Mechanisms of the different effects of sucrose, glucose, fructose, and a glucose–fructose mixture on wheat starch gelatinization, pasting, and retrogradation

Abstract

Abstract

The gelatinization, pasting, and retrogradation of starch influence texture, quality, and shelf‐life attributes of many foods. The purpose of this work was to document the effects of a 50:50 glucose:fructose (glc:fru) mixture and sucrose solutions on these starch traits to provide a fundamental basis to explain the different texture and shelf‐life attributes of baked goods formulated with these sugars. Differential scanning calorimetry, rapid visco analyzer, and oscillatory rheometry were used to quantify the effects of glucose, fructose, glc:fru mixture, and sucrose at different concentrations (0% to 60% w/w), on the gelatinization temperature, pasting, and retrogradation properties of wheat starch. Distinct differences were found between the effects of sucrose and those of the monosaccharides including the glc:fru mixture. Sucrose elevated T _gel and pasting temperature most and decreased other RVA parameters compared to the monosaccharides as concentration increased. Fructose and the glc:fru mixture promoted amylopectin retrogradation, while retrogradation was inhibited in sucrose and glucose solutions. The glc:fru mixture had similar effects on starch properties compared to fructose under static measurement conditions (DSC), and the effects were in between those of glucose and fructose under dynamic conditions when shear was applied (RVA and rheology). These effects are explained by the phase separation and/or solute partitioning of the monosaccharide constituents of the glc:fru mixture. Sugar solution physicochemical properties correlated strongly with starch gelatinization and retrogradation. The results substantiate the important relationship between sugar physicochemical properties and solution dynamics with starch thermal properties, which in turn affect the texture and structure of starch‐containing food products.

Practical Application

The quality attributes of starch‐containing baked goods are influenced by how different amounts and types of sugars affect starch cooking properties. The underlying mechanisms of the different sugar effects involve solution viscosity, intermolecular hydrogen bonding, and phase separation. Substituting one sugar for another has less effect on these starch properties in products with lower sugar concentrations than in products with more sugar. Mixtures of sugars behave differently than single sugars in different conditions due to phase separation. Baked goods made with glucose:fructose mixtures in place of sucrose likely have higher amounts of gelatinized starch and increased firmness (i.e., staling or retrogradation) over time.

1. INTRODUCTION

Wheat starch is the predominant component in cookie and pastry flour, and controlling its thermal properties (gelatinization, pasting, and retrogradation) is vital for creating baked goods with desirable textural properties (Kweon et al., 2014; Woodbury et al., 2021). It is well understood that sugars alter the thermal properties of starch, and the effects of single sugars on starch gelatinization, pasting, and retrogradation have been documented (Allan et al., 2018; Deffenbaugh & Walker, 1989; Gunaratne et al., 2007; Hoover & Senanayake, 1996). Different starch‐containing baked goods formulations vary in both water and sugar contents (Table 1); thus, it is important to examine a range of sugar concentrations to better understand starch functionality in different products. Here, again, the effects of single sugars across a range of concentrations on the thermal properties of starch have been documented (Allan et al., 2018; Renzetti et al., 2021; van der Sman & Mauer, 2019; Woodbury & Mauer, 2022; Woodbury et al., 2022, 2021). However, mixtures of sugars may be used in food products (Renzetti et al., 2021; Torley & van der Molen, 2005), and the quantitative and mechanistic differences in how mixtures of sugars affect starch properties compared to single sugars, across a range of concentrations (including the saturated solution conditions representative of low moisture products), have not received as much attention.

TABLE 1.

The solution concentration of sugars in various baked good formulations

Sugar solution concentration (% w/w)	Product examples	Sources
∼15	White bread, muffins	Cauvain and Young (2006)
∼30	Pastries, doughnuts	Cauvain and Young (2006)
∼45	Cakes	AACC International (2000)
∼60 and higher	Cookies	AACC International (2000)

An equimolar mixture of glucose and fructose is similar in composition to invert sugar after sucrose has been completely hydrolyzed by invertase. Compared to sucrose, invert sugar or a 50:50 mixture of glucose and fructose (glc:fru) is sweeter; caramelizes at lower temperatures; has a lower glass transition temperature (T _g), a lower solution viscosity, a higher aqueous solubility, and a lower deliquescence point; and absorbs and retains more moisture (Hartel et al., 2018; Saavedra‐Leos et al., 2012; Salameh & Taylor, 2006; van der Sman, 2017). Although invert sugar is most commonly used in confections, a glc:fru mixture may also be used in starch‐containing baked goods to create softer textures, boost sweetness, minimize surface cracking, and increase surface browning more than sucrose alone due to the reducing nature of glucose and fructose (Doescher & Hoseney, 1985; Manohar & Rao, 1997). Sucrose is also known to decompose to free glucose and fructose during the processing of baked goods, impacting product color and flavor attributes (Ameur et al., 2007). High fructose corn syrup (HFCS) is commonly used in baked goods and contains ratios of fructose to glucose (42% to 55% fructose) near to that of an equimolar mixture of glc:fru, but has other larger saccharides which impact functionality (Zargaraan et al., 2016). Given the presence of glc:fru mixtures in food products for which starch functionality is important, it is surprising that little information is available on how different concentrations of this monosaccharide mixture affect starch gelatinization, pasting, and retrogradation.

The objectives of this study were to quantify the effects of a 50:50 glc:fru mixture at different concentrations on the gelatinization, pasting, and retrogradation properties of wheat starch, and to compare its effects to those of its component monosaccharides in free form (glucose and fructose) and when covalently bound in a glycosidic linkage (sucrose). The physicochemical properties of sugars have been shown to have a direct influence on starch thermal properties (Allan et al., 2018; Renzetti et al., 2021; van der Sman & Mauer, 2019; Woodbury et al., 2022). Therefore, this study compared the physiochemical traits of the sugars (molecular weight, dry T _g, the number of hydrogen bonding sites effectively available within the sugar molar volume [N _OH,s /v _s]) and sugar solutions (water activity, viscosity, and the volumetric density of hydrogen bonds in the sugar solutions [Φ _w,eff]) with their effects on starch thermal properties (gelatinization, pasting, and retrogradation). Results were used to determine which sugar traits best explained starch response and how a glc:fru mixture affected starch compared to its constituent monosaccharides.

2. MATERIALS AND METHODS

2.1. Materials

The native wheat starch used in this study was Aytex^® P from ADM (Minneapolis, MN, USA) with the following composition: 25% amylose, 9.9% moisture, <0.2% ash, <0.2% protein, and <0.1% fat. The sugars investigated were analytical grade glucose and fructose from Acros Organics (Fair Lawn, NJ, USA), and sucrose from Mallinckrodt Chemicals (Phillipsburg, NJ, USA). These sugars were chosen to compare the effects of glucose and fructose by themselves, when combined in a 50:50 glc:fru mixture (similar in composition to invert sugar but composed of the same monosaccharides studied individually to facilitate direct comparisons of results), and when bound through a glycosidic linkage (sucrose) on wheat starch thermal properties responsible for the texture characteristics of low‐moisture baked goods. All of the water used to prepare samples in this study was ultrafiltered by a Barnstead E‐Pure Lab Water System (Dubuque, IA, USA) to >17.4 MΩ cm.

