Raw material

Tubers of three types of sweet potato (SP), Ipomoea batatas L., were used in this study based on their color: purple, yellow, and white (Figure 1). The SP are indigenous to the Mexican state of Michoacán and were procured from a market in the city of Morelia. For analysis, they were washed with distilled water and sliced into approximately 5-cm thick pieces, including the peel, and dehydrated at 50°C for 48 h in an oven (model 9023A, Ecoshel, Pharr, TX, USA). Subsequently, the samples were ground and passed through a 60-mesh sieve, resulting in a flour with a particle size of less than 260 μm. The dry matter (DM) content of the samples was calculated by the difference in weight before and after drying.

Figure 1

Appearance of fresh purple (left), yellow (center) and white (right) sweet potatoes.

Preparation of extrudates

The fresh SP were blanched for 10 min at boiling temperature, the peel was removed, and they were cut into 5-cm thick pieces. Subsequently, they were dehydrated in a gas oven at 55±5°C for up to 2 h at separate intervals to obtain varying levels of humidity. The final humidity levels achieved were 15–20% in 30 min, 10–15% in 60 min, and 5–10% in 120 min. The extruder equipment was assembled in a laboratory within the Universidad Michoacana de San Nicolás de Hidalgo, México. The screw-type extruder used had a helical worm of circular panel with a 1-cm diameter, and a blade and a die made of stainless steel which allowed to produce 1-cm diameter filament. It was powered by a 1 HP Marver model 10012 series 1009 motor (Crompton, Mumbai, India). The extrusion of SP was conducted without any additives. After extrusion, the products were placed in a gas oven set to 90±5°C for 1 h to remove excess moisture and achieve the desired texture.

Analysis of proximate composition, dietary fiber content and mineral contents

The proximate composition analysis of potatoes and extrudates was conducted according to AOAC International methods [AOAC, 2000]; moisture (method no. 934.01), protein (method no. 960.52), lipids (method no. 920.85), ash (method no. 942.05), and crude fiber (962.09) were determined. The content of dietary fiber (total, soluble and insoluble fractions) was assessed using the method of Prosky et al. [1988], with a total dietary fiber assay kit (Sigma-Aldrich, Saint Louis, MO, USA). The analyses were conducted in triplicate. The content of carbohydrates was calculated by difference from the data on moisture, protein, lipid, ash, and dietary fiber. The roots’ metabolizable energy was estimated using conversion factors of 4 kcal/g for proteins and carbohydrates, 9 kcal/g for lipids, and a value of 2 kcal/g for dietary fiber, as indicated in the data provided by the FAO [2003].

The determination of SP starch was conducted in accordance with the stipulations of the Mexican standard NOM-F321-S1978 [1978], which represents a modified iteration of the Lane-Eynon procedure. A quantity of 5 g of the sweet potato powder was placed in a container with 150 mL of water and 25 mL of concentrated HCl, and the contents were mixed thoroughly. The solution was heated for 75 min, cooled and neutralized with NaOH/H₂O (1:1, w/v) solution. Subsequently, the hydrolysate, after filling the volume to 250 mL was filtered. Then, the 10-fold diluted filtrate was transferred to a burette and was added drop-wise to the flask containing a mixture of CuSO₄×5H₂O solution and NaKC₄H₄O₆×4H₂O solution. Just prior to the complete reduction of the copper, 1 mL of the 0.2% aqueous solution of methylene blue was added. The titration was completed until the indicator lost its coloration. Starch content in the sample (g/100 g DM) was calculated utilizing a glucose-to-starch conversion factor of 0.9.

Certain minerals including calcium, copper, iron, potassium, magnesium, sodium, and zinc of SP were quantified using atomic absorption spectrometry. Digestion was performed using nitric acid, according to the procedure described by Uddin et al. [2016]. A quantity of 0.5 g of the sample was weighed, and 5 mL of 65% HNO₃ was added. The resulting mixture was then boiled gently over a water bath maintained at 90°C for a period of 1–2 h until a clear, fully digested solution was obtained. The solution was then filtered using Whatman 42 filter paper (2.5 μm). Enough deionized water was then added to the final sample’s volume to 50 mL. The analysis was conducted in triplicate using an atomic absorption spectrometer model AAnalyst 200 (Perkin Elmer, Waltham, MA, USA). The results were reported as mg/100 g DM.

