temperature dependence of the
juice powder respectively. After GAB model, smith model also ade- monolayer moisture [19].
quately fitted the experimental data of sorption showing minimum GAB model is based on the monolayer moisture concept and pro-
values of SE, RMSE and P. This suggests that GAB and Oswin sorption vides the value of monolayer moisture content of the material, consid-
equations could best describe the adsorption behavior of tamarind ered as the safe moisture for dried foods during preservation, while
pulp powder. The shape of moisture sorption isotherms (Fig. 1) and most other models lack this parameter. The monolayer moisture con-
the value for model parameter, C N 1 of the BET equation (Table 3) tent (Mo) indicates the amount of water that is strongly adsorbed to
imply that the moisture sorption isotherms for tamarind pulp powder specific sites at the food surface and is considered as the moisture con-
are sigmoid and based on the classification of isotherms given by tent affording the longest time period with minimum quality loss at a
Brunauer et al. [18] belong to type II class of isotherms. given temperature [11].
The fit of GAB equation to the water vapor sorption data of dried
3.2. Parameters of GAB equation tamarind pulp powder was performed by non-linear regression analysis
of the standard three parameter GAB equation (Eq. (5)) and the three
The GAB equation and the definition of its three parameters are parameters of the GAB equation (Mo, C and K) were determined at
given in Eqs. (5)–(8) each temperature (Table 3). Then the constants Co, Ko, M′, ΔHc, ΔHk
and ΔHx were calculated by using a successive regression of
M0 :C:K:aw Eqs. (6) to (8) (Table 4). The values of the monolayer moisture contents
M¼ ð5Þ
ð1−K:aw Þð1−K:aw þ C:K:aw Þ (Mo) decreased slightly as the temperature increased, whereas the pa-
rameter K and C first slightly decreased at 30 °C and then increased. A
C ¼ C 0 expðΔH c =RT Þ ð6Þ

Table 4
K ¼ K 0 expðΔHk =RT Þ ð7Þ
GAB constants calculated by linear regression of GAB parameters and
reciprocals of absolute temperature.
M0 ¼ M0 expðΔH x =RT Þ: ð8Þ
C0 8.44
ΔHc (J/mol) 3241.63
Where ‘M0,’ known as monolayer moisture content is the amount of K0 1.25
water that is adsorbed in a monolayer on the surface of the adsorbent ΔHk (J/mol) −620.31
and is a measure of the availability of active sorption sites. The parame- Mʹ (% dry basis) 0.54 or 5.4 × 10−1
ΔHx (J/mol) 3697.57
ter ‘C’ determines the strength of binding for water molecules to the
326 K. Muzaffar, P. Kumar / Powder Technology 291 (2016) 322–327

