The influence of extruded flour on water content and retrogradation process in muffins during storage: NMR relaxation study

Abstract

A comparative study of the moisture content, mobility of water protons and retrogradation process in muffin samples with the addition of extruded (ME) flour (10%) during storage (14 days) was carried out and compared to control samples without the addition (MO). A significant increase in water content during swelling of the extruded flour was found. The proton magnetic relaxation signal for both ME and MO showed a two-component character for the fast-decay region (free induction decay, FID experiment) and a three-component character for the slow-decay (CPMG experiment) one. The relaxation times T₂ characterizing the protons of both water and non-aqueous fraction decreased during the experiment except for the protons of the fat fraction. The values of relaxation times T₂ for ME exceeded the values for MO, which reflects the increased water content in ME. The amplitude of the signal of water protons (two components of CPMG) decreased for ME more slowly (by about 3–5 days), which can serve as a characterization the shelf life of the muffins. Estimation of the retrogradation rate according to the Avrami model showed a lower value of the rate for ME compared to MO, which may reflect the increased moisture content in ME.

Introduction

Extruded wheat flour is obtained by intensive action of high temperature, pressure, and shear stress on wheat grain, which leads to various changes of wheat components is widely used in the production of bread and bakery products. When storing bakery products, including cupcakes and muffins, numerous physical and chemical processes occur. As pointed out by several bread staling models (Zobel and Kulp 1996) the starch recrystallization (retrogradation), water loss and redistribution and storage conditions are the most important aspects that characterize the staling.

The transition of up to 30% of starch polysaccharides in bread from an amorphous to a crystalline state during retrogradation (Zobel 1988) characterizes its consistency. Numerous works (Baik and Chinachoti 2001; Shiraldi and Fessas 2001) mark the role of water in the storage of bread not only at the macroscopic level, when it passes from the crumb to the crust, but also at the molecular level, which manifests itself in changes its mobility.

Assessing the research methods of the retrogradation process, the authors (Abd Karim et al. 2000) note the advantages of an integrated approach to this task. Using DSC, NMR, rheological methods and X-ray diffraction method, the authors (Bosmans et al. 2013) established a linear relationship between the content of solid phase protons in the starch–water system and the melting enthalpy of the crystal phase. For bread samples, this dependence was established for the spin–spin relaxation time T₂ of mobile protons and rheological characteristics. Comparison of DSC data and X-ray diffraction allowed to estimate the amount of non-freezing water associated with the B-form of starch in retrogradation. A significant number studies of the water state in food products were conducted by NMR relaxation methods (Le Boltan et al. 1998; Ruan et al. 1999; Assifaoui et al. 2006; van Nieuwenhuijzen et al. 2010; Curti et al. 2011; Bosmans et al. 2013). This method allows one to study the mobility of certain nuclei (in our case, protons) by measuring their spin–lattice (T₁) and spin–spin (T₂) relaxation times and to determine the number of nuclei according their mobility.

For the so-called fast-relaxing protons (microsecond region), free induction decay (FID) of the induced signal is investigated. The FID of gelatinized starch and bread during storage was widely studied (Farhat et al. 2000; Curti et al. 2011). The authors interpreted the observed increase in the immobilization degree of fast-relaxing protons (decrease in T₂) as a consequence of a decrease in the mobility of the molecular structure of bread, which occurs both due to recrystallization of amylopectin and due to water redistribution. In addition to calorimetric, rheological, x-ray diffraction, optical methods, a lot of studies of recrystallization (retrogradation) of starch used NMR relaxation methods (Teo and Seow 1992; Farhat et al. 2000). These methods are based on isolating the proton signal of the crystalline part of amylopectin and observing changes in this signal during storage.

