**3. Results**

#### *3.1. Buckwheat Milling Fractions Composition*

The values of physico-chemical analyses of the buckwheat milling fractions, large (BL), medium (BM), and small (BS) are presented in Table 2. It can be seen that the moisture content ranged from 12.00% to 12.85% and the lowest value has been obtained for the medium particle size. The low moisture found for this particle size suggested its potential for higher storage stability and longer shelf life. Offia-Olua (2014) [42] mentioned that a moisture flour up to 12% indicated higher storage stability compared to the one with higher moisture. The results showed different moisture among the samples (*p* < 0.05). Similar data on the content of different buckwheat milling fractions was reported by Sciarini et al. (2020) [43] and Slukova et al. (2017) [44].


**Table 2.** Proximate composition and functional properties of buckwheat milling fractions.

<sup>a</sup> Mean ± SD. WAC—water absorption capacity; WRC—water retention capacity; SC—swelling capacity; VD volumetric density. Values followed by different lowercase letter (a, b, c) are statistically different at 95% confidence level.

The ash content for all fractions was between 1.01% and 4.23%. Similar results were found by Kasar et al. (2020) [26], Martin-Garcia et al. (2019) [45], and Bobkov S. (2016) [24]. The ash content of the studied milling fractions showed significant differences (*p* < 0.05) between all the samples. These findings can be attributed to the biochemical and morphological structure of the buckwheat seeds. In comparison with wheat flour, buckwheat is a good source of iron, zinc, copper and manganese [18,46], minerals that act as cofactors of antioxidant enzymes and serve as indirect antioxidants. A high level of magnesium, potassium, and phosphorus and a slightly low content of calcium was reported in buckwheat [18,47]. The high ash value (4.23%) observed in medium particle size (BM) can be related to the minerals such as magnesium, potassium, and phosphorus, which are stored in the embryo [48], and thus, BM was expected to have a high ash content. The supplementation of wheat flour with buckwheat flour fractions with high ash content could imply a rise in the mineral amount in the newly formulated flour.

Buckwheat flour milling fractions had a high protein content, which varied between 10.09% and 26.61% and was only different (*p* < 0.05) between fraction BM and the other two particles (BL and BS). Close results for protein content of various buckwheat milling fractions that differ from those studied in this work were reported in some studies [43,45,49]. Therefore, it can be concluded that noticeable differences in protein content can be associated with the presence of different embryo parts in the milling fractions. The protein content between BL and BS fractions did not show significant differences (*p* > 0.05), which can be possible because, during milling, the parts of the embryo where the protein is concentrated are separated as components of the BM fractions. In this case, the BM fraction, which is high in ash, may be a feasible option for use as a mineral-rich ingredient.

The fat content of the buckwheat milling fractions varied from 1.93% to 5.63% and was significantly different (*p* < 0.05) between the medium fraction and the other large and small fractions. Similar data for fat content on various buckwheat flour fractions were reported in some studies [24,50]. Most lipids in buckwheat seeds are located in the embryo and seed coat, followed by the pericarp and endosperm [51]. According to Ahmed et al. (2014) [46], the total lipid content in buckwheat grains ranged from 2.8% to 3.4%, while Dziadek et al. (2016) [19] found a significantly lower lipid content in whole common buckwheat grain samples (1.72–2.24%), a variation that can be attributed to the crop genetics and growing condition. As can be observed in Table 2, the fat content has an irregular trend with particle size variation. The highest values of fat (5.63%) were found for the medium fraction and can be related to the large amount presence of the lipidic part from buckwheat seed in this fraction.

The amount of carbohydrate in buckwheat fractions ranged from 51.52% to 74.02%, the milling fractions showing a significant (*p* < 0.05) effect on the carbohydrate content. The large particle size (BL) presented the highest carbohydrate content (74.02%), which can be related to the starchy endosperm. A decrease in carbohydrates in medium particle size was expected due to the rise in protein and ash content in this fraction.

#### *3.2. Functional Proprieties of Buckwheat Milling Fractions*

The water absorption capacity (WAC) of buckwheat milling fractions showed a significant reduction in a small fraction (BS) compared to medium (BM) and large particle size (Table 2). Data revealed that buckwheat milling fractions were statistically different (*p* < 0.05). The highest value (2.91%) observed at BM could be explained by the presence of a higher quantity of hydrophilic compounds from this fraction. The proteins and carbohydrates have a great influence on WAC due to the presence of such components as polar or charged side chains [52]. A high WAC implies high water incorporation in the dough, which can improve bread properties, especially texture and mouthfeel, with the knowledge that WAC is an index of the ability of protein and fiber to absorb and retain water. Thus, the WAC values confirmed that M particle size can absorb water well and swell for improved dough consistency, making it suitable for use in bread making. This WAC provides information on the strength of the starch inter-granular bond. A lower WAC was attributed to a close association of starch polymers in the native granule or can be attributed to the presence of the lower amounts of hydrophilic constituents in BF like carbohydrates and proteins, and influenced the cohesiveness of the food products.

