- Email: [email protected]
Study of the Drying Kinetics of Green Bell Pepper and Chemical Characterization
STUDY OF THE DRYING KINETICS OF GREEN BELL PEPPER AND CHEMICAL CHARACTERIZATION J. M. F. Faustino, M. J. Barroca and R. P. F. Guine´ Department of Food Engineering, ESAV, Polytechnic Institute of Viseu,Viseu, Portugal.
Abstract: The present work aimed, on one hand, the study of the drying of green peppers, in terms of drying kinetics evaluated at 308C, 408C, 508C, 608C and 708C, having the experimental data been fitted to different empirical kinetic models from literature. This kinetic study was then complemented with the modelling in terms of Fick’s diffusion equation. On the other hand, the chemical characterization in fresh and after drying at the lowest and highest temperatures was analysed, for evaluation of the effect of drying and drying temperature on the chemical composition of the product. In this way, the analyses made were: moisture content, sugar content, proteins, ash, fat, fibre and acidity. From the results obtained, it was concluded that the empirical models that best describe the dehydration kinetics were the Page and Newton models. From the experimental data was possible to estimate the diffusivities, which range between 9.0 10210 at 308C and 8.0 1029 m2 s21 at 708C. Moreover, it was verified that drying influences the chemical composition of the peppers, but, conversely, the influence of the drying temperature was not very significant. Keywords: bell pepper; capsicum annuum; dried bell pepper; drying kinetics; chemical characterization; nutritional value.
INTRODUCTION
Correspondence to: Professor Raquel P.F. Guine´, Quinta da Alagoa. Estrada de Nelas. Ranhados. 3500-606 Viseu. Portugal. E-mail: raquelguine@ esav.ipv.pt
DOI: 10.1205/fbp07009 0960–3085/07/ $30.00 þ 0.00 Food and Bioproducts Processing Trans IChemE, Part C, September 2007 # 2007 Institution of Chemical Engineers
They are fleshy and heavily seeded, and mature from green to red. They vary in size, colour and flavour and the level of heat as measured in Scoville heat units. The higher the number of Scoville heat units, the hotter the pepper. The fruit of most species of Capsicum contains capsaicin (methyl vanillyl nonenamide), a lipophilic chemical that can produce a strong burning sensation in the mouth. Bell peppers contain a recessive gene that eliminates the capsaicin in the fruit. The colour can be green, red, yellow, orange and, more rarely, white, purple, blue, black and brown, depending on when they are harvested. Green peppers are unripe bell peppers, while the others are all ripe, with the colour variation based on cultivar selection (Simonne et al., 1997). Because they are unripe, green peppers are less sweet and slightly more bitter than yellow, orange or red peppers, which all have a rather similar hot taste. The taste of ripe peppers can also vary with growing conditions and post-harvest storage treatment. The fruits allowed to ripen fully on the plant in full sunshine are the sweetest, while the fruits harvested green and after-ripened in storage are less sweet.
Capsicum is a genus of plants from the nightshade family (Solanaceae). Some of them are used as spices, vegetables or even medicines. The fruit of Capsicum plants can have a variety of names depending on place and type, being commonly called chilli pepper, red or green pepper or just pepper. The bell pepper refers to the actual fruit of the capsicum plant, being this term used in North America. In Britain it is simply referred to as ‘pepper’, whereas in many countries such as India, Malaysia and Australia it is called ‘capsicum’. In areas such as Scandinavia, Hungary, Germany and Indonesia bell peppers are commonly called ‘paprika’, and in France they are called ‘poivron’, with the same root as poivre (meaning black pepper). ‘Pimento’ or ‘pimenta˜o’ are Portuguese words for bell pepper, while ‘pimenta’ refers both to chilli peppers and to black pepper. The ‘Pimento’ is a variety of large, red, heartshaped chile pepper (Capsicum annuum) that measures 7–10 cm long and 5–7 cm wide (medium, elongate). These ‘pimentos’ are the familiar red stuffing found in green olives. They are eaten green or ripe and are used for salads, soups, stews, relishes and pickling. 163
Vol 85 (C3) 163–170
164
FAUSTINO et al.
Capsicums originated in Central and South America in pre-Columbian times, but the seeds were taken to Spain in 1493 and from there spread to other European and Asian countries, being now grown worldwide. Bell peppers present different nutritional compositions, depending on the variety and stage of maturity, but are naturally rich in ascorbic acid and provitamin A carotenoids, specially the red ripe ones (Kidmose et al., 2006; Lee et al., 2005; Niizu and Rodriguez-Amaya, 2005; Simonne et al., 1997). These two vitamins are powerful antioxidants which have a very important function neutralizing the free radicals in the human body, thus reducing the risk of diseases such as arthritis, cardiovascular disease (Kritchevsky, 1992), cancer (Olson, 1989; Ziegler, 1989; Krinsky, 1989), and delay in the aging process (Packer, 1996). The nutritional and health value of vitamin C is of great importance in the human diet, since it prevents diseases such as scurvy and reduces stress damage, besides its antioxidant activity. It maintains collagen, is important in the biosynthesis of amino acids, the formation of adrenaline and detoxification in the liver (Smirnoff, 1996). However, its amounts diminish with maturation and senescence (Yahia et al., 2001). Bell peppers are also a source of minerals such as calcium, phosphorous, potassium and iron. Its attributed medicinal properties are very diversified: it helps cicatrization, prevents arteriosclerosis (Mezzetti et al., 1995) and haemorrhage, avoids cholesterol and improves physical resistance. They are also rich in flavonoids (Lee et al., 1995, 2005) and other phytochmicals (Duke, 1992). Bell peppers are quite rich in HPO lyase (Shibata et al., 1995; Matsui et al., 1996), which is an enzyme involved in biosynthesis of volatile aldehydes and alcohols (Gardner, 1991), some of them with recognized anti-bacterial benefits (Croft et al., 1993). According to the USDA National Nutrient Database for Standard Reference (Release 19: 2006), the nutritional composition of green sweet peppers in raw state, per 100 g of edible portion is 93.89 g of water, 0.86 g of protein, 0.17 g total lipids (fat), 0.43 g of ash, 6.64 g of carbohydrate, 2.40 g total sugars and 1.70 g total dietary fibre, corresponding to an energy of 20 kcal. The most important minerals amounts are: 10 mg calcium, 175 mg potassium, 20 mg phosphorous and 0.34 mg iron, and the amounts of total ascorbic acid (vitamin C) is 80.4 mg and vitamin A is 370 IU, all per 100 g of edible portion. The bell pepper is used worldwide either as a food or as a condiment. However, like other vegetables, they are quite perishable, originating high losses due to storage problems, marketing and inappropriate processing technologies (Ade-Omowaye et al., 2002; Tunde-Akintunde et al., 2005). One of the most common methods of preservation is drying, which reduces water activity through the decrease of water content, thus preventing deterioration (Vega et al., 2007). The primary preservation method for pepper is drying, which can be achieved by solar drying, hot-air, freeze drying or osmotic dehydration (Abe-Omowaye et al., 2002; Tunde-Akintunde et al., 2005; Vega et al., 2007). Drying combines the effects of physical and chemical stability with the reduction in weight and transportations costs. The moisture removal and its dependence on the process variables is expressed in terms of the drying kinetics,
being essential for development of reliable process models (Guine´ and Fernandes, 2006). The drying process involves simultaneous coupled heat and mass transfer phenomena, and many studies have been conducted for many different food products, considering different types of kinetics: empirical or mechanistic models, or a combination of both (Salgado et al., 1994; Sun and Woods, 1997; Va´zquez and Chenlo, 1997; Madamba, 1997). The drying of red and green bell pepper has been object of only a few studies and even in those, the drying kinetics has only been modelled in terms of empirical model, like in the study by Doymaz and Pala (2002) in which the drying kinetics of red peppers under different pretreatment and air drying conditions was fitted to two different models, the Page and exponential equations. Chenlo et al. (2006) studied the osmotic dehydration kinetics of padro´n pepper. The lack of published work on the air-drying kinetics of green bell pepper, either in terms of empirical models or in terms of diffusivity model explains the interest for the present work. Furthermore, the evaluation of the changes in the nutritional value due to drying were not included in previous works of kinetic studies with peppers. The experiments conducted aimed, on one hand, the study of the hot-air drying of green bell peppers, in terms of drying kinetics evaluated at different temperatures: 308C, 408C, 508C, 608C and 708C. On the other hand, the chemical characterization in fresh and after drying at the lowest and highest temperatures was analysed, for evaluation of the effect of drying and drying temperature on the chemical composition of the product. In this way, the analyses made were: moisture content, total sugar content, protein, ash, fat, crude fibre and acidity.
MATERIALS AND METHODS Experimental Drying Procedure Fresh green bell peppers from a local market were selected, washed and cut to samples of approximately 2.5 cm diameter and 3–4 mm of thickness. These were left in ventilated ovens (WTB-Binder) with an air flux of 300 m3 h21, at constant temperatures of 308C, 408C, 508C, 608C and 708C, until they reached a moisture content under 5%. Each drying experiment was independent, and the green bell peppers used for all were from the same supplier and had the same average initial moisture content. Periodically the samples were removed from each oven in order to measure their average water content with a Mettler Toledo HG53 Halogen Moisture Analyser, which was previously calibrated in terms of optimal operating parameters for this type of food.