2.2. Sugar solution preparation

The sugar solutions examined were made on a % w/w dry basis and encompassed 0%, 15%, 30%, 45%, and 60% sugar concentrations, with the exception of glucose for which the highest concentration achieved was 45% due to solubility limits (≈50.5%) in water at 25°C (Alves et al., 2007). The sugar solutions were prepared in 50 ml centrifuge tubes and agitated with an HT Mini vortexer (OPS Diagnostics, Lebanon, NJ, USA) and a Roto‐Shake Genie (Bohemia, NY) until crystals were no longer evident upon visual examination. The solutions with higher solids contents (45% and 60%) were briefly placed on a heating block (<5 min) heated to 80°C to aid sugar dissolution during mixing. Once the sugar solutions were cooled and mixed sufficiently, they were either used immediately for wheat starch thermal property experiments (for most experiments) or stored in the refrigerator (at 4°C). Solutions stored at 4°C were briefly (1 to 2 min) placed on a heating block to ensure no crystals were present prior to use.

2.3. Dry sugar properties

The dry glass transition temperature (T _g) values of the pure individual sugars were previously determined and reported by Allan et al. (2018) using a PerkinElmer differential scanning calorimeter (DSC) 4000 (Waltham, MA) and a melt–quench method established by Finegold et al. (1989). The dry T_g of a 50:50 glc:fru mixture was calculated using the Gordon–Taylor equation for a binary mixture of sugars from Hancock and Zografi (1994):

$$T_{\mathrm{g}}=\frac{w_1{T}_{\mathrm{g}1}+w_{2}k{w}_2{T}_{\mathrm{g}2}}{w_1+k{w}_2}$$ (1)

where T _g is the glass transition temperature of the mixture, T _g1 and w ₁ are the dry T _g and weight fraction of component 1, respectively, and T _g2 and w ₂ for component 2. The fitting constant (k) was 1.33, based on experimental data from Arvanitoyannis et al. (1993).

N _OH,s and N _OH,w are the number of hydroxyl groups per molecule available for intermolecular hydrogen bonding for the solute/sugar (s) and water (w), respectively (Table 2). The N _OH,s value for each sugar was obtained from the following relationship established by van der Sman (2013):

$$\frac{1}{2}-\frac{1}{N_{\mathrm{OH},\mathrm{s}}}=\frac{1}{2}\frac{T_{\mathrm{g},\mathrm{s}}-T_{\mathrm{g},\mathrm{w}}}{T_{\mathrm{g},\infty }-T_{\mathrm{g},\mathrm{w}}}$$ (2)

The T _g,s was the dry T _g of the solute; the T _g,w was that of water (−134.15°C); and the T _g,∞ that of starch (201.85°C) all of which were obtained from previous publications or calculated using the Gordon–Taylor equation (Allan et al., 2018; Arvanitoyannis et al., 1993; van der Sman & Mauer, 2019). The number of equatorial hydroxyl groups within a molar volume (N _OH,s /v _s) of a sugar was shown to have a curvilinear relationship with starch pasting properties by Renzetti et al. (2021), and the N _OH,s /v _s for the sugars examined in this study are reported in Table 2.

TABLE 2.

The size and dry ingredient properties of sugars

	Dry ingredient properties
Sugars	M w (g/mol)	ρi (kg/m3)1	Dry T g onset (°C)2,3	N OH,s 1	N OH,s/v s (1000 × mol/cm3)1
Water	18	1000	—	2.00	111.11
Glucose	180	1540	38.30 ± 0.01B	4.11	35.15
Fructose	180	1540	15.16 ± 0.11C	3.60	30.80
Glc:Fru	180	1540	25.09*	3.80	32.53
Sucrose	342	1550	59.36 ± 0.56A	4.72	21.37

Uppercase letters indicate statistically significant differences (α = 0.05) between sugar types.

Abbreviation: Glc:Fru, 50:50 glucose:fructose mixture.

^*The dry T _g value for glc:fru was calculated using the Gordon–Taylor equation based on experimental data from Arvanitoyannis et al. (1993).

Sources: 1. van der Sman and Mauer (2019), 2. Experimental values for glucose, fructose, and sucrose from Allan et al. (2018).

2.4. Sugar solution properties

The water activity (a _w) values of all sugar solutions were measured in triplicate at 25°C with an AquaLab 4TE water activity meter (METER Group, Pullman, WA, USA) following the manufacturer's instructions. The calibration of the a _w meter was verified each week using the standard a _w salt solutions provided by the manufacturer with the following a _w values: 0.760, 0.920, 0.984, and 1.000.

The dynamic viscosity of sugar solutions was measured in triplicate at 25°C following a method from Allan et al. (2018) utilizing capillary viscometers (sizes 75 and 200) from Fungilab Cannon‐Fenske (Barcelona, ES). The viscometers were clamped onto a burette stand, leveled, and partially immersed in a temperature‐controlled water bath set to 25°C. The viscometers were primed by pipetting 3 ml of sugar solution into the viscometer and pushing the solution throughout the measuring portion of the viscometer using compressed air. Once primed, the priming solution was discarded and 5 ml of measuring solution was added. The viscosity of sugar solutions was determined using Equation (3) by multiplying the time (t) for the solution to travel between two marked points on the tube by the viscometer constant (k), which was 0.0091 for the size 75 viscometer and 0.012 for the size 200 viscometer, and the solution density (ρ).

$$\eta =tk\rho$$ (3)

The water activity, solution density, and solution viscosity were recorded in Table 3.

TABLE 3.

The concentrations and properties of sugar solutions

	Solution properties at 25°C
Sugars	Conc. (% w/w, db)	Solution density (g/ml)	Molarity	Water activity	Viscosity (mPa·s)
Water	0	0.9920	—	1.0045 ± 0.0035	1.293 ± 0.002
Glucose	15	1.0460	0.90	0.9808 ± 0.0009 Da	1.411 ± 0.002Ca
	30	1.1136	1.91	0.9560 ± 0.0005Cb	2.649 ± 0.005Bb
	45	1.1803	3.00	0.9138 ± 0.0005Bc	6.524 ± 0.016Bc
Fructose	15	1.0449	0.90	0.9828 ± 0.0003Ca	1.364 ± 0.001 Da
	30	1.1160	1.92	0.9578 ± 0.0003Bb	2.511 ± 0.003Db
	45	1.1926	3.06	0.9148 ± 0.0003Bc	5.637 ± 0.043Cc
	60	1.2638	4.48	0.8381 ± 0.0006Bd	22.68 ± 0.05Cd
Glc:Fru	15	1.0380	0.90	0.9843 ± 0.0008Ba	1.427 ± 0.002Ba
	30	1.1078	1.92	0.9585 ± 0.0005Bb	2.545 ± 0.003Cb
	45	1.1893	3.06	0.9148 ± 0.0010Bc	6.522 ± 0.006Bc
	60	1.2599	4.48	0.8375 ± 0.0002Bd	25.01 ± 0.08Bd
Sucrose	15	1.0410	0.45	0.9927 ± 0.0002Aa	1.467 ± 0.003Aa
	30	1.0913	0.96	0.9773 ± 0.0007Ab	2.920 ± 0.003Ab
	45	1.1807	1.54	0.9505 ± 0.0010Ac	8.150 ± 0.063Ac
	60	1.2388	2.13	0.8963 ± 0.0003Ad	46.18 ± 0.08Ad

Uppercase letters indicate statistically significant differences (α= 0.05) between sugar types for a given measurement at a certain concentration level (i.e., a _w of all sugar types at the 15% w/w concentration), whereas lowercase letters indicate statistically significant differences between solution concentrations of one sugar type for a given measurement (i.e., a _w of sucrose solutions at 15, 30, 45, and 60% w/w concentrations).

Abbreviation: Glc:Fru, 50:50 glucose:fructose mixture.

