Physicochemical and Sensory Evaluation Data to Drive the Development of a Green Chili Pepper Hot Sauce from Unexploited Raw Materials

The present study shows the set of analyses conducted during the development of a hot chili pepper sauce to valorize green peppers usually discarded in the Espelette region (France). A traditional production process was used as the inspiration for product development, and two different fermentation processes were assessed and characterized by measuring pH, sugar content, instrumental color, volatile composition, and conducting sensory (discriminant test) and microbiological analyses (total plate count). Significant differences were observed among pepper mash samples with respect to their physicochemical characteristics, but the products were considered similar from a sensory standpoint. Both sensory and physicochemical tests suggested that the ingredients added to make the sauces were determinant and had a higher impact on the organoleptic profile of the final product than the fermentation process. Finally, a Napping® test was conducted to determine the attributes that could differentiate the product from the hot sauces found in the current market. The results of the present research allowed the optimization of the elaboration process of the new product, saving time and ingredient costs. The procedures shown in the study could be used as an example of a new product development process in which physicochemical and sensory data are collected and used for decision making.


Introduction
The geographical origin of peppers (Capsicum spp.) has been reported to be tropical America [1] but, thanks to its adaptability to different environments, this plant has spread across different parts of the world, mainly warm-temperature regions [2,3]. Over 20 to 30 species of pepper have been identified, including five domesticated species: C. annuum, C. chinese, C. pubescens, C. frutescens, and C. baccatum [4,5]. This huge biodiversity has given rise to a great variety of fruits with unique morphologic (e.g., shape and size) and organoleptic characteristics (e.g., color, flavor, pungency), and different gastronomic cultures have developed diverse products and applications from peppers (e.g., fresh, dried, processed as spices, pickled, pastes, sauces, or roasted).

Pepper Samples and Sauce Making Samples
Green peppers (C. annuum var. Gorria) were obtained from a local producer from Espelette (France). The samples were harvested during the last week of November 2021, washed in tap water, and stored in a freezer (−20 °C) until processing. The list of ingredients and summary of the processes used to make and analyze the 3 prototypes is shown in Figure 1; the main differences among them were linked to the fermentation process, to assess the impact of this stage in the final product. The peduncles of the fruits were removed, and the rest of the peppers were mashed in a blender with different ingredients (Thermomix TM6, Wuppertal, Germany). The samples to be fermented (R1 and R2) were transferred to sterilized glass jars (500 g/jar); one batch was left to ferment using the autochthonous microorganisms (spontaneous fermentation, R1-S), and the other one was inoculated with Lactobacillus plantarum Harvest LB-1 (0.08% as suggested by the producer of the starter; Chr Hansen, Hvidovre, Denmark) (R2-I). L. plantarum was chosen due to being one of the multiple lactic-acid bacteria found in peppers, which has been previously studied in pepper fermentation [25,27,28]. The samples were covered with 20 mL extra virgin olive oil (Urzante, Tudela, Spain) to avoid the surface coming into contact with the oxygen of the head space, sealed tightly, and The peduncles of the fruits were removed, and the rest of the peppers were mashed in a blender with different ingredients (Thermomix TM6, Wuppertal, Germany). The samples to be fermented (R1 and R2) were transferred to sterilized glass jars (500 g/jar); one batch was left to ferment using the autochthonous microorganisms (spontaneous fermentation, R1-S), and the other one was inoculated with Lactobacillus plantarum Harvest LB-1 (0.08% as suggested by the producer of the starter; Chr Hansen, Hvidovre, Denmark) (R2-I). L. plantarum was chosen due to being one of the multiple lactic-acid bacteria found in peppers, which has been previously studied in pepper fermentation [25,27,28]. The samples were covered with 20 mL extra virgin olive oil (Urzante, Tudela, Spain) to avoid the surface coming into contact with the oxygen of the head space, sealed tightly, and stored at room temperature (21-24 • C) for 15 days. One of the samples (R3) was directly processed without fermentation into the final product; therefore, sugar was not added to this sample, and it was considered an example of the "original mash" at time 0 days. The three mash samples (R1-S, R2-I, and R3) were smoked (at 72 • C for 150 min) and mixed with other ingredients to finish the formula of the sauce (Figure 1). Then, samples were strained and thickened with xanthan gum (0.2%), obtaining 3 different prototypes of sauce.