Calibration of the Halogen Moisture Analyser The Halogen Moisture Analyser Mettler Toledo HG53 allows a various wide range of operating parameters, depending on the objectives of the analysis and the properties of the product to be analysed. Therefore, the selection of the most appropriate conditions should precede its application to a certain type of product. In this way, the moisture of the green bell peppers was determined by a standard method, to be considered the reference for the product at use. This was determined by drying until constant weight
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
DRYING KINETICS OF GREEN BELL PAPPER in a drying chamber at atmospheric pressure at 1058C, and a medium of three replicates was used. Later, samples of the same peppers were dried in the Halogen Moisture Analyser, varying two operating parameters: temperature and speed. The temperatures tested were 1058C, 1158C, 1258C and 1358C, and the speeds were 2, 3 and 4 (in a scale of 1 to 5, in which 1 is too fast and 5 is too slow).
Chemical Analysis The chemical composition was determined by analysis of moisture, protein, total sugars, ash, fat, crude fibre and acidity, made in duplicate and following standard methods, according to the Association of Official Analytical Chemists (AOAC, 1990). For the analyses made to the fresh or dehydrated product, the peppers were subject to fragmentation thus obtaining a homogeneous product.
Fitting of the Kinetic Data to Empirical Models The data obtained experimentally for the five different temperatures studied was plotted in the form of moisture ratio (MR) versus time, with MR ¼ (W We )=(W0 We )
(1)
where, W is the moisture content at any time t, We is the equilibrium moisture content and W0 is the initial moisture content, all expressed in dry basis (g water/g dry solids). The experimental sets of (MR, t) were fitted to five different empirical models from literature, shown in Table 1, using Sigma Plot, v 8.0 (SPSS, Inc.). To evaluate the quality of each estimation, the statistical parameters determined, apart from the correlation coefficient, R, were the reduced chi-square, x 2, the mean bias error, MBE, an the root mean square error, RMSE, defined by
x2 ¼
N 1 X (Vexp ,i Vprev,i )2 N n i¼1
(2)
N 1X (Vprev,i Vexp ,i ) N i¼1 vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u N u1 X (Vprev,i Vexp ,i )2 RMSE ¼ t N i¼1
(3)
MBE ¼
(4)
the model (Togrul and Pehlivan, 2003; Yaldiz and Erketin, 2001).
Modelling the Drying Kinetics in Terms of Diffusion Model Fick’s second law equation for non steady-state diffusion, assuming that the samples used can be approximated to cylinders, the diffusion is expressed by (Crank, 1970; Yang et al., 2001): @W 1 @ @W @ @W ¼ De r þ De r (5) @t r @r @r @z @z where De is effective moisture diffusivity, r is the cylinder radius, z is the height and t is the time, expressed in seconds. Assuming an uniform initial moisture content and a constant effective diffusivity throughout the sample, the analytical solution of equation (5) is given by 1 X W We 4 b 2n t ¼ exp De 2 r W0 We n¼1 b2n
(6)
where (W 2 We)/(W0 2 We) is the moisture ratio (MR) (Crank, 1970; Zogzas, 1994). Considering only the first term of the series in equation (6), the solution of the Fick’s equation becomes ! 2 b1 W We 4 ¼ t exp D (7) e W0 We r2 b21 For the evaluation of the diffusion coefficient, it was considered that the effective diffusivity varies with temperature according to an Arrhenius function, of the type: E (8) De ¼ D0e exp R(T þ 273:15) where D0e is the diffusivity for an infinite temperature, E is the activation energy for moisture diffusion, R is the gas constant (R ¼ 8.31451 J mol21 K21) and T is the drying temperature (expressed in 8C) (Konishi et al., 2001; Vega et al., 2007). The model described by equations (7) and (8) can be expressed in the simplified form: MR ¼ f (t, T ; b1 , D0e , E)
where Vexp,i and Vprev,i are respectively the experimental and predicted values for the observation i, N is the number of observations and n the number of parameters in
165
(9)
where t and T are variables and b1, D0e and E are the model parameters.
Fitting Procedure Used for the Diffusion Model Table 1. Kinetic models from literature (Togrul and Pehlivan, 2003). Model Newton Page Henderson and Pabis Logarithmic Wang and Singh
Equation MR ¼ exp(2kt) MR ¼ exp(2kt n) MR ¼ a exp(2kt) MR ¼ c þ a exp(2kt) MR ¼ 1 þ at þ bt 2
The parameters of the model were estimated from the experimental data using an explicit orthogonal distance regression (ODR) algorithm with the derivatives approximated by central finite differences, where the unknown errors are taken into account in both dependent and independent variables. If for a set of data (xi, yi), i ¼ 1, . . . , n, (where yi is supposed to be a non-linear function of xi and a set of parameters bk) both xi and yi contain unknown errors (di and 1i, respectively),
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
166
FAUSTINO et al.
Table 2. Results obtained for the operational calibration of the Halogen Moisture Analyser. Temperature (8C)
Speed
Moisture content (%)
2 3 4 2 3 4 2 3 4 2 3 4
84.84 92.26 92.71 88.19 89.18 93.47 19.17 93.38 92.41 90.42 92.50 92.93
105 115 125 135
then the observed value of yi satisfies (Boggs et al., 1992): yi þ 1i ¼ fi (xi þ di ; bk ); i ¼ 1, . . . , n
(10)
for some set of values bk, k ¼ 1, . . . , np. As a result of unknown errors in both the dependent and independent variables, the ODR method finds the best solution of the estimation problem by minimizing the n orthogonal distance from the curve f (x;b) to all data points, i.e. (Boggs et al., 1992): min b
n X
(d2i þ 12i )
Figure 1. Batch drying curves for green bell pepper at temperatures of 308C, 408C, 508C, 608C and 708C.
the estimation of the model parameters. The results of the estimation are complemented with the statistic information that characterises it: the number of observations (N), the sum of square deltas (S d2—referring to the independent variables, t and T), the sum of square epsilons (S 12— referring to the dependent variable, MR), the sum of square errors (Se 2) and the residual standard deviation (RSD).
(11)
j¼1
The software package used to compute the parameters is ODRPACK, developed by the Center for Computing and Applied Mathematics of the National Institute of Standards and Technology (USA) (Boggs et al., 1992). The experimental data for the five temperatures studied was treated together in the form of triplets (MR, t, T) allowing
RESULTS AND DISCUSSION Calibration of the Halogen Moisture Analyser The moisture of the green bell-peppers was first determined by oven drying until constant weight. The medium value obtained from two replicates (94.05% and 93.97%),
Table 3. Results of the fitting of the experimental data to the Newton model.
Parameters k (+sd) Statistical information R N x2 MBE RMSE
308C
408C
0.1274 (+0.0046)
0.2447 (+0.0136)
0.9373 74 5.15E-03 4.88E-04 7.13E-02
0.9320 40 6.19E-03 4.53E-03 7.87E-02
508C 0.5637 (+0.0495) 0.9364 23 6.50E-03 25.26E-03 8.06E-02
608C 0.5777 (+0.0416) 0.9671 18 4.25E-03 3.99E-03 6.52E-02
708C 1.0036 (+0.0706) 0.9864 11 2.23E-03 24.66E-03 4.73E-02
Table 4. Results of the fitting of the experimental data to the Page model.
Parameters k (+sd) n (+sd) Statistical information R N x2 MBE RMSE
308C
408C
508C
608C
708C
0.2876 (+0.0158) 0.6573 (+0.0218)
0.4539 (+0.0336) 0.6489 (+0.0390)
0.8205 (+0.0563) 0.6030 (+0.0529)
0.7438 (+0.0543) 0.7203 (+0.0643)
1.0758 (+0.0559) 0.7676 (+0.0693)
0.9814 74 1.59E-03 7.30E-03 3.93E-02
0.9732 40 2.49E-03 9.82E-03 4.99E-02
0.9789 23 2.21E-03 8.16E-03 4.70E-02
0.9836 18 2.13E-03 9.50E-03 4.62E-02
0.9936 11 1.06E-03 4.08E-03 3.25E-02
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
DRYING KINETICS OF GREEN BELL PAPPER
167
Table 5. Results of the fitting of the experimental data to the Henderson and Pabis model.
Parameters k (+sd) a (+sd) Statistical information R N x2 MBE RMSE
308C
408C
0.7645 (+0.0158) 0.0946 (+0.0029)
0.7773 (+0.0297) 0.1860 (+0.0106)
0.9832 74 1.43E-03 2.28E-04 3.73E-02
0.9712 40 2.68E-03 2.60E-03 5.18E-02
was 94.01% (wet basis) and was used as the reference moisture content for this food product. In Table 2 are presented the values obtained with the measurement made for the operational calibration of the Halogen Moisture Analyser (Mettler Toledo HG53). From the results shown is possible to conclude that the operational conditions that enable to obtain the closest values to the reference moisture content are: temperature of 1158C and speed 4 (showing 93.47%) or temperature 1258C and speed 3 (showing 93.38%). These two sets of conditions could be used, but the one that was selected was the first, because the time of the analysis was also an important criterion, and it was the one selected for taking the final decision. In fact, with 1158C/sp.4 the analysis took more than 30 min against the 20 min of the analysis made with 1258C/sp.3, representing a very important difference. Therefore, every analysis made for the kinetic study was done with the Halogen Moisture Analyser set to 1258C and speed 3.
