Pineapple Peel Flours: Drying Kinetics, Thermodynamic Properties, and Physicochemical Characterization

Abstract

Pineapple is a widely cultivated, consumed, and processed fruit by the industry. However, only 22.5% of the whole fruit is used, which constitutes economic waste and environmental impact. The objective was to determine the drying kinetics and characterize the residual peel flours of two pineapple varieties at four drying temperatures. Jupi and Pérola pineapple peels were dried at temperatures of 50, 60, 70, and 80 °C in a thin layer. Ten mathematical models were adjusted to the experimental data to characterize the drying process. Fresh samples and flours were characterized according to their physicochemical properties (water content, ash, water activity, total sugars, reducers, pH, acidity, proteins, lipids, carbohydrates, and total energy value—TEV). The Midilli model was chosen because it best represents the drying process with high values of determination coefficients (R²) and low mean squared deviations (MSD), Chi-square (χ²), and estimated mean error (EME). The increase in temperature led to an increase in the effective diffusivity coefficient and consequent reduction in drying time. The activation energy obtained from the Arrhenius equation was 24.59 and 26.25 kJ/mol for Jupi and Pérola, respectively. Differences were reported in the enthalpy and entropy decrease with the increasing temperature, contrary to the Gibbs free energy. The flours produced had good characteristics for conservation, being acidic with low water content and low water activity. High levels of total and reducing sugars, carbohydrates, and total energy value were observed, in addition to good protein content.

1. Introduction

Pineapple (Ananas comosus (L.) Merryl) is appreciated all over the world, possibly as the most economically important bromeliad fruit. Brazil is one of the leading countries in pineapple production, with a production of 1617.684 tons in 2019, with the main producers being the states of Pará (311,947 tons), Paraíba (307,116 tons), and Minas Gerais (179,287 tons) [1]. Pineapple farming in Paraíba has always been relevant in national production, with pineapple being one of the most attractive tropical fruits. In fruit growing in Paraíba, it is the crop that has the greatest economic significance, with a high demand for labor caused by the lack of mechanization of the crop [2].

Pineapple is usually consumed in its natural form or in the production of juices. However, other products are made in smaller proportions, such as pineapple syrup, jelly, jam, and dehydrated pineapple [3]. Pineapple waste flours are developed and tested in innovative food products, such as vinegar, vanillin [4], and bread [5]. Research shows that drying has effects on the physical–chemical and technofunctional properties in the preparation of pineapple peel flour, showing functional properties that enable its use in the form of flour rich in dietary fiber, thus contributing to the environment [6].

Fruit processing generates a significant amount of waste, which can be used in the production of new products, as they usually have significant nutritional value [7]. The pineapple canning industry is one of the many food industries that contribute to the accumulation of solid waste [8]. About three quarters of the pineapple, comprising the peel, stalk, and crown, are classified as waste by the fruit pulp processing industries [9]. This high amount of organic material receives inadequate disposal, thus generating economic and nutritional waste [10], since the composition of the different types of waste such as the bark, core, crown, stem, and residual pulp have a nutritional composition such as carbohydrates (50–80%), vitamin C, and beta-carotene, which makes it suitable for the development of value-added products [11,12].

Drying is a conservation and transformation technology that can be adopted to take advantage of pineapple residues, transforming them into a new product. Drying is conventionally defined as a process of removing water from agricultural products that involves the simultaneous transfer of heat and mass [13], providing conservation of the material during storage [14]. Through drying kinetics, the data obtained allow us to understand and describe the behavior of the product and its interactions with mass transfer and the drying agent, whose particularities associated with the diversity of materials of biological origin stimulate the interest of researchers in analyzing the drying behavior of the most varied products [15].

Lately, pineapple waste is dried in different ways: pineapple peels enriched by fermentation and dried via forced air circulation at different temperatures and air speeds [16]; dried via refraction at different temperatures [6], and dried in vacuum [17]. The best drying method depends on the purpose attributed to the product, since freeze-drying was more effective in preserving anthocyanins, and oven drying showed higher individual phenolic acids, as observed in dried “BRS Magna” grape skins [18]. Convective drying with forced circulation with hot air is the technique most used by chemical industries to dry foods [19]. This method has the advantage of not being pollutive, in addition to being highly efficient in the drying process [20]. The advantages and relative low cost associated with this type of drying juxtapose the transformations of the food, which can be minimized by associating techniques for different materials in particular [21]. It is an artificial method that makes it possible to control the temperature, air flow, and exposure time [22], depending on the final objective. Low temperatures can be used, which allows for the drying of thermosensitive products or products that undergo oxidation quickly [23]. Fruit waste, for example, must be dried under controlled temperature conditions to avoid nutrient degradation [24].

Among the phenomena studied in drying, effective diffusivity is an indication of the flow of water removed from the product during the process [25]. From the determination of the effective diffusivity, the main thermodynamic properties of the dehydration process are calculated, such as the enthalpy, entropy, and Gibbs free energy [26]. According to Corrêa et al. [27], knowledge of thermodynamic properties in the drying process is of fundamental importance as these sources of information are necessary to understand the processes involved such as equipment design, required energy, the properties of adsorbed water, the evaluation of the microstructure of products, and the study of physical phenomena that occur on their surfaces.

Given the above and considering the importance of studies that contribute to the reduction in waste from fruit processing, as well as the comparison of the physico-chemical characteristics of different varieties of pineapple in relation to the drying process, this study aimed to prepare flours from the peel of pineapples of the Pérola and Jupi varieties from convective drying at temperatures of 50, 60, 70, and 80 °C, to determine the drying kinetics, to adjust the mathematical models to the experimental data, to determine the effective diffusivity and thermodynamic properties, and to characterize the flours obtained.