Water retention capacity (WRC) for buckwheat flours fractions ranged between 3.81 and 5.91 g/g (Table 2), and are similar to the results obtained by Kasar et al. (2020) [26] for different buckwheat milling fractions. Differences were found among all the studied milling fractions (*p* < 0.05). These differences can be due to the different compositions of milling fractions and to the particle size distribution and its morphology. The increase of WRC in BM (5.91%) can be linked to a substantial content of ash and protein from this fraction.

Swelling capacity (SC) for the milling fractions of buckwheat flour studied ranged between 5.02 and 5.67 mL/g and does not present significant differences (*p* > 0.05) among the samples (Table 2). A similar trend was reported by Kasar et al. (2020) [26] when some functional properties of different buckwheat milled fractions were investigated. Buckwheat starch granules have a diameter of, on average, 5 μm and presented pores [53,54], facilitating, more or less, the penetration of water molecules into buckwheat starch granules, depending on particle size, but a regular trend was not observed. The values obtained for SC can be related to the extent of degree swelling during the heating process, and can also predict the buckwheat milling fractions behavior in further processing.

Volumetric density (VD) values for each particle size are showed in Table 2, which varied from 0.49 to 0.61 g/mL, showing a statistically difference (*p* < 0.05) for each particle size. The BS presented the highest volumetric density that can be explained, probably, by the positive association with the carbohydrates and negatively with the lipid content. Similar results (0.67 g/mL) were reported by Sindhu et Khatkar (2016) [55]. The lower VD

value found for medium particle size indicated a relatively good packaging characteristic of this milling fractions compared to other particle sizes.

It can be concluded that the compositional difference of carbohydrates and proteins between the particle sizes affected the functional properties of buckwheat particle size. The hydrophilic components present in milling fractions, like polysaccharides, have a good water retention capacity, whereas the polar amino acid residues of proteins have an affinity for water molecules, imparting water binding property.

#### *3.3. Fitting Models*

The investigation was conducted following the experimental design for determining the effects of the particle size and buckwheat milling fractions level added in wheat flour on falling number (FN) value and thermo-mechanical parameters. Mixolab rheological properties during the mixing stage in terms of water absorption (WA), dough stability (ST), dough development time (DT), minimum C2 torque, the difference between torques C1 and C2 (C1-2), and in the starch behaviour process as maximum C3 torque (C3), the difference between torques C3 and C2 (C3-2), minimum C4 torque (C4), the difference between torques C3 and C4, (C3-C4), maximum C5 torque (C5) and the difference between torques C5 and C4 (C5-4) [32] were assessed. The most fitting models were highly significant for almost all responses and were used for studying the influence of factors on the responses. These adequate models represented the experimental data well, showing high values (0.66–0.92) for coefficients of determination (*R*2) (Table 3). The ANOVA results, which comprise significant regression coefficients (*p < 0.05*), expressed in terms of coded value, quadratic determination coefficients (*R2*), and adjusted-*R*<sup>2</sup> (*Adj*.-*R*2), were calculated to assess the adequacy of quadratic models and are shown in Table 3.

**Table 3.** The coefficients in the predictive models for FN index and dough rheological properties during mixing and heating process of buckwheat-wheat flour dough.


*Note.* FN: Falling number index; WA: water absorption; DT: development time; ST: stability; C2: minimum torque during temperature increase; C1-2: protein weakening; C3: peak viscosity; C3-2: starch gelatinization; C4: cooking stability; C3-4: breakdown; C5: starch retrogradation; C5-4: setback; ns: non-significantly. <sup>a</sup> A: Particle size (μm); B: Addition of buckwheat flour in refined wheat flour (%); *R*2, *Adj.-R*2: is measures of fit of the model. *\*\*\**, *\*\**, *\** indicated significance at *p* < 0.0001, *p* < 0.001, and *p* < 0.05, respectively.

#### *3.4. Falling Number Index Estimation as Influenced by Buckwheat Milling Fractions and the Addition Level in Wheat Flour*

The Falling Number (FN) index values as the parameter used for estimation of amylolytic activity of grain enzymatic complex in the buckwheat-wheat composite flours formulated ranged from 299 to 369.5 s. The model can be adequate to predict the FN index, with a high level of significance (*p <* 0.001), defining the amylolytic activity of the buckwheat-wheat flour through the FN index. As in data showed in Table 3, the milling fractions caused a significant negative influence (*p <* 0.001) on FN, while the amount of buckwheat flour (BF) added was not-significant (*p >* 0.05). The negative effect of buckwheat milling fractions addition showed that the FN index of buckwheat-wheat flour increased with decreasing particle size of BF. This consequence can be probably associated to the phenolic compounds from the lowest fractions that bind to α-amylase and changing its conformation, and thus led to a rise in the FN index. A significant negative correlation was observed between FN and the interaction effect between the factors, while the quadratic

term of BF addition showed a significant positive correlation with FN (Table 3). The effect of factors, BF milling fractions, and BF addition level is presented in Figure 1a, showing that, with the increase of particle size and BF addition, the FN index decrease. This tendency suggested a rise in the α-amylase activity because an inversely correlation exists between the FN index and α-amylase activity in flour [56]. Consequently, buckwheat milling fractions can be used as an ingredient to adjust the α-amylase activity of wheat flour.

