All chemicals, reagents and solvents used were of analytical grade or HPLC and obtained from Sigma-Aldrich (St. Louis, MO, USA). Deionized water (> 18 MΩ.cm) was obtained from a Milli-Q system (Millipore, Brussels, Belgium).

The buru almonds were obtained from the Cerrado Biome region, located in the city of Barra do Garças-MT, during the harvest season (between August and September 2016). The baru almond has the following centesimal composition (g/100g): 6.63 moisture, 22.96 protein, 31.73 lipid, 14.44 dietary fiber and it wieghs 1.55 grams.

For this work, the completely randomized experimental design was used with three replications. Three treatments were evaluated: baru almonds “in natura” (T1 – which was the control for the drying experiment), baru almonds submitted to drying at 65 °C for 30 min (T2) and baru almonds subjected to drying at 105 °C for 30 min (T3). The temperatures were chosen because they are the most commonly used by the producers of these oleaginous plants. In addition, it is known that the roasting process compromises part of the nutritional and bioactive value of the food, and therefore to use lower temperatures which are effective in drying the baru almond preserve its nutrients can become a new form of use and consumption for this food.

The drying of the baru almonds was carried out in a forced air circulation oven (Marconi, MA0351, Piracicaba, Brazil). They were then ground (20 mesh - in experimental mill, Viti Molinos, VG 2000i, Itajaí, Brazil), packed in transparent polyethylene containers and stored at 25 °C (oven model Eletrolab, EL202). During the drying processes, the time and temperature were chosen because they are yet to be investigated. The moisture contents of the almonds were determined after the drying process (AOAC, 1990) and the averages were 6.63, 5.10 and 2.64 g·100 g^-1 respectively for T1 T2 and T3. All the results of the analytical determinations described below (including moisture) were expressed on a wet basis.

The contents of total phenolic compounds, anthocyanins, flavonoids, tannins and vitamin C were evaluated. The hydroalcoholic extract was prepared according to the methodology adapted from Milardovic et al., (2006).

Total phenolics were determined by the Folin–Ciocalteu reagent method, using gallic acid as the standard for the calibration curve. The absorbance was measured at 765 nm in a spectrophotometer (UV-Visible 50 Probe-Cary) and the results were expressed in mg of gallic acid equivalent (GAE) 100 g^-1 (Lemos, 2012).

The total phenolic compounds were also evaluated by the use of diazonium salt Fast Blue BB, using standard gallic acid for the calibration curve. The absorbance was measured at 420 nm in a spectrophotometer (UV-Visible 50 Probe-Cary) and the results were expressed in milligrams gallic acid equivalent (GAE) per 100 g^-1 (Palombini et al., 2016).

Monomeric anthocyanins were determined by the differential pH method described by Giusti and Wrolstad (2001), using a spectrophotometer (UV-Visible 50 Probe-Cary) for absorbance measurements of samples (510 and 700 nm) and the results were calculated as malvidin-3,5-diglucoside.

Flavonoids were determined following the methodology described by Fraguas et al., (2014). The reading was held at 415 nm spectrophotometer (UV-Visible 50 Probe-Cary) using a 2% solution of aluminum chloride in methanol. The total flavonoid values were expressed as equivalents of catechin.

Tannins were measured by the colorimetric method according to the Association of Official Analytical Chemists (AOAC, 1990). The method was based on the intensity of the blue color produced in reducing the Folin-Denis reagent for phenols, and was then measured in the spectrophotometer (UV-Visible 50 Probe-Cary) to 760 nm expressed as equivalents of catechin.

Vitamin C was evaluated by the colorimetric method using 2,4-dinitrophenylhydrazine and the results were read on a spectrophotometer (UV-Visible 50 Probe-Cary) to 520 nm and expressed in equivalents of ascorbic acid (Milardovic et al., 2006).