2. Materials and Methods

2.1. Samples Preparation

To prepare the samples, ripe pineapples of the Pérola and Jupi varieties were cultivated in the municipalities of Itapororoca (Geographical coordinates: 6°49′48″ S, 35°14′49″ W) and Pedras de Fogo (Geographical coordinates: 7°24′7″ S, 35°6′57″ W), respectively, in the state of Paraiba-Brazil. Pineapples were considered ripe if they had °Brix levels equal to or greater than 12° [28], specific weight around 1.012 g/mL [29], greater translucency, and low acidity levels, being correlated with the skin tones, presenting ripeness indices larger in colorful coloring [30]. The pineapples were washed and sanitized in a chlorine solution (200 ppm) for 15 min. The peels were removed using stainless steel knives and ground in a pulper (Laboremus^®, Campina Grande, Brazil). The crushed peels were packed in low-density polyethylene bags and stored in a freezer at −18 °C until use, remaining in the freezer for 2 months before starting to dry.

2.2. Drying

The pineapple residues (crushed peels) were thawed under refrigeration at 4 °C and after thawing, the samples were left at room temperature for 30 min. The residues were arranged in a thin layer (0.5 cm) on stainless steel trays and dried in an oven with forced air circulation-convective drying (FANEM LTDA, Sao Paulo, Brazil, model 320/5 200 °C) at temperatures of 50, 60, 70, and 80 °C and with a drying air speed of 1.0 m/s.

The drying kinetics were performed by weighing the trays with the samples at regular intervals of 5, 10, 20, 30, and 60 min until reaching the equilibrium water content, then the dry mass was determined according to the Adolfo Lutz Institute [31] and the water content ratio was calculated according to Equation (1).

$$\mathrm{MR}=\frac{X-X_e}{X_i-X_e}$$

(1)

where:

MR—ratio of the water content of the product (dimensionless);

X—water content at time t (% db);

X_i—initial water content of the product (% db);

X_e—product balance water content (% db).

The mathematical models in

Table 1. Mathematical models used to estimate the drying kinetics of pineapple skin.

Model Name	Equation	Reference
Diffusion Approach	$\mathrm{MR}=a \exp \left(-kt\right)+\left(1-a\right)\exp \left(-kbt\right)$	Sharaf-Elden et al. [32]
Two Terms	$\mathrm{MR}=a \exp \left(-k_{0}t\right)+b \exp \left(-k_{1}t\right)$	Henderson [33]
Two-term exponential	$\mathrm{MR}=a\exp \left(-kt\right)+\left(1-a\right)\exp \left(-kat\right)$	Sharaf-Elden et al. [32]
Midilli	$\mathrm{MR}=a \exp \left(-k{t}^n\right)+bt$	Midilli et al. [34]
Page	$\mathrm{M}=\mathrm{Rexp} \left(-k{t}^n\right)$	Page [35]
Henderson and Pabis	$\mathrm{MR}=a \exp \left(-kt\right)$	Henderson & Pabis [36]
Logarithmic	$\mathrm{MR}=a\exp \exp \left(-kt\right)+c$	Yagcioglu et al. [37]
Newton	$\mathrm{MR}=\exp \exp \left(-kt\right)$	Lewis [38]
Verma	$\mathrm{MR}=a \exp \left(-kt\right)+\left(1-a\right)\exp \left(-k_{1}t\right)$	Verma et al. [39]
Wang and Singh	$\mathrm{MR}=1+\left(at\right)+\left(b{t}^2\right)$	Wang & Singh [40]

MR—ratio of water content of the product (dimensionless); t—drying time (min); a, b, c, k, k₀, k₁, n—model constants.

were adjusted to the experimental data of the drying kinetics using the computational program Statistica version 7.0 via non-linear regression and the Quasi-Newton method.

As a criterion to evaluate the quality of the models’ adjustments to the experimental data, the coefficient of determination (R²), mean squared deviation (MSD) (Equation (2)), Chi-square (χ²) (Equation (3)), and the estimated mean error (EME) (Equation (4)) were used, with the best model being considered the one that presents a high R² value and low values of MSD, χ² e EME.

$${\mathrm{R}}^2=\frac{{{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{N}}{[({\mathrm{MR}}_{\exp ,\mathrm{i}}-\overline{{\mathrm{MR}}_{\exp ,\mathrm{i})}})({\mathrm{MR}}_{\mathrm{pred},\mathrm{i}}-\overline{{\mathrm{MR}}_{\mathrm{pred},\mathrm{i})}})]}^2}{{{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{N}}({{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{N}}[{\left({\mathrm{MR}}_{\exp ,\mathrm{i}}-\overline{{\mathrm{MR}}_{\exp ,\mathrm{i})}}\right)}^2)({{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{N}}({({\mathrm{MR}}_{\mathrm{pred},\mathrm{i}}-\overline{{\mathrm{MR}}_{\mathrm{pred},\mathrm{i})}})}^2}$$

(2)

where:

MSD—mean squared deviation;

RX_pred—ratio of water content predicted by the model;

RX_exp—experimental water content ratio;

n—number of observations.

$${\mathsf{\chi }}^2=\frac{1}{\mathrm{GLR}}{{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{N}}{\left({\mathrm{MR}}_{\mathrm{pred},\mathrm{i}}-{\mathrm{MR}}_{\exp ,\mathrm{i}}\right)}^2$$

(3)

where:

χ²—Chi-square;

RX_pred—ratio of water content predicted by the model;

RX_exp—experimental water content ratio;

N—number of experimental observations;

n—number of model constants.

$$\mathrm{DQM}=\frac{\sqrt{\mathsf{\Sigma } {\left({\mathrm{RX}}_{\mathrm{pred}}-{\mathrm{RX}}_{\exp }\right)}^2}}{\mathrm{n}}$$

(4)

where:

EME—estimated mean error.