**Figure 1.** 3D response surface plots showing the interaction between buckwheat flour particle and addition level on (**a**) the falling number index and (**b**) water absorption (WA), (**c**) development time (DT), and (**d**) dough stability (ST) achieved during mixing.

#### *3.5. Dough Rheological Properties Evaluation Using Mixolab Test*

When some part of wheat flour is replaced with different buckwheat milling fractions, various physical and chemical changes in composite flours can occur, modifying the performance during processing. Therefore, the Mixolab test was used to observe dough rheological properties of wheat flour enriched with different buckwheat milling fractions at different levels. The Mixolab presents the advantage measure dough behavior during mixing and heating in a single test, simulating the mixing and baking processes, and offer the possibility to monitor both the protein and starch behavior during processing. The protein characteristics of dough are expressed as WA, DT, ST, C2 torque, and C1–2. The starch behavior during heating is expressed as torques C3, C4, C5, and C3-2, C3-4, C5-4, quantifying the changes in dough structure caused by temperature rise and mechanical forces of mixing.

#### 3.5.1. Water Absorption Evaluation during Mixing

The substitution of wheat flour with BF determined changes in dough consistency and stability, which are mainly associated with the formation of the aggregates as a result of hydrogen bonding or proteins linking through disulfide or dityrosine bonds [31]. The changes in dough consistency due to added BF milling fractions in wheat flour can be

related to the water quantity resulted from BF milling fractions and its addition level. Water absorption capacity has an essential role in influencing the baking process affecting the volume efficiency of the baked goods. Besides, this parameter offers an indication of the potential for the protein molecules in the sample to absorb the added water. Thus, as reported by Liu et al. (2019) [20], water absorption is considered to be an indicator of baking quality. For the formulated samples, water absorption (WA) values ranged from 57.5 to 58.7%. The predictive model for WA indicated that particle size and BF level had a significant (*p* < 0.05) effect on WA (Table 3). The negative coefficient of particle size and BF level showed that WA increased significantly (*p < 0.05*) when the BF level and particle size decreased. Figure 1b may notice that PS and the addition level of buckwheat flour have a negative effect on water absorption (WA), which indicated that WA decreased as the addition and particle size of BF increased. Nedeljkovi´c et al. (2014) [7] and Gavurníková S. (2011) [57] also reported a negative relationship between the water absorption and the BF milling fractions. In this study, the medium fraction, followed by large particles, needed more water to swell the starch than the small particle size, a finding that can be associated with WAC values (Table 2). This tendency can be related to several factors, especially fractions of chemical composition. As Sapirstein et al. (2018) [58] stated, damaged starch represents one of the major factors that contribute to the water absorption alongside the proteins and starch. At the same time, an increase in WA with a decrease of milling fractions was reported in previous studies [28,29,31,32] on the impact of substituting wheat flour with different fraction sizes of non-gluten flours, revealing that the small particle size raised the WA values.

#### 3.5.2. Dough Development Time Evaluation

The effect of buckwheat milling fractions addition on dough development time (DT) is presented in the quadratic regression model shown in Table 3. The buckwheat flour (BF) addition level influenced the gluten network of the dough and significantly modified the dough development time depending on fraction sizes. The quadratic model was found to be adequate to predict DT, depending on the studied factors, particle size and BF addition level (Table 3). Describing the goodness of fit of the models, the quadratic determination coefficient (*R*2) was 0.75, which confirms the suitability of the model, showing that the model explained 75% of data variation. DT of buckwheat-wheat flour ranged between 0.9 to 4.65 min and it was affected (*p < 0.05*) by the particle size, BF addition level, and the interaction between fraction size and BF addition level. A higher increase in the DT was observed with a particle size increase compared to the BF level increase (Table 3). Thus, DT increased notably as the particle size increased (Figure 1c), showing that it took a long time from the time the water was added to the time when the dough reached the optimal consistency (C1 torque =1.1 N·m). An increase of DT of about 2.1–2.7 times in the buckwheat-wheat flour with the largest fraction and about 2.1–2.5 times in the medium fraction as the BF addition level increased by over 10% was obtained beside the wheat dough used as control. This fact may be due to the rise in water when the BF addition was high, thus requiring mixing. Larger and medium particle sizes rich in some compounds such as dietary fibers and proteins, respectively, required a longer time for hydration, which implies a higher DT. With the addition of gluten-free flour, the gluten network is affected and leads to a rise in DT. As the DT value quantifies the dough strength, a higher value indicated stronger dough. Increased DT was also found by Sedej et al. (2011) [22] and Nikolic et al. (2011) [59] when different levels of buckwheat flour were used to supplement wheat flour, but without taking into account different particle sizes.