The effective volume fraction of the solvent (Φ _w,eff), previously correlated to the intermolecular hydrogen bond density, was calculated for individual sugar solutions using the following equation from van der Sman and Mauer (2019):

$${\varphi }_{\mathrm{w},\mathrm{eff}}={\varphi }_{\mathrm{w}}+{\varphi }_{\mathrm{s}}\frac{N_{\mathrm{OH},\mathrm{s}}}{N_{\mathrm{OH},\mathrm{w}}}\frac{v_{\mathrm{w}}}{v_{\mathrm{s}}}$$ (4)

The variable ϕ_i represents the volume fraction of compound i, and ν_i is the molar volume of compound i.

2.5. Effects of sugars on the gelatinization and retrogradation properties of wheat starch

The gelatinization temperature (T _gel) of wheat starch in the presence of water controls and sugar solutions was measured with a PerkinElmer DSC 4000 (Waltham, MA) using a method adapted from Allan et al. (2018). The samples were prepared by combining wheat starch in a 1:2 (w/w as is basis) ratio with DI water or sugar solution in a 1 ml centrifuge tube. The controls were prepared using water at the same 1:2 starch:solution ratio as the sugar solutions, as well as in starch:water ratios containing the same amount of water as each of the sugar solutions. The samples were vortexed to form slurries and allowed to equilibrate overnight at room temperature (∼23°C). The next day, the samples were vortexed again before pipetting 15 to 20 mg into a DSC pan which was then hermetically sealed and placed into the DSC cell along with an empty reference pan. Samples were heated from 10 to 120°C at a rate of 10°C/min, and the purge gas was N₂ with a flow rate of 20 ml/min. Pyris software was used to calculate the onset temperature (T _gel), peak temperature, area under the curve, and enthalpy (∆H) of each sample (based on the weight of the starch‐solution slurry). An indium reference was used to calibrate the DSC. The DSC analyses will be referred to as static measurements because the measurement involved heating the samples with no agitation.

The retrogradation behavior of wheat starch in the presence of sugars was also measured using DSC. Samples (1:2 starch:solution ratio) were initially prepared the same as they were for gelatinization measurements prior to placement in the DSC as well as the gelatinization profile. After the samples had been gelatinized, they were cooled down to 30°C at 40°C/min, and were either stored at 4°C for 7 days or immediately rescanned (day 0). The day 0 and day 7 samples were reheated in the DSC from 30°C up to 120°C at 10°C/min. Pyris software was used to calculate the area under the curve, enthalpy (∆H), onset temperature, and peak temperature. All analyses were done in triplicate, and the averages ± standard deviations were reported.

2.6. Pasting properties of sugar–starch slurries

The effects of sugar type and concentration on the pasting properties of wheat starch were assessed using a Newport Scientific RVA 4 Rapid Visco Analyzer (Newport Scientific, Warriewood, Australia). Each sample contained 2.5 g of wheat starch (9.9% moisture content) and 25.5 g of solution which were combined directly in an aluminum canister and stirred with the plastic paddle until the slurry (8% starch db) appeared homogeneous and no clumps were visible. This was done within 2 min of the start of each run to maintain equal contact time between the starch and the sugar solutions before the pasting process. The paddle was zeroed every day prior running samples, and the rapid visco analyzer (RVA) runs were conducted using the Standard 1 method. The paddle mixed the slurry at 960 RPM for the first 10 s and at 160 RPM for the remaining 13 min. The temperature profile settings were to hold at 50°C for the first min, increase to 95°C (at 4 min and 42 s), hold at 95°C (until 7 min and 12 s), cool to 50°C (until 11 min), and hold at 50°C until the end (13 min). Samples were analyzed in triplicate for each sugar concentration. Upon completion of an RVA profile, the pastes from the first and second replicates were then used for rheology measurements. The RVA analyses will be referred to as dynamic measurements because the samples were continuously stirred while viscosities were recorded during heating and cooling.

2.7. Rheological behavior of retrograded sugar–starch pastes

The rheological properties (storage modulus [GʹGʹ] and loss modulus [Gʹʹ]) of the RVA samples were determined using a DHR‐3 rheometer from TA Instruments (New Castle, DE, USA) with a 40.0 mm parallel plate geometry. Pastes from the RVA replicates were subdivided into four water activity cups, at approximately 2 ml per cup, and samples were analyzed immediately after cooling to room temperature (day 0) and after storage for 7 days at 4°C to determine sugar effects on short‐term amylose retrogradation (day 0) and longer‐term amylopectin retrogradation (day 7). The stored samples were sealed with a lid and surrounded by parafilm to prevent moisture loss during storage. The rheometer standardized all of the samples by flattening the paste to a 1 mm disk prior to measuring the storage and loss moduli. If the samples had formed gels, then they were gently scooped out of the cup with a metal spatula and placed onto the Peltier plate of the rheometer. The samples which did not gel were simply poured onto the Peltier plate and then analyzed.

The rheometer was calibrated with both rotational and oscillatory mapping at the beginning of each day prior to running samples. The gap was zeroed, and the sample was placed on the center of the Peltier plate. Next, the “trim gap” function was used to trim any excess edges from samples which formed a gel. The rheological properties of the trimmed sample were then measured using a frequency sweep at a controlled temperature of 25°C. The test strain was set to 0.5% ensuring all samples were measured within the previously determined linear viscoelastic region, and the angular frequency was set to 0.1 to 100.0 rad/s.

The Gʹ curves for all samples (excluding 60% sucrose) from day 0 and day 7 measurements were fitted with a power law model (Equation 5) to allow for statistical comparisons between different sample types (water or sugar solutions) and concentration effects on the rheological properties of starch pastes over time.

$$K^{\prime }=\frac{{\omega }^{n^{\prime }}}{G^{\prime }}$$ (5)

The proportionality constant Kʹ was the consistency index of starch pastes, nʹ the slope or flow behavior index, ω the angular frequency of the test (between 0.1 and 15.849 rad/s), and Gʹ the measured storage modulus (Yoo & Yoo, 2005). The Kʹ values were reported as the average ± standard deviation of at least three replicate samples for each day (0 and 7).

2.8. Statistical analysis

The significant differences between the effects of the dry sugar and sugar solution properties on a given starch thermal event (T _gel, PT, PV, BD, SB, FV, ΔH _ret,0, ΔH _ret,7, Kʹ_ret,0, Kʹ_ret,7) were determined using one‐way ANOVA with a Tukey post hoc test (α = 0.05). SAS version 9.4 (SAS Institute, Cary, NC) was used for the ANOVA statistical analyses. All graphs, tables, and figures were created in Microsoft Excel 365 (Redmond, WA). Linear multivariate correlations were performed with JMP Pro 15 (SAS Institute) using the Pearson product–moment correlation coefficient (r) to determine the statistical significance (α = 0.05), correlation strengths and direction (+/−) between sugar physicochemical properties (dry and solution) and observed effects on starch response variables. Correlation strengths were classified as negligible (r = 0.00–0.10), weak (r = 0.10–0.39), moderate (r = 0.4–0.69), strong (r = 0.7–0.89), and very strong (r = 0.9–1.0) (Schober et al., 2018).

