Quality Differences between Fresh and Dried Buckwheat Noodles Associated with Water Status and Inner Structure

Buckwheat noodles are mainly sold in the form of fresh and dried noodles in China. Among the noodles with varied proportions of extruded buckwheat flour (20% to 80%), the cooking or textural qualities of fresh and dried buckwheat noodles (FBN and DBN, respectively) were significantly different, and FBN showed a lower cooking loss and breakage ratio and were more elastic than DBN. FBN-20% showed the highest sensory score, followed by DBN-50%. The mechanisms causing the quality differences were investigated using water mobility and the internal structures of the noodles were investigated with low-field nuclear magnetic resonance and scanning electron microscopy, respectively. Compared with FBN, DBN showed a denser internal structure, which explained its higher hardness. The water within FBN and DBN was mainly in the form of softly bound water and tightly bound water, respectively. FBN with highly mobile softly bound water (longer T22) and a more uniform internal structure had a lower breakage ratio, whereas the trends of water relation with texture properties were different for FBN and DBN. The drying process and added extruded buckwheat flour together contributed to the varied cooking and textural properties.


Introduction
Buckwheat (Fagopyrum esculentum Moench) is widely consumed in Asian countries because of its functional components, such as rutin, and high biological value due to its well-balanced amino acid composition (rich in lysine and arginine) [1]. In China, buckwheat noodles are mainly sold in the form of fresh and dried noodles. Dried noodles are often selected owing to their easy transportation and long shelf life [2], ensuring food supply during severe emergencies.
Consumers head for buckwheat noodles with higher buckwheat contents, which requires solving the problem that the predominant proteins within buckwheat (globulin and albumin) cannot form an effective gluten network. As indicated from our previous study [3], buckwheat noodles with a high percentage of buckwheat flour (80%) can be achieved using extruded buckwheat flours. Secondly, dried buckwheat noodles with good eating qualities are favourable for commercial consumption [4]. Fresh and dried noodles are intrinsically different [5][6][7], as excessive dehydration may deteriorate the colour, texture, and even nutritional properties of the noodles [8]. However, the factors causing the differences in the eating qualities of fresh and dried buckwheat noodles (FBN and DBN) and how the reconstituted composition of buckwheat noodles influence their eating qualities are still unknown.
Moisture is one of the possible factors leading to the different eating qualities between fresh and dried noodles. During the drying process in making dried noodles, free water and softly bound water gradually migrate from the interior of the noodles to the environment, and the residual moisture is more tightly bound to the noodles [9]. As a result, a denser inner structure within the noodles is formed, which may further influence the cooking time, water absorption ratio [10], and cooking and sensory qualities of dried noodles [6,11]. Additionally, the moisture content and water status may be different in fresh and dried buckwheat noodles considering their optimum cooking time and ability to combine with water (with varying amounts of extruded buckwheat flour). Besides this, the varying amounts of extruded buckwheat flour may also generate different gluten networks and food matrixes within FBN and DBN. As reported by Drechsler and Bornhorst [12], food matrixes with different structures can bring about different textures and even nutritional properties in cereal products [8,12].
Thus, to elucidate the reasons causing the cooking and eating differences between FBN and DBN regarding water status and inner structure, low-field nuclear magnetic resonance (LF-NMR) and scanning electron microscopy (SEM) were used. A targeted improvement in the eating qualities of buckwheat noodles is expected to be achieved by understanding the mechanism causing the varying qualities.
Buckwheat extrusion: Buckwheat flour was mixed with distilled water to obtain a moisture content of 18.0% and equilibrated overnight at 25 • C. The buckwheat flour was extruded through a twin-screw extruder (DSE-25, Brabender OHG, Duisburg, Germany). The ratio of screw length to diameter (L/D) is 20:1, and the diameter of the die nozzle is 5 mm. The extrusion condition was under the temperature of 148 • C, screw speed of 150 rpm, and feed rate of 35 g/min [3]. The extrudates were freeze-dried using a freeze dryer (ALPHA 1-2 LD plus, CHRIST, Osterode, Germany) and milled using a grinder (ZM 200, Retsch, Germany) then through a 90-mesh sieve.
Dried buckwheat noodles: FBN were dried in a humidity chamber (BCL-250-III, Beijing Luxi Technology Co., Ltd., Beijing, China) at 40 • C, 75% relative humidity for ten hours, then were further maintained at ambient conditions for 12 h for the preparation of dried buckwheat noodles.