508C
608C
0.8199 (+0.0538) 0.4446 (+0.0461)
0.8839 (+0.0476) 0.5065 (+0.0436)
0.9566 23 4.48E-03 24.25E-03 6.69E-02
0.9757 18 3.15E-03 9.63E-04 5.62E-02
708C 0.9589 (+0.0464) 0.9603 (+0.0823) 0.9874 11 2.06E-03 26.71E-03 4.54E-02
Empirical Kinetic Models Figure 1 shows the batch drying curves for the drying of green bell pepper at the five temperatures studied (308C, 408C, 508C, 608C and 708C). It is visible how the moisture (expressed in wet basis) follows a sigmoidal shape characteristic of the drying processes, and how the increase in temperature accelerates the drying, reducing the time from 36.5 h at 308C to 5.0 h at 708C. The initial moisture content was 94.01%, being the same for all situations, since the fresh green bell peppers were from the same lot. The drying kinetics data obtained for the five temperatures studied, in the form of moisture ratio versus time, was fitted to five different kinetic models commonly cited in literature, shown in Table 1. The results of such fittings are presented for each model in Tables 3 to 7, which show the values of the estimated parameters with the corresponding standard deviation, as well as the statistical information which characterizes each fitting.
Table 6. Results of the fitting of the experimental data to the logarithmic model.
Parameters k (+sd) a (+sd) c (+sd) Statistical information R N x2 MBE RMSE
308C
408C
20.0023 (+0.0134) 0.7653 (+0.0168) 0.0937 (+0.0059)
20.0237 (+0.0264) 0.7875 (+0.0315) 0.1691 (+0.0195)
0.9833 74 1.45E-03 23.02E-06 3.73E-02
0.9718 40 2.62E-03 25.61E-06 5.12E-02
508C 0.0191 (+0.0262) 0.8170 (+0.0574) 0.4898 (+0.0765) 0.9575 23 4.39E-03 23.89E-06 6.63E-02
608C
708C
20.0050 (+0.0319) 0.8862 (+0.0517) 0.4969 (+0.0730)
0.0267 (+0.0267) 0.9446 (+0.0511) 1.0647 (+0.1374)
0.9757 18 3.15E-03 7.73E-06 5.61E-02
0.9886 11 2.07E-03 1.42E-06 4.54E-02
Table 7. Results of the fitting of the experimental data to the Wang and Singh model.
Parameters a (+sd) b (+sd) Statistical information R N x2 MBE RMSE
308C
408C
508C
608C
708C
20.0768 (+0.0026) 0.0015 (+0.0001)
20.1436 (+0.0070) 0.0051 (+0.0005)
20.2761 (+0.0194) 0.0180 (+0.0022)
20.3334 (+0.0221) 0.0268 (+0.0033)
20.5647 (+0.0488) 0.0768 (+0.0120)
0.9851 74 1.42E-02 3.68E-02 1.17E-01
0.8249 40 1.51E-02 3.77E-02 1.23E-01
0.7850 23 2.03E-02 4.14E-02 1.42E-01
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
0.9037 18 1.20E-02 3.08E-02 1.10E-01
0.9274 11 1.16E-02 2.61E-02 1.08E-01
168
FAUSTINO et al.
Figure 2. Fitting of the experimental data at 308C with the different models tested.
Figure 4. Fitting of the experimental data at all temperatures with the Newton model.
Figure 2 shows the experimental points obtained for one selected temperature (308C, because was the curve with more data) together with the plots corresponding to the fittings obtained for all the models tested. It can be observed that the Logarithmic and Henderson and Pabis curves are perfectly coincident, and that, despite showing a good performance for t . 0, do not represent the initial state (t ¼ 0). Moreover, from Table 6 it is possible to see that although the correlation coefficients of the fittings are not too bad, the significance of the values estimated for the parameter k in the logarithmic model is poor, since the standard deviation is higher than the value itself. The Wang and Singh model does not show a good prediction capacity, since the corresponding curve is clearly away from the experimental points, and besides, the trend of this model is also not adequate to represent the decreasing function that the experimental data shows towards the end of drying. In this way, the models that seem best to describe the experimental behaviour observed are the Newton and Page models, and of these the Page is better either by observing the curves in Figure 1 or by comparing the statistical information from
Tables 3 and 4 (namely R, x2 and RMSE, since MBE is not squared). Considering that Page is one of the best models applicable to the drying of green bell pepper, in Figure 3 is represented the drying data for all temperatures studied, together with the fitting obtained for each temperature with the Page model. The curves show a fairly good performance over the entire range of time, which varies from according to the temperature. However, at the final stages the model tends to give higher predictions than the values observed. In this way, the Newton model, represented in Figure 4, shows a better adjustment at the final stages of drying, nevertheless does not allow to distinguish the curves at 508C and 608C.
Diffusion Model Table 8 shows the results of the fitting to the diffusion model expressed by equations (7) and (8). The values of the standard deviations of each parameter together with the statistical information allow to infer that the quality of the estimation is good, with low values of the sum of square errors (Se 2) and residual standard deviation (RSD). The value obtained for the diffusion coefficient at an infinite temperature, D0e was 0.1176 m2 s21, and the activation energy for moisture diffusion, E, was found to be 47.10 kJ mol21. This later is in accordance with other values reported in literature for some varieties of red bell pepper: 28.4 kJ mol21 (Turhan
Table 8. Results obtained with ODR program to fit the diffusion model [equations (7) and (8)]. Parameters
Figure 3. Fitting of the experimental data at all temperatures with the Page model.
b1 D0e (m2 s21) E(J mol21) Statistical information N Sd2 (t,T) S12 (MR) Se2 RSD
Value
Standard deviation
2.2319 0.1176 47 101
0.0219 0.0713 1 574 166 4.4104 1025 5.36119 1021 5.3615 1021 5.7352 1022
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
DRYING KINETICS OF GREEN BELL PAPPER
169
Table 10. Effect of drying on the nutritional value of green bellpeppers. Component Moisture (g/100 g product) Protein (g/100 g dry solids)1 Total sugars (g/100 g dry solids) Fat (g/100 g dry solids) Crude fibre (g/100 g dry solids) Ash (g/100 g dry solids) Acidity (cm3/100 g dry solids)
Fresh product
Dehydrated at 308C
Dehydrated at 708C
94.16
3.25
4.71
17.98
11.82
1.40
85.27
4.36
4.58
6.50 5.14
1.06 7.30
0.90 7.73
5.14 10.27
6.78 5.65
6.42 5.55
1
Factor used for the conversion of nitrogen in protein ¼ 6.25.
Figure 5. Variation of the diffusion coefficient with temperature.
and Turhan, 1997), 44.8 kJ mol21(Sanjua´n et al, 2003), 42.8 kJ mol21 (Kaymak-Ertekin, 2002), 41.95 (Gupta et al., 2002), 39.70 (Vega et al., 2007). If Figure 5 the effective diffusivity of green bell peppers is represented in the range of temperatures studied (between 308C and 708C). The experimental points were obtained for each temperature separately, by applying linear regression to the data in the form of ln(MR) versus time plots. The curve expressing the variation of diffusivity with temperature was obtained from equation (8) with the parameters estimated and presented in Table 8. From the graph in Figure 5 is possible to verify the increase in diffusivity when the temperature is raised. For range of temperatures studied, the effective diffusivity varies between 9.0 10210 m2 s21at 308C and 8.0 1029 m2 s21 at 708C. For example, at 508C the value obtained is 2.9 1029, which is very close to that found by Vega et al. (2007) for the same temperature for the drying of red bell pepper (3.2 1029 m2 s21).
Nutritional Evaluation Table 9 shows the chemical composition of fresh green peppers, both analysed and reference. It can be seen that in general terms the results obtained in the present study are in accordance with the reference values. However, some small differences can be pointed out when comparing
Table 9. Chemical characterization of fresh green peppers (g/100 g edible portion). Component
Reference1
Analysed
Moisture Protein Total sugars Fat Dietary fibre Ash
93.89 0.86 2.40 0.17 1.70 0.43
94.01 1.052 4.98 0.38 0.733 0.30
1 USDA National Nutrient Database for Standard Reference (Release 19: 2006). 2 Factor used for the conversion of nitrogen in protein ¼ 6.25. 3 Crude fibre.
the values obtained with the reference, in particular slightly higher contents of protein and sugar, and lower amounts of fibre and minerals, which can be attributed to differences in the variety analysed. In Table 10 the chemical composition of the peppers, fresh and dehydrated at the lowest and highest temperatures, are shown for comparison purposes. From the values is possible to see that, except for the protein content, the chemical compositions of both dehydrated peppers (at 308C and 708C) are very similar, meaning that for the range of temperatures studied, the value of the temperature chosen for the process does not influence significantly the nutritional value of the final product. On the other hand, the changes observed by comparing the fresh with the dehydrated products are very significant for all components evaluated, with the exception of fibre and ash. In this way, the drying operation induces reductions of 34% in proteins, 95% in sugars, 84% in fat and 45% in acidity. In fact, the heat provided during the drying process originates the degradation of sugars in a very important extension, favouring also a significant loss in the acids (especially volatile acids) and causing some protein denaturising.