3. RESULTS AND DISCUSSION

3.1. Effects of sugar type and concentration on the gelatinization temperature of wheat starch

The onset T _gel of wheat starch was measured in the presence of different sugar solutions and corresponding amounts of water, the controls (Figure 1). Increasing the sugar concentration significantly increased the T _gel regardless of sugar type (Figure 1, Table 4), although the different sugar types also had different effects on the extent of T _gel elevation (Figure 2). A strong correlation (r = 0.9348) was found between the onset T _gel of wheat starch and the % w/w sugar concentrations studied (Figure 3a). The T _gel of wheat starch in water was 60.71°C, altering the amount of water to be consistent with the amount of water in each concentration of sugar solution (the controls) had no significant effect on T _gel (ranging from 60.70 to 61.57°C), and in comparison the T _gel increased the most in a 60% solution of sucrose, up to 98.49°C (Table 4). Water plasticizes the amorphous regions of starch, initially decreasing the T _gel as moisture content increases (e.g., from 0% to 40% water); however, once a threshold moisture content is reached no further decreases in T _gel occur with further increases in moisture content (Donovan, 1979; Renzetti et al., 2021). The solutions containing sucrose consistently had higher T _gels than the other sugar solutions at the same concentration, even though at a given sugar concentration the amount of water would have been the same. For example, starch in a 30% sucrose solution had a T _gel of 72.23°C whereas starch in a 30% solution of the glc:fru mixture had a T _gel of 68.62°C. The solutions with the glc:fru mixture elevated the T _gel of starch similarly to pure solutions of fructose and glucose, except at the 45% concentration at which glucose elevated the T _gel (to 77.49°C) significantly more than fructose and the glc:fru mixture (75.96 and 76.91°C, respectively) (Table 4). These results differentiate T _gel elevation from the colligative properties of sugar solutions (e.g., boiling point elevation and freezing point depression) because more than the number of molecules in solution affects the extent of T _gel elevation, and sucrose elevates T _gel more than the monosaccharides at the same % w/w concentrations despite having fewer molecules in solution.

FIGURE 1.

The DSC thermograms of wheat starch heated in the presence of (a) glucose, (b) fructose, (c) Glc:Fru (50:50 glucose:fructose mixture), and (d) sucrose at difference concentrations (15−60%) compared to the control (wheat starch in water)

TABLE 4.

The effects of sugar solution type and concentration on the onset gelatinization temperature (T _gel) of wheat starch

	Onset T gel (°C) at sugar concentration (% w/w)
Sugars	0%	15%	30%	45%	60%
Control (water amount)	60.71 ± 0.41Ce	61.57 ± 0.14Ce	61.24 ± 0.13Ce	61.23 ± 0.05De	60.70 ± 0.17Ce
Glucose		64.31 ± 0.27Bc	69.15 ± 0.77Bb	77.49 ± 0.04Ba	n.d.
Fructose		63.31 ± .026Bd	68.35 ± 0.49Bc	75.96 ± 0.24Cb	85.38 ± 1.09Ba
Glc:Fru		63.61 ± 0.28Bd	68.62 ± 0.09Bc	76.91 ± 0.16Cb	86.77 ± 0.44Ba
Sucrose		65.80 ± 0.3Ad	72.23 ± 0.24Ac	82.35 ± 0.3Ab	98.49 ± 0.42Aa

Uppercase letters indicate significant differences between sugar types at a given concentration [columns] and lowercase letters indicate significant differences between concentration for a given sugar type or controls (varying water amounts) [rows].

Abbreviations: Glc:Fru, 50:50 glucose:fructose mixture; n.d., not determined.

FIGURE 2.

The onset T _gel of wheat starch in sugar solutions grouped by concentration: 15% (□), 30% (Δ), 45% (◊), and 60% (○) w/w; and the water control (○)

FIGURE 3.

Correlation plots between the onset T _gel of wheat starch and different sugar and solution traits: (a) sugar solution concentration, (b) sugar dry T _g, (c) sugar solution a _w, and (d) log sugar solution viscosity. The sugars examined were glucose (◇), fructose (○), glc:fru (50:50 glucose:fructose mixture) (□), and sucrose (×)

Select sugar physicochemical properties have been shown to correlate strongly with increases in the T _gel of starch, including: dry T _g, molecular weight or size, molar volume, dielectric relaxation time, solution a _w, number of equatorial hydroxyl groups, and solution viscosity (Allan et al., 2020, 2018; Evans & Haisman, 1982; Slade & Levine, 1988). The correlations between sugar physicochemical properties and the sugar solution effects on starch T _gel were determined in this study for comparison (as shown in Figures 3 and 4b). Across all sugar solution concentrations, the strongest correlations with T _gel were found for log sugar solution viscosity (r = 0.9821) and the effective volume fraction of the solvent Φ _w,eff (r = −0.9573) (Figures 3d and 4b, respectively), consistent with the findings from Allan et al. (2018). Strong positive correlations were found between sugar T _g and T _gel at the same sugar concentration (r = 0.9444 to 0.9929), but not across all sugar concentrations (r = 0.2337) (Figure 3b). The a _ws of the different sugar solutions had a strong negative correlation with the T _gel of starch (r = −0.8002) (Figure 3c). While water would move into dry starch granules until the point a _w in the starch and the a _w of the surrounding solution were equal, a thermodynamically driven process, the sugars in solution also had different effects on the starch that resulted in a lower correlation but still important relationship between solution a _w and starch T _gel. The physicochemical properties of the glc:fru mixture and its solutions were closer to those of the other monosaccharides than to sucrose, often intermediate between glucose and fructose, and the resulting effects of the glc:fru mixture solutions on the onset T _gel also tended to be between those of glucose and fructose solutions (numerically but not significantly) and significantly less than the effects of sucrose solutions on T _gel. A unique finding from this portion of the study was that at the higher solution concentration (45%), the glc:fru mixture resulted in a starch T _gel similar to the T _gel in fructose solutions and significantly less than the T _gel in glucose solutions, attributed to solute partitioning in the glc:fru mixture. Fructose is a more effective plasticizer of starch than glucose.

FIGURE 4.

The relationship between (a) onset T _gel and pasting temperature (PT) recorded for wheat starch, and (b) the effective volume fraction (Ø _w,eff) of the sugar solutions examined and the onset T _gel and PT. The PT of starch in water (control) was excluded due to an abnormally higher value relative to the sugar solutions and did not show a consistent trend with the onset T _gel data

Although the sugar solutions in this study resulted in higher T _gels than the T _gel of starch in water, all samples had a single “G” endotherm (Waigh et al., 2000) and the enthalpy of gelatinization (i.e., the energy involved with the melting of the crystalline regions) was not significantly different (results not shown). Therefore, the gelatinization event itself was generally the same in water versus sugar solutions except it occurred at higher temperatures in sugar solutions. Similar findings were reported in a recent study conducted on samples with higher sugar solution to starch ratios (4:1) (Renzetti et al., 2021). As starch is heated in solution, the solvent penetrates the amorphous regions of the granule, interacts with the starch, changes the thermodynamic state of the granule, and the presence of sugars elevated the melting temperature of the crystalline regions more than water alone (van der Sman & Mauer, 2019). The viscosity of sugar solutions is higher than pure water due to increased intermolecular hydrogen bonding between sugars and water molecules (Seuvre & Mathlouthi, 2010). The viscosity of sucrose solutions was higher and the Φ _w,eff was lower than water and the other sugar solutions examined, indicating sucrose displayed increased intermolecular interactions with water and a decreased volumetric density of hydrogen bonds. Solutes with a lower Φ _w,eff, indicative of having a reduced density of hydrogen bonds compared to other solutes, do not lower the T _gel of starch compared to the dry state as much as water and solutes with higher Φ _w,eff. Accordingly, the T _gel of starch in the presence of solutions containing solutes with a lower Φ _w,eff is higher than the T _gel in water. Taken together, these traits provide an explanation as to why sucrose solutions behaved as superior “antiplasticizers” (Slade & Levine, 1987) by limiting the molecular mobility of starch chains within the amorphous regions compared to water which led to increased stability of the starch granules and higher melting temperatures of the crystalline regions (i.e., higher T _gel).