Colour of Buckwheat Noodles
The colour of the uncooked DBN and FBN was measured using a Lab Colorimeter (CR400/410, KONICA MINOLTA, Tokyo, Japan). The lightness (L*), redness (a*), and yellowness (b*) were recorded as the average of 7 measurements. The total colour difference (∆E) was calculated by the following Equation (1).
L s *, a s *, and b s * are the standard values of a white calibration ceramic plate (97.13, 0.21, and 1.87, respectively). A higher ∆E with a more considerable value indicates the colour is darker, taken as buckwheat noodles with a higher content of buckwheat flour.

Cooking Properties of Buckwheat Noodles
The optimal cooking time (OCT), cooking loss (CL), and cooking breakage ratio (CBR) were measured following the methods of [13] and [4]. OCT is defined as the time when the white core within the noodles disappeared with two transparent glass plates squeezing the noodle strands. CL was calculated as the number of solid substances in the cooking water at OCT. CBR was expressed as the ratio of the number of the broken noodle strands and the original number of noodle strands after cooking. All the noodles were cooked at OCT in this research.

Textural Properties of Buckwheat Noodles
The texture of buckwheat noodles was analysed using a texture analyser (TA-XT2i, Stable Micro Systems, London, UK). Five cooked noodle strands were placed side by side on a loading platform and were compressed by a probe (A/LKB-F) under the strain of 75%. The pre-test speed, test speed, and post-test speed were set as 2.0 mm/s, 0.8 mm/s and 2.0 mm/s, respectively. The time between two compressions was 10 s, and the trigger force was 10 g. The textural parameters were averaged from 7 measurements.

Sensory Properties of Buckwheat Noodles
Sensory properties of buckwheat noodles were evaluated by a sensory panel (eight members, four females and four males) according to the method of Wang, Li, Chen, Hui, Tang, and Wei [3]. The sensory panel members were carefully selected with a preexperiment. A control noodle sample was used to test the panel's stability. Sensory parameters and their scores were shown in Table S1.

Water Status and Distribution in Uncooked Buckwheat Noodles
The water status and distribution of the uncooked FBN and DBN were measured according to the method of [2] using the LF-NMR (NMI20-030H-I, Niumag Analytical Instruments Co., Ltd., Suzhou, China), which was equipped with a 0.5 T permanent magnet corresponding to a proton resonance frequency of 21 MHz at 32 • C. Transverse relaxation times (T 2 ) was obtained from Free Induction Decay (FID) and Carr-Purcell-Meiboom-Gill (CPMG) experiments. FID sequence was used to set the system parameters using soybean oil. The number of sampling points (TD) was 10,104, the relaxation time (TR) was 1000 ms, echo time (TE) was 0.101 ms, while the number of echoes (NE) was 1000 and acquired as 64 scan repetitions. Uncooked noodle strands were put in an NMR tube with a 5 mm diameter. MultiExpInv analysis software (Suzhou Niumag Analytical Instrument Corporation, Suzhou, China) was used to fit CPMG decay curves to a multi-exponential function model.
The times of the peak position were recorded as T 21 , T 22 , and T 23 , which were termed as tightly bound water, softly bound water, and free water, respectively. The relative content of water in different status was calculated using the proportion of the corresponding peak area to the total area, recorded as A 21 , A 22 , and A 23 , respectively.

Scanning Electron Microscopy (SEM) Analysis
The morphology of the uncooked and cooked buckwheat noodles was characterised following the method of Wang, Li, Chen, Hui, Tang, and Wei [3]. Samples were lyophilised using a freeze-dryer (ALPHA 1-2 LD plus, CHRIST, Germany), fractured and mounted on a copper stub. Then the samples were coated using a Sputter coater (JFC-1600, JEOL, Tokyo, Japan). The morphologies of the fracture surface of noodles were observed using a scanning electron microscope (JSM-6510LV, JEOL, Tokyo, Japan), at a magnification of 40× and 500×.