2.3. Effective Diffusivity

To determine the effective diffusivity, the mathematical model of liquid diffusion was adjusted to the experimental data, considering the uniform initial water distribution and absence of thermal resistance. Equation (5) is the analytical solution to Fick’s second law, applied to products with approximate shape to an infinite flat plate. The Arrhenius-type equation (Equation (6) was applied to evaluate the influence of temperature on the effective diffusivity.

$$\mathrm{MR}=\frac{8}{{\mathsf{\pi }}^2}{{\displaystyle \sum }}_{\mathrm{n}=0}^{\infty }\frac{1}{{\left(2\mathrm{n}+1\right)}^2}\exp \left[-{\left(2\mathrm{n}+1\right)}^2{\mathsf{\pi }}^2{\mathrm{D}}_{\mathrm{ef}}\frac{\mathrm{t}}{{\mathrm{L}}^2}\right]$$

(5)

where:

MR—ratio of the water content of the product (dimensionless);

D_ef—effective diffusivity (m²/s);

n—number of terms;

L—characteristic dimension (half plate thickness);

t—time (s).

$${\mathrm{D}}_{\mathrm{ef}}={ \mathrm{D}}_0\exp \left(\frac{{\mathrm{E}}_{\mathrm{a}}}{{\mathrm{RT}}_{\mathrm{a}}}\right)$$

(6)

where:

D_ef—effective diffusivity (m²/s);

D_o—pre-exponential factor (m²/s);

E_a—activation energy (J/mol);

R—universal gas constant (8.314 J/mol K);

Ta—absolute temperature (K).

2.4. Thermodynamic Properties

The thermodynamic properties of enthalpy, entropy, and Gibbs free energy, related to the drying process of pineapple residues, were determined according to Equations (7)–(9), respectively [41].

$$\Delta {\mathrm{H}=\mathrm{E}}_{\mathrm{a}}-\mathrm{RT}$$

(7)

$$\Delta \mathrm{S}=\mathrm{R}\left[\ln \left({\mathrm{D}}_0\right)-\ln \left(\frac{{\mathrm{K}}_{\mathrm{B}}}{{\mathrm{h}}_{\mathrm{P}}}\right)-\ln \left(\mathrm{T}\right)\right]$$

(8)

$$\Delta \mathrm{G}=\Delta \mathrm{H}-\left(\mathrm{T}\right)\Delta \mathrm{S}$$

(9)

where:

ΔH—enthalpy (J/mol);

ΔS—entropy (J/mol K);

ΔG—Gibbs free energy (J/mol);

E_a—activation energy (J/mol);

R—universal gas constant (8.314 J/mol K);

K_B—Boltzmann constant (1.38 × 10⁻²³ J/K);

h_p—Planck constant (6.626 × 10⁻³⁴ J/s);

T—absolute temperature (K).

2.5. Physicochemical Evaluation and Centesimal Composition of in Natura Peels and Flours

After drying, the dehydrated residues were crushed in a knife mill to transform them into flour.

Samples of fresh pineapple skins and elaborated flours were characterized by means of physicochemical analyses and centesimal composition, in triplicate, with respect to the following parameters: the gravimetric method was used to determine the water content, which consists of the difference in the initial and final weight of the sample after weighing until a constant weight in a vacuum oven at 70 °C, with the results expressed as a percentage (%); the ash samples were incinerated at 550 °C and the results were measured in percentage; for total titratable acidity, the titrimetric method was used with a 0.1 mol/L NaOH solution and phenolphthalein as a turning point indicator and the result was expressed as a percentage of citric acid (%) and proteins, according to the methodologies of Instituto Adolfo Lutz [31]; the water activity (a_w) was determined at 25 °C (Aqualab 3TE); the dinitrosalicylic acid (DNS) method was used to determine the levels of reducing sugars in the samples [42]; total sugars were determined using an anthrone solution [43]; lipids were quantified using the method that uses a cold extraction of lipids [44] carbohydrates via difference (100% − (water content + proteins + lipids + ash)) [45]; and the total energy value (TEV) was calculated using the results of the following analyses multiplied by the conversion values: carbohydrates (4 kcal/g), proteins (4 kcal/g), and lipids (9 kcal/g), which were then added up [46].

The data from the physicochemical analyses and centesimal composition were submitted to an analysis of variance (ANOVA) in a factorial scheme of 5 treatments (in natura, 50, 60, 70, and 80 °C) × 2 varieties (Pérola and Jupi) with three repetitions, with the averages compared applying Tukey’s test at a 5% probability level using the program Assistat, version 7.7 [47].

3. Results and Discussion

Table 2. Values of mean square deviations (MSD), coefficients of determination (R²), Chi-square (χ²), and estimated mean errors (EME) of the models adjusted to experimental data on drying the peel of pineapple cv. Jupi.

