2.1 Sweet Potato Peel Flour (SPPF) elaboration

Purple sweet potatoes (70 kg) were used, showing a good visual appearance, smooth surface without imperfections and medium size. The whole sweet potatoes (Ipomoea batatas L. (Lam.)) were washed and immersed in a sodium hypochlorite solution, with a proportion of 8 ml for each liter of water. After 15 minutes, the tubers were rinsed again under running water. The peels (2 mm thick) were manually removed with the aid of a knife and dried in a dehydrator with forced air circulation (Pardal^®, PE 60, Brazil) at 60 °C for 24 hours. Peels were grounded in a mill (Tecnal^®, Tec mill TE-633, Brazil), yielding 1.4 kg of flour. The product was packed and stored at - 18 °C until analyzes were carried out.

2.2 Beef patties processing and cooking

Four formulations of hamburger were prepared, containing three independent replicates of each treatment: beef (shoulder clod) (F1: 77.9%, F2: 77.1%, F3: 76.4% and F4: 75.7%), SPPF (F1: 0%, F2: 0.75%, F3: 1.5% and F4: 2.25%), ice flakes (15%), pork fat (5%), sodium chloride (1.5%), onion powder (0.2%), garlic powder (0.2%) and black pepper. The percentages of each ingredient were defined by means of preliminary sensorial tests carried out with the product. To elaborate the hamburgers, the meat (approximately 10 kg) was ground in a meat grinder (C.A.F.^®, Brazil), on a 3 mm disk and with a temperature around 4 °C. Subsequently, the ground beef was then homogenized in a commercial blender (Super Cutter Sire^®, Brazil) for 1 minute at 9 ± 1 °C. The onion, garlic, pepper, sodium chloride, ice flake and pork fat were added to the mixture and homogenized again for 3 minutes at 9 ± 1 °C. SPPF was incorporated into the dough and homogenized for an additional 3 minutes at 9 ± 1 °C. Additional levels of ground beef and SPPF differed in each formulation as described above. The resulting dough of each formulation was shaped into hamburgers (100 g, 10 cm in diameter and 1 cm thick) using a hand-fed hamburger (Picelli^®, HP 128, Brazil). The products were stored in plastic bags of low-density polyethylene and frozen in a conventional freezer (- 18 °C) for 10 days.

The frozen hamburgers were grilled on an electric plate with grill on the upper and lower sides (Britania grill^® mega 2N, Brazil) heated to 200 °C. The internal temperature of the hamburger was controlled by a digital thermometer (Tp 101^®, Brazil) until reaching 71 °C at its geometric center (^{American Meat Science Association, 2015}). The average cooking time was 8 to 10 minutes.

2.3 Physicochemical composition

All analyses were performed on three replicates in triplicate for SPPF and for cooked hamburgers. Moisture, ash, protein, fat and dietary fiber content were determined by the ^{Association of Official Analytical Chemists (2011)}. The moisture content was determined by drying in a greenhouse (105 ± 2 °C). Fat content was determined according to the Soxhlet method, using petroleum ether. Protein was analyzed according to the Kjeldahl method. Factor 6.25 was used for the conversion of nitrogen to crude protein in hamburger and SPPF respectively. Ash was performed by a muffle furnace. Total, soluble and insoluble dietary fiber was determined by the enzymatic method. The carbohydrate content was evaluated by means of theoretical calculation (by difference) in the results of the triplicates, according to the Formula 1:

The total caloric value (kcal) was calculated theoretically using Atwater factors (^{Atwater & Woods, 1896}) for lipid (9 kcal g^-1), protein (4 kcal g^-1) and carbohydrate (4 kcal g^-1).

Water activity (Aw) was determined using the Aw analyzer (Novasina^®, Labswift model, Switzerland), at 20 °C (^{Association of Official Analytical Chemists, 2011}). pH was measured using a pH-meter (Tecnopon^®, MPA-210 model, Brazil) (^{Association of Official Analytical Chemists, 2011}). To establish the color, five hamburgers were used per treatment, evaluated in five different points of the hamburger. The color was evaluated by the system of the Commission Internationale de L'Eclairage (CIE), lightness (L*), redness (a*), yellowness (b*), colorimeter reading (Konica Minolta^®, Chroma Meter CR 4400 model, Japan) with illuminating calibration D65 and angle of observation 10º, previously calibrated.

2.4 Technological analyses

Five hamburgers from each formulation were cooked in the same procedure as mentioned previously then cooled to room temperature at 23 °C for 2 h. The following cooking characteristics were evaluated: cooking yield (2) and fat retention (3) (^{Murphy et al., 1975}), shrinkage (4) (^{Berry, 1992}) and moisture retention (5) (^{El-Magoli et al., 1996}). All experiments were done in triplicate. The hamburgers were measured according to the following Equations 2-5:

2.5 Sensorial analyses

Participated in sensory analyses 65 untrained volunteer subjects, hamburger usual consumers. Consumers had aged between 18 and 50 years and were recruited among students and staff of Universidade Estadual do Centro-Oeste, Guarapuava, Paraná, Brazil. For conducting the sensory test, hamburgers have been cooked as previously described. All samples were evaluated by means of an acceptance test using a nine-point hedonic scale, with extremes ranging from dislike extremely (1) to like extremely (9) (^{Meilgaard et al., 1999}). Attributes related to appearance, aroma, flavor, color and texture, beyond overall acceptance were evaluated. For the purchase intent test a 5-point attitude structured scale was used, varying from definitely would not buy it (1) to definitely would buy it (5) (^{Meilgaard et al., 1999}). The sensory Acceptability Index (AI) was calculated by multiplying the average score reported by consumers to the product by 100, dividing the result by the maximum average score given to the product within the hedonic scale of 9.0 points. Each sample was served to consumers in white plates coded with randomly selected 3-digit numbers in monadic form and using balanced design (^{Macfie et al., 1989}). Sensory evaluations were performed by consumers under fluorescence lighting. After consuming each sample, the consumer was instructed to drink water for palate cleansing. Samples were evaluated in triplicate in a separate session.

2.6 Statistical analysis

The results were analyzed using analysis of variance (ANOVA). The means were compared by Tukey’s test at 5% significance level (p ≤ 0.05). Software R was used to perform the statistical calculations.

2.7 Ethical issues

The study was approved by the Ethics in Research Committee of UNICENTRO, Brazil, under the case number of 608.950/2014.

3.1 Physicochemical composition

The physicochemical composition results of cooked hamburgers are presented in Table 1.

Sweet potato peel contains high carbohydrate and fiber content when compared to beef which is exempt in its composition (^{United States Department of Agriculture, 2014}). Studies have shown that adequate fiber consumption reduces the risk of developing pathologies such as cardiovascular disorders (^{Mirmiran et al., 2016}), systemic arterial hypertension (^{Evans et al., 2015}), diabetes mellitus (^{Wu et al., 2015}), among others.

Moisture, ash, carbohydrate and fiber contents increased with the addition of SPPF. The highest moisture content of the SPPF hamburger is explained by the water retention property of the fibers (^{Célino et al., 2014}), as previously reported. In addition, fibers interact with proteins of the meat, resulting in a network that prevents the translocation of water from the product to the surface (^{Song et al., 2016}). The higher ash, carbohydrate and fiber content in F2, F3 and F4 are due to the higher amount of these nutrients present in SPPF compared to meat. Similar results were verified after the addition of orange peel flour (5%) in bovine hamburger (^{Mahmoud et al., 2017}). Protein, lipid and caloric contents were lower for SPPF-added hamburgers since SPPF contains lower levels of these nutrients compared to meat. These results corroborate with other studies evaluating the addition of poppy seed (^{Gök et al., 2011}) and orange peel flour in bovine hamburger (^{Mahmoud et al., 2017}) reduces protein and lipid content. There was no significant difference (p> 0.05) between the pH and Aw results of the formulations, as reported in the literature (^{Longato et al., 2017}). The instrumental color results of cooked hamburgers are presented in Table 2.

The addition of SPPF in hamburger reduced (p < 0.05) the values ​​of L* and b*. The red content (a*) decreased only in the product with 2.25% SPPF. In general, hamburgers with SPPF addition were darker and less red and yellow, since the sweet potato peel has a light brown color. In addition, the sweet potato peel has catalytic chelating metals, such as iron and zinc, which favor oxidation of lipids and proteins present in meat (^{Ahn et al., 2002}; ^{Lund et al., 2007}; ^{Andrés et al., 2017}). These compounds alter the color of the product, which reduces consumer acceptability (^{Jha et al., 2007}). ^{Garcia et al. (2009)}, who evaluated hamburger with dry tomato peel (1.5% to 6.0%), reported similar results.

3.2 Technological analyses

The results of the cooking characteristics of hamburgers are shown in Table 3.

The shrinkage and moisture retention increased after the increase of SPPF in the hamburger, due to the high fiber content of the sweet potato peel, which retains water in the product, increasing the succulence (^{Anderson & Berry, 2001}). However, there was a reduction in cooking yield (p < 0.05) and fat retention of SPPF hamburgers, corroborating with the literature (^{Gök et al., 2011}). The preferential bonding of the fibers by water in detriment of fat may explain these findings (^{Anderson & Berry, 2001}), because the fibers form gels in aqueous solution, a process called myofibrillar protein gelation (^{Cordeiro, 2011}).

3.3 Sensorial analyses

The results of the hamburger sensory test added at different levels of SPPF are described in Table 4.

There was no statistical difference between the formulations F1, F2 and F3 (p > 0.05), which exhibited greater acceptability (p < 0.05) than F4 for all attributes. Thus, the addition of up to 1.5% of SPPF in hamburger can be considered well accepted by consumers. However, the increase in higher levels of SPPF (2.25%) reduces product acceptance, due to the residual and bitter taste of phenolic compounds present in large quantities in sweet potato peel (^{Anastácio et al., 2016}). Moreover, the addition of SPPF in the hamburger modified the texture of the dough making it more brittle, due to sweet potato peel high fiber content. Fiber hygroscopic capacity may explain this effect since they retain water inside the product (^{Célino et al., 2014}). All formulations showed high acceptance rates (≥ 70%), which demonstrate good sensorial acceptance of the products (^{Corradini et al., 2014}). Thus, it is demonstrated the feasibility of using SPPF as an ingredient in hamburger, which favors the consumption of healthier foods by the population.