Statistical Analysis
All the analyses were conducted in triplicate unless otherwise mentioned. ANOVA and Duncan's Multiple Range Test were used at a significance level of p < 0.05, accomplished using SPSS 18.0 (IBM, New York, NY, USA).

Cooking Properties
FBN showed a shorter optimal cooking time (OCT) than DBN (Table 1), which also showed better cooking properties than DBN. For instance, the CL of DBN was higher than that of FBN (16.60-44.40% and 11.56-22.31%, respectively) ( Figure 1). The cooking breakage ratio (CBR) for DBN was almost two times higher than FBN (37.5% and 21.3%, respectively) when the addition ratio of extruded buckwheat flour was 80% ( Figure 1). The cooking loss (CL) of buckwheat noodles was significantly increased (p < 0.05) when the amount of extruded buckwheat flour was increased from 20% to 80%. The morphology of the uncooked and cooked buckwheat noodles was characterised following the method of Wang, Li, Chen, Hui, Tang, and Wei [3]. Samples were lyophilised using a freeze-dryer (ALPHA 1-2 LD plus, CHRIST, Germany), fractured and mounted on a copper stub. Then the samples were coated using a Sputter coater (JFC-1600, JEOL, Tokyo, Japan). The morphologies of the fracture surface of noodles were observed using a scanning electron microscope (JSM-6510LV, JEOL, Tokyo, Japan), at a magnification of 40× and 500×.

Statistical Analysis
All the analyses were conducted in triplicate unless otherwise mentioned. ANOVA and Duncan's Multiple Range Test were used at a significance level of p < 0.05, accomplished using SPSS 18.0 (IBM, New York, NY, USA).

Cooking Properties
FBN showed a shorter optimal cooking time (OCT) than DBN (Table 1), which also showed better cooking properties than DBN. For instance, the CL of DBN was higher than that of FBN (16.60-44.40% and 11.56-22.31%, respectively) ( Figure 1). The cooking breakage ratio (CBR) for DBN was almost two times higher than FBN (37.5% and 21.3%, respectively) when the addition ratio of extruded buckwheat flour was 80% ( Figure 1). The cooking loss (CL) of buckwheat noodles was significantly increased (p < 0.05) when the amount of extruded buckwheat flour was increased from 20% to 80%.

Textural Properties
The textural properties of DBN and FBN varied with the ratio of extruded buckwheat flour. Hardness and elasticity are the most critical texture parameters for con-

Textural Properties
The textural properties of DBN and FBN varied with the ratio of extruded buckwheat flour. Hardness and elasticity are the most critical texture parameters for consumers. In this study, the hardness of DBN and FBN was decreased with an increased ratio of extruded buckwheat flour ( Table 1). The hardness of DBN-20% was significantly higher than that of FBN-20%, while FBN-80% showed a higher hardness than DBN-80% (p < 0.05). The elasticity of FBNs was significantly higher than that of DBNs. The elasticity was high and relatively stable for FBN, but a noticeable increase in the elasticity of DBN was observed with increased extruded buckwheat flour ratios (p < 0.05). The resilience of DBNs was also higher than FBN regardless of extruded buckwheat flour ratios. Similar results were also found regarding the cohesiveness and chewiness of DBNs and FBNs. DBN with increased proportions of extruded buckwheat flour showed increased cohesiveness while FBN showed insignificant changes. For chewiness, FBN showed a decrease when the ratios of extruded buckwheat flour increased from 20% to 80%, whereas the chewiness of DBN was the highest when the amount of extruded buckwheat flour was 50%.

Colour of Uncooked FBN and DBN
The colour of DBN was darker than that of FBN (Table 2), with a higher ∆E for DBN (values ranged from 36.56 to 58.81 of DBN, and 28.79 to 53.11 of FBN, respectively) (p < 0.05). DBN showed lower lightness (L*) and yellowness (b*) values than those of FBN. The L* and a* of FBN were higher than those of DBN with the same ratio of extruded buckwheat flour, whereas b* showed an opposite trend. Besides this, buckwheat noodles with 20% and 80% of extruded buckwheat flour showed significantly lower b* values than that with 50% extruded buckwheat flour (24.74 and 21.62 for FBN-50% and DBN-50% respectively).