CONCLUSIONS From the results obtained, it was concluded that the empirical models that best describe the dehydration kinetics for green bell peppers are Page and Newton. For the drying process modelled in terms of Fick’s diffusion law, the values obtained for the diffusion coefficient at an infinite temperature, D0e, and activation energy for moisture diffusion, E, were, respectively, 0.1176 m2 s21and 47.10 kJ mol21. These values allow us to estimate the effective diffusivity, which ranges between 9.0 10210 and 8.0 1029 m2 s21, for the temperatures in the range 308C to 708C. From the chemical analyses made was possible to possible to conclude that drying influences the chemical composition of the peppers (due to volatilization of some components, oxidation processes and protein denaturizing). On the other hand, the influence of the drying temperature was not very significant on the chemical composition, and has the advantage of largely reducing the drying time.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
170
FAUSTINO et al. REFERENCES
Ade-Omowaye, B.I.O., Rastogi, N.K., Angersbach, A. and Knorr, D., 2002, Osmotic dehydration of bell peppers: influence of high intensity electric field pulses and elevated temperature treatment, Journal of food Engineering, 54: 35–43. Association of Official Analytical Chemists (AOAC), 1990, Official Methods of Analysis, 15th edition, (AOAC, Arlington). Boggs, P.T., Byrd, R.H., Rogers, J.E. and Schnabl, R.B., 1992, User’s Reference Guide for ODRPACK. Software for Weighted Orthogonal Distance Regression, Version 2.01. (National Institute of Standards and Technology, Gaithersburg, USA). Chenlo, F., Changuri, L., Santos, F. and Moreira, R., 2006, Osmotic dehydration/impregnation kinetics of padro´n pepper (Capsicum annuum L. Longum) with sodium chloride solutions: process modelling and colour analysis, Food Science and Technology International, 12(3): 221–227. Crank, J., 1970, The Mathematics of Diffusion (Oxford University Press, London, UK). Croft, K.P.C., Juttner, F. and Slusarenko, A.J., 1993, Volatile products of the lipoxygenase pathway evolved from phaseolus vulgaris (L.) leaves inoculated with pseudomonas syringae pv phaseolicola, Plant Physiol, 101: 13– 24. Doymaz, I. and Pala, M., 2002, Hot-air drying characteristics of red pepper, Journal of Food Engineering, 55(4): 331 –335. Duke, J.A., 1992, Handbook of Biologically Active Phytochemicals and Their Activities (CRC Press, Boca Raton, Florida, USA). Gardner, H.W., 1991, Recent investigations into the lipoxygenase. Pathway of plants, Biochim Biophys Acta, 1084: 221–239. Guine´, R.P.F. and Fernandes, R.M.C., 2006, Analysis of the drying kinetics of chestnuts, Journal of Food Engineering, 76: 460–467. Gupta, P., Ahmed, J., Shivhare, U.S. and Raghavan, G.S.V., 2002, Drying characteristics of red chilli, Drying Technology, 20: 1975–1987. Kaymak-Ertekin, F., 2002, Drying and rehydrating kinetics of green and red bell peppers, J Food Sci, 67(1): 168 –175. Kidmose, U., Yang, R.Y., Thilsted, S.H., Christensen, L.P. and Brandt, K., 2006, Content of carotenoids in commonly Asian vegetables and stability and extractability during frying, Journal of Food Composition and Analysis, 19: 562–571. Konishi, Y., Oriuchi, J.-I. and Kobayashi, M., 2001, Dynamic evaluation of the dehydration response curves of food characterized by a poultrice-up process using a fish-paste sausage. II. A new tank model for computer simulation, Drying Technol, 19(7): 1271– 1285. Krinsky, N.I., 1989, Carotenoids and cancer in animal models, The Journal of Nutrition, 119: 123–126. Kritchevsky, D., 1992, Antioxidant vitamins in the prevention of cardiovascular disease, Nutrition Today, 27: 30– 33. Lee, J.J., Crosby, K.M., Pike, L.M., Yoo, K.S. and Lescobar, D.I., 2005, Impact of genetic and environmental variation of development of flavonoids and carotenoids in pepper (Capsicum spp.), Scientia Horticulturae, 106: 341 –352. Lee, Y., Howard, L.R. and Villalon, B., 1995, Flavonoid and ascorbic acid content and antioxidant activity of fresh pepper (capsicum annuum) cultivars, IFT Abstract, 55: 79. Madamba, P.S., 1997, Optimization of the drying process: an application to the drying of garlic, Drying Technol, 15(1): 117– 136. Matsui, K., Shibutani, M., Hase, T. and Kajiwara, T., 1996, Bell pepper fruit fatty acid hydroperoxide lyase is a cytochrome P450 (CYP74B), FEBS Lett, 394: 21– 24. Mezzetti, A., Lapenna, D., Pierdomenico, S.D., Calafiore, A.M., Constantini, F., Riario-Sforza, G., Imbastaro, T., Neri, M. and
Cuccurullo, F., 1995, Vitamin E, C and lipid peroxidation in plasma and arterial tissue of smokers and non-smokers, Atheroscletrosis, 112: 91–99. Niizu, P. and Rodriguez-Amaya, D.B., 2005, New data on the carotenoid composition of raw salad vegetables, Journal of Food Composition and Analysis, 18: 739–749. Olson, J.A., 1989, Biological actions of carotenoids, The Journal of Nutrition, 119: 94–95. Packer, L., 1996, Possible factors influencing aging, United Nations and World Health Organization First Joint Conference on Healthy Aging, New York, NY, 29 April –1 May. Salgado, M.A., Lebert, A., Garcia, H.S., Muchnik, J. and Bimbenet, J.J., 1994, Development of the characteristic drying curve for cassava chips in monolayer, Drying Tecnology, 12(3): 685 – 696. Sanjua´n, N., Lozano, M., Garcı´a-Pascual, P. and Mulet, A., 2003, Dehydration kinetics of red bell pepper (Capsicum annuum L var Jaranda), Journal of The Science of Food and Agriculture, 83: 697– 701. Shibata, Y., Matsui, K., Kajiwara, T. and Hatanaka, A., 1995, Purification and properties of fatty acid hydroperoxide lyase from green bell pepper fruits, Plant Cell Physiol, 36: 147– 156. Simonne, A.H., Simonne, E.H., Eitenmiller, R.R., Mills, H.A. and Green, N.R., 1997, Ascorbic acid and provitamin A contents in unusually colored bell peppers (Capsicum annuum L.), Journal of Food Composition and Analysis, 10: 299– 311. Smirnoff, N., 1996, The function and metabolism of ascorbic acid in plants, (Annals of Botany, 7: 661– 669. Sun, D.-W. and Woods, J.L., 1997, Simulation of the heat and moisture transfer process during drying in deep grain beds, Drying Technol, 15(10): 2479–2508. Togrul, I.T. and Pehlivan, D., 2003, Modelling of drying kinetics of single apricot, Journal of Food Engineering, 58(3): 23–32. Tunde-Akintunde, T.Y., Afolabi, T.J. and Akintunde, B.O., 2005, Influence of drying methods on drying of belle pepper (Capsicum annuum), Journal of Food Engineering, 68: 439– 442. Turhan, M. and Turhan, K.N., 1997, Drying kinetics of red bell pepper, Journal of Food Processing and Preservation, 21: 209–223. Va´zquez, G. and Chenlo, F., 1997, Grape drying in a pilot plant with a heat pump, Drying Technology, 15(3&4): 899– 920. Vega, A., Fito, P., Andre´s, A. and Lemus, R., 2007, Mathematical modeling of hot-air drying kinetics of red bell pepper (var. Lamuyo), Journal of Food Engineering, 79: 1460– 1466. Yahia, E.M., Contreras-Padilla, M. and Gonzalez-Aguilar, G., 2001, Ascorbic acid content in relation to ascorbic acid oxidase activity and polyamine content in tomato and bell pepper fruits during development, maturation and senescence, Lebensm-Wiss U-Technol, 34: 452–457. Yaldiz, O. and Erketin, C., 2001, Thin layer solar drying of some different vegetables, Drying Technology, 19(3): 583– 596. Yang, H., Sakai, N. and Watanabe, M., 2001, Drying model with non-isotropic shrinkage deformation undergoing simultaneous heat and mass transfer, Drying Technol, 19(7): 1441– 1460. Ziegler, R.G., 1989, A review of epidemiologic evidence that carotenoids reduce the risk of cancer, The Journal of Nutrition, 119: 116–122. Zogzas, N.P., Maroulis, Z.B. and Marinos-Kouris, D., 1994, Moisture diffusivity methods of experimental determination. A review, Drying Technology, 12(3): 483– 515. The manuscript was received 30 January 2007 and accepted for publication after revision 7 June 2007.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
Abstract: The present work aimed, on one hand, the study of the drying of green peppers, in terms of drying kinetics evaluated at 308C, 408C, 508C, 608C and 708C, having the experimental data been fitted to different empirical kinetic models from literature. This kinetic study was then complemented with the modelling in terms of Fick’s diffusion equation. On the other hand, the chemical characterization in fresh and after drying at the lowest and highest temperatures was analysed, for evaluation of the effect of drying and drying temperature on the chemical composition of the product. In this way, the analyses made were: moisture content, sugar content, proteins, ash, fat, fibre and acidity. From the results obtained, it was concluded that the empirical models that best describe the dehydration kinetics were the Page and Newton models. From the experimental data was possible to estimate the diffusivities, which range between 9.0 10210 at 308C and 8.0 1029 m2 s21 at 708C. Moreover, it was verified that drying influences the chemical composition of the peppers, but, conversely, the influence of the drying temperature was not very significant. Keywords: bell pepper; capsicum annuum; dried bell pepper; drying kinetics; chemical characterization; nutritional value.