Models	T (°C)	R2	DQM	χ2	EME
Diffusion Approach	50	0.9992	0.0142	0.0002	0.0148
	60	0.9986	0.0186	0.0004	0.0196
	70	0.9993	0.0128	0.0002	0.0135
	80	0.9989	0.0158	0.0003	0.0168
Two Terms	50	0.9956	0.0327	0.0012	0.0347
	60	0.9924	0.0434	0.0021	0.0464
	70	0.9940	0.0374	0.0016	0.0403
	80	0.9924	0.0417	0.0021	0.0453
Page	50	0.9992	0.0143	0.0002	0.0147
	60	0.9989	0.0166	0.0003	0.0171
	70	0.9995	0.0112	0.0001	0.0116
	80	0.9993	0.0131	0.0002	0.0136
Newton	50	0.9935	0.0398	0.0016	0.0404
	60	0.9888	0.0525	0.0028	0.0534
	70	0.9907	0.0465	0.0022	0.0473
	80	0.9890	0.0503	0.0026	0.0513
Henderson and Pabis	50	0.9956	0.0327	0.0011	0.0336
	60	0.9924	0.0434	0.0020	0.0448
	70	0.9940	0.0374	0.0015	0.0387
	80	0.9924	0.0417	0.0019	0.0434
Two Terms Exponential	50	0.9932	0.0409	0.0018	0.0421
	60	0.9886	0.0530	0.0030	0.0547
	70	0.9905	0.0470	0.0024	0.0487
	80	0.9886	0.0511	0.0028	0.0532
Logarithmic	50	0.9968	0.0279	0.0008	0.0292
	60	0.9944	0.0373	0.0015	0.0392
	70	0.9957	0.0316	0.0011	0.0333
	80	0.9949	0.0343	0.0013	0.0365
Midilli	50	0.9983	0.0203	0.0005	0.0215
	60	0.9993	0.0136	0.0002	0.0146
	70	0.9996	0.0097	0.0001	0.0105
	80	0.9995	0.0110	0.0001	0.0120
Verma	50	0.9992	0.0142	0.0002	0.0148
	60	0.9985	0.0191	0.0004	0.0200
	70	0.9907	0.0465	0.0024	0.0491
	80	0.9989	0.0157	0.0003	0.0167
Wang and Singh	50	0.9625	0.0951	0.0099	0.0993
	60	0.9795	0.0709	0.0056	0.0745
	70	0.9761	0.0744	0.0059	0.0771
	80	0.9822	0.0638	0.0044	0.0664

and

Table 3. Values of mean square deviations (MSD), coefficients of determination (R²), Chi-square (χ²), and estimated mean errors (EME) of the models adjusted to experimental data on drying the peel of pineapple cv. Pérola.

Models	T (°C)	R2	DQM	χ2	EME
Diffusion Approach	50	0.9994	0.0121	0.0002	0.0127
	60	0.9988	0.0169	0.0003	0.0177
	70	0.9987	0.0180	0.0004	0.0190
	80	0.9992	0.0136	0.0002	0.0144
Two Terms	50	0.9969	0.0272	0.0008	0.0288
	60	0.9937	0.0394	0.0018	0.0420
	70	0.9918	0.0448	0.0023	0.0482
	80	0.9928	0.0405	0.0019	0.0441
Page	50	0.9993	0.0129	0.0002	0.0133
	60	0.9990	0.0155	0.0003	0.0160
	70	0.9991	0.0148	0.0002	0.0153
	80	0.9995	0.0109	0.0001	0.0113
Newton	50	0.9954	0.0331	0.0011	0.0336
	60	0.9904	0.0484	0.0024	0.0491
	70	0.9874	0.0553	0.0032	0.0563
	80	0.9891	0.0499	0.0026	0.0508
Henderson and Pabis	50	0.9969	0.0272	0.0008	0.0279
	60	0.9937	0.0394	0.0016	0.0406
	70	0.9918	0.0448	0.0022	0.0464
	80	0.9928	0.0405	0.0018	0.0422
Two Terms Exponential	50	0.9951	0.0342	0.0012	0.0352
	60	0.9899	0.0498	0.0026	0.0514
	70	0.9868	0.0565	0.0034	0.0585
	80	0.9889	0.0504	0.0028	0.0525
Logarithmic	50	0.9978	0.0231	0.0006	0.0241
	60	0.9952	0.0342	0.0013	0.0359
	70	0.9940	0.0381	0.0016	0.0403
	80	0.9950	0.0339	0.0013	0.0360
Midilli	50	0.9995	0.0114	0.0001	0.0120
	60	0.9993	0.0135	0.0002	0.0144
	70	0.9993	0.0130	0.0002	0.0140
	80	0.9996	0.0096	0.0001	0.0104
Verma	50	0.9994	0.0122	0.0002	0.0127
	60	0.9988	0.0169	0.0003	0.0178
	70	0.9987	0.0180	0.0004	0.0191
	80	0.9992	0.0137	0.0002	0.0145
Wang and Singh	50	0.9498	0.1085	0.0124	0.1115
	60	0.9739	0.0795	0.0067	0.0821
	70	0.9789	0.0713	0.0055	0.0739
	80	0.9773	0.0718	0.0056	0.0747

show the mean squared deviations (MSD), coefficients of determination (R²), Chi-square (χ²), and estimated mean errors (EME) of the models adjusted to the experimental data on the drying of the peel of pineapples Jupi and Pérola, respectively.

