The Fate of Physico-Chemical Characteristics and Antioxidant Properties in Dried and Roasted ‘Super Hot’ Chili Fruit

Dried and roasted chilies are used in many recipes due to higher pungency, especially in Asia. However, the roasting process is usually done on a pan, or wok without temperature control. Thus, this study aimed to evaluate the effects of different hot air oven drying temperatures (60, 80 and 100 °C), roasting temperatures (120, 140 and 160 °C) and times (30, 60, 120 min and at 12–13% moisture content (MC)) on the physicochemical changes in dried ‘Super Hot’ chili fruits. High antioxidant compounds that have health benets were detected in dried chili samples such as capsaicinoids, limonene, pinene, tocopherol, and oleic acid regardless of temperature and time of drying. Chilies dried at 60 °C until reaching 12–13% MC (1440 min) had better color retention and DPPH • values. In contrast, the contents of bioactive compounds were the highest at 160 °C, despite having a darker red color. This result showed that chilies dried or roasted at high temperatures allowed the bioactive compounds to be released from the bound state and can be easily absorbed during consumption. Thus, hot air oven roasting at 160 °C can be applied to improve bioactive compounds and antioxidant activity in ‘Super Hot’ chili for health benets and taste. a chain the vanillin into the secondary metabolites still roasting still possess nutritive

A Minolta CR-400 colorimeter (New Jersey, USA) was used to determine the surface color of the samples (L*, a*, and b* values). The values were recorded as the mean of three determinations at three different sample locations. Total color differences were calculated according to equation 2:

Extraction
The dried chilies were extracted using the method described by Maokam et al. (2014). The extracts were used for determining metabolites, phenolic and avonoid compounds, total phenolic content (TPC), total avonoid content (TFC), and biological activity analyses.
Determination of metabolites and capsaicinoid derivatives by gas chromatography-mass spectrometry (GC-MS) The extraction solution was analyzed and identi ed for metabolites and capsaicinoid derivatives by GC-MS using an Agilent 7890A gas chromatograph (Agilent Technologies, Santa Clara, CA, USA). The metabolites were separated using a silica capillary column HP5-MS (30 m length; 0.25 mm i.d.; 0.25 µm lm thickness; Agilent J&W Santa Clara, CA, USA). Helium was used as the carrier gas at a constant ow of 2.0 mL/min. The temperature of the injector and detector was 250 °C and 280 °C, respectively. The initial temperature condition was 35 °C for 10 min; increased to 95 °C at 3 °C/min, then to 270 °C at a rate of 10 °C/min (constant for 10 min) and nally to 300 °C at a rate of 3 °C/min (constant for 10 min). The mass scan range used was 30 to 550 amu at 310 °C and 5 μL extracts were directly injected. The compounds were compared with Wiley 275 and the NIST library database using ChemStation Data Analysis software at 80% quality match.

Total phenolic content (TPC)
The Folin-Ciocalteau method (Singleton and Rossi 1965) was used to determine the TPC of extracts expressed in mg gallic acid equivalents per g dry weight (DW) (mg GAE/g DW).
Total avonoid content (TFC) The aluminum chloride colorimetric method (Tacouri et al. 2013) was used to determine the TFC of the extracts expressed as mg rutin equivalents per g dry weight (mg RUE/g DW).

Pro ling and quanti cation of phenolic and avonoid compounds
The separation of phenolic and avonoid compounds was performed by high-performance liquid chromatography (HPLC) using an Agilent 1200 series (Agilent Technologies, Santa Clara, CA, USA) equipped with a diode array detector (G1315B 1200) and an RP-18 GP Mightysil column, 250 mm × 4.6 mm × 5 µm (Kanto Corp, Portland, Oregon, USA). The mobile phase used was deionized water adjusted to pH 2.5 with tri uoroacetic acid (solvent A) and acetonitrile (solvent B). For phenolic pro ling, a gradient ow for 70 min was used with a starting ratio of 95:5 (A:B) at injection volume of 10 μL. For avonoid pro ling, a gradient ow for 33 min was used with a starting ratio of 50:50 (A:B) at injection volume of 20 μL. Both pro ling used a ow rate of 0.6 mL/min and wavelength at 280 nm (phenolic) and 350 nm ( avonoids). Phenolic compounds standards consisted of gallic acid, chlorogenic acid, vanillic acid, caffeic acid, vanillin, 2-hydroxycinnamic acid, syringic acid, p-coumaric acid, epicatechin, and avonoid compounds standards of rutin and quercetin. Quanti cation was performed with calibration curves using external standards and expressed in dry weight basis.