Sensory Properties
The sensory evaluation reflects the direct conceiving of the consumers, which cannot be easily replaced by instrumental analysis, as some of the parameters measured by instruments cannot cover the complex feeling of humans [14]. In this study, the FBN-20% showed the highest sensory score, followed by DBN-50%, as the colour and appearance were much higher for noodles with 20% extruded buckwheat flour than for other noodles. The hardness and total scores showed a decreased trend for FBN with increased ratios of extruded buckwheat flour, while DBN showed the highest value with 50% extruded buckwheat flour. With 80% extruded buckwheat flour, the noodles tended to be soft and sticky, which showed the worst sensory quality among all noodles, with a lower score in the colour, hardness, stickiness, and smoothness.
Interestingly, there were insignificant differences observed in terms of elasticity, stickiness, smoothness, and flavour between FBN and DBN with the same ratio of extruded buckwheat flour ( Table 3). Note that the TPA hardness was not increased with a higher sensory score, as the preferable hardness was given the highest score, and the preferred TPA hardness of the buckwheat noodles was approximately 167 g and 150 g. FBN-20/50/80% is fresh buckwheat noodles with 20/50/80% of extruded buckwheat flour; DBN-20/50/80% is dried buckwheat noodles with 20/50/80% of extruded buckwheat flour. The flavour score is in the range of 1 to 5; the colour, appearance, and smoothness score range from 1 to 10; the hardness and stickiness score is in the range of 1 to 20; the elasticity score is in the range of 1 to 25. Detailed meaning for an individual score can be found in Table S1 in the supplementary information. a , b , and c in a column indicates values are significantly different at p < 0.05.

Water Status and Distribution
The water within noodles distributes differently during dough mixing and the drying process [2]. The transverse relaxation time of proton (T 2 ) expresses the interaction between water and substances, suggesting the mobility of the water. In this study, the T 2 peaks representing the water under different status appeared in three positions, i.e., about 0.16 to 0.66 ms, 4.04 to 17.34 ms, and 30.30 to 127.31 ms, respectively (Table 4, Figure 2A). The water was termed as tightly bound water (TBW), softly bound water (SBW), and free water (FW), which was similar to the results of previous studies [15,16].

Morphology
The morphology of fracture surfaces of uncooked and cooked FBN and DBN shown in Figure 3. For uncooked noodles, DBN-80% (Figure 3f)  For noodles with the same ratio of extruded buckwheat flour, the T 2 for uncooked FBN and DBN was significantly different (Table 4) (p < 0.05) and the T 21 and T 22 of FBN were smaller than those of DBN, indicating tight bonding among water molecules (TBW and SBW) and non-water substances within FBN. However, the T 23 of DBN was smaller, indicating that drying induced a stronger interaction between the free water and solids [17]. However, FBN and DBN displayed different trends with varying amounts of extruded buckwheat flour in terms of T 21 , T 22 , and T 23 . T 21 of FBN decreased from 0.21 ms to 0.16 ms, whereas that of DBN increased with a higher addition ratio. T 22 of FBN-20% was insignificant to that of FBN-50% and similar for DBN with 20% and 50% extruded buckwheat flour. However, as the amounts of extruded buckwheat flour up to 80%, FBN showed a downward T 22 , while DBN showed an upward T 22 . T 23 of FBN and DBN increased with increased proportions of extruded buckwheat flour.
The water within FBN was mainly in the form of SBW, while the proportion of TBW was the highest in DBN ( Figure 2B). As the main form of water, the proportion of SBW (A 22 ) of FBN decreased from 81.60 to 72.51% (Table S2) with the increasing ratios of extruded buckwheat flour from 20% to 80%. In comparison, A 22 of DBN remained unchanged (about 3.70%). The ratio of TBW (A 21 ) of DBN decreased with increased proportions of extruded buckwheat flour (93.58-89.86-87.65%), whereas the A 21 of FBN increased from 18.03% to 24.88%. A 23 of DBN showed a significant increase from 2.78 to 8.50% with increased ratios of extruded buckwheat flour, higher than those of FBN. There was no simple pattern with the T 2 and the A 2 under a different status. The relation between the water mobility and noodle quality will be discussed in the discussion section.