The individual identification of phenolic compounds was made following the methodology described by Ramaiya et al., (2013). The quantification and identification of these phenols compounds were performed in a liquid chromatography (HPLC-DAD/UV-Vis) model Shimadzu (Shimadzu Corp., Kyoto, Japan) equipped with a gradient pump (2487 Serie), a valve injector with a loop of 50 μL, a degasser (Waters 200 Series) and an integrator-plotter with software (Total Chrom, Waters). Phenolic compounds were separated in a C18 reversed-phase column (150 mm × 4.6 mm I.D., 5 μM), (Phenomenex, CA, USA) with a C18 (20 mm × 4.6 mm I.D.) pre-column cartridge. The mobile phase consisted of 2% (v/v) acetic acid in deionized water (mobile Phase A) and 70:28:2 (v/v) methanol/water/acetic acid (mobile phase B), and phenolics were detected at 280 nm. Phenolic compounds were identified by comparison of retention times with standards (gallic acid, catechin, chlorogenic acid, caffeic acid, vanillin, p-coumaric acid, ferulic acid, m-coumaric acid, o-coumaric acid, trans-cinnamic acid, quercetin and rutin). The results were expressed as mg of phenolic compound in 100 g^-1 of fresh weight.

Test for elimination of free radicals (DPPH•). The antioxidant activity was evaluated by the DPPH• scavenging ability of the antioxidant activity. Absorbance was measured using UV-Visible spectrophotometer (50 Probe–Cary) at 517 nm expressed as % of free radical sequestration (Milardovic et al., 1943).

Antioxidant activity via β-carotene⁄linoleic acid system. Antioxidant activity determination by β-carotene/linoleic acid system was conducted according to the methodology described by Miller (1971) and absorbance was measured at 470 nm with a spectrophotometer (UV-Visible 50 Probe-Cary). The results were expressed as percentage of inhibition of β-carotene oxidation.

Oxygen Radical Antioxidant Capacity (ORAC) assay. The ORAC method used, with fluorescein (FL) as the “fluorescent probe”, was applied as described by Aazza et al., (2011).

The levels of total sterols, tocopherols and fatty acids present in baru almonds were determined. To carry out these analyses, it was necessary to extract the oil present in the baru almond.

The results were submitted to analysis of variance (ANOVA) and test for comparison of averages (5% probability of error according to the Tukey test). Statistical calculations were performed using the program R^® version 5.0. The differences were considered significant when p < 0.05.

The results of the determinations of total phenolics, monomeric anthocyanins, flavonoids, tannins and vitamin C of baru almonds are displayed in Table 1.

It was observed that the drying processes did not alter the total phenolic or tannins because these compounds remained unchanged (p > 0.05), thus showing the thermal stability to these temperatures. When applied to the process of 65 °C for 30 min (T2), a significant reduction (p < 0.05) was detected only in the presence of monomeric anthocyanins, which proves these molecules are sensitive to such a drying process. When using 105 °C for 30 min (T3) they were reduced (p < 0.05) in addition to the anthocyanins and vitamin C, although the level of flavonoids increased (p < 0.05).

In reference to the means obtained from the determination of phenolic compounds, it the predominance of these substances in baru almonds was demonstrated, which in turn, exceeded the values reported for other nuts (with an averages between 32 and 420 mg·100 g^-1 for macadamias, cashews, walnuts, hazelnuts and peanuts) (Kornsteiner et al., 2006). The process of drying was found not to influence this content and this behavior can be associated with the absence of oxidative reactions by reactive oxygen species. Discrepant behavior was found by Lin et al., (2016), who studied the effect of drying on almonds (150 °C for 20 min) and found that the phenolic compounds were reduced. Pasqualone et al., (2018), when studying the drying process of almond by-products, also found a significant reduction in the presence of phenolic compounds in these foods.

On the other hand, the use of the Fast Blue BB method provided a reduction in the detection of these phenolic substances, demonstrating a difference between the two processes. This evidence is attributed to the fact that this procedure is specific and unique to the quantification of phenols, for pure reagents which are not mixed with other compounds such as proteins, sugars or ascorbic acid reducers (as occurs in the Folin Ciocalteau’s method) (Naczk and Shahidi, 2004). It was noted that this methodology presented greater precision for determining these substances.

Concerning the quantification of monomeric anthocyanins in baru almonds, a considerable reduction (p < 0.05) was observed (15.78% and 34.21%, respectively in T2 and T3). It was observed that the higher the temperature of the drying process, the greater the degradation of these substances. A negative effect of the temperature/oxygen interaction was evidenced, which was detrimental to the maintenance of these molecules.