3.2. The effects of sugar type and concentration on wheat starch pasting parameters

The effects of water (control) and the sugar solutions on the pasting behaviors of wheat starch are summarized in Figure 5 and Table 5. When water was replaced with sugar solution, all of the pasting parameters (PV, TV, SB, and FV) changed dramatically depending on the type and concentration of sugar present.

FIGURE 5.

The effect of sugar solution concentration (15% to 60%) of the pasting behavior of wheat starch measured in the RVA for (a) glucose, (b) fructose, (c), glc:fru (50:50 glucose:fructose mixture), and (d) sucrose

TABLE 5.

The effects of different types and concentrations of sugars on the pasting behavior of wheat starch

(a) Pasting temperature (PT in °C)
	Concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	87.4 ± 0.4	—	—	—	—
Glucose		72.4 ± 0.5ABd	76.9 ± 0.5Bc	84.8 ± 0.1Bb	—
Fructose		71.6 ± 0.4Be	75.3 ± 0.5Cd	82.8 ± 0.5Cc	94.2 ± 0.4Aa
Glc:Fru		72.4 ± 0.4ABe	76.15 ± 0.4BCd	84.0 ± 0.0BCc	95.1 ± 0.6Aa
Sucrose		73.2 ± 0.4Ac	78.3 ± 0.1Ab	88.3 ± 0.9Aa	>95

(b) Peak viscosity (PV in cP)
	Concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	1534 ± 17	—	—	—	—
Glucose		2930 ± 15Ac	4041 ± 36Bb	4836 ± 66Ca	—
Fructose		2846 ± 3Ad	4189 ± 32Ac	5655 ± 51Ab	5830 ± 49Aa
Glc:Fru		2912 ± 9Ad	4140 ± 21ABc	5285 ± 33Ba	4716 ± 106Bb
Sucrose		2714 ± 39Bc	3423 ± 48Ca	3047 ± 58Db	122 ± 128Ce

(c) Trough viscosity (TV in cP)
	Concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	1305 ± 20	—	—	—	—
Glucose		2679 ± 15Ac	3849 ± 59Bb	4802 ± 46Ca	—
Fructose		2578 ± 18ABc	4018 ± 15Ab	5560 ± 32Aa	n.d.
Glc:Fru		2649 ± 12Ac	3965 ± 14ABb	5249 ± 34Ba	n.d.
Sucrose		2496 ± 51Bc	3309 ± 80Ca	3040 ± 60Db	n.d.

(d) Breakdown viscosity (BD in cP)
	Concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	228 ± 9	—	—	—	—
Glucose		251 ± 4Aa	192 ± 36Ab	34 ± 20Ac	—
Fructose		268 ± 16Aa	171 ± 18ABb	55 ± 20Ac	n.d.
Glc:Fru		263 ± 9Aa	175 ± 11Ab	36 ± 3Ac	n.d.
Sucrose		218 ± 12Aa	115 ± 33Bb	7 ± 2Ac	n.d.

(e) Setback viscosity (SB in cP)
	Concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	250 ± 8	—	—	—	—
Glucose		442 ± 14Aab	485 ± 54Aa	392 ± 42Cb	—
Fructose		463 ± 27Ac	568 ± 8Ab	818 ± 68Aa	n.d.
Glc:Fru		435 ± 9Ab	529 ± 18Aa	561 ± 32Ba	n.d.
Sucrose		376 ± 13Ba	373 ± 33Ba	67 ± 23Dc	n.d.

(f) Final viscosity (FV in cP)
	Concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	1555 ± 28	—	—	—	—
Glucose		3121 ± 7Ac	4334 ± 15Bb	5194 ± 36Ca	—
Fructose		3040 ± 10Ad	4586 ± 23Ac	6418 ± 50Ab	7388 ± 101Aa
Glc:Fru		3084 ± 20Ad	4494 ± 28Ac	5809 ± 15Ba	5510 ± 94Bb
Sucrose		2872 ± 63Bb	3681 ± 48Ca	3107 ± 77Db	252 ± 233Cd

Uppercase letters indicate significant differences between sugar types at a given concentration and lowercase letters indicate significant differences between concentration for a given sugar.

Abbreviations: Glc:Fru, 50:50 glucose:fructose mixture; n.d., not determined.

3.2.1. Pasting temperature

The PT occurred at higher temperatures than the T _gel in the sugar solutions, and a strong positive correlation (r = 0.9649) was found between these two temperatures (Table 5a and Figure 4). The PT occurs after gelatinization because large increases in granule swelling only occur after the amylopectin double helices have melted (Balet et al., 2019; Biliaderis, 2009). In samples for which the PT was measurable (those which had T _gel < 95°C, the maximum temperature of the RVA analysis), the granules were assumed to be fully gelatinized. The PT of wheat starch in the presence of water was 87.4°C, and in sugar solutions ranged from 71.6 to >95°C (Table 5a). Increasing sugar concentration significantly increased PT, with a strong negative correlation observed (r = −0.9879) between sugar solution Φ _w,eff and PT (Figure 4). Sugar type was also a significant factor: sucrose increased PT more than the monosaccharides, glucose increased PT more than fructose, and the PT in the glc:fru mixture solutions tended to be between the PTs in glucose and fructose solutions.

Although sugars at all concentrations elevated the T _gel of starch compared to the T _gel of starch in water, their effects on PT followed a different trend. The presence of sugars at lower concentrations (15% and 30%) decreased the PT of starch compared to the control PT, but higher sugar concentrations (60%) resulted in PTs that exceeded the control PT (Table 5a). The decreased PT indicates that viscosity increased earlier in the presence of the more dilute sugar solutions than in water, attributed to increased amylose leaching after the T _gel was reached, intermolecular interactions between this amylose and the sugars in solution, and increased granule packing in the presence of sugars (BeMiller, 2011; Doublier et al., 1987; Renzetti et al., 2021; Richardson et al., 2003; Woodbury et al., 2022). The increase in PT as sugar solution concentration increased was attributed to elevation of the T _gel due to starch–sugar hydrogen bonding interactions and phase separation which would increase the effective concentration of starch (Renzetti et al., 2021; van der Sman & Mauer, 2019; Woodbury et al., 2022).

3.2.2. Peak viscosity

The peak viscosity (PV) of starch‐–sugar pastes varied widely depending on the type and concentration of the sugar (Table 5b and Figures 5 and 6). Compared to the control, the PV increased in the presence of all sugar solutions except the 60% sucrose solution (Table 5b). The control PV was 1534 cP, and the highest PV was found in the 60% solution of fructose (5830 cP). Increasing the monosaccharide sugar concentrations increased the PV, but PV followed a different trend in the disaccharide, sucrose‐containing solutions. The PV in sucrose solutions increased up to the 30% concentration (3423 cP), but then decreased as the sucrose concentration increased to 45% (3047 cP) and 60% (122 cP) (Table 5b). When comparing sugar types, the monosaccharides increased PV more than sucrose across all solution concentrations. At the higher concentrations, fructose solutions resulted in the highest PVs, followed by the glc:fru mixture then glucose. These trends are opposite of the sugar type effects on elevating the onset T _gel and PT.

FIGURE 6.

