• Sorption Isotherm Equations of Potato Flakes and Sweet Potato ...

  • FileName: p0034.pdf [read-online]
    • Abstract: Flakes and Sweet Potato Flakes. Isotermas de Adsorção de Purê de Batata e. Purê de Batata ... sweet potato flakes are shown in Figure 2. The isotherms gave. the characteristic S shaped ...

Download the ebook

SUMMARY Facultad de Industrias Alimentarias Universidade Federal Sweet Potato Flakes sweet potato flakes Tecnologia de Alimentos moisture content Potato Flakes Instituto de Ci Universidad Nacional Agr water activities Sorption Isotherm Equations of Potato
Flakes and Sweet Potato Flakes
Isotermas de Adsorção de Purê de Batata e
Purê de Batata-Doce em Flocos
Facultad de Industrias Alimentarias, Universidad Nacional
Agrária La Molina, Lima, Peru
The water adsorption isotherms of potato and sweet potato flakes at 15, 20, 25 and
30°C were determined, using the static gravimetric method. The equilibrium moisture content
Caciano Pelayo ZAPATA-NORENA
at water activities of up to 0.80 decreased as the temperature increased. At higher water
Instituto de Ciência e Tecnologia de Alimentos,
Universidade Federal do Rio Grande do Sul activities the moisture content increased sharply as the temperature increased, resulting in an
91501-970 Porto Alegre, Brasil inversion of the isotherm curves. The BET, GAB, Henderson-Thompson, Halsey-modified (Iglesias
and Chirife) and Chung-Pfost equations were used to fit the experimental moisture sorption
data of both products. The experimental data showed the best fit with the GAB model for the
range of relative humidity values (11 to 92%). Only the GAB model explained the increase in
Instituto de Ciência e Tecnologia de Alimentos,
Universidade Federal do Rio Grande do Sul moisture content with increasing temperature at water activities above 0.80, caused by sugar
Av. Bento Gonçalves 9500 dissolution. The Henderson-Thompson, Halsey-modified and Chung-Pfost models did not fit
91501-970 Porto Alegre, Brasil the experimental data, due to their inability to explain the inversion of the isotherms.
Fax: +55-51-316-7048
e-mail: [email protected]
As isotermas de adsorção de purê de batata e purê de batata-doce em flocos foram
determinadas a 15, 20, 25 e 30°C, utilizando o método gravimétrico. A umidade relativa
diminuiu com o aumento da temperatura para atividades de água até 0,80. Em valores superiores
de atividade de água a umidade relativa aumentou rapidamente com o aumento da temperatura,
resultando na inversão das isotermas. Foram utilizados os modelos de BET, GAB, Henderson-
Thompson, Halsey-modificado (Iglesias e Chirife) e Chung-Pfost para ajustar os valores de
sorção de equilíbrio dos produtos. O modelo de GAB foi o que melhor ajustou a informação
experimental para todo o intervalo de umidade relativa (11 a 92%). Este modelo foi o único
capaz de explicar o fenômeno de inversão das isotermas para atividades de água superiores a
0,80, causado pela dissolução de açúcares. Os modelos de Henderson-Thompson, Halsey-
modificado e Chung-Pfost não ajustaram adequadamente os dados experimentais, devido à sua
incapacidade para explicar a inversão das isotermas.
Sorption isotherm; Water activity; Potato; Sweet potato;
Dehydrated food; Models / Isoterma de adsorção;
Atividade de água; Batata; Batata-doce; Alimento
desidratado; Modelos teóricos.
Braz. J. Food Technol., 3:53-57, 2000 53 Recebido / Received: 25/05/1999. Aprovado / Approved: 31/05/2000.