INTRODUCTION
Correspondence to: Professor Raquel P.F. Guine´, Quinta da Alagoa. Estrada de Nelas. Ranhados. 3500-606 Viseu. Portugal. E-mail: raquelguine@ esav.ipv.pt
DOI: 10.1205/fbp07009 0960–3085/07/ $30.00 þ 0.00 Food and Bioproducts Processing Trans IChemE, Part C, September 2007 # 2007 Institution of Chemical Engineers
They are fleshy and heavily seeded, and mature from green to red. They vary in size, colour and flavour and the level of heat as measured in Scoville heat units. The higher the number of Scoville heat units, the hotter the pepper. The fruit of most species of Capsicum contains capsaicin (methyl vanillyl nonenamide), a lipophilic chemical that can produce a strong burning sensation in the mouth. Bell peppers contain a recessive gene that eliminates the capsaicin in the fruit. The colour can be green, red, yellow, orange and, more rarely, white, purple, blue, black and brown, depending on when they are harvested. Green peppers are unripe bell peppers, while the others are all ripe, with the colour variation based on cultivar selection (Simonne et al., 1997). Because they are unripe, green peppers are less sweet and slightly more bitter than yellow, orange or red peppers, which all have a rather similar hot taste. The taste of ripe peppers can also vary with growing conditions and post-harvest storage treatment. The fruits allowed to ripen fully on the plant in full sunshine are the sweetest, while the fruits harvested green and after-ripened in storage are less sweet.
Capsicum is a genus of plants from the nightshade family (Solanaceae). Some of them are used as spices, vegetables or even medicines. The fruit of Capsicum plants can have a variety of names depending on place and type, being commonly called chilli pepper, red or green pepper or just pepper. The bell pepper refers to the actual fruit of the capsicum plant, being this term used in North America. In Britain it is simply referred to as ‘pepper’, whereas in many countries such as India, Malaysia and Australia it is called ‘capsicum’. In areas such as Scandinavia, Hungary, Germany and Indonesia bell peppers are commonly called ‘paprika’, and in France they are called ‘poivron’, with the same root as poivre (meaning black pepper). ‘Pimento’ or ‘pimenta˜o’ are Portuguese words for bell pepper, while ‘pimenta’ refers both to chilli peppers and to black pepper. The ‘Pimento’ is a variety of large, red, heartshaped chile pepper (Capsicum annuum) that measures 7–10 cm long and 5–7 cm wide (medium, elongate). These ‘pimentos’ are the familiar red stuffing found in green olives. They are eaten green or ripe and are used for salads, soups, stews, relishes and pickling. 163
Vol 85 (C3) 163–170
164
FAUSTINO et al.
Capsicums originated in Central and South America in pre-Columbian times, but the seeds were taken to Spain in 1493 and from there spread to other European and Asian countries, being now grown worldwide. Bell peppers present different nutritional compositions, depending on the variety and stage of maturity, but are naturally rich in ascorbic acid and provitamin A carotenoids, specially the red ripe ones (Kidmose et al., 2006; Lee et al., 2005; Niizu and Rodriguez-Amaya, 2005; Simonne et al., 1997). These two vitamins are powerful antioxidants which have a very important function neutralizing the free radicals in the human body, thus reducing the risk of diseases such as arthritis, cardiovascular disease (Kritchevsky, 1992), cancer (Olson, 1989; Ziegler, 1989; Krinsky, 1989), and delay in the aging process (Packer, 1996). The nutritional and health value of vitamin C is of great importance in the human diet, since it prevents diseases such as scurvy and reduces stress damage, besides its antioxidant activity. It maintains collagen, is important in the biosynthesis of amino acids, the formation of adrenaline and detoxification in the liver (Smirnoff, 1996). However, its amounts diminish with maturation and senescence (Yahia et al., 2001). Bell peppers are also a source of minerals such as calcium, phosphorous, potassium and iron. Its attributed medicinal properties are very diversified: it helps cicatrization, prevents arteriosclerosis (Mezzetti et al., 1995) and haemorrhage, avoids cholesterol and improves physical resistance. They are also rich in flavonoids (Lee et al., 1995, 2005) and other phytochmicals (Duke, 1992). Bell peppers are quite rich in HPO lyase (Shibata et al., 1995; Matsui et al., 1996), which is an enzyme involved in biosynthesis of volatile aldehydes and alcohols (Gardner, 1991), some of them with recognized anti-bacterial benefits (Croft et al., 1993). According to the USDA National Nutrient Database for Standard Reference (Release 19: 2006), the nutritional composition of green sweet peppers in raw state, per 100 g of edible portion is 93.89 g of water, 0.86 g of protein, 0.17 g total lipids (fat), 0.43 g of ash, 6.64 g of carbohydrate, 2.40 g total sugars and 1.70 g total dietary fibre, corresponding to an energy of 20 kcal. The most important minerals amounts are: 10 mg calcium, 175 mg potassium, 20 mg phosphorous and 0.34 mg iron, and the amounts of total ascorbic acid (vitamin C) is 80.4 mg and vitamin A is 370 IU, all per 100 g of edible portion. The bell pepper is used worldwide either as a food or as a condiment. However, like other vegetables, they are quite perishable, originating high losses due to storage problems, marketing and inappropriate processing technologies (Ade-Omowaye et al., 2002; Tunde-Akintunde et al., 2005). One of the most common methods of preservation is drying, which reduces water activity through the decrease of water content, thus preventing deterioration (Vega et al., 2007). The primary preservation method for pepper is drying, which can be achieved by solar drying, hot-air, freeze drying or osmotic dehydration (Abe-Omowaye et al., 2002; Tunde-Akintunde et al., 2005; Vega et al., 2007). Drying combines the effects of physical and chemical stability with the reduction in weight and transportations costs. The moisture removal and its dependence on the process variables is expressed in terms of the drying kinetics,
being essential for development of reliable process models (Guine´ and Fernandes, 2006). The drying process involves simultaneous coupled heat and mass transfer phenomena, and many studies have been conducted for many different food products, considering different types of kinetics: empirical or mechanistic models, or a combination of both (Salgado et al., 1994; Sun and Woods, 1997; Va´zquez and Chenlo, 1997; Madamba, 1997). The drying of red and green bell pepper has been object of only a few studies and even in those, the drying kinetics has only been modelled in terms of empirical model, like in the study by Doymaz and Pala (2002) in which the drying kinetics of red peppers under different pretreatment and air drying conditions was fitted to two different models, the Page and exponential equations. Chenlo et al. (2006) studied the osmotic dehydration kinetics of padro´n pepper. The lack of published work on the air-drying kinetics of green bell pepper, either in terms of empirical models or in terms of diffusivity model explains the interest for the present work. Furthermore, the evaluation of the changes in the nutritional value due to drying were not included in previous works of kinetic studies with peppers. The experiments conducted aimed, on one hand, the study of the hot-air drying of green bell peppers, in terms of drying kinetics evaluated at different temperatures: 308C, 408C, 508C, 608C and 708C. On the other hand, the chemical characterization in fresh and after drying at the lowest and highest temperatures was analysed, for evaluation of the effect of drying and drying temperature on the chemical composition of the product. In this way, the analyses made were: moisture content, total sugar content, protein, ash, fat, crude fibre and acidity.
MATERIALS AND METHODS Experimental Drying Procedure Fresh green bell peppers from a local market were selected, washed and cut to samples of approximately 2.5 cm diameter and 3–4 mm of thickness. These were left in ventilated ovens (WTB-Binder) with an air flux of 300 m3 h21, at constant temperatures of 308C, 408C, 508C, 608C and 708C, until they reached a moisture content under 5%. Each drying experiment was independent, and the green bell peppers used for all were from the same supplier and had the same average initial moisture content. Periodically the samples were removed from each oven in order to measure their average water content with a Mettler Toledo HG53 Halogen Moisture Analyser, which was previously calibrated in terms of optimal operating parameters for this type of food.