Determination of ascorbic acid and total xanthophyll contents

Ascorbic acid in potatoes was quantified using the titrimetric method with 2,6-dichloroindophenol according to the AOAC International procedure (967.21) [AOAC, 2000] and its content was expressed as mg/100 g DM.

The total xanthophyll content was determined in accordance with the AOAC International method no. 970.64 [AOAC, 2000]. A sample weighing 50 mg was mixed with 3 mL of a solution composed of hexane, ethanol, acetone and toluene (10:6:7:7, v/v/v/v), and then 2 mL of 40% KOH in 80% methanol was added. The mixture was then incubated at 20°C for 16 h, brought to a volume of 10 mL with 10% NaSO₄ and shaken for 2 min in a vortex mixer. It was then allowed to stand for 1 h until the epiphase had clarified in the dark. Subsequently, the solution was transferred to a centrifuge tube and centrifuged at 3,000×g for 5 min. The absorbance was determined at a wavelength of 474 nm using a spectrophotometer, with hexane used as the blank. The total xanthophyll content was calculated using extinction coefficient for trans-lutein (236 L/(g×cm)) and expressed in μg/g potato sample.

Determination of total phenolic content, total flavonoid content, and antioxidant capacity

An extraction of each of the three SPs and three extrudates was conducted in order to determine the content of total phenolics, total flavonoids (only extrudates), and antioxidant capacity via assays with 2,2-diphenyl-1-picrylhydrazyl (DPPH) radicals and with 2,2’-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical cations. The powdered samples were treated as follows: 10 mL of ethanol (75%) was added to 1 g of each sample and stirred for 24 h in the dark. The samples were then centrifuged at 3,910×g for 10 min, after which the supernatants were collected and kept at 4°C until used for analyses.

The total phenolic content was determined using the colorimetric method with the Folin-Ciocalteau reagent, following the procedure established by Singleton et al. [1999] with minor modifications suggested by Treviño-Gómez et al. [2017]. A 250 μL aliquot of the extract was combined with 250 μL of 2 N Folin-Ciocalteu reagent and 250 μL of 20% Na₂CO₃, agitated, and incubated at 40°C for 30 min. Subsequently, 2 mL of distilled water were added to the mixture, which was then vortexed. The samples were subjected to absorbance measurement at 750 nm using a spectrophotometer (Genesys 20, Thermo Fisher Scientific Inc.). Gallic acid was used as a standard and the results were reported as mg gallic acid equivalent (GAE)/100 g DM.

The total flavonoid content of the extrudates was evaluated in accordance with the methodology described by Zhishen et al. [1999]. A 150 μL aliquot of each extract was combined with 150 μL of 5% NaNO₂, 150 μL of 10% AlCl₃, and 1 mL of 0.1 M NaOH. Subsequently, the samples were analyzed using a spectrophotometer (Genesys 20, Thermo Fisher Scientific Inc.) at an absorbance of 510 nm. The standard used was quercetin, and the results were expressed as mg quercetin equivalent (QE)/100 g DM.

The determination of DPPH radical scavenging activity was conducted according to the method developed by Brand-Williams et al. [1995] with modifications proposed by Ru et al. [2019]. In summary, 200 μL of each extract was added to 3 mL of the DPPH radical solution in methanol (100 μM). The samples were homogenized for 10 s, then the reaction mixture was incubated in the dark at room temperature for 30 min, after which its absorbance was measured at 517 nm using a spectrophotometer (Genesys 20, Thermo Fisher Scientific Inc.). The results were expressed as μmol Trolox equivalent (TE)/g DM using a standard curve of Trolox.

The ABTS assay was conducted in accordance with the methodology outlined by Re et al. [1999]. Firstly, a solution consisting of ABTS (7 mM) and potassium persulfate (2.45 mM) was mixed at room temperature in the dark for 20 h to generate ABTS^•+. Then, methanol was used to dilute the ABTS^•+ mixture to an absorbance around 0.70 at 734 nm. Subsequently, 0.1 mL of each extract was combined with 3.9 mL of the aforementioned ABTS^•+ solution. The reaction mixture was kept at room temperature for 6 min, then the absorbance was recorded at 734 nm. The results were calculated based on the calibration curve plotted with Trolox and were expressed as μmol Trolox equivalent (TE)/g DM.