Characterization of the Fermentation Process and Products: Green Chili Pepper Mashes and Sauces
The original mix, the fermented mash, and the final products were analyzed using different physicochemical analyses. Microbiological analyses were performed to verify that the fermentation procedures used in R1-S and R2-I were driven by different microorganisms and, therefore, that using the starter added in R2-I could result in a different product. All analyses were run in triplicate. In addition, sensory tests were performed to determine the differences among samples.

Determination of pH
The pH was determined at 20 ± 0.5 • C using a digital pH meter (Crison Basic 20, Crison instruments, Barcelona, Spain) calibrated with standard buffer solutions.

Microbiological Analysis
The mash samples were diluted in tenfold series in saline peptone solution (0.85% NaCl and 1% peptone) and plated onto different culture media to quantify the presence of different microorganisms using a traditional Plate Count Agar (PCA) method (ISO 4833-1:2013) [29]: nutrient agar for total viable colonies in aerobic conditions (30 • C for 48 h) and MRS agar for lactic acid bacteria (37 • C for 72 h, in anaerobic conditions).

Determination of Sugars
Sugars determinations were conducted as reported by Razola-Díaz et al. [30] with some modifications. Freeze-dried mash samples were diluted 1:100 (v/v) with ultrapure water at 60 • C, then shaken for 30 min, and centrifugated at 2367× g for 10 min. The supernatant was filtered using a 25 mm, 0.45 µm nylon VWR ® Syrenge Filter (VWR Inc., Darmstadt, Germany), and 50 µL were injected in the same equipment as Deba-Rementeria et al. [8], using the same column and conditions indicated by these authors. Calibration curves for glucose, fructose, and sucrose, were used to quantify the presence of these compounds, identified by retention time.

Analysis of Volatile Composition
The volatile compositions of the samples were determined by headspace solid phase micro-extraction (HS-SPME) using the same conditions, fiber, column, and equipment reported by Deba-Rementeria et al. [31]. A total of 0.5 g of freeze-dried sample was weighted and ultrapure water (10 mL) and NaCl (1.0 g) were added into the 40 mL vial with polypropylene caps and PTFE/silicone septa for the extraction and processing as reported by these authors. Retention indexes of a commercial alkane standard mixture (Sigma-Aldrich, Steinheim, Germany) were used to identify the compounds, as well as the NIST 17 Mass Spectral and Retention Index Libraries [32]. The identification was considered tentative when only based on mass spectral data; a linear retention similarity filter was set at ±10 units. The relative abundance of each compound was expressed as the percentage (%) of the total arbitrary area units.