similar trend for GAB parameters was also observed by Sormoli and Table 5. The data shows that except for the powder samples filled in
Langrish [15] for orange juice powder. According to Lewicki [20] in glass there was a gradual increase in moisture content of the samples
order to represent a sigmoid type of sorption isotherm, the parameters during storage. As per ANOVA results, the effect of packaging material
of the GAB equations should be in the following ranges: C N 5.67 and and storage time on moisture content was significant (Table 6). The pos-
0 b K b 1 and the same is depicted by GAB parameters for tamarind sible reason for increase in moisture content of the powder during stor-
pulp powder. A positive value for ΔHc was expected as a result of the age is due to the ingress of moisture through the packages which have
exothermic interaction of the water molecules with the sorption sites different degree of permeability to water vapor.
on the surfaces of the powders. A negative value is usually expected Titratable acidity of the powder samples showed an increase during
for ΔHk due to the weaker bonding of the multilayer molecules [21] storage, suggesting the interaction between various constituents and
which has been the case here. However, a positive value of the ΔH2 resulting chemical changes (Table 5). ANOVA results indicated that
has been also reported for fruits at high water activities and has been re- the type of packaging material and storage time significantly affected
lated to the endothermic dissolution of fruit sugars in the absorbed the titratable acidity of tamarind pulp powder (Table 6). The increase
water [19]. in titratable acidity may be due to the reaction of basic amines to form
compounds of lower basicity and to the degradation of sugars into
3.3. Storage study acids during the Maillard reaction [22]. These results are in agreement
with the findings of Chauhan and Patil [23] and Liu et al. [24], found dur-
Changes in physicochemical properties of the food materials occur ing storage of mango milk powder and tomato pulp powder
during storage. The packaging material is one of the important factor respectively.
that influences the physical, chemical and sensory properties of dried Bulk density of tamarind pulp powder samples increased with stor-
food powders. The effect of packaging material on physicochemical age period depending on the extent of moisture gain of the powder
properties of tamarind pulp powder during storage is discussed below. samples packed in different packaging materials (Table 5). Type of Pack-
aging material and storage time had a significant effect (P b 0.05) on
3.3.1. Color bulk density values (Table 6). Increase in bulk density with increasing
Color is a very important quality characteristic of fruit and vegetable moisture content during storage was also observed by Chauhan and
products which influences the consumer acceptability. Color changes of Patil [23] in mango milk powder. An increase in bulk density may also
tamarind pulp powder samples during storage using different packag- be attributed to increased cohesiveness between powder particles
ing material are presented in Table 5. During storage there was a de- caused by absorption of moisture during storage.
crease in L⁎ and b⁎ value and an increase in a⁎ value for color. The
ANOVA results showed that both packaging material and storage period 3.3.3. Flowability
significantly (P b 0.05) affected the color values of tamarind pulp pow- Flowability of tamarind pulp powder samples was measured in
der (Table 6). With respect to the package used it was observed that terms of cohesion index. Higher the cohesion index, higher is the cohe-
change in color values was minimum for the powder samples stored siveness of the powder, thus poorer flowability. On the basis of cohesive
in glass while there was a considerable change in color values for the index, the powders are categorized as follows: 19+, extremely cohe-
powder packed in LDPE. The minimum color change for the powder in sive; 16–19, very cohesive; 14–16, cohesive; 11–14, easy flowing and
glass bottle may be due to impermeability of glass to air and moisture 11, free flowing (Stable Micro Systems Ltd., UK). Table 5 shows the
which favor the reactions responsible for color change. The change in changes in cohesive index of the powder samples packed in various
color values can be attributed to Maillard reaction. packaging materials during storage. The effect storage of period and
packaging material on the cohesion index of the powder samples is re-
3.3.2. Moisture content, acidity and bulk density lated to their effect on moisture content of the powder. Increase in
Moisture content, acidity and bulk density values of tamarind pulp moisture content increases the cohesiveness of the powder and thus
powder packed in different packaging materials are summarized in leads to reduction in flowability [25]. Both packaging material and

Table 5
Changes in physicochemical properties of spray dried tamarind pulp powder at ambient conditions (25 °C temperature and 45% relative humidity) during storage using three different
packaging materials.

Packaging material Storage period (months) Color Moisture (%) Acidity (%) Bulk density (g/ml) Cohesion index (mm)