Morphology
The morphology of fracture surfaces of uncooked and cooked FBN and DBN is shown in Figure 3. For uncooked noodles, DBN-80% (Figure 3f) showed no granular structure, where the extruded buckwheat flour formed a continuous phase similar to the protein matrix, which improved the internal network [3,18]. After drying treatment, the wheat starch granules were less exposed to the surface in DBN, and the internal structure was denser for DBN (Figure 3a-f). After cooking, the integral starch granules swelled and lost their granular structure (Figure 3m-r). The penetration of water created holes with different sizes and numbers which were finer from the outside than the inside of the cooked DBN, while the fracture surface of FBN was looser and the holes were more homogeneously distributed (Figure 3g-l). Additionally, the number of irregular holes was less for noodles with increased ratios of extruded buckwheat flour. DBN-20% and DBN-50% showed denser and less porous internal structure than the cooked FBN with the same proportion of extruded buckwheat flour (Figure 3m

Discussion
Drying is the main processing procedure that is different for DBN and FBN; it changes the water status and distribution of noodles. It is well known that moisture is one of the reasons for noodles'e different eating qualities. However, the moisture content of cooked DBN and FBN were not significantly different (Table 4), although the OCT was different for DBN and FBN in this study (Table 1). On the other side, longer cooking time may result in a higher CL and CBR, as more degraded starch molecules due to extrusion may dissolve in the cooking water. Besides, processing can also induce varied water mobility and food structure, which can also affect the eating qualities. The water channels due to water migration and denser network due to moisture loss may influence the structure of the noodles [19]. For fresh buckwheat noodles with different contents of extruded buckwheat flour, the viscous extruded buckwheat flour together with the wheat gluten may create varied network structure, contributing to buckwheat noodles with a higher proportion of buckwheat flour. To verify the above inference, the correlation between the cooking properties and eating properties of buckwheat noodles with the features of water of different statuses was performed (Table 5).

Roles of Water Status/Distribution and Internal Structure on the Cooking Properties of FBN and DBN
CBR was correlated well with the T 22 of uncooked noodles and the moisture content after cooking (p < 0.05) ( Table 5). T 22 showed a negative correlation with the CBR of FBN, but a positive correlation with the CBR of DBN. The higher T 22 of FBN was correlated with a lower CBR. A possible reason for this is that the SBW (the water that existed inside the gluten network [20]) tended to be more mobile, and highly mobile water can lead to a more uniform and developed protein network, leading to a lower breakage ratio. This uniform inner structure was confirmed by the SEM images (Figure 3a-c). In comparison, SBW was not the main form of water within DBN, and the longer T 22 implied that the SBW could not be bound tightly in the structure of dried noodles (Table 4). DBN with a higher ratio of extruded buckwheat flour showed a longer T 22 and a less uniform but denser internal structure, leading to a larger CBR. Thus, FBN-20% with a relatively more homogenous structure (Figure 3m) showed a higher resistance to breakage during cooking. The moisture content was negatively correlated with the CBR of FBN (Table 5), which was mainly due to the loose structure of FBN-80% (with the lowest moisture content).
The CL of FBN showed a positive relationship with the A 23 (Figures 1 and 2B). However, this is more likely to be due to the dissolution of the degraded amylopectin molecules [21] from extruded buckwheat flour during noodle cooking. For FBN with a higher ratio of extruded buckwheat flour, the small molecules from the extruded buckwheat flour could be released more easily into the cooking soup [22], causing the greater CL of FBN. Moreover, a weak or discontinuous protein matrix results in a protein network that is too loose and permits a large amount of extrudate to dissolve in the water during cooking [23]. DBN-80% and FBN-80% showed denser but less uniform internal structures (more holes), which explained the higher cooking loss (Figures 1 and 3). Chillo, Laverse, Falcone, Protopapa, and Del Nobile [13] also found that an internal structure with more or fewer holes inhibited or improved the absorption of water, which contributed to the breakage susceptibility and cooking resistance of the noodles. However, further evidence is required to justify if it is a coincidence that the CL was correlated to the A 23 or if the CL could be attributed to other parameters relating to the water status.