Higher flavonoid contents were found in treatment T3 when compared to T1 and T2, indicating that the process of drying at 105 °C for 30 min increased (p < 0.05) the presence of these molecules by 72%. The increase in bioactive compounds, such as flavonoids, may be related not only to the degradation of polymeric polyphenols and flavonoid glycosylated hydrolysis (Lemos et al., 2012) but also to the Maillard products reaction (Liu, Kitts, 2011). In addition, the intracellular water evaporation could have changed the lignocellulosic structure, in addition to promoting the denaturation of proteins, resulting in an increased availability of active compounds in the array (Lemos et al., 2012). Similar behavior was found by Lin et al., (2016), who concluded that the roasting process (200 °C for 20 min) contributed to the increase of 124% in the content of flavonoids in almonds.

With respect to tannins, a considerable presence was observed in baru almonds and the drying processes were found not to promote a significant reduction (p > 0.05) thereof. The abundant consumption of this substance should be controlled because the exploitation of minerals and proteins causes antinutritional effects. However, as the consumption of “in natura” baru almonds is not a common practice, mostly due to its sensory characteristics in this condition (bitter and astringent), the application of heat becomes an important feature that aims to improve the sensory aspects and inactivate the antinutritional substances. Recent researches has indicated that these substances present anti mutagenic potential, which, in turn, are related to their antioxidant potential, and are effective in protection from oxidative damage (Macáková et al., 2014).

The concentration of vitamin C was fond to decrease (p < 0.05) by 9.98% due to the drying process at 105 °C for 30 min in T3, which can be assigned to a likely oxidation of this vitamin, which by its biologically active nature, is unstable and reversibly oxidized to Ldehidroascórbic acid. Even with this, it was found that the treatments corresponded to a high content of vitamins. According to the Dietary Guidelines for Americans (McGuire, 2011), the recommendation of daily vitamin C intake is 75 mg for women and 90 mg for men over 30 years old. These levels are based on their physiological functions and antioxidants needed to benefit the physiological organism. In this regard, when 100 g of baru almonds are consumed, treatments T1, T2 and T3 may represent 52.2, 46.9 and 50.5%, respectively, of the recommendation for women, and 41.4, 43.5 and 39.2% for men. It is found that this almond is a food with high levels of this nutrient. It is important to note that Fatin and Azrina (2017) have found lower results for vitamin C in fresh lime (27.78 mg·100 g^-1), which is one of the referenced foods for this vitamin. Therefore, the consumption of baru almonds may be beneficial in the case of possible vitamin deficiencies.

Test of elimination of free radicals (DPPH•). The results obtained for the total antioxidant levels analyzed by the DPPH• method were 69.02, 79.68 and 84.38% free radical scavenging capacity for T1, T2 and T3, respectively (Table 1). The drying process at 105 °C for 30 min promoted a significant reduction (p < 0.05) in T3, decreasing by 36.05% the DPPH• radical scavenging capacity in comparison with baru almonds “in natura”. However, almond T2 remained stable when compared to T1 (control), therefore, it is reasonable to conclude that the drying process held at 65 °C for 30 min did not reduce the presence of antioxidants in almonds. It was found that more intense heat treatments could be responsible for the loss in antioxidants. Lemos et al., (2012), explained that during heating, part of the moisture was lost through evaporation and the Maillard reaction increased the antioxidant capacity of the almonds. Higher values for free radical scavenging capacity indicate a lower antioxidant activity. Some compounds do not react against the DPPH• free radical because they are lipophilic and, therefore, it is necessary to apply different methodologies that seek to characterize such components.

Antioxidant activity via β-carotene⁄linoleic acid system. The total antioxidant activity was also measured by the β-carotene/linoleic acid system and the average value obtained among treatments was 89.24% protection without any statistical differences (p > 0.05) (Table 1). These results indicated that baru almonds presented high levels of these components and heat-resistant behavior. These data present promising results because oxidative processes can be avoided through the use of antioxidants with prevention or decrease in triggering oxidative reactions. Moreover, this method is suitable for the investigation of lipophilic antioxidants and fits the constitution of baru almonds especially due to their high lipid content.