M. LIENDO-CÁRDENAS Sorption Isotherm Equations of Potato
et al. Flakes and Sweet Potato Flakes
Water activity has an important effect on the stability 2.1 Materials
of low-moisture foods. The water sorption characteristics are
of fundamental importance in food materials, influencing every Potato flakes and sweet potato flakes were obtained
aspect of the drying process and the storage stability of the from local supermarkets. The dehydratation process was
dried products (LABUZA et al., 1970). Sorption equilibrium effected on a single drum drier, and the final drying on a
data are an extremely important tool in food science, because vibrating bed drier at an air temperature of 110°C and velocity
they can be used to predict changes in food stability and to of 0.6 m/s, resulting in a product with a bulk density of about
select appropriate packaging materials and ingredients. For 802 kg/m3. The products were kept in air-tight containers at
this reason, many studies concerning experimental approaches room temperature and analysed for up to 20 days after they
to obtain sorption isotherms have been described, as well as were produced. The proximate composition was determined
using AOAC methodology (AOAC, 1984) and expressed as
how to obtain a suitable correlation with equilibrium moisture
percentage (dry basis). Potato: humidity (7.75), protein (8.29),
and water activity.
fat (0.54), fiber (2.59), ash (3.29), carbohydrate (85.29), sugar
A large number of correlations have been proposed (2.60). Sweet potato: humidity (3.82), protein (6.24), fat (0.59),
to represent the hygroscopic behaviour of food products fiber (3.35), ash (2.60), carbohydrate (87.21), sugar (3.70).
(VAN DEN BERG, BRUIN, 1981). These equations were
derived from surface science, such as the classical BET
equation (BRUNAUER et al., 1938), or from food systems, 2.2 Equilibrium moisture data
such as the Henderson’s correlation (HENDERSON, 1952)
and the Chung-Pfost equation (CHUNG, PFOST, 1967). To A static gravimetric technique was used for the
design, simulate or model a process, data input and equilibrium moisture data determination. Samples of
calculations are needed on a continuous, rather than a approximately 1 g were placed in wire baskets, which were
discrete, representation. Therefore, it is more useful to have then placed in desiccators. Each desiccator contained a
correlations where temperature is the variable. Some of the saturated salt solution. A set of nine solutions was selected to
best known approaches are the Henderson-Thompson provide adequate screening for water activities from 0.11 to
correlation (THOMPSON, 1972), the Chung-Pfost equation, 0.92 at different temperatures (SARAVACOS et al., 1986). The
and the Iglesias and Chirife modification (CHIRIFE, IGLESIA S, desiccators were incubated at 15, 20, 25, and 30° C for 15
1978) of Halsey ’s equation (HALSEY, 1948). The days. Equilibrium was reached within 15 days, as evidenced
Guggenhein-Anderson-de Boer (GAB) equation is a popular by a constant weight after successive weighings of the sample
(SARAVACOS et al., 1986; LIM et al., 1995). After equilibrium,
theoretical model for foods (LOMAURO et al., 1985a,
the samples were analyzed for moisture content by the
LOMAURO et al., 1985b).
vacuum oven method (AOAC, 1984).
According to VAN DEN BERG, BRUIN (1981), a
general sigmoid-shaped sorption isotherm is divided into
three different parts. In the first region, an almost constant 2.3 Isotherm equations
water adsorption up to water activity values of 0.5 is
observed. In this low water activity range the moisture The isotherm equations are listed in Table 1. The BET
content is low, because water can be adsorbed only to equation parameter constants were calculated by regression
surface -OH sites of crystalline sugar. In the second zone, using the linear form of the isotherm equation for each
there is an increase in water adsorption. This can be temperature. The GAB parameter constants were obtained by a
explained by the sorption or diffusion of water molecules non-linear regression analysis. A multiple non-linear regression
into newly created pores in the already swollen structure. analysis was developed to obtain parameter values from the
The water molecules are mechanically entrapped in the Henderson-Thompson, Halsey’s-modified, and Chung-Pfost
void spaces of the starch. At higher water activities, the equations. The data were processed using the Statistics Analysis
water uptake appears to be markedly influenced by the System 6.0 software (SAS Institute, Cary, North Carolina).
stability of the microporous structure of the starch (BOKI et
al., 1990). At high water activity, dissolution of sugar occurs, TABLE 1. Isotherm equations of moisture sorption dataa.