The study of slow-relaxing protons (millisecond regions) is usually performed using the multi-pulse sequence Carr-Purcell-Meiboom-Gill (CPMG) (Meiboom and Gill 1958). Many researchers (Tang et al. 2000; Choi and Kerr 2003; Bosmans et al. 2013) note that the relaxation of protons in complex systems, such as food, has a multicomponent character, in which each component characterizes the corresponding fraction of protons and the diffusion exchange between different fractions in the system.

By studying the crispness of model bread, the authors (van Nieuwenhuijzen et al. 2010) note the greater accuracy of the NMR relaxation method in detecting the point of transition from a glassy to a more ductile state compared to the methods of DSC and phase transition analysis (PTA). The benefits of NMR-relaxation should be attributed to non-destructiveness, which allows to study the process of retrogradation in the vial of the NMR device (Bosmans et al. 2013). The presence in the magnetization decay the signals from all protons allows simultaneously to study the mobility of protons in solid- and liquid-like phases. This explains our attempt NMR investigation of the changes in the state of water and retrogradation in muffins during storage. The result of the work highlights the opportunity storage time estimation since shelf life matters.

Materials and methods

Materials

Samples of muffins were provided by IREKS LLC. The main components of the recipe for 119 g of the dough were: baking wheat flour—40 g, sugar—20 g, butter—16 g, whole milk—28 g, egg mélange—12 g, moisture content—36%.

In experimental samples of muffins, 10% of flour was replaced with extruded flour. This composition is the result of an optimization study (Balaeva 2013). Extruded flour was obtained at the KMZ-2U extruder with a raw material humidity of 13.7%. Moisture content was determined by drying (24 h at 105 °C).

NMR measurements

All studied samples of muffins weighing 0.4 g, taken from the central part of the product, were placed in a teflon ampoule with a sealed lid and the measurements were carried out on a spectrometer. Samples for retrogradation measurements were taken from the same muffin (3 samples every day) during all storage period (14 days). After the measurements, the water content was determined in the samples.

NMR studies were performed using three methods. Using the FID method, we studied less mobile protons in the decay range of 10–100 µs, the CPMG method—more mobile protons in the range of 0.5–75 ms, and the EDM-110A program was used to estimate the signal amplitude of protons in the crystal phase. Measurements were carried out on a spectrometer MINISPEC PC120 (Brucker, Germany) with an operating frequency 20 MHz.

The experimental conditions were:

FID: 90°pulse duration—2.5 µ, the decay digitization frequency—1 MHz, the repetition time—3 s, the number of scans—16.

CPMG: 90°pulse duration—2.5 µ, the number of experimental points on the curve—1500, the number of scans—16, the time between 90° and 180° pulses—25 µs.

EDM 110A program: the amplitude protons was measured at 11 µ (L1), and at 70 µ (L2).

The obtained magnetization decay curves were analyzed using the ORIGIN 9 program. A discrete multi-exponential model was used for mobile protons and the sum of Gauss exponents for solid-phase protons. All measurements were made in triplicate and results were reported as an average of the three data set. The standard deviation did not exceed 6%.

Result and discussion

Water in muffins

Changes in the moisture content of wheat flour muffins (MO) and muffins with 10% addition of extruded flour (ME) are shown in Fig. 1a

Fig. 1.

Changes in the moisture content (a), short component (FID) relaxation time T₂ (b), long component (FID) contribution (c) and long component (CPMG experiment) contribution (d) during storage of the muffins. Muffins with addition of extruded flour (ME) and without (MO)

The moisture content for muffins ME was higher (about 5%) than wheat flour muffins (MO) during all 14 days of experiments. It reflects the high water-holding capacity of muffins with extruded flour addition. Moisture loss is one of the factors that result in bread staling (Sereno et al. 2007) and studying the molecular mobility of both water and non-water protons can help us understand the processes involved.

Two methods were used to study muffin samples by NMR relaxation: single-pulse-FID and multi-pulse-CPMG. FID characterize the protons of the so-called “solid” phase. Several studies in literature (Farhat et al. 2000; Sereno et al. 2007) which studied the storage of corn starch and bread associate these changes with a decrease in the structure mobility of bread due to recrystallization of amylopectin and loss/redistribution of moisture in the bread crumb.