Texture analysis

A texture analysis by compression was conducted on the three extrudates using a Brookfield CT3 texture analyzer with TexturePro CT v1.7 Build 28 software (Brookfield Engineering Lab. Inc, Middleboro, MA, USA). The apparatus was operated with a TA11/1000 probe TA-SB element. The analysis was conducted on a single cycle with the target at 6.0 mm and an activation load of 0.07 N at a speed of 0.5 mm/s and 0.5 mm/s lap speed with a sampling rate of 50 points per s. The sample was 16.0 mm in length and 6.0 mm in height. The analysis was performed at room temperature (25±2°C) with 10 samples per extrudate.

Sensory analysis

Forty untrained participants were used to perform a sensory analysis of the extruded products using a hedonic test. The acceptability of the extrudates was determined using a 5-point hedonic scale, where 5 corresponded to “I like very much”; 4 to “I like”; 3 to “I neither like nor dislike”; 2 to “I dislike”; and 1 to “I dislike very much”. Color, odor, texture, and taste of the samples were evaluated.

Color parameter analysis

The color was evaluated in fresh, blanched, and extruded sweet potatoes from the three varieties. A Spectro-guide 45/0 colorimeter was employed for this purpose (BYK-Gardner GmbH, Geretsried, Germany). Values were obtained based on the CIELab scale with a brightness of D65, viewing angle of 10° and aperture diameter of 8 mm. The instrument provided darkness/brightness (L^*), green/red intensity (a^*), and blue/yellow intensity (b^*). Three replicates of measurements were conducted for each sample, and the instrument was calibrated with a white ceramic plate standard prior to each procedure.

Statistical analysis

Analyses of nutritional composition, mineral content, bioactive compound content, antioxidant capacity, and color parameters were performed in triplicate. For texture parameters and sensory scores, the number of repetitions were 10 and 40, respectively. Results were subjected to a one-way analysis of variance (ANOVA) using JMP software version 8.0 (JMP Statistical Discovery LLC, Cary, NC, USA). Means were compared using Tukey’s test, and differences were considered significant when p<0.05.

Macronutrients of sweet potatoes

The results of proximate analysis of SP are presented in Table 1. Moisture content was 60.80 g/100 g in purple, 72.44 g/100 g in yellow, and 68.54 g/100 g in white potatoes, with an average of 67.26 g/100 g. The moisture content of yellow and white varieties did not differ significantly (p≥0.05).

The three varieties under examination had an average dry matter content of 32.74 g/100 g. Grüneberg et al. [2017] conducted a comprehensive study of SP with 1,174 clones analyzed in various environments and determined an average dry matter content of 34.9 g/100 g. They also noted that genotypic factors had a greater impact on variability than environmental factors. Rowena et al. [2009] reported a dry matter content of 32.46 g/100 g after analyzing five varieties from the Philippines, while Teow et al. [2007] found a value of 30.2 g/100 g after studying nineteen clones from the USA.

Carbohydrates were found to be the most abundant nutrient of SP (Table 1), with an average content of 77.92 g/100 g DM. The highest content was found in the yellow variety, while the lowest content was in the purple variety. These findings were statistically significant (p<0.05). Additionally, the white variety did not differ statistically (p≥0.05) from the yellow and purple varieties. Armijos et al. [2020] reported an average of 81.66 g/100 g DM in carbohydrate content for seven varieties native to Ecuador with yellow, orange, and purple colors. Nascimento et al. [2015] studied three Brazilian SP varieties (white and orange) from organic cultivation and reported that their carbohydrate content was 57.21 g/100 g DM, being higher in the white variety, which is consistent with the results of our study. The content of starch, the main carbohydrate fraction of SP, ranged from 43.87 to 59.67 g/100 g DM (Table 1), and a significantly (p<0.05) higher value was noted for the purple variety compared to the other varieties. Armijos et al. [2020] reported an average of 60.56 g/100 g DM in starch content of SP with yellow, orange, and purple colors. Grüneberg et al. [2017] found that starch content of different clones of I. batatas was 66.0 g/100 g DM, while the content of sucrose, glucose, and fructose in DM was 10.3 g/100 g, 2.2 g/100 g, and 1.7 g/100 g, respectively. Nascimento et al. [2015] reported a starch content of 29.3 g/100 g DM in Brazilian SP. In turn, Hossain et al. [2022] who analyzed eight SP varieties from Bangladesh, reported an average starch content of 52.72 g/100 g in powdered dry roots. It is also noteworthy that the purple varieties exhibit the highest starch content, which is comparable to that determined in the present study.

