Combination of gamma irradiation and storage condition for improving mechanical and physical postharvest characteristics of fresh garlic cloves

Abstract The aim of this study was the discrimination and optimization of irradiation effect under physical and mechanical experiments on garlic. The samples were irradiated with 0, 75, and 150 Gy doses and stored at 4 and 18°C for 5 months. Physical, mechanical, and color properties were measured in the period of storage. Based on the results, all irradiated garlic samples had less quality variation than control samples. Response surface methodology (RSM) optimized dose, storage time, and temperature of the stored garlic which was 75 Gy, 2 months, and 17°C, respectively. In addition, after finding the optimal dose, time, and temperature, the most effective factor as weight loss was obtained and the data were classified by the principal component analysis (PCA) approach. The results showed that the PCA method had a high ability to classify and separate the data obtained from measuring the physicochemical properties of garlic and cover 99% variance of data. Moreover, partial least square (PLS) was applied for predicting weight loss data with R2 0.9999. As well, a mechanical test was investigated for finding the best situation and duration of storage condition. Finally, irradiation prevented the destruction of garlic and saved garlic in the best quality as compared with control or nonirradiated samples. After all this, it can be decided to keep garlic in warehouses and transfer this product with minimum damage.


| INTRODUC TI ON
Preservation of garlic as an export basic crop for a long time is very important. Garlic is rich in folic acid, vitamin C, calcium, iron, magnesium, potassium, zinc, and vitamins B 2 , B 1 , and B 3 (Balmori et al., 2019;Eugster et al., 2018). It is necessary to maintain the quality of this type of food that is rich in nutrients for a long time (Noda et al., 2018). In order to preserve foods, there are different ways, such as thermal, freezing, drying, fermentation, coating, packaging film, irradiation, atmosphere control, etc. (Li et al., 2016).
Irradiation is an advanced technical approach to preserve the physical and chemical properties of the food product (Bearth & Siegrist, 2019;Güler et al., 2017). Gamma irradiation is a nonthermal method and has emerged for the preservation of fruits and vegetables to prevent the use of chemical preservatives (Olanya et al., 2015). Many studies have described the useful effect of low-dose irradiation (50-150 Gy) for sprouting inhibition (Zhang et al., 2016). Some studies showed that storage of irradiated garlic with100 Gydoseinarefrigeratorenvironmentreducesweightloss and spoilage .
Designing experiments using response surface methodology (RSM) was performed for the first time in 1950. It was initially used for chemical industries, but recently, RSM has been widely used to improve quality, design product, and analyze uncertainty. RSM is a set of statistical and applied mathematics techniques to build experimental models. The aim of this type of design is to optimize the response (output variable) which is influenced by several independent variables (input variables) (Mohammad-Razdari et al., 2021).
To better understand the data of chemistry, the application of statistical sciences, computers, mathematics, and graphics is necessary. Chemometrics methods are used to gain chemical information obtained in the laboratory, in such a way that useful information is extracted by analyzing the obtained chemical data. Based on this information, the desired experiments can be designed with high efficiency. For data analysis, researchers use PCA (principal component analysis),LDA(linearDiscriminantAnalysis),ANN(ArtificialNeural Network), K-NN (Nearest Neighbor), and multivariate regression (Tazi et al., 2019).
Duetothecostofkeepinggarlicinwarehouses,limitedstorage environment, high garlic production in Iran, and its high consumption, it is very important to try to improve the quality of garlic. Also, irradiation is considered as an effective way to preserve garlic and improve its safety (Hassan et al., 2019). Research on the application of irradiation treatment on the garlic product has been carried out to control pests and stop microbial activity, but the effect of treatment on the quality properties, especially the mechanical properties of this product had not been done. Herein, based on our knowledge, for the first time was investigated discrimination of the results based on the chemometric approach and optimization by response surface method in gamma-irradiated garlic.