L⁎ a⁎ b⁎

LDPE 0 66.56 ± 0.61 10.87 ± 0.47 27.21 ± 0.52 2.62 ± 0.02 9.33 ± 0.06 0.43 ± 0.01 10.19 ± 0.32
1 65.32 ± 0.49 10.94 ± 0.98 27.09 ± 0.61 3.16 ± 0.06 9.41 ± 0.04 0.46 ± 0.01 10.64 ± 0.06
2 62.19 ± 0.82 11.36 ± 0.49 26.93 ± 0.45 4.19 ± 0.24 9.55 ± 0.03 0.49 ± 0.03 11.56 ± 0.13
3 59.12 ± 0.40 11.42 ± 0.97 26.66 ± 0.40 5.11 ± 0.15 9.71 ± 0.07 0.52 ± 0.03 14.52 ± 0.12
4 56.62 ± 0.33 12.57 ± 1.37 25.97 ± 0.79 6.93 ± 0.06 9.92 ± 0.08 0.58 ± 0.03 15.36 ± 0.13
5 53.67 ± 0.57 12.99 ± 0.46 25.83 ± 0.74 7.39 ± 0.50 10.15 ± 0.06 0.64 ± 0.02 16.77 ± 0.29
6 46.34 ± 1.58 13.52 ± 0.61 24.75 ± 0.77 7.87 ± 0.67 10.26 ± 0.10 0.66 ± 0.03 17.88 ± 0.63
ALP 0 66.56 ± 0.61 10.87 ± 0.47 27.21 ± 0.52 2.62 ± 0.02 9.33 ± 0.06 0.43 ± 0.01 10.19 ± 0.32
1 66.01 ± 0.31 10.99 ± 0.91 27.16 ± 0.73 2.70 ± 0.02 9.35 ± 0.06 0.44 ± 0.01 10.34 ± 0.04
2 65.93 ± 1.05 10.96 ± 0.75 27.15 ± 0.51 2.83 ± 0.03 9.40 ± 0.02 0.46 ± 0.02 10.87 ± 0.09
3 65.86 ± 0.32 11.13 ± 0.86 27.14 ± 0.67 3.06 ± 0.14 9.43 ± 0.04 0.48 ± 0.01 11.40 ± 0.13
4 62.59 ± 0.41 11.42 ± 0.61 27.09 ± 0.67 3.59 ± 0.26 9.65 ± 0.11 0.49 ± 0.02 11.97 ± 0.39
5 61.39 ± 0.44 11.72 ± 0.66 27.05 ± 0.93 4.36 ± 0.51 9.72 ± 0.09 0.51 ± 0.02 12.49 ± 0.48
6 58.78 ± 2.41 12.06 ± 0.65 26.94 ± 1.01 4.95 ± 0.08 9.97 ± 0.23 0.52 ± 0.03 13.18 ± 0.35
Glass 0 66.56 ± 0.61 10.87 ± 0.47 27.21 ± 0.52 2.62 ± 0.02 9.33 ± 0.06 0.43 ± 0.01 10.19 ± 0.32
1 66.43 ± 0.27 10.87 ± 0.88 27.20 ± 0.37 2.62 ± 0.03 9.33 ± 0.13 0.43 ± 0.02 10.19 ± 0.24
2 66.39 ± 0.44 10.93 ± 0.68 27.20 ± 0.70 2.62 ± 0.05 9.36 ± 0.08 0.43 ± 0.02 10.20 ± 0.10
3 66.12 ± 0.18 10.95 ± 0.39 27.18 ± 0.63 2.65 ± 0.02 9.38 ± 0.08 0.43 ± 0.02 10.22 ± 0.08
4 66.02 ± 0.46 10.95 ± 0.42 27.17 ± 0.54 2.64 ± 0.03 9.40 ± 0.05 0.44 ± 0.03 10.22 ± 0.14
5 65.81 ± 0.67 10.97 ± 0.85 27.15 ± 0.35 2.65 ± 0.03 9.66 ± 0.05 0.44 ± 0.04 10.24 ± 0.03
6 65.27 ± 0.66 11.03 ± 0.12 27.18 ± 0.22 2.62 ± 0.07 9.89 ± 0.10 0.46 ± 0.02 10.31 ± 0.18
 K. Muzaffar, P. Kumar / Powder Technology 291 (2016) 322–327 327

Table 6
ANOVA for physicochemical properties of spray dried tamarind pulp powder during storage.

Source of variation df MSS

L⁎ a⁎ b⁎ Moisture content Acidity Bulk density Cohesive index

Packaging material (PM) 2 315.49⁎⁎ 5.54⁎⁎ 4.46⁎⁎ 40.04⁎⁎ 0.455⁎⁎ 0.056⁎⁎ 70.86⁎⁎
Storage period (SP) 6 108.44⁎⁎ 2.41⁎⁎ 0.97⁎⁎ 8.81⁎⁎ 0.630⁎⁎ 0.019⁎⁎ 17.72⁎⁎
PM × SP 12 34.56⁎⁎ 0.79⁎ 0.70⁎ 3.46⁎⁎ 0.034⁎⁎ 0.005⁎⁎ 7.12⁎⁎
Error 42 0.66 0.52 0.40 0.06 0.007 0.0005 0.07
⁎⁎ Mean significant at P b 0.05.
⁎ Mean non-significant at P b 0.05.