Sensory Analysis of the Samples
The protocol for the consumer study was approved by the ethics committee of Mondragon Unibertsitatea (IEB-20221115). All articles from the Declaration of Helsinki and the 2016/679 EU Regulation on the protection of natural persons regarding the processing of personal data and on the free movement of such data were met. The experimental procedure was explained to and a written consent indicating voluntary participation was obtained from each participant prior to beginning the study. The tasting sessions were conducted in a sensory lab with individual booths and controlled temperature and relative humidity (21 ± 2 • C; 55 ± 5% RH); the illumination was a combination of natural and nonnatural light (fluorescent). To determine if sensory differences could be perceptive between products, a triangle test was performed following the UNE-EN ISO 4120:2022 procedure, considering a = 0.05 and the recommended randomization design (ABB, AAB, ABA, BAA, BBA, BAB) [33]. A panel of 25 assessors trained for discriminant tests participated in the study and conducted a series of triangle tests in different sessions: (1) R1-S mash sample vs. R2-I mash sample; (2) R1-S sauce vs. R2-I sauce; and (3) R1-S and R2-I sauces vs. R3 sauce. Cream and breadsticks were provided for palate cleansing between samples.
After observing that the prototypes were considered similar from a sensory standpoint, a Napping ® test [34] was conducted with the R3 sample to identify similitudes and differences between the developed sauce and similar products found in the local supermarkets (13 samples; shown in Table 1). A total of 21 professional chefs/gastronomes participated in the session to ensure familiarity with the product category. Approximately 10 mL of each sample were served in 20 mL disposable cups coded with 3-digit random numbers. Participants were asked to smell and taste the sauces and place them in an A3 paper (297 × 420 mm), close together or further apart, depending on their similarities/differences. In addition, assessors were instructed to include descriptors for each sauce/group of sauces. Cream and breadsticks were provided for palate cleansing between samples.

Data Analysis
Physicochemical and microbiological determinations were analyzed using one-way Analysis of Variance (ANOVA) using "recipe" as factor, followed by post-hoc Tukey's HSD test (α = 0.05). The results of the triangular test were analyzed as suggested by the UNE-EN ISO 4120:2008 procedure [33]; the significance was determined by considering the minimum number of correct responses. A multiple-factor analysis was carried out with the Napping ® test [35]. The results were analyzed using XLSTAT (Version, 2021.5, Addinsoft, Denver, CO, USA) [36]. Table 2 shows the initial pH of the original green chili pepper mash without fermentation, corresponding to sample R3 (5.2), and both R1-S and R2-I mash samples after 15 days of fermentation, which had a significantly lower pH value (p ≤ 0.05). The inoculated sample (R2-I) reached a lower pH (3.7) than the spontaneously fermented sample (4.8).