Drying kinetics are usually very well predicted by the models available in the literature, obtaining adjustments with determination coefficients greater than 0.99 and Chi-squares close to zero. Among the adjusted mathematical models, the one by Wang and Singh had the worst results, highlighting the temperature of 50 °C for the Pérola variety, with R² lower than 0.98 and higher values of MSD, χ², and EME, for both pineapple varieties.

It was found that, with the exception of the Wang and Singh model, all the applied models satisfactorily adjusted to the experimental data, with emphasis on Diffusion Approximation, Page, and Midilli, with the highest determination coefficients (R² ≥ 0.998) and the lowest MSD (≤0.0203), χ² ≤ 0.0005, and EME ≤ 0.0215, thus indicating an excellent description of the drying process of pineapple skins under the studied conditions. A similar behavior was verified by Barbosa and Lobato [48], in studies of the drying kinetics of pineapple slices, observing R² values greater than 0.990 when using the Page model in the drying kinetics at temperatures of 60, 65, and 70 °C; and by Olanipekun et al. [49], in drying pineapple slices in a convective dryer with an air velocity of 1.5 m/s at temperatures of 50, 60, and 70 °C; also noting that the Page and Two Terms models presented satisfactory adjustments (R² > 0.980 and χ² ≤ 0.0011). Santos et al. [50], when studying the drying of grapefruit peels, reported on the eleven adjusted models, with emphasis on the Page, Logarithmic, Diffusion Approximation, and Midilli models for all temperatures (60, 70, 80, and 90 °C). However, due to its simplicity, they adopted Page’s model as the best representation.

Figure 1 shows the experimental points with the adjustment curves obtained with the Midilli model.

The effect of increasing the drying temperature on the faster reduction in the water content ratio is observed in both varieties with very different curves, demonstrating higher drying rates at higher temperatures. The gradual effect of the increase in temperature on the increase in rates is reported for most varied products, such as the drying of pomegranate by-products at temperatures of 50, 60, 70, and 80 °C with an air velocity of 2 m/s [51]; jabuticaba (Myrciaria jaboticaba) peels with an air velocity of 5.6 m/s at temperatures of 40, 50, 60, and 70 °C [52]; and dried passion fruit peels at 40, 50, 60, and 70 °C with an air velocity of 1 m/s [53].

The drying time of pineapple peels varied between 630 and 250 min for the Jupi variety and between 690 and 250 min for Pérola between temperatures of 50 and 80 °C, respectively, with very similar values between the varieties.

The parameters of the Midilli model adjusted to the pineapple peel drying kinetics data are presented in

Table 4. Parameters of the Midilli model adjusted to pineapple peel drying kinetics data.

Variety	Temperature (°C)	a	k	b	n
Jupi	50	1.0136	0.0052	−0.000025	1.1380
	60	0.9695	0.0017	−0.000016	1.4498
	70	0.9819	0.0036	−0.000015	1.3729
	80	0.9768	0.0038	−0.000029	1.4245
Pérola	50	0.9808	0.0030	−0.000010	1.2501
	60	0.9757	0.0022	−0.000014	1.3940
	70	0.9782	0.0024	−0.000021	1.4558
	80	0.9840	0.0043	−0.000023	1.4156

. The Midilli model was also the one that best adjusted the experimental data of pomegranate peel drying (cv. Wonderful) at drying temperatures of 40, 50, and 60 °C [54].

Parameter “k” represents the drying rate and parameter “n” reflects the internal resistance to drying. It is observed that the constants “k” and “n” of the Midilli model did not show definite trends between each subsequent temperature. However, between the extreme temperatures, 50 and 80 °C, there is an increase in the value of the parameter “k” in the Pérola variety. The same is observed in relation to the constant “n” in both varieties. In the Pérola variety, where “n” increases with the increase in temperature in the range of 50 to 70 °C, then decreases at 80 °C, but generally maintains the increasing trend. Moreira et al. [55], studying the drying kinetics of mandacaru at temperatures of 40, 50, and 60 °C, obtained, in general, an increase in the constant “n” with the increasing temperature, as observed in this research.

The values of the effective diffusivity as a function of the drying temperature of the pineapple peels are shown in

Table 5. Effective diffusivity values as a function of drying air temperature (40, 50, 70, and 80 °C) for Jupi and Pérola pineapple peels.

Variety	Temperature (°C)	Effective Diffusivity (m2/s)	R2
Jupi	50	3.24 × 10−10	0.9699
	60	4.29 × 10−10	0.9623
	70	5.62 × 10−10	0.9661
	80	7.01 × 10−10	0.9642
Pérola	50	3.19 × 10−10	0.9740
	60	4.21 × 10−10	0.9651
	70	5.55 × 10−10	0.9607
	80	7.33 × 10−10	0.9645

. An increase in the effective diffusivity is observed with the increase in temperature, explained by the decrease in the viscosity of the water with the increase in temperature, which promotes the ease of its removal [56]. Ojediran et al. [57] described effective diffusivity as the rate at which material water is removed from the center of the geometry to the surface. D_ef values close to those of pineapple peels were obtained by Mphahlele, Pathare, and Opara [54] when drying pomegranate peels, obtaining effective diffusivity values of 4.05 × 10⁻¹⁰, 5.06 × 10⁻¹⁰ and 8, 10 × 10⁻¹⁰ m²/s at temperatures of 40, 50, 60 °C, respectively; by Nascimento et al. [53] in passion fruit peel drying with and without ultrasound, observing D_ef values that ranged from 0.6 to 2.0 × 10⁻¹⁰ m²/s without ultrasound and from 1.2 to 2.6 × 10⁻¹⁰ m²/s with ultrasound at drying temperatures of 40, 50, 60, and 70 °C; and by Azeez et al. [58], who reported D_ef values of 2.53 × 10⁻¹⁰, 3.21 × 10⁻¹⁰ and 5.00 × 10⁻¹⁰ m²/s for temperatures of 50, 60, and 70 °C, respectively, in the drying of tomato slices.