Comparison of sugar type effects on the pasting behavior of wheat starch in the RVA at different solution concentrations: (a) 15%, (b) 30%, (c) 45%, and (d) 60% w/w, db

The PV occurs during pasting when the rate of granule disruption equals the rate of granular swelling, and PV represents the thickening power of the starch paste (Batey, 2007). The granular swelling of starch in sugar solutions is also known to increase in the order of fructose > glucose > sucrose as the slurry temperature is elevated up to 95°C (Ahmad & Williams, 1999; Hoover & Senanayake, 1996). The increased granular swelling and PV in sugar–starch pastes has been attributed to sugar–starch hydrogen bonding interactions and/or sugar competition for water (Cheer & Lelievre, 1983; Martínez et al., 2015; Sun et al., 2014). Recently, Renzetti et al. (2021) reported a strong correlation between the PV of sugar–starch dispersions and the amount of effective hydroxyl groups within a molar volume (i.e., N _OH,s /v _s), with PV increasing more in the presence of low M _w sugars (glucose and xylitol) with higher N _OH,s /v _s than in solutions of larger M _w sugars (sucrose and fructan oligosaccharides) that have lower N _OH,s /v _s. While this PV–N _OH,s /v _s relationship held for comparing the relative effects of mono‐ versus di−saccharides on PV in this study (Tables 2 and 5b), it did not describe the different effects of the monosaccharides on PV. Glucose had the highest N _OH,s /v _s but did not raise PV as much as fructose. The PV–N _OH,s /v _s relationship also did not describe the effects of sorbitol on PV relative to those of other mono‐ and disaccharides in a previous report (Woodbury et al., 2022).

3.2.3. Breakdown and trough viscosity

Following the PV, continued heating and stirring in the RVA often creates conditions in which the rate of starch granule disruption and breakdown exceeds the rate of granular swelling, and a period of decreased viscosity ensues. The minimum viscosity reached during this period is called the trough viscosity (TV), and breakdown (BD) is calculated by subtracting the TV from the PV. Both sugar type and sugar concentration significantly affected TV and BD (Tables 4cd). At the highest solution concentration (60%), the paste viscosity did not drop following PV and hence no TV or BD values were recorded (Figures 5 and 6). All other sugar solutions (15% to 45%) increased the TV compared to the control, the monosaccharide solutions had higher TV than sucrose, and at the higher solution concentrations (30% and 45%) the TV was higher in fructose solutions than glucose solutions, with the glc:fru mixture resulting in intermediate TVs. These trends were similar to the trends found for the sugar effects on PV.

BD is an indicator of granule rigidity (Batey, 2007), with smaller BD values indicative of more rigid granule structures. Increasing sugar concentrations decreased BD, and BD was less in sucrose solutions than in the monosaccharide solutions. The decreased BD in increasingly concentrated sugar solutions could be attributed to increasing sugar hydrogen bonding with amylose chains attempting to leave the granule during pasting subsequently causing an increase in granule rigidity and stability (Blazek & Copeland, 2008; Nagano et al., 2008).

3.2.4. Setback and final viscosity

Both the amount of amylose that leaches out of starch granules and the propensity for the leached amylose to form junction zones influence SB and FV (Ai & Jane, 2015). All sugar solutions (15% to 45% concentrations) except the 45% sucrose solutions resulted in higher SB viscosities than the control (250 cP), the monosaccharide solutions tended to have higher SB than the sucrose solutions, and fructose solutions resulted in the highest SB (Tables 5e and 4f). At the 45% concentration, the descending order of SB occurred in fructose solutions (818 cP), followed by the glc:fru mixture (561 cP), glucose (392 cP), and then sucrose (67 cP). In sucrose and glucose solutions, SB was lower at the 45% concentration than in the 15% and 30% solutions, whereas SB continued to increase as fructose concentration increased from 15% to 45%. All sugar solutions (15–60% concentrations) except the 60% sucrose concentration resulted in higher FV values than the control, and FV in sucrose solutions was lower than in the monosaccharide solutions. The sugar concentration had variable effects on FV: the maximum FV in sucrose solutions occurred at 30% sucrose concentration, the maximum FV in glucose, and the glc:fru mixture solutions occurred at the 45% concentration, and increasing fructose concentrations significantly increased FV across the entire 15–60% concentration range studied.

More amylose has been reported to leach out of starch granules in the presence of fructose and glucose solutions (10% to 20%) than in sucrose solutions (Ahmad & Williams, 1999), which would account for some of the differences found between the SB and FV values. Compared to the monosaccharide solutions, sucrose elevated the T _gel and PT more, in part by forming stabilizing intermolecular hydrogen bonds in the amorphous regions of starch (Allan et al., 2018), and reduced the other RVA pasting parameters presumably by limiting starch granule swelling and reducing the amount of leached amylose. The increased SB and FV in the monosaccharide solutions resulted from an increased amount of amylose in solution (Mua & Jackson, 1997) and possibly an enhancement of amylose recrystallization (Cairns et al., 1991).

The effects of the 50:50 glc:fru mixture on the pasting properties of wheat starch were in between those of pure glucose and fructose at higher solution concentrations, consistent with findings from previous studies (Sopade et al., 2004; Torley & van der Molen, 2005). Possible reasons for the different trends observed for the glc:fru mixture effects on pasting and T _gel could be explained by the mixing of the solution during pasting in the RVA, which would counteract the solute partitioning/phase separation tendency of the solution under the static conditions of the DSC. At high concentrations where mixing is applied, as in the RVA during pasting, the glucose and fructose molecules of the glc:fru mixture had equal access to amylose and/or amylopectin hydrogen bonding sites both inside and external to the granule resulting in an effect that was in between that of pure fructose and glucose.

3.3. Sugar type and concentration effects on the retrogradation behavior of wheat starch after 7 days of storage at 4°C

Two methods were used to evaluate different aspects of the sugar solution effects on starch retrogradation over time: DSC and oscillatory rheometry. The DSC technique determined the enthalpy associated with the melting of recrystallized short external branches of amylopectin (Miles et al., 1985; Yoshimura et al., 1996), and the Kʹ values determined from rheometer measurements were indicative of the three‐dimensional structure and texture changes of starch pastes and gels over time.

3.3.1. Retrogradation assessed using DSC

The melting enthalpy of retrograded amylopectin occurs at a lower temperature than the onset T _gel and is used as a parameter to describe the extent of amylopectin recrystallization or retrogradation in various treatment conditions (Durán et al., 2001; Karlsson & Eliasson, 2003). As expected, the enthalpies of all samples increased between day 0 and day 7 (Table 6), indicating amylopectin retrogradation occurred over time. There were very few significant differences and no clear trends observed for the day 0 retrogradation measurements (Table 7a). After 7 days of storage, the presence of lower amounts of sugars (15% and 30%) tended to increase retrogradation compared to the control, while higher sugar concentrations had varying effects on ΔH _ret,7. Increasing concentrations of fructose and the glc:fru mixture above 30% resulted in further increases in ΔH _ret,7, while ΔH _ret,7 decreased in the higher concentrations of glucose and sucrose. At these higher concentrations, fructose and the glc:fru mixture promoted amylopectin retrogradation during storage while sucrose and glucose inhibited retrogradation compared to the control.

TABLE 6.