and crystalline sugar is converted into amorphous sugar. Model Isotherm equation
Consequently, the amount of water to be adsorbed increases
BET aw = 1 + aw(C-1)
greatly because of the increase in the number of adsorption (1-aw)X XmC
sites, resulting from the breakage of the crystalline structure GAB X= Xm C k aw
of the sugar. (1- k aw)(1- k aw + C k aw)
The objective of this investigation was to determine Henderson-Thompson X = (Ln [(1-R) / -a (t + b)]) 1/c
the parameters of the BET, GAB, Henderson-Thompson, Halsey-modified (Iglesias & Chirife) X = (exp [a - bT] / - Ln [RH])1/c
Halsey-modified, and Chung-Pfost equations for potato flakes Chung-Pfost X = -1/b Ln (RT Ln [RH]/a)
and sweet potato flakes, as well as to determine the fit for the a
aw = water activity; X = moisture content; T= absolute temperature; t= temperature
sorption isotherm in each case. in °C; R=8.314 J/mol K; a, b, c, Xm, C and k are parameters for the isotherm equations.
Braz. J. Food Technol., 3:53-57, 2000 54
M. LIENDO-CÁRDENAS Sorption Isotherm Equations of Potato
et al. Flakes and Sweet Potato Flakes
3. RESULTS AND DISCUSSION higher than 0.80, an inversion of the temperature effect was
observed, i.e., there was an increase in moisture content with
3.1 Isotherms an increase in temperature at a given water activity. This effect
may be explained by dissolution of sugar in the water, as
Experimental moisture sorption data for potato flakes observed for carbohydrate-rich products (MAZZA, 1984,
at several temperatures are shown in Figure 1. The data for LABUZA et al., 1985, BOKI et al., 1990), and was also similar to
sweet potato flakes are shown in Figure 2. The isotherms gave that reported for high sugar content foods (AYRANCI et al.,
the characteristic S shaped curve of normal moisture adsorption 1990, TSAMI et al., 1990, LIM et al., 1995). The addition of
isotherms (LABUZA et al., 1985), and contained three zones, sugar to potato slices also resulted in a similar effect
similar to those observed for starches and starch-rich products (MAZZA, 1982).
(LABUZA et al., 1985; AJIBOLA 1986; BOKI, OHNO, 1991). The sugar content in raisins, currants, figs, prunes
and apricots ranges from 51 to 82 % (SARAVACOS et
al.,1986, AYRANCI et al., 1990, TSAMI et al., 1990). In a
35 15oC (G.A.B.)
way similar to the isotherms of these high sugar fruits, the
15oC (experimental) isotherms of potato and sweet potato flakes intersected,
20oC (G.A.B.)
30 but at a higher water activity level (about 0.8). The
Moisture content (g / 100 g d.b.)
20oC (experimental)
25oC (G.A.B.) intersection point depends on the type of sugar present,
25oC (experimental)
30oC (G.A.B.) the sugar size distribution and the composition of the foods
20 30oC (experimental) (WEISSER, 1985). The sugar content of the potato tuber is
about 1.5%, which may increase on processing
(SCHWIMMER, BURR, 1967). During the preparation of sweet
10 potato flakes, the natural amylases react with the freshly
gelatinized starch, decreasing the iodine blue value and
producing dextrins, maltose, and glucose (DANIEL,
0 WHISTLER, 1985). The skim milk powder, used in the
0,0 0,2 0,4 0,6 0,8 1,0
preparation of the potato flakes (KNAACK, 1976), also
Water activity
contributed to increasing the sugar content of this product.
FIGURE 1. Experimental moisture sorption values for potato
flakes and predicted isotherms using the GAB equation. 3.2 Fitting sorption data to isotherm equations
The experimental data were fitted to the isotherm
70 equations. The parameter constants and standard errors
15oC (G.A.B.) for the different models are shown in Tables 2 and 3. The
60 15oC (experimental) BET and GAB parameters are listed in Table 2, and the
20oC (G.A.B.)
20oC (experimental) Henderson-Thompson, Halsey’s-modified and Chung-Pfost
Moisture content (g / 100 g d.b.)
25oC (G.A.B.)