Figure 2 shows the values of the effective diffusivity (Ln D_ef) as a function of the inverse of the drying temperature (1/T) adjusted via an Arrhenius-type equation.

Table 6. Arrhenius equation adjustment parameters.

Variety	D0 (m2/s)	Ea (kJ/mol)	R2
Jupi	3.08 × 10−6	24.60	0.9991
Pérola	5.55 × 10−6	26.25	0.9992

shows the adjustment parameters of the Arrhenius equation applied to the effective diffusivity data of pineapple peels, observing satisfactory results of the determination coefficients (R² > 0.9990).

The activation energies (E_a) found for drying the skins of pineapples cv. Jupi and Pérola are lower than that reported by [59] in infrared drying of pineapple slices at temperatures ranging from 50 to 100 °C, with an E_a value of 33.63 kJ/mol. The E_a varies from product to product and is influenced by hygroscopicity, morphology, and environmental conditions [60]. Pineapple peels are within the activation energy range of agricultural products, from 12.7 to 110 kJ/mol, according to Zogzas, Maroulis, and Marinos-Kouris [61].

Table 7. Thermodynamic properties of pineapple skins subjected to drying at temperatures from 50 to 80 °C.

Variety	T (°C)	ΔH (kJ/mol)	ΔS (kJ/mol K)	ΔG (kJ/mol)
Jupi	50	21.9086	−0.3511	135.3628
	60	21.8255	−0.3513	138.8749
	70	21.7424	−0.3516	142.3896
	80	21.6592	−0.3518	145.9066
Pérola	50	23.5656	−0.3462	135.4377
	60	23.4825	−0.3464	138.9009
	70	23.3993	−0.3467	142.3666
	80	23.3162	−0.3469	145.8347

ΔH—enthalpy; ΔS—entropy; ΔG—Gibbs free energy.

shows the main thermodynamic properties obtained from pineapple peel drying. In both pineapple varieties, increasing the temperature leads to a decrease in enthalpy (ΔH), indicating that higher drying temperatures require less thermal energy to remove water bound to the product during the drying process [62]. The values remained in the range of 21.9086 to 21.6592 kJ/mol for Jupi pineapple and 23.5656 to 23.3162 kJ/mol for Pérola pineapple. Close values were found by Almeida et al. [63] in the drying of achachairu (Garcinia humilis) peels at temperatures ranging from 40 to 80 °C with an air velocity of 1.5 m/s, with ΔH ranging from 24.96 to 24.63 kJ/mol.

It is noticed that entropy (ΔS), which measures the degree of disorder of the system, decreased with the increasing temperature, indicating less agitation of water molecules and a greater degree of order between these molecules [64]. Negative entropy values indicate the presence of chemical adsorption and/or structural modifications of the adsorbent, as well as that the drying processes are entropically unfavorable [65]. Similar behavior was also observed in the drying of achachairu (Garcinia humilis) peels [63], mesocarp of baru (Dipteryx alata Vogel) [66], and mesocarp of pequi (Caryocar brasiliense Cambess) [67].

The Gibbs free energy (ΔG) is the total amount of energy involved in the system, and positive values (ΔG > 0) mean that the drying process was not spontaneous; that is, it required external energy (heated air) for its effectiveness, resulting in the change in the liquid phase from water to steam [68,69]. In addition to the gradual increases brought about by the increase in drying temperature, the samples of the two pineapple varieties showed similar ΔG values, ranging from 135 to 145 kJ/mol, indicating that for the drying of both varieties, the total energy required is same. The increase in ΔG with the increasing temperature was also reported by Resende et al. [62], in the drying of baru fruits (Dipteryx alata Vogel) at temperatures from 40 to 100 °C, and by Morais et al. [70], on the drying of bacaba pulp (Oenocarpus bacaba Mart) at temperatures of 40, 50, and 60 °C.

Values close to ΔG were reported by Tavone et al. [71], varying between 127.4250 and 134.8996 kJ/mol for calabura fruits (Muntingia calabura L.) dried at temperatures from 40 to 60 °C.

The results of the centesimal composition of the two pineapple varieties are presented in

Table 8. Proximate composition of fresh peels of Jupi and Pérola pineapples and flours obtained from the peels after drying at temperatures ranging from 50 to 80 °C.