Quanti cation of capsaicin (CAP) and dihydrocapsaicin (DHC) contents
The extraction and cleanup procedure using solid phase extraction (SPE) followed the method by Maokam et al. (2014). The mobile phase consisted of 60% acetonitrile in distilled water and 0.5% formic acid in distilled water with a ow rate of 1 mL/min and an isocratic ow at ambient temperature for 20 min.
Quanti cation was performed with calibration curves using CAP and DHC external standards.

Determination of antioxidant activities
Determination of ferric reducing antioxidant power (FRAP) The FRAP assay was performed based on the method of Benzie and Strain (1999) and the results were expressed as mg Trolox equivalents (TE) per g of dry weight (mg TE/g DW).
Determination of 2,2-diphenyl-1-picrylhdrazyl (DPPH) radical scavenging capacity The DPPH radical-scavenging capacity of the extract was determined, and the results are expressed as percentage inhibition (Brand-Williams et al. 1995).

Statistical analysis
The experiment was arranged in a 6 × 4 factorial in a randomized complete block design (RCBD) with three replications. All parameters were analyzed in triplicate. The data obtained were subjected to analysis of variance (ANOVA). The mean comparisons were separated by using Duncan multiple range test (DMRT) when the F-test was signi cant at p≤0.05.

Results And Discussion
Moisture content and moisture ratio The drying curves for whole chilies dried at different drying temperatures are shown in Figure 1 as a plot of moisture ratio versus drying time. In this study, a hot air oven was used for dehydration of whole chilies using convection, and a constant rate and falling rate period are exhibited in the drying curves. During the constant rate period, water from the chili surface was removed and rapidly replaced by moisture from the chili's interior, thus maintaining the wet surface (Barbosa-Cánovas and Vega-Mercado 1996), and the constant rate period was governed by the drying temperature. As the drying temperature increased, a shorter constant rate period was observed, and vice versa. The falling rate period indicated that the moisture at the chili surface continued to diminish until the surface was completely dry, which was indicated by the end drying times for each drying temperature (Sanjuán et al. 2003). A constant rate period was obtained because of the resistance in water evaporation caused by the whole chili pericarp ( Barbosa-Cánovas et al. 1996), wax content, and thickness of the pericarp (Sanjuán et al. 2003). In contrast, another study only found a falling rate period during whole chili dehydration (Li et al. 2009). The difference in the results might be due to the chili variety, which had less resistance to water evaporation (thinner pericarp and less wax). A shorter time was needed to dehydrate the whole chili at a higher temperature (160 °C) compared to the time required at a lower temperature (60 °C) (almost 15-fold less). This was due to faster water evaporation from a larger driving force for evaporation of the water from samples (Tunde-Akintunde and Afolabi 2009). Based on these results, the best drying temperature was 160 °C, which requires the shortest drying time (120 min) to achieve 12-13% MC.

Surface color
The measured color values are presented in Table 1 for chilies dried at different temperatures and times. All drying temperatures showed a reduction in the L* (lightness) value of the chili when dried to 12-13% MC. The a* value, which describes the redness of the chilies, also showed a signi cant decrease during drying. ∆E, which compares the color between fresh and dried chilies at different temperatures and times, increased signi cantly regardless of drying temperatures due to the darkening of the chilies during drying. The changes of L*, a* and ∆E values for LTLT were slower than those of HTST which may be due to the rate of nonenzymatic browning (Maillard reaction and caramelization) (Benjakul et al. 2005) and carotenoid degradation ) that increased with drying temperature and time. Based on these results, a drying temperature of 60 °C showed the best L* and a* values at 12-13% MC, with the lowest ∆E value for maintaining the red color of the dried chilies.