| Physical and chemical parameters
The garlic clove color was measured with the portable colorimeter China). In addition, the humidity and weight loss of garlic cloves were investigated through storage (Borchert et al., 2014). In order to estimate the allicin, the amount of total pyruvate and nonenzymatic pyruvate were calculated. After removing the contaminants, 50 g of thesamplewascrushedwellwith100 mlofdistilledwaterinanelectric mixer (German Yuro-Sonic). After 10 min, the resulting solution was filtered through Erlenmeyer filter paper. From the filtered extract mixture, 50 μl was taken and poured into the tube with 2 ml of distilled water and 2 ml of dinitrophenylhydrazine solution. The resulting solu-tionwasplacedinahotwaterbathat37°Cfor10min,then2mlof 1.5 M sodium hydroxide was added to each tube. After adding NaOH, the samples must be read quickly. The absorption of the samples at a wavelengthof515 nmwasreadwithaspectrophotometer(Cary100 model, manufactured by Varian Company, USA). Sodium pyruvate was alsousedasastandardintheconcentrationrangeof0-60 mMinstead of garlic extract (Gruhlke et al., 2019;Reiter et al., 2020).

| Optimization by RSM
The effect of three independent variables (dose, storage time, and temperature) was investigated by Box Behnken design on the dependent variables (humidity, weight loss, color, and allicin). Assumed that there are three mathematical functions fk for yk which are in equation (1): where, ε 1 , ε 2 , and ε 3 are natural variables. ε 1 , ε 2 , and ε 3 are dose, temperature, and storage time, respectively. In the response surface method, natural variables change to coded variables (x 1 , x 2 , …) equation (2): This research used a quadratic model for modeling equation (3) (Sukumar & Athmaselvi, 2019): In equation (3), y k is the predicted response which was considered as dependent variables (k = 1, 2, …, 8). Xi is the input coded variable or the same independent variable (i = 1, 2, 3). The values of the independent variables were coded between −2 and + 2. All coefficients are parameters of regression coefficients. Using the quadratic model, five mathematical models were evaluated for each dependent variable. Each factor in the design was measured at three different levels (2−, 0, +2), two axial points, and six repetitions at the central point (Table 1). Experimental design as well (1) y k = f k ε 1 , ε 2 , ε 3 (2) yk = fk x 1 , x 2 , x 3 , … (3) ware. Analysis of variance on quadratic model coefficients was also performed using this software. Significant sentences in the model were obtained using analysis of variance for each response.  The basic requirement of designing machines for the processing of fruit and food is having information about the physical characteristics of the fruit. In the lines of transportation and packing, the fruit is subjected to different loads that may damage its tissue. The stress and strain state under static and dynamic loads that are related to the nature and mechanical behavior of the material is considered as the first step in quantitative analysis of the characteristics of agricultural products (Mahiuddin et al., 2018). One of the best models for investigating load damages of fruit is the general Maxwell model stress relaxation test design. To overcome this drawback in agricultural products, springiness with coefficient E e was added to the general Maxwell model (Mohsenin, 1986). Equation (4) shows the expression of Maxwell's mathematical model:

|
where the consideration σ i = ε 0 E i , σ e = ε 0 E e and can be written as equation (5): The obtained model coefficients were determined and evaluated from relaxation stress curves. Residues were determined using the sequential model. All stress relaxation time models were calculated by MATLAB R.13 software.

| Sheartest
In order to determine enough shear force for the garlic product, an experimental shear tool with a commercial single sickle knife section and a counter shear was used as the twofold shear test. Finally, thetestwasdoneusingaZwick/Roelltexturometer.Thediameter cutter probe, preload, distance of probe to the bottom of the page, andspeedatstartpositionare0.5 mm,0.3N,65,mm,and300 mm. min −1 , respectively ( Figure 1).
The shear stress was calculated in MPa using equation (6) (Mohsenin, 1986): where F max is the maximum shear force (N) and A is the cross-sectional area of the stalk at shear planes (mm 2 ). The shear energy was calculated

Independent variables Coded variables Natural variables Levels
Storage time (month) by the area under these curves to the maximum force of the curve (Chen et al., 2004). The cross-sectional area for calculating the shear test was calculated by Solid Works R.2015 software. For this purpose, garlic cloves were designed in the software to obtain the main dimensions (length, width, and thickness) of the samples. By drawing a three-dimensional image, the rectangular area was measured in these dimensions ( Figure 1).