storage time significantly affected the cohesion index value (Table 6). [5] K. Muzaffar, P. Kumar, Parameter optimization for spray drying of tamarind pulp
using response surface methodology, Powder Technol. 279 (2015) 179–184.
Tamarind pulp powder packed in LDPE showed higher decrease in pow- [6] K. Muzaffar, P. Kumar, Effect of process parameters on extraction of pulp from tam-
der flowability due to considerable increase in moisture content and the arind fruit, in: G.C. Mishra (Ed.), Conceptual Framework & Innovations in Agroecol-
powder became cohesive after a period of two months, with cohesive ogy and Food Sciences, Krishi Sanskriti Publications, New Delhi 2015, pp. 65–67.
[7] B.R. Bhandari, N. Datta, R. Crooks, T. Howes, S. Rigby, A semi-empirical approach to
index greater than 14. Flowability of the powder packed in ALP also de- optimize the quantity of drying aids required to spray dry sugar rich foods, Dry.
creased but showed easily flowing behavior after six months of storage Technol. 15 (1997) 2509–2525.
period. However powder packed in glass showed almost no decrease in [8] L. Greenspan, Humidity fixed points of binary saturated aqueous solutions, J. Res.
Natl. Bur. Stand. A. Phys. Chem. 81 (1977) 89–96.
powder flowability and remained in a free flowing state throughout the
[9] AOAC, Official Methods of Analysis, 19th ed. Association of Official Analytical Chem-
storage period, with cohesive index less than 11. ists, Washington, 2012.
[10] S. Rangana, Hand Book of Analysis and Quality Control for Fruits and Vegetable
Products, second ed. Tata McGraw Hall Publishing Company Lid., New Delhi, 1986.
3.4. Sensory quality
[11] A.M. Goula, T.D. Karapantsios, D.S. Achilias, K.G. Adamopoulos, Water sorption iso-
therms and glass transition temperature of spray dried tomato pulp, J. Food Eng.
Packaging material and storage time had marked influence on the 85 (2008) 73–83.
sensory quality of tamarind pulp powder. Tamarind pulp powder [12] Z. Yang, E. Zhu, Z. Zhu, Moisture sorption isotherms and Net isosteric heats of sorp-
tion of green soybean, Int. J. Food Eng. 8 (2012)http://dx.doi.org/10.1515/1556-
packed in LDPE was acceptable up to 2 months of storage period. Pow- 3758.2757.
der samples packed in ALP and glass showed acceptable sensory quality [13] Z.B. Maroulis, E. Tsami, D. Marinos-Kouris, G.D. Saravacos, Application of the GAB
throughout 6 months of storage period. However glass offered the model to the moisture sorption isotherms for dried fruits, J. Food Eng. 7 (1988)
63–78.
highest protection against loss in sensory scores of tamarind pulp [14] F. Kaymak-Ertekin, A. Gedik, Sorption isotherms and isosteric heat of sorption for
powder. grapes, apricots, apples and potatoes, LWT Food Sci. Technol. 37 (2004) 429–438.
[15] M.E. Sormoli, T.A.G. Langrish, Moisture sorption isotherms and net isosteric heat of
sorption for spray-dried pure orange juice powder, LWT Food Sci. Technol. 62
4. Conclusion (2015) 875–882.
[16] K.B. Palipane, R.H. Driscoll, Moisture sorption characteristics of inshell macadamia
The moisture sorption isotherms of spray dried tamarind pulp pow- nuts, J. Food Eng. 18 (1992) 63–76.
[17] K.D. Foster, J.E. Bronlund, A.H.J. Paterson, The prediction of moisture sorption iso-
der followed a sigmoid isotherm curve, typical of the type II BET classi-
therms for dairy powders, Int. Dairy J. 15 (2005) 411–418.
fication shape. Temperature affected the sorption behavior, with [18] S. Brunauer, L.S. Deming, W.E. Deming, E. Teller, On a theory of the Van der Waals