25oC (experimental)
data can be found in Table 3. The BET equation is known
40 30oC (G.A.B.) to hold for water activities of up to about 0.5 (CHIRIFE,
30oC (experimental)
IGLESIAS, 1978). In this low water activity region, the fit of
30 the sorption data to the BET equation was satisfactory
(Table 2). The GAB model fitted well in the whole range of
20 water activity, giving a low standard error and r 2 values
higher than 0.99. The Henderson-Thompson, Halsey-
modified and Chung-Pfost models gave high standard
errors (Table 3). On fitting the data from the Henderson-
0,0 0,2 0,4 0,6 0,8 1,0 Thompson and Halsey-modified equations, in the whole
Water activity
range of water activity and at the different temperatures,
the four isotherms were superimposed. This was observed
FIGURE 2. Experimental moisture sorption values for sweet
for the two products tested. Although the r2 values ranged
potato flakes and predicted isotherms using the GAB equation.
from 0.98 to 0.99, the high standard errors and the
significant F test revealed by the variance analysis, indicated
that these models did not fit the experimental data well.
Moisture content is often expected to decrease with Indeed, the r 2 values did not warrant an adequate fitting
increasing temperature at a given water activity (aw). This was by non-linear equations (BROTONS et al., 1986). With
observed for the two products in the aw region below 0.80 respect to this, the optimum range of relative moisture for
(Figures 1 and 2). In this region, it is clearly evident from the application of Halsey’s model is from 30 to 75 % (BOKI,
figures that for any constant aw, an increase in temperature OHNO, 1991). Thus, the GAB equation best represented
decreases the moisture content. However, above this region, the experimental moisture sorption data for potato and
the reverse was observed. In the third region, with water activity sweet potato flakes.
Braz. J. Food Technol., 3:53-57, 2000 55
M. LIENDO-CÁRDENAS Sorption Isotherm Equations of Potato
et al. Flakes and Sweet Potato Flakes
TABLE 2. Comparison of the equilibrium relative moisture were in agreement with those stated by VAN DEN BERG, BRUIN
according to the BET and GAB models for potato and sweet (1981), which indicated that the monolayer moisture content
potato flakes. in foods was lower than 10 g/100g dry matter.
On the other hand, the heat of sorption (Qs) is an
Parameters Standard
Model Product important thermodynamic term, which measures the
r2 error (%)
T (°C) Xmb C k interaction between water vapour and the adsorbent food
BETa Potato 15 3.8760 19.9180 0.9824 5.87 material. Heat of sorption values were also obtained by the
flakes 20 3.8352 10.7326 0.9853 4.25 BET and GAB models (Table 4). Generally, the Qs values
25 3.5839 7.7277 0.9720 5.00 decreased with increasing temperature, similar to those
30 3.4178 7.0469 0.9396 6.77
reported for corn meal and fish flour (LABUZA et al., 1985).
Sweet 15 7.4535 4.7443 0.9600 5.11
The Qs values found in both products indicated that the
potato 20 7.3000 3.5125 0.9960 1.03
flakes 25 6.1140 4.3602 0.9700 3.65
sorption type corresponded to physical adsorption, since the
30 5.5740 4.5083 0.9800 3.16 heats of sorption were lower than 41.67 KJ/mol. Potato flakes
GAB Potato 15 4.6668 10.8705 0.8834 0.9994 2.74
presented higher Qs values than sweet potato flakes by both
flakes 20 4.4390 8.1940 0.9013 0.9989 2.34
the BET and GAB models, indicating stronger interactions
25 4.1610 5.9168 0.9206 0.9974 3.16 among water molecules and the -OH, -COOH and -NH 2
30 3.7041 6.1770 0.9542 0.9981 3.95 groups in potato flakes.
Sweet 15 9.6500 4.1401 0.8193 0.9990 3.28
potato 20 8.8536 3.5802 0.8607 0.9985 3.45
TABLE 4. Heat of sorption values for potato flakes and sweet
flakes 25 7.1500 3.6299 0.9172 0.9995 2.79
30 6.1452 3.8233 0.9592 0.9992 1.19
potato flakes determined by the BET and GAB equations a.
aw range of 0.11-0.53 for BET equation and 0.11-0.92 for GAB equation.
Xm = monolayer moisture content is expressed in g / 100 g dry matter.