Comparison between the retrogradation measurements conducted in the DSC (ΔH) and rheometer (Kʹ) immediately after gelatinization or pasting pretreatments (day 0), and after being stored at 4°C for 7 days

		DSC		Rheometer
Sugars	% w/w Conc.	ΔH ret,0 (J/g°C)	ΔH ret,7 (J/g°C)	Kʹret,0 (Mpa·s × 103)	Kʹret,7 (Mpa·s ×103)
Water	0	0.047 ± 0.032A	1.567 ± 0.147B	2.18 ± 0.08A	2.22 ± 0.16A
Glucose	15	0.123 ± 0.047A	1.889 ± 0.158B	2.78 ± 0.17B	3.57 ± 0.24A
	30	0.223 ± 0.358A	1.616 ± 0.518B	4.68 ± 0.14B	6.28 ± 0.39A
	45	0.300 ± 0.408A	1.237 ± 0.341B	4.16 ± 0.75B	15.5 ± 0.3A
Fructose	15	0.044 ± 0.030A	2.110 ± 0.398B	3.30 ± 0.16A	3.37 ± 0.23A
	30	0.129 ± 0.056A	3.104 ± 0.902B	5.27 ± 0.27A	5.81 ± 0.58A
	45	0.023 ± 0.024A	2.824 ± 0.127B	4.61 ± 0.59B	15.0 ± 0.6A
	60	0.199 ± 0.048A	4.259 ± 1.559B	2.42 ± 0.35B	21.9 ± 1.4A
Glc:Fru	15	0.036 ± 0.030A	2.054 ± 0.557B	3.04 ± 0.34B	3.77 ± 0.28A
	30	0.068 ± 0.033A	2.667 ± 0.246B	4.21 ± 0.19B	5.77 ± 0.65A
	45	0.049 ± 0.043A	3.183 ± 0.811B	5.27 ± 0.13B	13.8 ± 0.7A
	60	0.038 ± 0.028A	3.592 ± 0.332B	1.98 ± 0.24B	15.5 ± 1.4A
Sucrose	15	0.064 ± 0.018A	1.717 ± 0.468B	3.02 ± 0.39B	3.57 ± 0.24A
	30	0.062 ± 0.036A	1.751 ± 0.298B	3.58 ± 0.20B	4.85 ± 0.10A
	45	0.057 ± 0.049A	0.830 ± 0.164B	3.32 ± 0.18B	7.98 ± 0.74A
	60	0.091 ± 0.036A	0.776 ± 0.058B	n.d.	n.d.

Uppercase letters indicate statistically significant differences between day 0 and day 7 for a given sample at a particular concentration.

Abbreviations: Glc:Fru, 50:50 glucose:fructose mixture; n.d., not determined.

TABLE 7.

The effects of different sugar types and concentrations on the retrogradation enthalpy of wheat starch (ΔHret) measured in the DSC after (a) 0 and (b) 7 days of storage at 4°C

(a) Day 0
	ΔH ret,0 (J/g) at sugar concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	0.047 ± 0.032	—	—	—	—
Glucose		0.123 ± 0.047Aa	0.223 ± 0.358Aa	0.300 ± 0.408Aa	n.d.
Fructose		0.044 ± 0.030Bbc	0.129 ± 0.056Aab	0.023 ± 0.024Ac	0.199 ± 0.048Aa
Glc:Fru		0.036 ± 0.030Ba	0.068 ± 0.033Aa	0.049 ± 0.043Aa	0.038 ± 0.028Ba
Sucrose		0.064 ± 0.018ABa	0.062 ± 0.036Aa	0.057 ± 0.049Aa	0.091 ± 0.036Ba

(b) Day 7
	ΔH ret,7 (J/g) at sugar concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	1.567 ± 0.147	—	—	—	—
Glucose		1.889 ± 0.158Aa	1.616 ± 0.518Ca	1.237 ± 0.341Ba	n.d.
Fructose		2.110 ± 0.398Ab	3.104 ± 0.902Aab	2.824 ± 0.127Aab	4.259 ± 1.559Aa
Glc:Fru		2.054 ± 0.557Abc	2.667 ± 0.246ABabc	3.183 ± 0.811Aab	3.592 ± 0.332Aa
Sucrose		1.717 ± 0.468Ab	1.751 ± 0.298BCb	0.830 ± 0.164Ba	0.776 ± 0.058Ba

Uppercase letters indicate significant differences between sugar types at a given concentration and lowercase letters indicate significant differences between concentration for a given sugar.

Abbreviations: Glc:Fru, 50:50 glucose:fructose mixture; n.d., not determined.

Others have reported that fructose has a tendency to increase starch retrogradation more than glucose or sucrose (Biliaderis & Prokopowich, 1994). The extended CH₂–OH group off the anomeric carbon of fructose was thought to increase in the effective concentration of amylopectin chains due to solute partitioning and disruption of the tetrahedral hydrogen bonding network of water, thereby enhancing the rate of amylopectin recrystallization (Biliaderis & Prokopowich, 1994). The lower amounts of retrogradation in glucose and sucrose solutions were attributed to the higher dynamic hydration numbers (DHN), the average number of water molecules “attached” to the solute (Starzak et al., 2000) of these sugars compared to fructose (Biliaderis & Prokopowich, 1994). Solutes with higher DHNs reportedly limit molecular diffusion and amylopectin recrystallization by increasing the structure of water and facilitating sugar access to the hydration sphere of starch chains (Biliaderis & Prokopowich, 1994). There is a strong correlation between DHN and the number of effective hydroxyl groups (N _OH,s) for different sugars and polyols (van der Sman, 2013). Sugars in this study with higher N _OH,s values (Table 2) tended to decrease ΔH _ret,7 retrogradation when present in more concentrated solutions (45% and 60%), while those with lower N _OH,s values promoted retrogradation (higher ΔH _ret,7). Water itself has the lowest N _OH,s value of the solution components studied, and the ΔH _ret,7 of sugar‐containing starch slurries varied above and below that found in the water control (Table 7), therefore N _OH,s itself is not an absolute predictor of solution effects on retrogradation.

The glc:fru mixture solutions had similar effects to those of the fructose solutions on ΔH _ret,7, which were significantly different than the effects of glucose or sucrose solutions especially as sugar concentration increased (Table 7). At the higher concentrations, fructose and the glc:fru mixture promoted amylopectin recrystallization, while glucose and sucrose inhibited retrogradation. Phase separation is known to occur more readily in highly concentrated (i.e., >30%) monosaccharide solutions (Renzetti et al., 2020). Under the static conditions of the DSC, the fructose phase of a highly concentrated glc:fru mixture solution could have exerted a greater destructuring effect on water than the glucose phase resulting in an increase in the effective starch concentration and higher retrogradation enthalpy values compared to water alone.

3.3.2. Retrogradation assessed using a rheometer

The Gʹ of starch gels and pastes after the RVA treatment were greater than Gʹʹ for most samples (excluding the 60% sucrose solution), indicating that these samples were more solid like than liquid like except for the 60% sucrose solutions in which the starch had not gelatinized (Figures 7 and 8). The Gʹ of sugar‐–starch pastes on day 0 was attributed to amylose recrystallization, whereas the increase in Gʹ between days 0 and 7 was mainly due to amylopectin retrogradation. Differences in the storage modulus consistency index (Kʹ) were found between sugar types, concentrations, and storage periods for most samples excluding the control (Tables 6 and 8). Sugar solutions tended to increase Kʹ compared to the control, and sucrose solutions tended to produce lower Kʹ values than the monosaccharide solutions when considering both sugar concentration and length of storage. On day 0 (Kʹ_ret,0), increasing sugar concentrations tended to increase Kʹ_ret,0 up to a point, followed by decreasing Kʹ_ret,0 values with further increases in sugar concentration. The highest Kʹ_ret,0 values in each type of sugar solution occurred at 30% glucose, 30% fructose, 45% glc:fru mixture, and 30% sucrose. The Kʹ_ret,0 values in the 60% fructose and the glc:fru mixture solutions fell to the Kʹ_ret,0 of the control indicating a weak and/or incomplete amylose network had likely formed despite observing the largest increases in paste viscosity parameters (PV and FV). Conversely, after 7 days of storage (Kʹ_ret,7), higher Kʹ_ret,7 values were found with increasing sugar solution concentration, with the highest Kʹ_ret,7 values occurring in the 60% fructose and 60% invert sugar solutions.