Mash Characterizations: Role of the Fermentation Process
Similar results have been reported by other authors; Di Cagno et al. [25] showed that spontaneously fermented red and yellow peppers (C. annuum L.) had higher pH values (pH above 4.8) than those inoculated with autochthonous bacteria (L. plantarum PE21, L. curvatus PE4, and W. confuse PE36) (approximately 3.6). Aryee et al. [17] reported a decrease of pH from 5.24 to 4.87 after 14 days of fermentation in habanero peppers (C. chinese) fermented spontaneously in a brine with a 5% salt concentration. The pH of Tabasco pepper mash (C. fructescens grinded with 8% salt), changed from 4.7 to 3.9-3.7, depending on the material of the container (plastic and wood, respectively), after the first month of a spontaneous fermentation [37]. The results of the present study suggested that using a low salt concentration, together with the inoculation of a commercial L. plantarum strain, favored the pH decrease. These results support the use of microorganisms well-adapted to the medium to promote the fermentation process. Reaching low pH values is recommended to ensure food safety while avoiding refrigeration of the product, because shelf stable hot sauces must have a pH below 4.6 [38]. Significant differences were found in total aerobic counts (TAC) and total anaerobic counts (TanC) among the samples (Table 2). TAC significantly decreased during fermentation in both the R1-S and R2-I samples, probably due to the lack of a proper environment for aerobic microorganism development. TanC increased in both samples during fermentation, starting from 4.6 log CFU ml −1 and reaching a significantly different amount in R2-I T15 (8.8 log CFU mL −1 ), the sample inoculated with L. plantarum. No significant differences were found between the pepper mashes T0 and R1-S at T15, although a slight increase was observed after 15 days of fermentation. Di Cagno et al. [25] showed an increment in lactic acid bacteria counts (L. plantarum, L. curvatus, and W. confusa) in peppers fermented at 35 • C during 15 days in a brine (1% NaCl), passing from approximately 4 log CFU/g to 9 log CFU/g; these authors reported higher increments in the samples in which starters were used. An increase in LAB count was also reported by Janiszewska-Turak et al. [27], who studied the differences between a spontaneous fermentation of green peppers (C. anuumm L.) vs. a fermentation driven by an inoculated starter of Levilactobacillus brevis, Limoslactobacillus fermentum, and L. plantarum. Fermentation time was different between the present study and the one reported by Janiszewska-Turak et al. [27] (15 days and 7 days, respectively), which may have led to a different result than the one obtained in the present study.
All color parameters varied significantly (p ≤ 0.05) during the fermentation process (Table 3). Lightness (L*), the blue-yellow coordinate (b*), chroma, and hue increments were significantly higher in the R2-I mash sample. On the contrary, the red-green coordinate (a*) and the total color transformation, represented by ∆E*, were higher in the mash sample of the spontaneous fermentation (R1-S). The fermented samples' colors were significantly different from the original mash, ∆E* being over 5 in both the R1-S (T15) and R2-I (T15) samples; ∆E* has been reported to show differences between treated and untreated samples when being over 0.5 in citrus products [31,39]. Janiszewska-Turak et al. [27] reported that, in general, the L*, a*, and b* coordinates increased during the fermentation of green peppers driven by different microorganisms, with a higher increase of L* in the samples fermented by L. plantarum, and a higher increase of a* in the spontaneously fermented samples. The volatile compositions of the mash samples are shown in Table 4. In general, the volatile compositions of the mash samples were mainly represented by compounds found in garlic such as diallyl disulfide or allyl methyl trisulfide [40,41], and other compounds previously found in chili peppers (e.g., 2-Isobutyl-3-methoxypyrazine, several aldehydes and alcohols [42]). Significant differences were found among samples, suggesting that the fermentation process influenced the aromatic profile of the samples.
To assess if the differences found in the physicochemical compositions of the mashes were perceived by the human senses, triangle tests were conducted with a trained panel of 25 assessors. Results showed that no significant differences were perceived between fermented mashes (correct answers = 10; at least 13 correct answers were needed to consider samples to be significantly different considering a = 0.05), suggesting that the different processes led to products that could be considered different from a sensory standpoint. The pungency of the mash samples could have influenced the perception of the differences identified in their volatile compositions or sugars profiles. Therefore, because other ingredients would be later incorporated to finish the product, and these could have a significant effect on their perception, additional triangle tests were conducted to determine if the three samples of sauce prototypes were perceived as different. No significant differences were detected among sauce samples, suggesting that, in general, the ingredients had a higher effect on the flavor of the sauces than the production process. The results of the sensory tests leads to the conclusion that the fermentation stage, although typically conducted in other products of the same category [18,19], was unnecessary to develop the proposed new product of the present research, probably because the contribution of the fermentation to the organoleptic properties of the sauce was limited in this product. The use of different sensory tests, such as discriminant tests, combined with physicochemical data has proven to be useful in the decision-making process of different product developments such as ice cream recipe reformulations [44], testing the application of degreening treatments in citrus [45], or estimating the shelf lives of fish broth products [46]. The present study supports the need to use at least simple sensory tests, as well as physicochemical data, in the design process of a product. Tables 5 and 6 show the physicochemical characteristics and volatile composition of the final sauce sample. The mixture of the mashes with the additional ingredients significantly modified their characteristics, resulting in what could be considered a kind of product with homogeneous sensory characteristics. The pH of the final product was significantly lower (p < 0.05) than the pH of the mashes, probably due to the addition of lemon juice, comprising 8%. The pH reached after adding this ingredient (pH = 3.3) ensured the absence of pathogen bacteria growth because it is under 4 [47], and supported the results of the sensory tests, confirming that the fermentation stage was not useful to differentiate the flavor of the product, nor was it necessary from a technical standpoint.  Sugars and color parameters were also heavily influenced by the additional ingredients. Although green chili peppers were the main ingredient of the sauce, the presence of other ingredients such as lemon juice and red fruits completely modified the sugar content and color of the final product, having higher a* coordinate values (redder than the mash samples) and lower b* coordinate values (bluer than the mash samples).