Calibration of the Halogen Moisture Analyser The Halogen Moisture Analyser Mettler Toledo HG53 allows a various wide range of operating parameters, depending on the objectives of the analysis and the properties of the product to be analysed. Therefore, the selection of the most appropriate conditions should precede its application to a certain type of product. In this way, the moisture of the green bell peppers was determined by a standard method, to be considered the reference for the product at use. This was determined by drying until constant weight
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
DRYING KINETICS OF GREEN BELL PAPPER in a drying chamber at atmospheric pressure at 1058C, and a medium of three replicates was used. Later, samples of the same peppers were dried in the Halogen Moisture Analyser, varying two operating parameters: temperature and speed. The temperatures tested were 1058C, 1158C, 1258C and 1358C, and the speeds were 2, 3 and 4 (in a scale of 1 to 5, in which 1 is too fast and 5 is too slow).
Chemical Analysis The chemical composition was determined by analysis of moisture, protein, total sugars, ash, fat, crude fibre and acidity, made in duplicate and following standard methods, according to the Association of Official Analytical Chemists (AOAC, 1990). For the analyses made to the fresh or dehydrated product, the peppers were subject to fragmentation thus obtaining a homogeneous product.
Fitting of the Kinetic Data to Empirical Models The data obtained experimentally for the five different temperatures studied was plotted in the form of moisture ratio (MR) versus time, with MR ¼ (W We )=(W0 We )
(1)
where, W is the moisture content at any time t, We is the equilibrium moisture content and W0 is the initial moisture content, all expressed in dry basis (g water/g dry solids). The experimental sets of (MR, t) were fitted to five different empirical models from literature, shown in Table 1, using Sigma Plot, v 8.0 (SPSS, Inc.). To evaluate the quality of each estimation, the statistical parameters determined, apart from the correlation coefficient, R, were the reduced chi-square, x 2, the mean bias error, MBE, an the root mean square error, RMSE, defined by
x2 ¼
N 1 X (Vexp ,i Vprev,i )2 N n i¼1
(2)
N 1X (Vprev,i Vexp ,i ) N i¼1 vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u N u1 X (Vprev,i Vexp ,i )2 RMSE ¼ t N i¼1
(3)
MBE ¼
(4)
the model (Togrul and Pehlivan, 2003; Yaldiz and Erketin, 2001).
Modelling the Drying Kinetics in Terms of Diffusion Model Fick’s second law equation for non steady-state diffusion, assuming that the samples used can be approximated to cylinders, the diffusion is expressed by (Crank, 1970; Yang et al., 2001): @W 1 @ @W @ @W ¼ De r þ De r (5) @t r @r @r @z @z where De is effective moisture diffusivity, r is the cylinder radius, z is the height and t is the time, expressed in seconds. Assuming an uniform initial moisture content and a constant effective diffusivity throughout the sample, the analytical solution of equation (5) is given by 1 X W We 4 b 2n t ¼ exp De 2 r W0 We n¼1 b2n
(6)
where (W 2 We)/(W0 2 We) is the moisture ratio (MR) (Crank, 1970; Zogzas, 1994). Considering only the first term of the series in equation (6), the solution of the Fick’s equation becomes ! 2 b1 W We 4 ¼ t exp D (7) e W0 We r2 b21 For the evaluation of the diffusion coefficient, it was considered that the effective diffusivity varies with temperature according to an Arrhenius function, of the type: E (8) De ¼ D0e exp R(T þ 273:15) where D0e is the diffusivity for an infinite temperature, E is the activation energy for moisture diffusion, R is the gas constant (R ¼ 8.31451 J mol21 K21) and T is the drying temperature (expressed in 8C) (Konishi et al., 2001; Vega et al., 2007). The model described by equations (7) and (8) can be expressed in the simplified form: MR ¼ f (t, T ; b1 , D0e , E)
where Vexp,i and Vprev,i are respectively the experimental and predicted values for the observation i, N is the number of observations and n the number of parameters in
165
(9)
where t and T are variables and b1, D0e and E are the model parameters.
Fitting Procedure Used for the Diffusion Model Table 1. Kinetic models from literature (Togrul and Pehlivan, 2003). Model Newton Page Henderson and Pabis Logarithmic Wang and Singh
Equation MR ¼ exp(2kt) MR ¼ exp(2kt n) MR ¼ a exp(2kt) MR ¼ c þ a exp(2kt) MR ¼ 1 þ at þ bt 2
The parameters of the model were estimated from the experimental data using an explicit orthogonal distance regression (ODR) algorithm with the derivatives approximated by central finite differences, where the unknown errors are taken into account in both dependent and independent variables. If for a set of data (xi, yi), i ¼ 1, . . . , n, (where yi is supposed to be a non-linear function of xi and a set of parameters bk) both xi and yi contain unknown errors (di and 1i, respectively),
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
166
FAUSTINO et al.
Table 2. Results obtained for the operational calibration of the Halogen Moisture Analyser. Temperature (8C)
Speed
Moisture content (%)
2 3 4 2 3 4 2 3 4 2 3 4
84.84 92.26 92.71 88.19 89.18 93.47 19.17 93.38 92.41 90.42 92.50 92.93
105 115 125 135
then the observed value of yi satisfies (Boggs et al., 1992): yi þ 1i ¼ fi (xi þ di ; bk ); i ¼ 1, . . . , n
(10)
for some set of values bk, k ¼ 1, . . . , np. As a result of unknown errors in both the dependent and independent variables, the ODR method finds the best solution of the estimation problem by minimizing the n orthogonal distance from the curve f (x;b) to all data points, i.e. (Boggs et al., 1992): min b
n X
(d2i þ 12i )
Figure 1. Batch drying curves for green bell pepper at temperatures of 308C, 408C, 508C, 608C and 708C.
the estimation of the model parameters. The results of the estimation are complemented with the statistic information that characterises it: the number of observations (N), the sum of square deltas (S d2—referring to the independent variables, t and T), the sum of square epsilons (S 12— referring to the dependent variable, MR), the sum of square errors (Se 2) and the residual standard deviation (RSD).
(11)
j¼1
The software package used to compute the parameters is ODRPACK, developed by the Center for Computing and Applied Mathematics of the National Institute of Standards and Technology (USA) (Boggs et al., 1992). The experimental data for the five temperatures studied was treated together in the form of triplets (MR, t, T) allowing
RESULTS AND DISCUSSION Calibration of the Halogen Moisture Analyser The moisture of the green bell-peppers was first determined by oven drying until constant weight. The medium value obtained from two replicates (94.05% and 93.97%),
Table 3. Results of the fitting of the experimental data to the Newton model.
Parameters k (+sd) Statistical information R N x2 MBE RMSE
308C
408C
0.1274 (+0.0046)
0.2447 (+0.0136)
0.9373 74 5.15E-03 4.88E-04 7.13E-02
0.9320 40 6.19E-03 4.53E-03 7.87E-02
508C 0.5637 (+0.0495) 0.9364 23 6.50E-03 25.26E-03 8.06E-02
608C 0.5777 (+0.0416) 0.9671 18 4.25E-03 3.99E-03 6.52E-02
708C 1.0036 (+0.0706) 0.9864 11 2.23E-03 24.66E-03 4.73E-02
Table 4. Results of the fitting of the experimental data to the Page model.
Parameters k (+sd) n (+sd) Statistical information R N x2 MBE RMSE
308C
408C
508C
608C
708C
0.2876 (+0.0158) 0.6573 (+0.0218)
0.4539 (+0.0336) 0.6489 (+0.0390)
0.8205 (+0.0563) 0.6030 (+0.0529)
0.7438 (+0.0543) 0.7203 (+0.0643)
1.0758 (+0.0559) 0.7676 (+0.0693)
0.9814 74 1.59E-03 7.30E-03 3.93E-02
0.9732 40 2.49E-03 9.82E-03 4.99E-02
0.9789 23 2.21E-03 8.16E-03 4.70E-02
0.9836 18 2.13E-03 9.50E-03 4.62E-02
0.9936 11 1.06E-03 4.08E-03 3.25E-02
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
DRYING KINETICS OF GREEN BELL PAPPER
167
Table 5. Results of the fitting of the experimental data to the Henderson and Pabis model.
Parameters k (+sd) a (+sd) Statistical information R N x2 MBE RMSE
308C
408C
0.7645 (+0.0158) 0.0946 (+0.0029)
0.7773 (+0.0297) 0.1860 (+0.0106)
0.9832 74 1.43E-03 2.28E-04 3.73E-02
0.9712 40 2.68E-03 2.60E-03 5.18E-02
was 94.01% (wet basis) and was used as the reference moisture content for this food product. In Table 2 are presented the values obtained with the measurement made for the operational calibration of the Halogen Moisture Analyser (Mettler Toledo HG53). From the results shown is possible to conclude that the operational conditions that enable to obtain the closest values to the reference moisture content are: temperature of 1158C and speed 4 (showing 93.47%) or temperature 1258C and speed 3 (showing 93.38%). These two sets of conditions could be used, but the one that was selected was the first, because the time of the analysis was also an important criterion, and it was the one selected for taking the final decision. In fact, with 1158C/sp.4 the analysis took more than 30 min against the 20 min of the analysis made with 1258C/sp.3, representing a very important difference. Therefore, every analysis made for the kinetic study was done with the Halogen Moisture Analyser set to 1258C and speed 3.