Roles of Water Status/Distribution and Internal Structure on the Textural Properties of FBN and DBN
Water mobility (Table 5) and inner structure shall be combined to explain the changes in textural properties [19]. In this study, T 21 showed a positive correlation with the hardness of FBN, whereas T 23 showed a negative correlation with the hardness of DBN. Both the FBN with a lower ratio of extruded buckwheat flour (longer T 21 ) and the DBN with a lower percentage of extruded buckwheat flour (shorter T 23 ) exhibited a more uniform but less dense protein network (Figure 3g,j,m,p), contributing to the higher hardness. The more compact internal structure led to a higher hardness, which was consistent with the study [4]. Additionally, a more continuous and denser internal structure was found for DBN-20%, with more tightly bound water and free water.
The elasticity of DBN was negatively correlated with A 21 but positively correlated with A 23 . The reason for the changes in the elasticity of DBN and FBN is different. Firstly, the elasticity of FBN was not significantly different, and there was no correlation found between the elasticity and the parameters reflecting water status changes for FBN. Secondly, as suggested by the improvement in rubber elasticity using linear or branched polymers [24], a possible presumption is that a more degraded branched/bulk structure was maintained in DBN-20% and DBN-80%, and thus a higher proportion of free water (higher A 23 ) was retained in the bulk structure. Meanwhile, the polymers formed from a higher content of extruded buckwheat flour may entangle with each other and/or with the proteins within the buckwheat noodles during drying, which caused increased elasticity. Without the drying treatment, FBN delivered a relatively stable elasticity (about 1.0) compared to DBN (about 0.7-1.0).
The cohesiveness of the noodles is of particular interest to us, and it can reflect the integrity of the soft food's structure. For FBN, the cohesiveness was negatively correlated with the T 21, A 21 , and A 22 , but positively correlated with the T 23 and A 23 . Interestingly, a shorter T 21 , smaller A 21 and A 22 , longer T 23 , and larger A 23 were observed in FBN with a higher ratio of extruded buckwheat flour. The higher addition of extruded buckwheat flour could stick different components, including starch granules and protein, together, as indicated previously [3], leading to an improved internal structure. Besides this, the moisture content was negatively correlated with the resilience of cooked DBN (Table 5), which explained the lower resilience of DBN-80% (with a higher moisture content) after cooking.
From this study, the water distribution/status cannot solely explain the changes in the cooking or eating properties without understanding the changes in the inner structure of noodles. The relationship between the water and the texture extensively relied on the specific food system, as the form of water taking the dominant proportion might be different. Since the water distribution/status correlated well with the texture of seafood [25], if a similar relationship can be found between the water distribution/status and the internal structure, then the texture of other starch-based foods shall be further confirmed. Besides this, a joint modification of water mobility and inner structure shall be considered for the improvement of food qualities. The water absorption ability of extruded buckwheat flour and gluten still merits further investigation to help explain the water mobility within the buckwheat noodle system.

Conclusions
The cooking and eating differences between FBNs and DBNs are reflected differently in their water status and inner structure. FBN exhibited more desirable cooking qualities (lower cooking loss and breakage ratio) and a more elastic texture than DBN. The highly mobile softly bound water of FBN helps gluten network formation and results in a lower breakage ratio. The denser inner structure of DBN results in a higher hardness, and the increased cooking loss of DBN was mainly due to the smaller molecules from the extruded buckwheat flour and the less uniform and looser internal structure of the noodles. The elasticity is more likely to be attributed to the bulky structure or entanglement with the branched polymers.

Conflicts of Interest:
The authors declare that there is no conflict of interest regarding the publication of this paper.

FBN
fresh buckwheat noodles DBN dried buckwheat noodles TBW tightly bound water SBW Softly bound water FW free water T 21 the transverse relaxation time of tightly bound water T 22 softly bound water T 23 free water A 21 the proportion of tightly bound water A 22 softly bound water A 23 free water OCT optimal cooking time CL cooking loss CBR cooking breakage ratio MC moisture content