The lipid content of SP was minimal with an average of 0.53/100 g DM, and the yellow variety had a significantly (p<0.05) higher lipid content (1.20 g/100 g DM) than the other varieties (Table 1). Armijos et al. [2020] and Hossain et al. [2022] reported lipid contents of 1.35 and 1.51 g/100 g DM, respectively. The latter author noted that the highest content of lipids was found in the yellow variety, a finding that aligns with the results of our research.

The average protein content was 8.25 g/100 g DM, with the white variety having the highest content (9.73 g/100 g) and differing significantly (p<0.05) in this respect from the other varieties (Table 1). Armijos et al. [2020] reported a protein content of 6.50 g/100 g DM, while Grüneberg et al. [2017] reported a range from 2.7 to 8.9 g/100 g DM for different clones of I. batatas. Leonel et al. [2023] analyzed data from multiple authors and reported values between 1.3 and 9.5 g/100 g DM.

The dietary fiber content was observed to be an average of 10.51 g/100 g DM, with 9.09 g/100 g DM being insoluble fiber and 1.42 g/100 g DM being soluble fiber. The white variety had the highest amount of soluble fiber, while the purple one had the highest amount of insoluble fiber (Table 1). Armijos et al. [2020] reported a soluble fiber content of 2.82 g/100 g DM and an insoluble fiber content of 10.17 g/100 g DM in differently colored Ecuadorian SP. Additionally, Leonel et al. [2023] reported a dietary fiber content of 19.36 g/100 g DM.

The average metabolized energy value of the three types of sweet potato was 371 kcal/100 g DM, according to theoretical data (Table 1). However, there were statistically significant (p<0.05) differences between the three varieties, being higher in the yellow variety and lower in the purple one. Vergun et al. [2020] found that the energy values of SP from Ukraine, when determined using calorimetric equipment, ranged from 342 to 362 kcal/100 g

To sum up, the macronutrient composition of sweet potatoes varied depending on their variety. Yellow potatoes had a higher content of carbohydrates and lipids, purple ones – dietary fiber, and white ones – proteins and ash.

Micronutrients of sweet potatoes

The mineral analysis results displaying the content of calcium, iron, potassium, magnesium, sodium, and zinc in dry matter of sweet potatoes are presented in Table 2. The copper was also analyzed but not detected. The level of calcium ranged from 44.4 to 139.8 mg/100 g DM, with the yellow variety having almost twice as much of this element as the other varieties. Ourresults were lower than those obtained by Armijos et al. [2020], who reported an average calcium content of 229 mg/100 g DM in Ecuadorian potatoes. Musilová et al. [2017] also determined a higher calcium content in the pulps of SP from the Slovak Republic (average of 468 mg/100 g DM). The average potassium content in Mexican potatoes in our study was found to be 2,778 mg/100 g DM, which was lower than the values reported by Armijos et al. [2020] (average of 3,501 mg/100 g DM) and higher than those reported by Musilová et al. [2017] (1,204 mg/100 g DM in pulp). The sodium levels varied significantly (p<0.05) among the three samples in the range of 54.0–106.9 mg/100 g DM, (Table 2). The average value found in this study was higher than those reported by other authors in the field; with Armijos et al. [2020] reporting 75.7 mg/100 g DM, while Musilová et al. [2017] reporting 63.4 mg/100 g DM in flesh.

Iron was determined in this study within a range from 10.5 to 12.8 mg/100 g DM (Table 2). Armijos et al. [2020] and Grüneberg et al. [2017] reported the iron content of 13.7 and 15.6 mg/100 g DM, respectively. Magnesium was recorded within a range from 75.3 to 117.4 mg/100 g DM with an average of 98.9 mg/100 g DM (Table 2). Lower magnesium levels were found in yellow, orange, and purple varieties from Ecuador with an average of 50.0 mg/100 g DM [Armijos et al., 2020] and in pulps of Slovak sweet potatoes with an average of 75.2 mg/100 g DM [Musilová et al., 2017]. Zinc was not detected in the white variety and its content in the yellow and purple SPs was 0.40 and 1.17 mg/100 g DM, respectively. Armijos et al. [2020] reported an average zinc content of 1.57 mg/100 g DM; Grüneberg et al. [2017] reported 0.93 mg/100 g DM, and Musilová et al. [2017] determined it at 0.92 mg/100 g DM.