Parameter	Variety		Drying Temperature (°C)
		In Natura	50	60	70	80
Water content (% wb)	Jupi	72.49 ± 0.32 bA	9.57 ± 0.32 aB	7.79 ± 0.27 aC	7.80 ± 0.19 aC	7.57 ± 0.46 aC
	Pérola	73.42 ± 0.21 aA	8.67 ± 0.33 bB	7.36 ± 0.16 aC	7.33 ± 0.10 bC	6.36 ± 0.09 bD
Ashes (% db)	Jupi	3.15 ± 0.09 bB	3.52 ± 0.02 aA	3.19 ± 0.08 bB	3.30 ± 0.01 bB	3.60 ± 0.15 aA
	Pérola	3.65 ± 0.07 aA	3.56 ± 0.03 aA	3.48 ± 0.07 aA	3.47 ± 0.13 aA	3.64 ± 0.13 aA
Proteins (% db)	Jupi	4.63 ± 0.01 bC	4.77 ± 0.01 bA	4.68 ± 0.01 bB	4.68 ± 0.00 bB	4.67 ± 0.00 bB
	Pérola	4.78 ± 0.03 aD	5.20 ± 0.01 aA	5.12 ± 0.01 aB	5.12 ± 0.01 aB	5.08 ± 0.01 aC
Lipids (% db)	Jupi	1.17 ± 0.02 bD	2.22 ± 0.02 bBC	2.26 ± 0.03 bB	2.15 ± 0.03 bC	2.35 ± 0.03 bA
	Pérola	1.67 ± 0.04 aC	2.42 ± 0.03 aB	2.87 ± 0.01 aA	2.87 ± 0.06 aA	2.88 ± 0.02 aA
Carbohydrates (% db)	Jupi	91.05 ± 0.13 aA	89.49 ± 0.02 aC	89.86 ± 0.07 aB	89.87 ± 0.01 aB	89.39 ± 0.12 aC
	Pérola	89.89 ± 0.08 bA	88.83 ± 0.03 bB	88.53 ± 0.09 bC	88.54 ± 0.08 bC	88.40 ± 0.14 bC
TEV (kcal/ 100 g db)	Jupi	393.21 ± 0.65 aC	397.02 ± 0.10 aB	398.53 ± 0.50 bA	397.54 ± 0.29 bAB	397.36 ± 0.28 bAB
	Pérola	393.76 ± 0.75 aC	397.87 ± 0.43 aB	400.46 ± 0.19 aA	400.44 ± 0.83 aA	399.83 ± 0.57 aA

db—dry basis; wb—wet base; TEV—Total Energy Value. Means followed by the same lowercase letter in columns and uppercase in rows do not differ statistically by Tukey’s test at 5% probability.

, corresponding to the averages and standard deviations obtained for the in natura sample and the dehydrated samples at temperatures from 50 to 80 °C.

The water content of the Jupi variety in natura was slightly higher than that of the Pérola variety, but with drying, the relationship was reversed, with Jupi contents resulting in lower values at most drying temperatures. Also, the highest water content of the flours occurred at a drying temperature of 50 °C and the lowest at 80 °C. The final water contents are considered adequate for the storage of agricultural products in general and, in particular, flours, whose values are considered satisfactory when lower than the maximum limit of 15% [72]. Similar water content values were studied in products such as kiwi peel flour [73] and tropical fruit residues [74].

The ash content in the in natura samples was higher in the Pérola variety, however, in a percentage difference that was not very expressive, so that in the flours, this difference only remained statistically significant in the samples obtained after drying at 60 and 70 °C. According to Storck et al. [75] the mineral content (ash) can be influenced by the grain size of the sample, which would explain the small variation in the values of the present research. The ash content detected in the flours was lower than those reported by Erkel et al. [76] and Santos et al. [77] on dehydrated pineapple peels, whose values were 4.70 and 4.66% db, respectively, and by Brito et al. [78], who reported a value of 6.45% db in pineapple crown flour.

The Pérola variety had the highest protein content in the in natura sample and this difference, in relation to Pérola, was maintained in all flours. The samples with the highest levels were obtained after drying at 50 °C, with values gradually decreasing with the increasing temperature. Although fruits do not represent a good source of protein, pineapple peel flours proved to be an alternative for incorporation into other products for protein purposes and may be a possibility of food for needy populations, which have difficult access to rich foods in proteins [75]. The results presented here are higher than the protein content of apple flour (2.7 to 3.5% db) [75] and close to the pineapple residues (peel and crown) of 4.56% before the fermentation process [79].

The flours produced showed low levels of lipids, with a slight advantage of the Pérola variety, in the in natura sample maintained in the flours obtained at all temperatures. Damasceno et al. [80] obtained lower levels (1.17 ± 0.08%) in Pérola pineapple peel flour, as well as Vieira et al. [81] on the dehydrated residues of pineapple (0.78% db) and cashew (1.63% db). Low levels of lipids in flours contribute to the enhancement of the product, given that lipids have a high caloric content, and their concentration is taken into account in products with a functional purpose or reduced caloric value [82].

The highest proportion in the composition of the samples, as expected, is composed of carbohydrates, with higher values in the in natura sample of the Jupi variety, but with a relatively small difference for the Pérola sample. The effect of temperature was not very expressive and did not result in consistent variations. The high values of total carbohydrates include dietary fiber, which is a percentage of important components of fruit peels, which, according to Garcia-Amezquita et al. [83], the total dietary fiber contents of orange, mango, and prickly pear peels dried via convection and freeze-drying correspond to 54.7, 54.8, and 49.2% db, respectively. Carbohydrate values lower than pineapple peel flours were quantified by Reis et al. [84], in passion fruit and orange peel flours with contents ranging between 72.12 and 73.37% db, and by Silva et al. [85], in avocado seed flour with a value of 79.06% db. Fresh pineapple peels had higher carbohydrates than tropical fruit pulp residues (guava, acerola, pineapple, soursop, bacuri, and cupuaçu) with values of 81.20, 65.02, 80.78, 77.14, 64, 41, and 9.77% db, respectively [86].