Metabolites pro ling and capsaicinoid derivatives
A total of 33 compounds were identi ed in both fresh and dried chili samples ( Table 2). Compounds such as 2-methoxy-4-vinylphenol, undecanoic acid, oleic acid, and pentadecanoic acid identi ed this study using GC-MS were also reported in other studies (Pérez-Gálvez et al. 1999). The compounds found can be divided into several classes such as alkaloids, aldehydes, fatty acids, pyrazines, and terpenes. The largest group of compounds detected in this study is terpenes (β-pinene, limonene, D-limonene, and γ-terpinene) and fatty acids (e.g. tetradecanoic acid, undecanoic acid, and pentadecanoic acid). The alkaloid group decreased at the initial drying time before increasing at the end when the MC was 12-13% w.b. (further discussed in later section). In this group, capsaicinoids were higher than fresh chili and increased at HTST and increased the pungency. Interestingly, the compound (S)-2-propylpiperidine, an alkaloid was also detected in this study even at HTST. This compound found in roasted chili can be a potential source of antioxidants. The same trend was also observed in the aldehyde and fatty acid groups and can be attributed to the reaction between fatty acids and oxygen or the thermal oxidation of polyunsaturated triacylglycerols (Zhang et al. 2012). The formation of pyrazine group in this study was due to Maillard reaction which facilitated the condensation between amino compounds and sugar fragments (Yu et al. 2013) during the drying or roasting processes. Studies have shown that pyrazine is an important aroma compound and safe for consumption (Fadel et al. 2018). In the terpene group, β-pinene and limonene showed a decreasing trend with drying time irrespective of temperature. However, the total amount of both limonene and terpene was higher at LTLT compared to HTST. This is because HTST accelerated the chemical reactions (autoxidation) of terpenes (Lee et al. 2007) and, consequently, more degradation occurred at HTST. Chili roasted at 160 °C still possessed high limonene and pinene which can be applied as natural insecticides (Hollingsworth 2005). Additionally, tocopherol was also found to increase in chili samples, except for 160 °C. Reports have shown that tocopherol is a strong antioxidant and can protect against β-carotene oxidation (Zhang and Omaye 2000). A small percentage of vanillin was also detected; however, the amount did not differ much among temperatures at 12-13% w.b. The presence of vanillin could be due to thermal cleavage of capsaicinoids between an amide bond and vanillin moiety (Henderson and Henderson 1992), or from the oxidation of ferulic acid (Kundu 2017). Capsaicin which consists of a vanillin moiety bind with fatty acid side chain by an amide can be destroyed by heat treatment and release the vanillin into the surroundings. These results showed that secondary metabolites were still detected at roasting temperature of 160°C in dried chilies and still possess nutritive values.