This study found that, among the analyzed SP varieties, the yellow variety was the rich source of calcium, iron, and magnesium. On the other hand, the purple variety had the highest content of sodium and zinc but a lower potassium content.

Table 2 also shows the percentages of the dietary reference intakes (DRI) of the analyzed minerals by consuming 100 g of fresh sweet potato, which were calculated based on the recommendations of the National Institutes of Health (NIH) in USA, [NIH, 2022] (the average content of each mineral for the three varieties was used). The reference values are for people aged 19 to 50 years. It is noted that sweet potatoes were a micronutrient-rich food, providing high levels of potassium and iron, as well as magnesium. In addition, the sodium content was very low, providing only 1.8% of the recommended daily intake.

Contents of ascorbic acid, total xanthophylls and total phenolics in sweet potatoes

The contents of ascorbic acid, total xanthophylls, and total phenolics in different varieties of SP are shown in Table 3. The ascorbic acid content ranged from 60.6 to 106.0 mg/100 g DM. The yellow variety had the highest content, being 1.7 times higher than that of the other varieties. According to literature data, ascorbic acid content of Ecuadorian yellow, orange, and purple varieties of sweet potatoes was on average 67.2 mg/100 g DM [Armijos et al., 2020], while Ukrainian SP contained approximately 41.5 mg of this compound in 100 g of pulp [Vergun et al., 2020]. Vidal et al. [2018] compiling information from various authors, reported an ascorbic acid content ranging from 2.4 to 25.0 mg/100 g wet matter (WM), and Gichuhi et al. [2014] reported a value ranging from 6.4 to 15.5 mg/100 g WM for organic sweet potatoes from the USA. Considering the average content of ascorbic acid in the samples in our study and DRI of vitamin C recommended by NIH regulations [NIH, 2022], women and men over 19 years would receive 32.8% and 27.3% of DRI, respectively, by consuming 100 g of fresh sweet potato.

Total xanthophyll content of three SP varieties ranged from 175 to 395 μg/g DM with an average of 252 (Table 3). The yellow variety showed its highest quantity, followed by white and purple varieties. Escobar-Puentes et al. [2022] analyzing studies of several authors, reported that the carotenoid content of sweet potatoes was 509 μg/g DM, with the highest content found in orange, followed by yellow, white, and purple potatoes. The authors also noted that these contents could be higher than those found in carrots and mangoes, which are typically recognized as reliable sources of carotenoids.

The average total phenolic content of SP was 395 GAE mg/100 g DM with a range from 216 to 581 mg GAE/100 g DM. Sweet potatoes cultivated in the Philippines had an average total phenolic content of 567 mg GAE/100 g DM, with a higher content found in the purple varieties, followed by yellow and white ones [Rowena et al., 2009]. Escobar-Puentes et al. [2022] in a review publication reported values ranging from 140 to 1,230 mg GAE/100 g DM, with the highest noted for purple potatoes, followed by orange, yellow, and white ones. The findings of this study indicated that the purple variety exhibited the highest total phenolic content, followed by the yellow and then the white variety, which was consistent with the above-mentioned literature data. Musilová et al. [2017] reported that the content of total phenolics in the peel of SP was higher than in the pulp. On the other hand, Murnihati et al. [2020] found that the main antioxidant compounds in purple sweet potatoes were anthocyanins and other polyphenols, while orange and yellow sweet potatoes mainly contained carotenoids. Our research findings are consistent with these conclusions.

Antioxidant capacity of sweet potatoes

The results of the antioxidant capacity of potatoes measured as the antiradical activity against DPPH^• and ABTS^•+ and reported as μmol TE/g DM are shown in Table 3. The DPPH assay results showed homogeneity of values with no statistical difference (p≥0.05) between differently colored potatoes. However, the ABTS assay indicated that the purple potatoes had significantly (p<0.05) higher antioxidant capacity compared to the other samples.

The average DPPH^• radical scavenging activity of Mexican sweet potatoes was 40.7 μmol TE/g. This value was consistent with that reported by Zhang et al. [2022] for Chinese pigmented sweet potatoes (average of 39.3 μmol TE/g) and was slightly higher compared to that noted by Tang et al. [2015] (average of 25.1 μmol TE/g). In the second mentioned study, unlike in our study, higher DPPH^• radical scavenging activity was observed in the purple and orange sweet potatoes than in the white and yellow ones. Makori et al. [2020] found that antiradical activity against DPPH^• was dependent on the part of the potato tuber and reported higher values for the skin than for the flesh.