| Statistical analysis
These experiments were designed based on a factorial design with threefactorsofgammaray(control,75,and150 Gydoses),storage time(1-5months),andthestoragetemperature(4and18°C).The analysis of variance was carried out on data using SPSS software (IBM SPSS Statistics 22, IBM, NY). Also, differences between means were determined with Tukey post hoc comparison tests. p-values of 0.05 or less were considered significant.

| Optimization
According to the results of ( Figure  According to the results of the analysis of variance (Table 3), the storage time on the L* index was not significant but had a great effect on the main effects of dose and temperature on the L* sample. Among the interactions, the interaction of the AB parameter on L* was not significant but the effects of AC and BC on L* were F I G U R E 1 3Dcoutureofhumidityand weight loss changing with dose, time, and temperature significant. The results of ( Figure 2) show the three-dimensional con-tourL*.Accordingtotheresults,thecontrolsampleand150 Gyhad higher brightness.
According to the analysis of variance (Table 3), the main effect of storage time on the parameter a* is significant and the main effects of B and C on a* are not significant. Also, the interactions of AC and BC on a* are significant. The effects of the square on a* are significant. According to (Figure 2),inthecontrolsampleand150 Gy,a*is morethaninthe75 Gysample.
With increasing storage time, the value of a* increased, which is due to the reduction of humidity in the sample tissue and the shrinkage process (Kainthola et al., 2020). With increasing the amount of radiation intensity, the value of a* increased compared to nonirra- According to the results of analysis of variance (Table 3) is also proportional to the intensity used (Olanya et al., 2015). In a study, by increasing the amount of radiation intensity to a certain valueandputtingthesamplesat4°C,theamountofb*oftheirradiated sample increased less than the control sample (Durante et al., 2020). This means that the L* index decreases with increasing storage time and increasing the amount of irradiation intensity (Tables 1 and 2). This is due to the decrease in humidity and the reduction of water between cells during the storage period (Noda et al., 2018). On the other hand, with increasing radiation intensity, the intracellular water, as mentioned above, decreases and evaporation takes place (Sukumar & Athmaselvi, 2019). Similar results were obtained in another study that decreased the value of L* with increasing the amount of radiation intensity and storage time (Duranteetal.,2020). Table 4  can be said that at the beginning of storage, the allicin was high, but after 5 months, the allicin has decreased sharply, and this reduction has occurred in all samples and treatments.
As the shelf life increased, the stiffness of the samples decreased

| Introducing optimal points and desirability index
Based on the results of the desirability index, the best and most optimal answer is obtained for the variables of time, temperature, and dose so that it can be decided based on the results of the variables of humidity, weight loss, color, and allicin to maintain garlic samples.
Desirability index is the way to find the best points that are 0.76.
According to (Figure 3  parameters. Also, all parameters except BC are directly related to b* and as these parameters increase, b* also increases.
Finally, in the study of the degree of allicin, the independent parameters A, C, AC, BC, and A 2 have an inverse effect on the allicin and if these parameters increase independently, the reverse downward trend occurs for the allicin and the parameters B, AB, B 2 , and C 2 are also related. It has a steep downward trend and increases allicin with their increase.

| Classification of weight loss by PCA
Duetothelargeeffectofindependentvariablesonweightloss,PCA results of weight loss changes have been investigated. The results of PCA on weight loss show a good and significant difference between the control samples, 75 and 150 Gy at two temperatures of 4 and 18°C and in different months. (Figure 4a) of PC1 and PC2 covers the ability of 71% and 21% of the data, respectively, and observed a total of 92% of the variance between the data, and is able to make a good distinction between the samples. It is noteworthy that the control groups, 75 and 150 Gy, have been identified together and can be grouped together. In the second month of storage, PC1 and PC2 were able to distinguish 66% and 31% of the variance between the data and 97% of the total variance between the data, respectively, and were able to differentiate the data. This indicates the high accuracy during testing and the ability of PC detection in data classification. Also in months 3, 4, and 5, PC values have 99%, 95%, and 85% variance between data, respectively.
The best classification was done in month 3. In other words, it can be said that from month 3 onwards and with the aging process of garlic samples, the data in almost all groups are closer to each other and the weight loss of samples is obvious. Also, (Figure 5) shows the PC results for separating the samples at 5 months of storage for the controlsamples,75and150 Gy.PCwasabletocoveratotalof93% of the variance between the data and in the samples of irradiated garlicwithadoseof75and150 Gy,thePCwasabletocover94% and 85% of the variance between the data and it can be said that the overall optimization results with RSM and high PCA accuracy, thebestradiationdoseis75 Gytoseethebestweightlossresults.