Product T (°C) Qs (KJ/mol)
TABLE 3. Comparison of equilibrium relative moisture models Potato flakes 15 7.1624 5.7125
for potato and sweet potato flakes. 20 5.7806 5.1233
25 5.0656 4.4041
Model Parameters Standard 30 4.9182 4.5864
error (%)
a b c r2 Sweet potato 15 4.8091 3.4014
Potato flakes 20 4.8419 3.1065
Henderson-Thompson 8.00 x 10-6 1.26 x 104 1.0121 0.9927 13.71
25 4.4853 3.1937
Modified-Halsey 1.0099 7.70 x 10-5 0.9952 0.9830 33.90
Chung-Pfost 5.37 x 10 3
0.1431 0.9826 23.00
30 4.3276 3.3781
Sweet potato a
Heat of sorption (Qs) was determined by the equation Qs=RT Ln C,
Henderson-Thompson 5.00 x 10-6 1.25 x 104 0.9999 0.9927 10.86
where: R=8.314 J/mol K; T= absolute temperature; C= parameter
Modified-Halsey 1.0115 -1.50 x 10-3 0.9952 0.9850 30.60
constant for the isotherm equation.
Chung-Pfost 5.04 x 103 8.61 x 10-2 0.9847 25.00
Figures 1 and 2 show the sorption isotherms fitted by 4. CONCLUSION
the GAB model for potato flakes and sweet potato flakes,
respectively. The inversion of the temperature effect is observed, Sorption isotherms of potato flakes and sweet potato
similar to that described by SARAVACOS et al. (1986) and AYRANCI flakes showed an increase in moisture with increasing
et al. (1990) for high sugar foods, indicating that the GAB temperature at water activity values higher than 0.80, a
equation is capable of explaining the effect of sugar dissolution characteristic effect obser ved in products with high
at high water activity values. carbohydrate contents. The GAB model best fitted the
experimental data for both products for the whole range of
3.3 Monolayer moisture content and heat of sorption water activity (0.11 to 0.92), presenting r2 values higher than
by the BET and GAB equations 0.99. It was capable of explaining the inversion of the isotherms
in the water activity range from 0.80 to 0.92.
The monolayer moisture content (Xm) is an important
parameter in food deterioration studies. The values obtained
for this parameter by the BET and GAB models were compared REFERENCES
(Table 2). The Xm values decreased with increasing temperature,
as previously reported for high-carbohydrate foods (LABUZA et AJIBOL A, O.O. Desorption isotherms for plantain at several
al., 1970, AYRANCI et al., 1990). In addition, the GAB values temperatures. Journal of Food Science, 51:169-171, 1986.
were slightly higher than the BET values, as observed by AOAC. Official Methods of Analysis. 17th edition. Washington:
LABUZA et al. (1985) and AYRANCI et al. (1990). The Xm values Association of Official Analytical Chemists, 1984.
Braz. J. Food Technol., 3:53-57, 2000 56
M. LIENDO-CÁRDENAS Sorption Isotherm Equations of Potato
et al. Flakes and Sweet Potato Flakes
AYRANCI, E., AYRANCI, G., DOGANTAN, Z. Moisture sorption isotherms LABUZA, T.P KAANANE, A., CHEN, J.Y. Effect of temperature on
of dried apricot, fig and raisin at 20°C and 36°C. Journal of Food the moisture sorption isotherms and water activity shift of two
Science, 55:1591-1593, 1990. dehydrated foods. Journal of Food Science, 50:385-391, 1985.
BOKI, K., OHNO, S. Equilibrium isotherm equations to represent LIN, L.T., TANG, J., HE, J. Moisture sorption characteristics of freeze
moisture sorption on starch. Journal of Food Science, dried blueberries. Journal of Food Science, 60:810-814, 1995.
56:1106-1110, 1991. LOMAURO, C.J., BAKSHI, A.S., L ABUZA, T.P Evaluation of food
moisture sorption isotherm equations. I. Fruit, vegetable and meat
BOKI, K., OHNO, S., SHINODA, S. Moisture sorption characteristics of
products. Lebensmittel-Wissenschaft und Technologie,
kudzu starch and sweet potato starch. Journal of Food Science,

Use: 0.0683