The mean value in the ABTS assay obtained in our study was 23.4 μmol TE/g. Armijos et al. [2020] and Zhang et al. [2022] reported the averages of 33.6 and 28.2 μmol TE/g for Ecuadorian and Chinese sweet potato varieties, respectively.

Proximate composition of sweet potato extrudates

In preliminary studies on obtaining extruded sweet potatoes, the fresh potatoes were dried at 50–60°C for various times to achieve different moisture contents of 5–10% (120 min), 10–15% (60 min), and 15–20% (30 min), and subsequently extruded. The dried products with a 5–10% moisture content had good color and flavor; however, they became excessively stiff over time. On the other hand, the extruded products with a moisture content of 10–15% exhibited favorable organoleptic properties and a crunchy consistency. Furthermore, these products demonstrated a seamless and uniform flow within the equipment. In contrast, the extruded products with a moisture content of 15–20% exhibited a pasty consistency and an indeterminate shape, which resulted in equipment blockages. Based on these findings, SP with a humidity range of 10–15% were selected for further production. According to the preliminary sensory evaluation, the products had a sweet taste, pleasing color and odor, high hardness, medium breaking, and high dryness, without being adhesive to the palate.

The results of a proximate analysis conducted on extrudates of each sweet potato variety are shown in Table 4. The moisture contents of the extruded purple, yellow, and white potatoes were 6.41 g/100 g, 6.86 g/100 g, and 5.97 g/100 g, respectively, and no significant (p≥0.05) differences were found between them. The extrudates had a protein content with an average of 7.00 g/100 g DM, with no significant (p≥0.05) difference between the three products. Extruded potatoes were also rich in carbohydrates (88.43–89.76 g/100 g DM), while the constituent which was devoid of any significant content was lipid (0.13–0.50 g/100 g DM). The average energy value was 389 kcal/100 g DM, theoretical data, being significantly (p<0.05) higher for the extruded yellow potatoes compared to the other varieties. The macronutrient content and energy value of the extrudates were consistent with the literature data, e.g., the carbohydrate, protein and lipid contents of extruded purple sweet potatoes reported by Palupi et al. [2024] were 85.57, 5.51 and 0.56 g/100 g DM, respectively, and provided 369.28 kcal/100 g.

Total phenolic content, total flavonoid content, and antioxidant capacity of sweet potato extrudates

The contents of total phenolics and total flavonoids, as well as the antioxidant capacity of extruded Mexican sweet potatoes are shown in Table 3. A significant difference was observed in the total flavonoid content among the extrudates obtained from the three sweet potato varieties, with the product from purple variety exhibiting the highest total flavonoid content and the extruded potatoes of white variety exhibiting the lowest. Simultaneously, the purple extrudate exhibited the highest total phenolic content, followed by yellow, and finally white one, with a notable distinction (p<0.05). Regarding the results of the DPPH assay, the yellow extrudate exhibited the highest value, demonstrating a statistically significant difference (p<0.05) compared to the other varieties; conversely, the purple and white extrudates did not differ statistically significantly (p≥0.05) from one another. The extruded potatoes of the purple variety exhibited the highest values in the ABTS assay, while no statistically significant difference (p≥0.05) was observed between the yellow and white varieties.

A decrease was observed in all extruded SP with respect to fresh SP when total phenolic content was subjected to analysis. A similar trend was observed in antioxidant activity, as determined by both the DPPH and ABTS assays. The blanching and extrusion process has a direct impact on the content of bioactive compounds and antioxidant capacity. Tang et al. [2015] reported that thermal treatment can reduce the quality of SP due to chemical degradation and the reduction of bioactive compounds such as phenolics and carotenoids. Boiling is a better heating method for keeping carotenoids, while steaming preserves more phenolic substances (excluding anthocyanin), and roasting keeps more anthocyanins.