The change in the T₂ relaxation times of solid-phase protons and the relative content of mobile-phase protons obtained by FID method for muffin samples during storage are shown in Fig. 1b,c.

A decrease in the relaxation time (from ~ 14 µs till ~ 7 µs) characterizes a decrease in the mobility of solid-phase protons. According to the authors (Choi and Kerr 2003) such changes can be explained by the plasticizing role of water in starch- water system. The decrease in the relative content of mobile phase protons muffin samples reflects the process of moisture loss and redistribution during storage (Fig. 1c). It should be noted that the relaxation time of the T₂ solid component for ME is always longer compared to MO, which indicates a more mobile state of protons in these samples. Since the relaxation decay of mobile protons reflects the molecular movement of several fractions such as water, fat fraction, sugars solutions etc., the water proton fraction should be separated to characterize moisture loss.

Using a multi-pulse CPMG sequence, the mobile phase proton magnetization for MO and ME were studied. Decomposition of the decay curves showed the existence of 3 fractions of mobile protons in the first 4–5 days of storage, and then these curves were the sum of 2 components. In MO samples, the first two components with relaxation times of T₂ (1)≈1–2 ms and T₂ (2)≈6–8 ms decreased their percentage over time and after 5 days they joined to 1 component (T₂≈1.8 ms), which disappeared at 10th day. The third component with T₂≈120–140 ms existed during the entire storage time, constantly increasing its relative content. Obviously, this component characterizes the fat fraction. The decrease in the relative content of the first two components obviously reflects their belonging to water. NMR investigation of biscuit dough (Assifaoui et al.) showed the similar results. In CPMG experiment long decay component corresponded to the fat fraction and two intermediate components were the intra- and inter-granular protons, which were more sensitive to moisture content. Investigating the change in the water content of bread during staling, the authors (Curti et al. 2011) found 3 components: T₂ (A)≈0.09–4 ms,T₂ (B)≈6–20 ms and T₂ (C)≈100 ms where the A fraction belongs to water. Thus, the obtained two components from the total number of mobile protons reflect the state of water protons in our experiment. Changes in the total relative contribution of these fractions in the process of storing muffins are shown in Fig. 1d. Up to 5 days, the content of water protons in the MO and ME samples is approximately the same (with a slight excess of ME over MO). The absolute value of the moisture content at this time (5 days), as seen in Fig. 1a is about 0.17 g of water/g of dry weight. In work (Le Boltan et al. 1998), the authors indicate that this value corresponds to a single molecular layer of water in the starch–water mixture, and a further decrease in the moisture content occurs due to the loss of bound water. On the ninth day the water protons content in the MO was below the limit of measurement, whereas on the fourteenth day this value was about 15% for ME. The same values of water protons content for ME were observed on average 3–4 days later than for MO, it may characterize of more slow process of staling for ME samples and its shelf life increasing.

Retrogradation in muffins

The process of studying starch retrogradation by NMR relaxation is based on the fact that the magnetization signal after 90° pulse reflects a two-component proton system—solid-like and liquid-like decays. Solid-like is the signal from CH, CH₂, –OH immobilized protons of the crystalline and amorphous part of starch with a relaxation time of the order of 10–20 µs. Liquid-like is of CH protons, CH₂ polysaccharides in mobile state (for example, after gelatinization), H₂O, chemically exchanged protons. The relaxation times of these two fractions differ significantly (1–2 orders of magnitude), which makes it possible to determine their content. Evaluation of FID signals in Fig. 2 showed that the fast-decay signal does not affect in the area with a time greater than 70 µs. This indicates that a signal in this region is only from the slow-relaxing protons. The total signal from protons is equal to the signal immediately after 90° of the pulse, but its measurement is technically impossible due to the characteristic “dead time” of the recording system, which is approximately 10 µs. The L₁ signal measured at τ = 11 µs is proportional to the total number of protons, while the L₂ signal measured at τ = 70 µs is proportional to the mobile phase protons. Their difference (L₁–L₂) is proportional to the number of protons of the “solid-like” phase.