Total phenolic content (TPC)
The TPC in chili extracts was ascertained for different drying temperatures and times (Table 3). Drying at 120 °C for 30 min (63.54 mg GAE/g DW) and 60 min (67.15 mg GAE/g DW) gave a signi cantly higher TPC value than the other treatments. Compared to fresh chilies, the TPC was higher when the chilies were subjected to various drying temperatures during the rst 30 min, with drying temperature 120 °C resulted in the highest TPC values. For LTLT, the TPC showed a gradual decrease with drying time, but at 12-13% MC, the TPC was not different from that of fresh chilies. Moreover, TPC values at drying temperatures of 140 °C and 160 °C were higher than those of fresh chilies by 27.9% and 33.6%, respectively. The overall trend of TPC during the drying time showed an initial increase, followed by a decrease and nally an increase.
During drying, several chemical processes occur concurrently which can either reduce or increase the TPC as the heat from high temperatures produced by the hot air oven caused damage to the cellular structure (Deng et al. 2018). The different types of phenolic compounds found in chili determine the TPC values.
Glycoside-and ester-bound fractions of phenolic acids can decrease due to cleavage by heat treatment (Materska and Perucka 2005). Additionally, increased contact between oxidative enzymes and its substrates (phenolic compounds) as the membrane loss its selective permeability could cause a reduction in the TPC due to oxidation (Raynal et al. 1989).
Furthermore, the accumulation of phenolic compounds due to the disassociation of the bound phenolics from proteins or the cell wall (Asano et al. 1982) could also occur. The disassociation of bound phenolics is due to their hydrophobic benzenoid rings and hydrogen-bonding potential of the phenolic hydroxyl groups, which are heat labile (Asano et al. 1982) and thus release of bound phenolics. Additionally, byproducts and intermediates from the Maillard reaction could contribute to an increase in TPC, as shown in a study on bird chilies, by which the TPC increased at a drying temperature of 121 °C compared to drying at a lower temperature (Wangcharoen and Morasuk 2009). The results from this study indicated that drying chilies at high temperatures (up to 160 °C) potentially improved TPC values.
Pro ling and quanti cation of phenolic compounds A total of nine phenolics (gallic acid, chlorogenic acid, vanillic acid, caffeic acid, syringic acid, vanillin, p-coumaric acid, epicatechin, and 2-hydroxycinnamic acid) were detected in the chili extracts by HPLC. In the phenolic fraction, a total of nine compounds were found alongside 18 unknown compounds. The overall trend observed was a higher increment in phenolic compounds at HTST compared to LTLT at 12-13% MC by 1. Total avonoid content (TFC) The TFC tended to decrease irrespective of drying time and temperature compared to that of fresh chilies (Table 3). When comparing the drying temperatures at 12-13% MC, the TFC values were in the order of 160 °C > 140 °C >120 °C = 100 °C > 80 °C < 60 °C. However, the trend of most drying temperatures was a decrease in the TFC during the rst 60 min before an increase was seen at 12-13% MC, but the levels were still lower than that in fresh chilies. These results were due to avonoid degradation from the interaction between avonoids with oxygen and degradative enzymes, such as peroxidase (Buchner et al. 2006 The results obtained showed that in fresh chilies, both rutin (4.22 mg/g DW) and quercetin (0.03 mg/g DW) were detected alongside 15 unknown compounds.
Rutin (1.62-5.17 mg/g DW) was observed to be more thermostable compared to quercetin and was detected in all chili extracts, irrespective of time and temperature. Interestingly, rutin and quercetin were detected at the roasting temperature of 160 °C (Figure 2b). Since rutin also has a catechol group, it also can regenerate in methanol solvent. However, the rutin showed a decreasing trend with drying time and was 2-to 3-fold lower than fresh chilies. As for quercetin, only a small concentration can be detected in fresh (0.03 mg/g DW) and dried samples (0.01-0.08 mg/g DW) with higher degradation at HTST than at LTLT. Quanti cation of capsaicin (CAP) and dihydrocapsaicin (DHC) contents Both CAP and DHC are the most abundant capsaicinoids in chili fruits (Maokam et al. 2014) that contributed the most to the pungency and spiciness in hot chilies. Therefore, only CAP and DHC were quanti ed in the chili samples (Table 4). In this study, the highest CAP contents could be found at the drying temperatures of 60 °C (60 and 120 min), 80 °C (60 min), 140 °C (30 min) and 160 °C (30 min), whereas drying at 80 °C for 60 min resulted in the highest DHC content. Both the CAP and DHC contents were also observed to be higher than those of the fresh chilies irrespective of temperature and time. Additionally, dried Habanero Red Savina chilies had a high CAP content, which was related to its high pungency (Popelka et al. 2017). In 'Super Hot' chilies, both the CAP and DHC contents were on average ve times higher in the dried than in the fresh chilies, which showed that the pungency increased upon drying. The results showed that CAP and DHC started their release from the cell matrix within 30 min of drying due to the dissociation of CAP and DHC from the cell structures.
Overall, at 12-13% MC, the CAP contents increased by an average of 6-fold compared to the control, with the greatest increase at 140 °C (7-fold) and the smallest increase at 80 °C (3-fold). Similarly, it was also observed that their contents were higher than those of fresh chilies irrespective of drying temperature and time with an average of a 5-fold increase. The CAP and DHC contents in dried chilies also increased at HTST.