Color parameters of fresh, blanched and extruded sweet potatoes

The color of fresh, blanched, and extruded products from the three types of sweet potatoes were evaluated, and the values of parameters L^*, a^* and b^* according to the CIELab scale are shown in Table 6. Figure 1 depicts the images of fresh sweet potatoes. At the L^* parameter level, the purple variety exhibited a value of 29.13, while the yellow variety demonstrated a value of 71.38 and the white variety exhibited a value of 84.58. Armijos et al. [2020] reported L^* values ranging from 13.89 to 71.61 for a variety of pigmented sweet potatoes. In turn, Tang et al. [2015] reported a value of 49.25 for the purple variety, 85.53 for the orange, 87.39 for the yellow, and 86.74 for the white one. The findings align with the values recorded in our research. The parameter a^*, in which positive values correlate with red pigments and negative values with green pigments, showed 26.38 for the purple sample, 28.78 for the yellow sample, and 1.58 for the white one (Table 6). Armijos et al. [2020] referred to values from 13.07 to 34.30, which were reported for different Ecuadorian varieties. Conversely, Tang et al. [2015] reported 23.83 for purple, 9.05 for orange, 5.72 for yellow, and 0.08 for white potatoes; with the data for the purple and white varieties aligning with our findings, whereas the orange and yellow varieties exhibiting lower values, suggesting that our varieties possess a greater degree of red pigmentation than those reported by the aforementioned author. Regarding the b^* parameter, where positive values are associated with yellow tones and negative values with blue, a significant difference was observed among the three sweet potato varieties (Table 6). The values of 37.13 and 18.20 were measured for yellow and white varieties, respectively. The color of the purple variety was characterized by a negative value of a^* parameter, −5.82, indicating a tendency towards blue. Armijos et al. [2020] referred to values from −8.89 to 57.81, whereas Tang et al. [2015] reported −9.56 for purple, 23.21 for orange, 25.83 for yellow, and 15.41 for white potatoes. These values are in accordance with those measured in this study.

It is noteworthy that the purple pigmentation of sweet potatoes is attributed to anthocyanins, which impart a dark hue and correspond with the lowest value of the L^* parameter [Tang et al., 2015]. The presence of anthocyanins in purple sweet potatoes could also be the reason for differences in the L^* value among the three sweet potato varieties included in our study. Additionally, purple pigmentation coincided with the highest content of phenolic compounds (Table 3). In contrast, the highest value of the L^* parameter (whiteness) was observed in the white sweet potato. This may be attributed to the absence of colored pigments in the sample, which exhibited the lowest total phenolic content. Similarly, the color of the yellow sweet potato was characterized by the highest b^* value and the highest total xanthophyll content.

A color analysis was also conducted on blanched and extruded sweet potatoes, noting that blanching represents a preliminary step preceding the preparation of the extrudates. Results are shown in Table 6. In the purple sweet potatoes, it was observed that the blanching process caused a lightening of the color of fresh sweet potatoes, significantly (p<0.05) increasing the L^* value, and this phenomenon was more pronounced in the extruded samples. Regarding parameter a^*, which concerns green/red, the fresh-blanched-extruded sample displayed a significant decrease (p<0.05) in purple color in accordance with the prevailing tendency to transition from red to green. With respect to parameter b^*, which refers to blue/yellow, no perceptible differences were observed between the blanched and extruded samples (p≥0.05). However, these samples differed from the fresh sample, exhibiting a tendency towards yellow and a loss of purple color.

In yellow products, the L^* values demonstrated a proclivity towards darkening, fresh sweet potatoes exhibited a higher L^* value than blanched ones, with this effect being further accentuated in extruded sweet potatoes. On the a^* parameter, no distinction was observed between the fresh and blanched samples; however, the extruded samples exhibited a loss of red coloration. Conversely, in the case of the b^* parameter, the blanched product exhibited a greater yellow than the fresh sample, whereas the extruded product displayed a diminished yellow pigmentation in comparison to the fresh and blanched sweet potatoes.

In the white sweet potatoes, it was observed that the brightness tended to decrease with blanching, with a more pronounced decrease evident in the extruded sample. Furthermore, both samples exhibited statistically significant differences (p<0.05) with respect to fresh sweet potatoes. The a^* parameter revealed that blanching caused the red color to shift towards green in comparison to fresh sweet potatoes, but extrusion resulted in the opposite effect, with extruded sweet potatoes exhibiting a reddish hue. Additionally, on the b^* parameter, a tendency from yellow to blue was observed, with a more pronounced one in the extruded product. There was no statistically significant difference (p≥0.05) between the fresh and blanched SP samples, but a notable discrepancy was observed when comparing them to the extruded products.