| Prediction by PLS
In this section, (Figure 6) shows the PLS results of weight loss data in the control and irradiated samples at 5 months of storage at 4 and 18°C. Predicted weights versus actual weights are shown and the results show that PLS has the ability to predict data in 5 months of high-performance R 2 storage.  (Calado et al., 2018).

| Relaxation test
Relaxation time is different based on the characteristics of the viscoelastic or viscous substances, but this time is longer in elastic material. The results of chemical studies (Calado et al., 2018)  F I G U R E 3 2Dcoutureofdesirability, color parameters, and allicin changing with dose, time, and temperature and best optimal point of the product and is reduced because of the humidity (Danalache et al., 2015). However, critical stress decreases with an increase in storage time, but this change does not occur in irradiated samples. Thus, ir-radiatedgarlicsampleswith150 Gydosehavemorecriticalstresstoo.

| Shear test
As shown in Figure 7a, shear stress in garlic is reduced with increasing storage time, but irradiation keeps the shear stress almost constant. does not allow the shear stress to reduce, and as a result, irradiated andcontrolsamplesstoredat4°Chavehighervaluesofshearstress.
Shear strain was significantly reduced, because of the increase in storage time and storage temperature (Figure 7b). But in contrast, irradiation prevents shear strain reduction, because of the tissue damage and softening of the stored sample (Calado et al., 2018).
According to the results (  Figure 7c shows the preserved tissues in irradiated samples and they have higher values. As mentioned, increasing storage time can decrease the shear stress, because the garlic samples lose their intracellular water with increasing storage time. Of course, due to biological activity, breathing decreased humidity content (Pérez et al., 2007).
Compressibility is most of the original strains because the cells are full of water in fresh garlic (Zhou et al., 2010). In a study that had been done on garlic, there was a decrease in the shear stress and energy of rupture when tissue was destroyed (Llamas et al., 2013).

| CON CLUS IONS
Irradiation preserves mechanical and physical properties of the food.
In this study, irradiated samples stored at 18 and 4°C had lower changes in color parameters (a*, b*, and L*). Also, based on the results, irradiation has a significant effect on garlic tissues and preserves the tissues of the sample, but a high dose has the reverse effect on shear stress, shear strain, and energy of rupture. Therefore, the storage time had a significant effect on mechanical properties. Likewise, results showed that chemometrics methods like PCA and PLS can classify and predict storage condition data exactly. Then, in a relaxation test, two components of the Maxwell model were obtained for garlic cloves. Moreover, with increasing storage time, the relaxation time average increased in irradiated samples than in the control samples.
Finally, the results showed that the irradiation preserved some mechanical and physical properties of garlic during 5 months of storage.

ACK N OWLED G M ENTS
Bu-Ali Sina University is appreciated for supporting this research in terms of supplying laboratory materials and equipment.

FU N D I N G I N FO R M ATI O N
This study was supported by the office of vice chancellor for research at Bu-Ali Sina University (Thesis No. 4375).

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author.

PR AC TI C A L A PPLI C ATI O N S
Irradiation is an effective way for increasing the shelf life of food products.Irradiatedgarlicstoredfor5 monthsundercontrolledcondition and optimization of effective parameters was done by RSM.
Also, discrimination of irradiated garlic was done by chemometric approaches. Moreover, a mechanical test was completed for the storage simulation by the General Maxwell model and shear test.