Fig. 2.

Avrami function dependence for describing the retrogradation process in muffins with the addition of extruded flour (ME) and without the addition (MO)

The phenomenon of retrogradation reflects the transition of a part of the mobile amorphous phase of starch obtained during gelatinization to a crystalline state. Crystals are formed from centers of crystallization, their number at a given time is a function of the nucleation density and the rate of crystallization. The change in L = (L₁–L₂)/L₁, which characterizes the normalized intensity of crystalline protons, reflects the process of retrogradation.

Several authors (Teo and Seow 1992; Farhat et al. 2000) used the Avrami (1941) equation to describe the kinetics of retrogradation.

$$U=e^{(-k{t}^n)}$$ 1

where U is the amount of non-crystalline phase, k is an overall rate constant that reflexes both nucleation and growth process and n is a dimensionless parameter, which depends on nucleation kinetics and the dimensionality of growing crystals. The retrogradation process has the non-equilibrium nature and that is why this equation does not describe recrystallization kinetics carefully, but it is a practical way to quantify this process in starch systems.

The equation is given as below

$$ln\left(U\right)=ln\left[\frac{\left(L_{\infty }-L_t\right)}{\left(L_{\infty }-L_o\right)}\right]=-k{t}^n$$ 2

where L_∞ is the relative signal of the crystal proton fraction at the end of the storage period, L_t is the relative signal of the crystal protons at time t, and L₀ is the relative signal of the crystal protons at the beginning of the experiment (in our case, 1 day). The parameters k and n can be obtained from the intercept and the slope of linear dependence of log[-lnU] on log (t) Fig. 2 shows this type of relationship for (ME) and (MO) muffin samples.

The values calculated from these dependencies for ME samples k = 0.27 ± 0.07 day⁻ⁿ, n = 0.3 ± 0.01 (R² = 0.978) and for MO samples k = 0.55 ± 0.03 day⁻ⁿ, n = 0.32 ± 0.01 (R² = 0.989). Teo and Seow (1992) demonstrated that the decrease in the parameter k in of starch-containing systems reflects a decrease in the rate of retrogradation during storage. Consequently, lower k values for ME samples reflect slower retrogradation compared to MO samples. As Farhat et al. (2000) showed, the rate of retrogradation obtained from the dependence of the water protons relaxation rate 1/T₂ in the Avrami equation decreases with increasing water content in the starch–water system. Therefore, the increased water content in ME samples can serve as an explanation the decrease of the retrogradation rate.

Conclusions

Unfortunately, we did not find in the literature data on the study of muffins by DSC and X-ray diffraction. However, comparison of our data with studies of this kind for other systems such as bread or starch (Bosmans et al. 2013; van Nieuwenhuijzen et al. 2010; Curti et al. 2011) allows us to draw certain conclusions.

Enriched extruded flour muffins (ME) demonstrated an increase of ¹H molecular mobility as compared with the control samples (MO). These increases were found in relaxation times T₂ of solid-like protons, in more mobile proton system which reflects water state. The excess of proton mobility for MO samples saved for all period of storage (14 days). The sum of magnetization signals, which correlated with moisture content, can serve as an indicator of shelf life increasing. The retrogradation rate was evaluated by Avrami model during the storage. It was found that its values for ME were less than for MO and the higher degree of moisture content may be one of the reasons of such decrease. Reducing the rate of retrogradation for extruded samples also indicates an increase in the storage time of muffins. Our studies confirm the rheological data (Balaeva 2013) on the increase in the shelf life of muffins when adding 10% extruded wheat flour.