508C
608C
0.8199 (+0.0538) 0.4446 (+0.0461)
0.8839 (+0.0476) 0.5065 (+0.0436)
0.9566 23 4.48E-03 24.25E-03 6.69E-02
0.9757 18 3.15E-03 9.63E-04 5.62E-02
708C 0.9589 (+0.0464) 0.9603 (+0.0823) 0.9874 11 2.06E-03 26.71E-03 4.54E-02
Empirical Kinetic Models Figure 1 shows the batch drying curves for the drying of green bell pepper at the five temperatures studied (308C, 408C, 508C, 608C and 708C). It is visible how the moisture (expressed in wet basis) follows a sigmoidal shape characteristic of the drying processes, and how the increase in temperature accelerates the drying, reducing the time from 36.5 h at 308C to 5.0 h at 708C. The initial moisture content was 94.01%, being the same for all situations, since the fresh green bell peppers were from the same lot. The drying kinetics data obtained for the five temperatures studied, in the form of moisture ratio versus time, was fitted to five different kinetic models commonly cited in literature, shown in Table 1. The results of such fittings are presented for each model in Tables 3 to 7, which show the values of the estimated parameters with the corresponding standard deviation, as well as the statistical information which characterizes each fitting.
Table 6. Results of the fitting of the experimental data to the logarithmic model.
Parameters k (+sd) a (+sd) c (+sd) Statistical information R N x2 MBE RMSE
308C
408C
20.0023 (+0.0134) 0.7653 (+0.0168) 0.0937 (+0.0059)
20.0237 (+0.0264) 0.7875 (+0.0315) 0.1691 (+0.0195)
0.9833 74 1.45E-03 23.02E-06 3.73E-02
0.9718 40 2.62E-03 25.61E-06 5.12E-02
508C 0.0191 (+0.0262) 0.8170 (+0.0574) 0.4898 (+0.0765) 0.9575 23 4.39E-03 23.89E-06 6.63E-02
608C
708C
20.0050 (+0.0319) 0.8862 (+0.0517) 0.4969 (+0.0730)
0.0267 (+0.0267) 0.9446 (+0.0511) 1.0647 (+0.1374)
0.9757 18 3.15E-03 7.73E-06 5.61E-02
0.9886 11 2.07E-03 1.42E-06 4.54E-02
Table 7. Results of the fitting of the experimental data to the Wang and Singh model.
Parameters a (+sd) b (+sd) Statistical information R N x2 MBE RMSE
308C
408C
508C
608C
708C
20.0768 (+0.0026) 0.0015 (+0.0001)
20.1436 (+0.0070) 0.0051 (+0.0005)
20.2761 (+0.0194) 0.0180 (+0.0022)
20.3334 (+0.0221) 0.0268 (+0.0033)
20.5647 (+0.0488) 0.0768 (+0.0120)
0.9851 74 1.42E-02 3.68E-02 1.17E-01
0.8249 40 1.51E-02 3.77E-02 1.23E-01
0.7850 23 2.03E-02 4.14E-02 1.42E-01
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
0.9037 18 1.20E-02 3.08E-02 1.10E-01
0.9274 11 1.16E-02 2.61E-02 1.08E-01
168
FAUSTINO et al.
Figure 2. Fitting of the experimental data at 308C with the different models tested.
Figure 4. Fitting of the experimental data at all temperatures with the Newton model.
Figure 2 shows the experimental points obtained for one selected temperature (308C, because was the curve with more data) together with the plots corresponding to the fittings obtained for all the models tested. It can be observed that the Logarithmic and Henderson and Pabis curves are perfectly coincident, and that, despite showing a good performance for t . 0, do not represent the initial state (t ¼ 0). Moreover, from Table 6 it is possible to see that although the correlation coefficients of the fittings are not too bad, the significance of the values estimated for the parameter k in the logarithmic model is poor, since the standard deviation is higher than the value itself. The Wang and Singh model does not show a good prediction capacity, since the corresponding curve is clearly away from the experimental points, and besides, the trend of this model is also not adequate to represent the decreasing function that the experimental data shows towards the end of drying. In this way, the models that seem best to describe the experimental behaviour observed are the Newton and Page models, and of these the Page is better either by observing the curves in Figure 1 or by comparing the statistical information from
Tables 3 and 4 (namely R, x2 and RMSE, since MBE is not squared). Considering that Page is one of the best models applicable to the drying of green bell pepper, in Figure 3 is represented the drying data for all temperatures studied, together with the fitting obtained for each temperature with the Page model. The curves show a fairly good performance over the entire range of time, which varies from according to the temperature. However, at the final stages the model tends to give higher predictions than the values observed. In this way, the Newton model, represented in Figure 4, shows a better adjustment at the final stages of drying, nevertheless does not allow to distinguish the curves at 508C and 608C.
Diffusion Model Table 8 shows the results of the fitting to the diffusion model expressed by equations (7) and (8). The values of the standard deviations of each parameter together with the statistical information allow to infer that the quality of the estimation is good, with low values of the sum of square errors (Se 2) and residual standard deviation (RSD). The value obtained for the diffusion coefficient at an infinite temperature, D0e was 0.1176 m2 s21, and the activation energy for moisture diffusion, E, was found to be 47.10 kJ mol21. This later is in accordance with other values reported in literature for some varieties of red bell pepper: 28.4 kJ mol21 (Turhan
Table 8. Results obtained with ODR program to fit the diffusion model [equations (7) and (8)]. Parameters
Figure 3. Fitting of the experimental data at all temperatures with the Page model.
b1 D0e (m2 s21) E(J mol21) Statistical information N Sd2 (t,T) S12 (MR) Se2 RSD
Value
Standard deviation
2.2319 0.1176 47 101
0.0219 0.0713 1 574 166 4.4104 1025 5.36119 1021 5.3615 1021 5.7352 1022
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
DRYING KINETICS OF GREEN BELL PAPPER
169
Table 10. Effect of drying on the nutritional value of green bellpeppers. Component Moisture (g/100 g product) Protein (g/100 g dry solids)1 Total sugars (g/100 g dry solids) Fat (g/100 g dry solids) Crude fibre (g/100 g dry solids) Ash (g/100 g dry solids) Acidity (cm3/100 g dry solids)
Fresh product
Dehydrated at 308C
Dehydrated at 708C
94.16
3.25
4.71
17.98
11.82
1.40
85.27
4.36
4.58
6.50 5.14
1.06 7.30
0.90 7.73
5.14 10.27
6.78 5.65
6.42 5.55
1
Factor used for the conversion of nitrogen in protein ¼ 6.25.
Figure 5. Variation of the diffusion coefficient with temperature.
and Turhan, 1997), 44.8 kJ mol21(Sanjua´n et al, 2003), 42.8 kJ mol21 (Kaymak-Ertekin, 2002), 41.95 (Gupta et al., 2002), 39.70 (Vega et al., 2007). If Figure 5 the effective diffusivity of green bell peppers is represented in the range of temperatures studied (between 308C and 708C). The experimental points were obtained for each temperature separately, by applying linear regression to the data in the form of ln(MR) versus time plots. The curve expressing the variation of diffusivity with temperature was obtained from equation (8) with the parameters estimated and presented in Table 8. From the graph in Figure 5 is possible to verify the increase in diffusivity when the temperature is raised. For range of temperatures studied, the effective diffusivity varies between 9.0 10210 m2 s21at 308C and 8.0 1029 m2 s21 at 708C. For example, at 508C the value obtained is 2.9 1029, which is very close to that found by Vega et al. (2007) for the same temperature for the drying of red bell pepper (3.2 1029 m2 s21).
Nutritional Evaluation Table 9 shows the chemical composition of fresh green peppers, both analysed and reference. It can be seen that in general terms the results obtained in the present study are in accordance with the reference values. However, some small differences can be pointed out when comparing
Table 9. Chemical characterization of fresh green peppers (g/100 g edible portion). Component
Reference1
Analysed
Moisture Protein Total sugars Fat Dietary fibre Ash
93.89 0.86 2.40 0.17 1.70 0.43
94.01 1.052 4.98 0.38 0.733 0.30
1 USDA National Nutrient Database for Standard Reference (Release 19: 2006). 2 Factor used for the conversion of nitrogen in protein ¼ 6.25. 3 Crude fibre.
the values obtained with the reference, in particular slightly higher contents of protein and sugar, and lower amounts of fibre and minerals, which can be attributed to differences in the variety analysed. In Table 10 the chemical composition of the peppers, fresh and dehydrated at the lowest and highest temperatures, are shown for comparison purposes. From the values is possible to see that, except for the protein content, the chemical compositions of both dehydrated peppers (at 308C and 708C) are very similar, meaning that for the range of temperatures studied, the value of the temperature chosen for the process does not influence significantly the nutritional value of the final product. On the other hand, the changes observed by comparing the fresh with the dehydrated products are very significant for all components evaluated, with the exception of fibre and ash. In this way, the drying operation induces reductions of 34% in proteins, 95% in sugars, 84% in fat and 45% in acidity. In fact, the heat provided during the drying process originates the degradation of sugars in a very important extension, favouring also a significant loss in the acids (especially volatile acids) and causing some protein denaturising.