equilibrium moisture content decreased with increasing temperature adsorption of gases, J. Am. Chem. Soc. 62 (1940) 1723–1732.
[19] E.J. Quirijns, A.J.B. van Boxtel, W.K.P. van Loon, G. van Straten, Sorption isotherms,
at constant water activity. Within the temperature range investigated,
GAB parameters and isosteric heat of sorption, J. Sci. Food Agric. 85 (2005)
the kinetic three-parameter GAB model and the empirical two parame- 1805–1814.
ter Oswin model were found best to represent the experimental data [20] P.P. Lewicki, The applicability of the GAB model to food water sorption isotherms,
throughout the entire range of water activity. Therefore the GAB Int. J. Food Sci. Technol. 32 (1997) 553–557.
[21] A.L. Gabas, F.C. Menegalli, J. Telis-Romero, Water sorption enthalpy entropy com-
model and Oswin model were chosen as the preferred models for pensation based on isotherms of plum skin and pulp, J. Food Sci. 65 (2000) 680.
predicting the moisture sorption isotherms of tamarind pulp powder. [22] J. Beck, F. Ledl, M. Sengl, T. Severin, Formation of acid lactones and esters through
In general, it can be concluded that among the three packages used, the Maillard reaction, Z. Lebensm. Unters. Forsch. 190 (1990) 212–216.
[23] A.K. Chauhan, V. Patil, Effect of packaging material on storage ability of mango milk
the glass is recommended for long term storage of tamarind pulp powder and the quality of reconstituted mango milk drink, Powder Technol. 239
powder. (2013) 86–93.
[24] F. Liu, X. Cao, H. Wang, X. Liao, Changes of tomato powder qualities during storage,
Powder Technol. 204 (2010) 159–166.
Acknowledgments [25] D. Janjatovic, M. Benkovic, S. Srecec, D. Jezek, I. Spoljaric, I. Bauman, Assessment of
powder flow characteristics in incoherent soup concentrates, Adv. Powder Technol.
The first author is thankful to University Grant Commission, New 23 (2012) 620–631.
[26] C. Van den Berg, S. Bruin, Water activity and its estimation in food systems, in: L.B.
Delhi for financial support.
Rockland, G.F. Stewart (Eds.), Water Activity: Influences on Food Quality, Academic
Press, New York 1981, pp. 1–61.
References [27] S. Brunauer, P.H. Emmett, E. Teller, Adsorption of gases in multimolecular layers, J.
Am. Chem. Soc. 60 (1938) 309–320.
[1] P. Kumar, H.N. Mishra, Moisture sorption characteristics of mango-soy fortified yo- [28] C.R. Oswin, The kinetics of package life III. The isotherm, J. Chem. Ind. 65 (1946)
gurt powder, Int. J. Dairy Technol. 59 (2006) 22–28. 419–421.
[2] R.V. Tonon, A.F. Baroni, C. Brabet, O. Gibert, D. Pallet, M.D. Hubinger, Water sorption [29] S.E. Smith, The sorption of water vapour by high polymers, J. Am. Chem. Soc. 69
and glass transition temperature of spray dried acai (Euterpe oleracea Mart.) juice, J. (1947) 646.
Food Eng. 94 (2009) 215–221. [30] M. Caurie, A new model equation for predicting safe storage moisture stability of
[3] J.V. García-Perez, J.A. Carcel, G. Clemente, A. Mulet, Water sorption isotherms for dehydrated foods, J. Food Technol. 5 (1970) 301–307.
lemon peel at different temperatures and isosteric heats, LWT Food Sci. Technol. [31] H. Iglesias, J. Chirife, Handbook of Food Isotherms: Water Sorption Parameters for
41 (2008) 18–25. Food and Food Components, Academic Press Inc., New York, 1982.
[4] S.N. Bhusari, K. Muzaffar, P. Kumar, Effect of carrier agents on physical and micro-
structural properties of spray dried tamarind pulp powder, Powder Technol. 266
(2014) 354–364.