Oxygen radical activity capacity (ORAC) assay. In the ORAC method it was observed that baru almond “in natura” showed significantly (p > 0.05) more antioxidant capacity when compared to the other treatments (Table 1). As they are heat-sensitive substances, the drying processes have contributed to the reduction in this attribute. As this essay is based on a hydrogen atom transfer reaction using a free radical source which is predominant in human biology (peroxyl radical), the ORAC test is relevant to express the antioxidant capacity of a given substance under in vivo conditions.

A total of twelve phenolic compounds were detected in the three treatments, and the results associated with the identification and quantification of these molecules are shown in Table 2.

Among the phenols identified by high-performance liquid chromatography, gallic acid was the main phenolic compound in baru almonds. It should be noted that the high contents of these substances in food, more specifically in almonds, are desirable because they provide health benefits. Gallic acid, which has a strong free radical capacity, is effective in preventing disease, and research indicates that it is able to induce apoptosis and cytotoxic and antiproliferative effects among different strains of tumor cells (Guimarães et al., 2007). As the drying process increased these substances (which was an increase of 1.07 times), we can infer that the benefits to be achieved through the ingestion of these molecules can be enhanced. Lin et al., (2016) have also observed a significant increase in the presence of gallic acid in almonds dried at 150 °C for 30 min. However, this increase was 13 times greater (when compared to raw almonds) and the end content of this substance was 52 mg·100 g^-1.

It was found that other components, such as catechin, vanillin, trans-cinnamic acid, m-coumaric acid and rutin, were increased (p < 0.05) with the drying process (when compared to the “in natura” treatment), and this increase was 1.22, 1.34, 1.19, 1.17 and 1.03%, respectively. One possible explanation for this increase is associated with the changes that might have occurred in the protein of phenolic compounds, causing their exposure and the consequent increased availability (Lemos et al., 2012). Thermal processing is responsible for increasing the phenolic content due to the increase in the number of free phenolic groups resulting from the hydrolysis of glycosylated flavonoids that are released from the phenolic cell walls (D’archivio et al., 2010). Rodríguez-Bencomo et al., (2015), when studying the effects of the roasting process (160 °C for 20 min) in pistachios, found an increase of 17, 79 and 81% in the presence of chlorogenic acid, rutin and catechin. It was also found that the drying (when comparing T1 to T3) promoted a reduction (p < 0.05) in the presence of chlorogenic acid, caffeic, p-coumaric acid, quercetin, ferulic and o-coumaric acid of 7.2, 41.3, 30.2, 9.4, 74.3 and 7.57%, respectively. The decreases in these substances can be attributed to the effects of heat, which may have caused possible protein denaturation resulting from the disruption of covalent links (Shahidi and Yeo, 2016).

Total sterols. The determination of the total sterol contents of baru almonds subjected to different drying processes of drying is shown in Table 3.

The total sterol content of baru almonds was measured at an average of 432.66 mg·100 g^-1 without statistical difference (p > 0.05) among the treatments, and these results may demonstrate that the drying process does not influence these molecules. For health benefits, it is recommended to eat between 1 to 2 g of plant sterols a day. In this regard, the ingestion of 100 g of baru almonds can meet more than 40% of this recommendation (considering the consumption of 1 g/day). This result is favorable since these components have different physiological effects mainly in the treatment of hypercholesterolemia, as well as in the control of cholesterol by secondary causes in diabetics and patients with metabolic syndrome (Lemos et al., 2016). In addition, baru almonds present total sterol values greater than other oilseeds considering that Brazil’s almonds, walnuts, cashews, hazelnuts, macadamia nuts, pistachios and walnuts have average total sterols of 192, 160, 154, 132, 105, 189 and 197 mg·100 g^-1, respectively (Miraliakbari and Shahidi, 2008).