CONCLUSIONS From the results obtained, it was concluded that the empirical models that best describe the dehydration kinetics for green bell peppers are Page and Newton. For the drying process modelled in terms of Fick’s diffusion law, the values obtained for the diffusion coefficient at an infinite temperature, D0e, and activation energy for moisture diffusion, E, were, respectively, 0.1176 m2 s21and 47.10 kJ mol21. These values allow us to estimate the effective diffusivity, which ranges between 9.0 10210 and 8.0 1029 m2 s21, for the temperatures in the range 308C to 708C. From the chemical analyses made was possible to possible to conclude that drying influences the chemical composition of the peppers (due to volatilization of some components, oxidation processes and protein denaturizing). On the other hand, the influence of the drying temperature was not very significant on the chemical composition, and has the advantage of largely reducing the drying time.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170
170
FAUSTINO et al. REFERENCES
Ade-Omowaye, B.I.O., Rastogi, N.K., Angersbach, A. and Knorr, D., 2002, Osmotic dehydration of bell peppers: influence of high intensity electric field pulses and elevated temperature treatment, Journal of food Engineering, 54: 35–43. Association of Official Analytical Chemists (AOAC), 1990, Official Methods of Analysis, 15th edition, (AOAC, Arlington). Boggs, P.T., Byrd, R.H., Rogers, J.E. and Schnabl, R.B., 1992, User’s Reference Guide for ODRPACK. Software for Weighted Orthogonal Distance Regression, Version 2.01. (National Institute of Standards and Technology, Gaithersburg, USA). Chenlo, F., Changuri, L., Santos, F. and Moreira, R., 2006, Osmotic dehydration/impregnation kinetics of padro´n pepper (Capsicum annuum L. Longum) with sodium chloride solutions: process modelling and colour analysis, Food Science and Technology International, 12(3): 221–227. Crank, J., 1970, The Mathematics of Diffusion (Oxford University Press, London, UK). Croft, K.P.C., Juttner, F. and Slusarenko, A.J., 1993, Volatile products of the lipoxygenase pathway evolved from phaseolus vulgaris (L.) leaves inoculated with pseudomonas syringae pv phaseolicola, Plant Physiol, 101: 13– 24. Doymaz, I. and Pala, M., 2002, Hot-air drying characteristics of red pepper, Journal of Food Engineering, 55(4): 331 –335. Duke, J.A., 1992, Handbook of Biologically Active Phytochemicals and Their Activities (CRC Press, Boca Raton, Florida, USA). Gardner, H.W., 1991, Recent investigations into the lipoxygenase. Pathway of plants, Biochim Biophys Acta, 1084: 221–239. Guine´, R.P.F. and Fernandes, R.M.C., 2006, Analysis of the drying kinetics of chestnuts, Journal of Food Engineering, 76: 460–467. Gupta, P., Ahmed, J., Shivhare, U.S. and Raghavan, G.S.V., 2002, Drying characteristics of red chilli, Drying Technology, 20: 1975–1987. Kaymak-Ertekin, F., 2002, Drying and rehydrating kinetics of green and red bell peppers, J Food Sci, 67(1): 168 –175. Kidmose, U., Yang, R.Y., Thilsted, S.H., Christensen, L.P. and Brandt, K., 2006, Content of carotenoids in commonly Asian vegetables and stability and extractability during frying, Journal of Food Composition and Analysis, 19: 562–571. Konishi, Y., Oriuchi, J.-I. and Kobayashi, M., 2001, Dynamic evaluation of the dehydration response curves of food characterized by a poultrice-up process using a fish-paste sausage. II. A new tank model for computer simulation, Drying Technol, 19(7): 1271– 1285. Krinsky, N.I., 1989, Carotenoids and cancer in animal models, The Journal of Nutrition, 119: 123–126. Kritchevsky, D., 1992, Antioxidant vitamins in the prevention of cardiovascular disease, Nutrition Today, 27: 30– 33. Lee, J.J., Crosby, K.M., Pike, L.M., Yoo, K.S. and Lescobar, D.I., 2005, Impact of genetic and environmental variation of development of flavonoids and carotenoids in pepper (Capsicum spp.), Scientia Horticulturae, 106: 341 –352. Lee, Y., Howard, L.R. and Villalon, B., 1995, Flavonoid and ascorbic acid content and antioxidant activity of fresh pepper (capsicum annuum) cultivars, IFT Abstract, 55: 79. Madamba, P.S., 1997, Optimization of the drying process: an application to the drying of garlic, Drying Technol, 15(1): 117– 136. Matsui, K., Shibutani, M., Hase, T. and Kajiwara, T., 1996, Bell pepper fruit fatty acid hydroperoxide lyase is a cytochrome P450 (CYP74B), FEBS Lett, 394: 21– 24. Mezzetti, A., Lapenna, D., Pierdomenico, S.D., Calafiore, A.M., Constantini, F., Riario-Sforza, G., Imbastaro, T., Neri, M. and
Cuccurullo, F., 1995, Vitamin E, C and lipid peroxidation in plasma and arterial tissue of smokers and non-smokers, Atheroscletrosis, 112: 91–99. Niizu, P. and Rodriguez-Amaya, D.B., 2005, New data on the carotenoid composition of raw salad vegetables, Journal of Food Composition and Analysis, 18: 739–749. Olson, J.A., 1989, Biological actions of carotenoids, The Journal of Nutrition, 119: 94–95. Packer, L., 1996, Possible factors influencing aging, United Nations and World Health Organization First Joint Conference on Healthy Aging, New York, NY, 29 April –1 May. Salgado, M.A., Lebert, A., Garcia, H.S., Muchnik, J. and Bimbenet, J.J., 1994, Development of the characteristic drying curve for cassava chips in monolayer, Drying Tecnology, 12(3): 685 – 696. Sanjua´n, N., Lozano, M., Garcı´a-Pascual, P. and Mulet, A., 2003, Dehydration kinetics of red bell pepper (Capsicum annuum L var Jaranda), Journal of The Science of Food and Agriculture, 83: 697– 701. Shibata, Y., Matsui, K., Kajiwara, T. and Hatanaka, A., 1995, Purification and properties of fatty acid hydroperoxide lyase from green bell pepper fruits, Plant Cell Physiol, 36: 147– 156. Simonne, A.H., Simonne, E.H., Eitenmiller, R.R., Mills, H.A. and Green, N.R., 1997, Ascorbic acid and provitamin A contents in unusually colored bell peppers (Capsicum annuum L.), Journal of Food Composition and Analysis, 10: 299– 311. Smirnoff, N., 1996, The function and metabolism of ascorbic acid in plants, (Annals of Botany, 7: 661– 669. Sun, D.-W. and Woods, J.L., 1997, Simulation of the heat and moisture transfer process during drying in deep grain beds, Drying Technol, 15(10): 2479–2508. Togrul, I.T. and Pehlivan, D., 2003, Modelling of drying kinetics of single apricot, Journal of Food Engineering, 58(3): 23–32. Tunde-Akintunde, T.Y., Afolabi, T.J. and Akintunde, B.O., 2005, Influence of drying methods on drying of belle pepper (Capsicum annuum), Journal of Food Engineering, 68: 439– 442. Turhan, M. and Turhan, K.N., 1997, Drying kinetics of red bell pepper, Journal of Food Processing and Preservation, 21: 209–223. Va´zquez, G. and Chenlo, F., 1997, Grape drying in a pilot plant with a heat pump, Drying Technology, 15(3&4): 899– 920. Vega, A., Fito, P., Andre´s, A. and Lemus, R., 2007, Mathematical modeling of hot-air drying kinetics of red bell pepper (var. Lamuyo), Journal of Food Engineering, 79: 1460– 1466. Yahia, E.M., Contreras-Padilla, M. and Gonzalez-Aguilar, G., 2001, Ascorbic acid content in relation to ascorbic acid oxidase activity and polyamine content in tomato and bell pepper fruits during development, maturation and senescence, Lebensm-Wiss U-Technol, 34: 452–457. Yaldiz, O. and Erketin, C., 2001, Thin layer solar drying of some different vegetables, Drying Technology, 19(3): 583– 596. Yang, H., Sakai, N. and Watanabe, M., 2001, Drying model with non-isotropic shrinkage deformation undergoing simultaneous heat and mass transfer, Drying Technol, 19(7): 1441– 1460. Ziegler, R.G., 1989, A review of epidemiologic evidence that carotenoids reduce the risk of cancer, The Journal of Nutrition, 119: 116–122. Zogzas, N.P., Maroulis, Z.B. and Marinos-Kouris, D., 1994, Moisture diffusivity methods of experimental determination. A review, Drying Technology, 12(3): 483– 515. The manuscript was received 30 January 2007 and accepted for publication after revision 7 June 2007.
Trans IChemE, Part C, Food and Bioproducts Processing, 2007, 85(C3): 163– 170