Final Product Characterization
Finally, the volatile composition of the sauce was significantly affected by the presence of the additional ingredients (Table 6). Approximately 18 volatile compounds of the mash samples were also present in the sauce samples, but limonene and eugenol represented over 50% of the total area of the volatile profile of the sauce samples, suggesting the importance of the additional ingredients (lemon juice and 4 épices) in the aromatic profile of the product. The final sauce had over 45 different volatile compounds from different chemical groups (terpenes, benzene derivatives, sulfide derivatives, etc.) showing the complexity of the product and confirming the difficulty of perceiving the flavors coming from the potential fermentation stage.
With the aim of mapping the new product in the food market, and understanding the similarities and differences among competitors, a Napping ® test was also developed by food experts (chefs and gastronomes) who could aid in positioning and describing the new product. Figure 2 shows the symmetric plot of all the sauces listed in Table 1, explaining the 39% variability of the samples.  Table 1. Note: cubes refer to sauce samples and dots to the descriptors for each sauce/group of sauces.
The sample developed during the present research was described as "sour", "spicy", "smoky", or "balanced", features shared with commercial sauces such as Valentina (664) and Tabasco chipotle (383), and that could be used to communicate its profile and encourage the selection of the developed product by consumers who prefer these attributes over other ones such as "sweet", "dense", or "bitter", which were used to describe other products such as Spanish sriracha (765), Gochujang (332) and Soya-based chili sauce (587). Different studies have shown that the intensive consumption of spicy foods could favor the ability to discriminate flavors in this type of product [12,48]; therefore, it is possible that heavy consumers or sensory experts in the corresponding food category (hot sauces) could detect subtle differences among the products. Further studies should be conducted to determine if consumers from different food cultures (e.g., gastronomies with high vs. low use of spicy ingredients) have different discriminatory abilities for these hot sauces.

Conclusions
Results of the present research suggest that the fermentation stage of the production process of some hot chili pepper sauces could be avoided depending on the additional ingredients used to finish the product. Physicochemical characteristics should always be analyzed and considered for food product development, but sensory tests are also key to avoid investing time and resources when a new product is developed (e.g., imitating traditional methods). Considering the gastronomy culture for which the product was developed in the present study, the fermentation stage of the process could be avoided, reducing the number of unit operations and ingredients (starter), and optimizing the processing time (from 15 days to 1 day).  Table 1. Note: cubes refer to sauce samples and dots to the descriptors for each sauce/group of sauces.
The sample developed during the present research was described as "sour", "spicy", "smoky", or "balanced", features shared with commercial sauces such as Valentina (664) and Tabasco chipotle (383), and that could be used to communicate its profile and encourage the selection of the developed product by consumers who prefer these attributes over other ones such as "sweet", "dense", or "bitter", which were used to describe other products such as Spanish sriracha (765), Gochujang (332) and Soya-based chili sauce (587). Different studies have shown that the intensive consumption of spicy foods could favor the ability to discriminate flavors in this type of product [12,48]; therefore, it is possible that heavy consumers or sensory experts in the corresponding food category (hot sauces) could detect subtle differences among the products. Further studies should be conducted to determine if consumers from different food cultures (e.g., gastronomies with high vs. low use of spicy ingredients) have different discriminatory abilities for these hot sauces.

Conclusions
Results of the present research suggest that the fermentation stage of the production process of some hot chili pepper sauces could be avoided depending on the additional ingredients used to finish the product. Physicochemical characteristics should always be analyzed and considered for food product development, but sensory tests are also key to avoid investing time and resources when a new product is developed (e.g., imitating traditional methods). Considering the gastronomy culture for which the product was developed in the present study, the fermentation stage of the process could be avoided, reducing the number of unit operations and ingredients (starter), and optimizing the processing time (from 15 days to 1 day).