Identification of tocopherols by HPLC-DAD. The total contents of tocopherols (α and γ) (Table 3) obtained among the treatments showed significant differences (p < 0.05) indicating that the drying process (65 and 105 °C for 30 min) can contribute to an increase of 8.12 and 63.95% in these compounds, respectively. When evaluated individually, the tocopherols also presented the same behavior, and these data indicate that these substances may be enhanced through the use of heat. γ-tocopherol was the most abundant, with corresponding values for T1, T2 and T3 of 74.6, 71.2 and 61.9%, respectively of total tocopherol content. α-tocopherol presented values of 25.3, 38.1 and 43.7% for the same treatments. Factors such as the composition in fatty acids, the presence of compounds with antioxidant activity (especially those from lipophilic), variety, degree of ripeness and care in almond production can be responsible for the maintenance of these molecules (Lemos et al., 2016). The quantification of the levels of tocopherols, specifically in food, becomes necessary due to the action of this vitamin in preventing oxidative damage to DNA, which act as antioxidant structures and assist in peroxidation inhibition of lipids (Lemos et al., 2016).

Profile of fatty acids by CG-FID. Table 4 presents the identification, quantification and relations among the saturated, polyunsaturated and unsaturated fatty acids present in baru almond subjected to different drying temperatures, as well as the atherogenic index (AI) and thrombogenic index (TI), and the hypocholesterolemic and hypercholesterolemic fatty acids (h/H).

It was found that the unsaturated fatty acids, including oleic and linoleic, were predominant. The high proportion of monounsaturated fatty acids and polyunsaturated was also observed. The Dietary Guidelines for Americans (AHA, 2016) states that the daily intake of PUFAs is 20 g^-1 and, in this sense, it is recommended to ingest 63.5, 65.1 and 64.1 g^-1 baru almonds corresponding to T1, T2 and T3 to provide 100% of this recommendation.

Among the saturated fatty acids, palmitic acid was the major representative in all treatments. The relation between PUFA/SFA was higher than 1.9, 1.8 and 1.8 respectively for T1, T2 and T3, and T2 and T3 are statistically similar (p < 0.05). For a balanced intake of these elements, this value must be at least 0.45. In relation to this profile (low levels of SFA and high of PUFAs), it is estimated that baru almonds can be effective in the control of the traditional risk factors for atherosclerotic cardiovascular disease. This is because the presence of polyunsaturated fatty acids may promote a hypocholesterolemic effect.

As to the atherogenic index (AI) and thrombogenic index (TI) and the hypocholesterolemic fatty acids and hypercholesterolemic (h/H), which indicate the potential for stimulating the aggregation platelet and coronary artery disease, the values obtained for T1, T2 and T3 were respectively 0.28, 0.09, 11.31, 0.29, 0.08, 12.01, and 0.27, 0.07, 12.58 with statistical difference (p < 0.05) among the treatments. Although there is no established parameter for these indices, the smaller the result for AI and TI and greater for h/H, the less likely are the changes as mentioned above and the healthier the food is. This is due to the greater concentration of anti-atherogenic fatty acids (Ω3 and Ω6) (Turan et al., 2007). Foods such as cheeses and meats that are significant sources of fatty acids, have indexes of AI, TI and respective h/h 2.32, 3.11, 1.23 and 0.54, 1.15, 1.76 (Faria et al., 2015), and comparing these figures with those of baru almonds, the superiority and greater possibility of cardiovascular protection on the part of the almond is demonstrated.

With regards to the drying process (T3), it was found that heat promoted a reduction (p < 0.05) in different fatty acids (relative to T1) in C16:0 and C18:2 of 1.4 and 6.6%. By analyzing all the reductions and correlating them to the degree of unsaturation, one can infer that polyunsaturated fatty acids were the most affected. This may be due to the fact that this compound has a greater number of double bonds, which results in increased susceptibility to oxidation.

Conversely, this same process of drying promoted increases (p < 0.05) in different fatty acids, C18:0, C18:1, C22:0 and C22:2 with expansion of 18.9, 4.1, 4.9 and 7.5%, respectively. The saturated fatty acids were the most favored by the drying process (when compared to T1 with T3), because they represented together a total increase of 134.6%. This occurs because heating may increase the oxidation rate of these molecules. Rodríguez-Bencomo et al., (2015), when analyzing pistachio oil subjected to heating (160 °C for 20 min), found that heat is not promoted and there is no significant increase in the levels of SFA, PUFA and MUFA.