Observing the results of the Total Energy Value (TEV), the flours had a higher value when compared to the in natura samples, explained by the lower water content of the samples and, therefore, the higher energy density. In addition, flours had higher levels of proteins and lipids, also contributing to the increase in calories. The calorific results obtained for the flours were higher than those found for wheat flour (360 kcal/100 g) and slightly lower for soy flour (404 kcal/100 g) [87]. Pequi peel flour (225.42 kcal/100 g db) was also superior [88].

The results of the physicochemical characterization of the two pineapple varieties are presented in

Table 9. Physicochemical characterization of the in natura peels of Jupi and Pérola pineapples and the flours of the peels obtained after drying at temperatures from 50 to 80 °C.

Parameter	Variety		Drying Temperature (°C)
		In Natura	50	60	70	80
Water activity (25 °C)	Jupi	0.984 ± 0.001 aA	0.291 ± 0.00 bB	0.278 ± 0.00 aC	0.241 ± 0.00 bD	0.235 ± 0.00 bD
	Pérola	0.988 ± 0.001 aA	0.305 ± 0.00 aB	0.270 ± 0.00 bC	0.252 ± 0.00 aD	0.243 ± 0.001 aD
Acidity (% citric ac. db)	Jupi	2.66 ± 0.02 aBC	2.59 ± 0.10 bC	2.71 ± 0.10 bBC	2.83 ± 0.10 bAB	2.99 ± 0.10 bA
	Pérola	2.13 ± 0.05 bC	3.09 ± 0.10 aB	3.10 ± 0.00 aB	3.27 ± 0.00 aB	3.64 ± 0.10 aA
Total sugars (g/100 g db)	Jupi	45.10 ± 0.72 aE	52.12 ± 1.27 aD	54.92 ± 0.89 aC	62.78 ± 0.16 aB	71.73 ± 0.81 aA
	Pérola	35.49 ± 0.31 bD	41.38 ± 0.09 bC	41.95 ± 0.16 bC	44.45 ± 0.09 bB	48.81 ± 0.33 bA
Reducing sugars (g/100 g db)	Jupi	41.45 ± 0.65 aA	40.99 ± 0.09 aA	40.81 ± 0.08 aA	36.72 ± 0.21 aB	32.33 ± 0.97 aC
	Pérola	32.78 ± 0.46 bC	36.84 ± 0.56 bA	35.68 ± 1.04 bAB	35.05 ± 0.27 bB	30.50 ± 0.12 bD

db—dry base. Means followed by the same lowercase letter in columns and uppercase in rows do not differ statistically by Tukey’s test at 5% probability.

, corresponding to the averages and standard deviations obtained for the in natura sample and the dehydrated samples at temperatures from 50 to 80 °C.

The water activity (a_w) did not follow the water content in most samples, presenting statistically similar values between the in natura samples and statistically different values between the flours of the two varieties. A tendency in the reduction in aw with the increase in drying temperature was also observed. The low values of a_w and the water content in the dehydrated samples are adequate to avoid biochemical and microbial reactions responsible for the loss of in natura products, making them suitable for storage [89,90,91]. Other studies addressing the drying of residues from fruit processing also correlated the reduction in a_w with the drying temperature: peels and stalks of pineapple cv. Pérola dried at 50, 60, and 70 °C [92]; white-fleshed pitaya peels dehydrated at 50, 60, and 70 °C [93]; and acerola agro-industrial residues dried at 60 and 80 °C [94].

The acidity content was higher in the Pérola variety in the flours obtained at all temperatures, surpassing the Jupi variety samples by about 0.5% or more. The increase in drying temperature resulted in a trend towards increased acidity in all samples. Similar behavior was also verified in taro flour with acidity contents of 0.77, 0.92, and 0.98% db for drying temperatures of 70, 80, and 90 °C, respectively [95].

The total sugar content was higher in the Jupi variety, thus remaining in the samples obtained at all drying temperatures. With the increase in temperature, the contents also increased, with increases between 37 and 57% between in natura and dried samples at 80 °C in the Jupi and Pérola varieties, respectively. The increase in the sugar content of the material subjected to drying can be attributed to a possible hydrolytic activity caused by the effect of time and temperature [96]. A similar behavior was described by Alves, Machado, and Queiroga [97] when they obtained higher levels of total sugars in cashew apple flour when compared to the in natura sample. The total sugars of the pineapple peel flours were much higher than that quantified by Queiroz et al. [98] in lychee peel flour with an average value of 11.08 g/100 g db.

The reducing sugars initially accompanied, like the total sugars, the highest value for the Jupi variety, surpassing the value of the Pérola in natura sample by more than 25%. With the drying, the behaviors diverge, reducing the values consistently in the Jupi variety between 50 and 80 °C, while in Pérola, an increase was observed among the in natura sample and the others, except for the sample obtained at 80 °C. According to Santos et al. [93], the reduction in reducing sugars with heating can be explained by the Maillard reaction, which results in the degradation of reducing sugars when complexing with free amino acids.

4. Conclusions

Of the ten drying kinetics adjustment models used for pineapple residues, nine showed good prediction, with emphasis on the Diffusion Approximation, Page, and Midilli models.

The effective diffusivity values were between 3.1 × 10⁻¹⁰ and 7.3 × 10⁻¹⁰ m²/s, with similar results for the samples of both varieties and a good correlation with temperature via the Arrhenius-type equation, with a higher value of activation energy in the sample of the Pérola variety; the residues of this variety showed greater enthalpy and entropy variation in the drying process.

The residual pineapple peels of the Jupi and Pérola varieties were sources of minerals, sugars, and proteins, with higher levels of acidity and proteins in the Pérola variety and more sugars in the Jupi.