Cell rupture is unavoidable during the drying process due to the removal of water, with a higher percentage of rupture occurring at higher temperatures ). During cell rupture, compounds are released from the cells and encounter degradative enzymes such as peroxidase, which causes a reduction in CAP and DHC contents. Consequently, the longer time needed to remove moisture ( Figure 1) due to LTLT also facilitated the oxidation reaction between CAP, DHC, and degradative peroxidase, which are in close proximity. In addition to oxidation, thermal decomposition of CAP and DHC caused by the cleavage between an amide bond and vanillin moiety, hydrogenation or the arrangement of methylnonenamide and deamination of 8-methyl-6-nonemide may also occur at a high temperature (Henderson et al. 1992), e.g., 160 °C.
The effect of peroxidase on CAP and DHC can be reduced by applying HTST with faster water removal. A higher reduction in peroxidase activity at a temperature of 60 °C was observed in another study compared to that at 50 °C with low water content (Korbel et al. 2013). The availability of CAP and DHC can also be increased by being released from the bound state into the free state when high heat is applied (Asano et al. 1982). Also, HTST and faster removal of water from whole chili fruit will result in higher concentrations of metabolites found in chilies, namely, CAP and DHC, due to their hydrophobicity and high boiling point. It was also observed that CAP and DHC contents were higher and stable throughout the drying time at a drying temperature of 160 °C. This suggested that high temperatures (e.g., 160 °C) can be used to dry chilies without compromising the CAP and DHC contents. This result also implied that whole chili fruit dried at HTST (e.g., 160 °C) induced metabolites by the disassociation from the cell wall (placenta) and faster removal of water that increased the concentration.  , and 120 °C was an average of 0.8-fold lower than that of fresh chilies. This can be explained by the degradation of polyphenol structures, such as glycosylation and alkylation, which caused a reduction in FRAP values (Chaaban et al. 2016). However, the FRAP values at high temperatures (140 °C and 160°C ) at 12-13% MC were 1.1-fold and 1.2-fold higher than those of fresh chilies, respectively. These increments can be contributed by the degradation products of avonoids such as rutin also contained antioxidant activity (Saito et al. 2005). Contribution from the degradation products of vanillin and p-coumaric acid such as 2-methoxy-4-vinylphenol and 4-vinylphenol, respectively were also reported to possess antioxidant activities (Rubab et al. 2020;Terpinic et al. 2011).
In this study, rutin and chlorogenic acid were detected in all samples irrespective of drying temperature and time which probably contributed to the antioxidant activity at HTST. Both CAP and DHC contents (Table 3) were higher at HTST and this also indicated the contribution to the antioxidant activity. Additionally, the DDMP compound identi ed from GCMS was also reported to possess strong antioxidant compound in glucose-histidine Maillard reaction products in a study by Yu et al. (2013). These results demonstrated that HTST (e.g., 140 °C and 160 °C) in uenced high antioxidant activity, especially the FRAP analysis.
These high FRAP values also indicated the in ux of antioxidants as the bonds between the polyphenols and cellular structures were destroyed by heat as drying time increased (Barthelmebs et al. 2000). Overall, drying at 160 °C for 120 min (12-13% MC) resulted in a high FRAP value of 7.20 mg TE/g DW.
In contrast, both the ABTS •+ and DPPH • scavenging activity decreased with time irrespective of drying temperature, with average losses of 0.6-fold and 0.9fold, respectively. The reduction in both assays showed that the ability of the antioxidants, such as capsaicinoids to donate hydrogen was weakened, probably due to the cleavage of hydroxyl groups from the structure by heat (Chaaban et al. 2016), although the content of these antioxidants increased. Additionally, in the case of rutin, the loss of hydroxyl in position 3 of the C-ring caused a decrease in antioxidant activity (Saito et al. 2005). However, the authors reported that the antioxidant activity has either higher, lower or the same activity as the native compounds depending on the number of hydroxyl groups in their structures due to a change in or the degradation of polyphenol structures, which gives rise to new products such as protocatechuic acid (Saito et al. 2005) that may or may not possess antioxidant properties. Stable and acceptable antioxidant activity in terms of the FRAP, ABTS •+ and DPPH • assays was observed when chilies were dried at 160 °C for a shorter time (120 min) compared to the effects observed at other temperatures.

Conclusion
Drying chilies at HTST increased capsaicinoid contents, phenolic, and avonoid contents. Additionally, the presence of compounds such as pinene, limonene, tocopherol, and oleic acid in chili samples regardless of temperature and time showed that chilies dried at HTST still contained metabolite with high antioxidant properties. Drying whole 'Super Hot' chilies at a roasting temperature of 160 °C (120 min) resulted in a rapid reduction of MC to 12-13% w.b, higher retention of capsaicin, metabolites, antioxidant activities, and pungency which are good for health. This new nding using the hot air drying is an alternative and inexpensive method to the traditional roasting method for improving the quality of dried chilies that can be utilized by the pharmaceutical and food industries that are interested in obtaining the maximum amount of secondary metabolites such as terpenes and natural capsaicinoid from chilies. This study is also relevant to suppliers who are looking to diversify their products. Additionally, the consumption of roasted chilies can bring health bene ts to consumers due to their high bioactive compound contents.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.