FIGURE 7.

The storage (Gʹ, black lines) and loss modulus (G″, grey lines) of pasted wheat starch samples in the presence of water (×) and different sugar types [(a–d): glucose, fructose, glc:fru (50:50 glucose:fructose mixture), and sucrose] with increasing solution concentrations [15% (□), 30% (Δ), 45% (◇), and 60% (○)], (1) immediately after pasting (day 0) and (2) after 7 days of storage at 4°C. Each curve represents the average of at least triplicate measurements

FIGURE 8.

The (1) storage (Gʹ) and (2) loss modulus (G″) of pasted starch samples after 7 days of storage at 4°C in the presence of water (×) and increasing concentrations of glucose (Δ), fructose (◇), glc:fru (50:50 glucose:fructose mixture) (□), and sucrose (○) solutions (a–d: 15% to 60% w/w). Each curve represents the average of at least triplicate measurements

TABLE 8.

The effects of different sugar types and concentrations on the power law derived consistency index (Kʹ) from the storage modulus of retrograded wheat starch pastes after (a) 0 and (b) 7 days of storage at 4°C

(a) Day 0
	K‵ (Pa·sn) at sugar concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	2.18 ± 0.08	—	—	—	—
Glucose		2.78 ± 0.17Ab	4.68 ± 0.14Ba	4.16 ± 0.75BCa	—
Fructose		3.30 ± 0.16Ab	5.27 ± 0.27Aa	4.61 ± 0.59ABa	2.42 ± 0.35Ac
Glc:Fru		3.04 ± 0.34Ac	4.21 ± 0.19Cb	5.27 ± 0.13Aa	1.98 ± 0.24Ad
Sucrose		3.02 ± 0.39Ab	3.58 ± 0.20 Da	3.32 ± 0.18Cab	n.d.

(b) Day 7
	K‵ (Pa·sn) at sugar concentration % w/w
Sugars	0%	15%	30%	45%	60%
Water	2.22 ± 0.16	—	—	—	—
Glucose		3.57 ± 0.24Ac	6.28 ± 0.39Ab	15.5 ± 0.3Aa	–
Fructose		3.37 ± 0.23ABd	5.81 ± 0.58ABc	15.0 ± 0.6ABb	21.9 ± 1.4Aa
Glc:Fru		3.77 ± 0.28Ac	5.77 ± 0.65ABb	13.8 ± 0.7Ba	15.5 ± 1.4Ba
Sucrose		2.87 ± 0.46Bc	4.85 ± 0.10Bb	7.98 ± 0.74Ca	n.d.

Uppercase letters indicate significant differences between sugar types at a given concentration and lowercase letters indicate significant differences between concentration for a given sugar.

Abbreviations: Glc:Fru, 50:50 glucose:fructose mixture; n.d., not determined.

Amylose retrogradation outpaces amylopectin retrogradation during the cooling of heated starch pastes (Miles et al., 1985) and contributed to the RVA setback (Table 5e) and rheometer day 0 (Table 8) measurements. The low ΔH _ret,0 values (Table 7) indicated little amylopectin retrogradation had occurred. Both SB and Kʹ_ret,0 values were higher in the sugar solutions (15% and 30%) than the control, both increased and then decreased with increasing concentrations of glucose and sucrose, and the highest SB and Kʹ_ret,0 values were found in fructose and the glc:fru mixture solutions. It is evident that both sugar type and concentration had a significant influence on amylose retrogradation, with fructose and the glc:fru mixture promoting more amylose retrogradation and associated viscosity increase than glucose or sucrose.

Amylopectin retrogradation is known to occur in starch pastes over time, and storage at 4°C is the temperature at which crystal nucleation and propagation are enhanced (Marsh & Blanshard, 1988). Changes in both ΔH _ret,7 and Kʹ_ret,7 compared to day 0 measurements are indicative of the progression of amylopectin retrogradation. While Kʹ_ret,7 increased in sugar solutions as sugar concentration was increased, the same was not true for the ΔH _ret,7 of glucose‐ and sucrose‐containing samples; however, in both measurements the sucrose‐containing samples consistently had the lowest values. Sucrose has the highest mean number of equatorial hydroxyl groups among the sugar types examined in this study (Hisashi et al., 1990; Katsuta et al., 1992). The equatorial hydroxyl groups of sugars stabilize the structure of water within a starch gel, minimize the mobility of starch, and thereby inhibit starch retrogradation (Katsuta et al., 1992). Glucose has a higher mean number of equatorial hydroxyl groups than fructose, which could account for the lower ΔH _ret,7 values in the presence of glucose than fructose; however, no significant differences were found between the Kʹ_ret,7 values of glucose and fructose, so the trend did not hold. Katsuta et al. (1992) found that galactose, which has more equatorial hydroxyl groups than fructose, stabilized rice starch gels less than fructose. Interestingly, the ΔH _ret,7 of the glc:fru mixture samples was more similar to the ΔH _ret,7 of fructose samples than those of glucose samples; however, the Kʹ_ret,7 values for the glc:fru mixture were statistically lower than Kʹ_ret,7 for glucose and similar to Kʹ_ret,7 for fructose (albeit numerically lower) only at the 45% concentration.

There are other factors in addition to the molecular conformation of a sugar that affect starch retrogradation. The pretreatment method is known greatly affect retrogradation (Wang et al., 2015). Gelatinization in the DSC simply results in the melting of crystalline amylopectin double helices and slight swelling of the granule with minimal amylose leaching. In pasting, however, starch granules can swell to the point of disruption and leach greater amounts of amylose leading to different micro and macro structural changes over time compared retrogradation observed in the DSC. In this study there was only a moderate correlation (r = 0.5629) between ΔH _ret,7 and Kʹ_ret,7, but a strong correlation (r = 0.9051) was observed between FV and Kʹ_ret,7. The physicochemical properties of the sugar solutions also dictate sugar effects on starch properties (Hoover & Senanayake, 1996; Renzetti et al., 2021; Woodbury et al., 2021). Sugar solution a _w (r = −0.9394), Φ _w,eff (r = −0.8886), viscosity (r = 8087) all strongly and significantly correlated with Kʹ_ret,7, indicating sugar solution dynamics played a significant role in determining the extent of retrogradation, presumably by altering the mobility of amylopectin chains attempting to recrystallize and form more solid‐like three‐dimensional networks over time.

3.4. Summary of the influential structural and physicochemical properties of the sugars

In comparison to the effects of pure water, sugar solutions altered the gelatinization, pasting, and retrogradation of wheat starch differently. The effects of sugar solutions on starch properties have been attributed to sugars possessing unique structural properties (number of equatorial hydroxyl groups, dry T _g, N _OH,s, N _OH,s /v _s, molar volume, DHN) and/or exhibiting distinct solution properties (Φ _w,eff, a _w, viscosity, dielectric relaxation time). For example, starch pastes with monosaccharide solutions had increased granular swelling, paste viscosity, amylose retrogradation (setback), but the monosaccharide solutions did not elevate the T _gel and PT as much as disaccharide and sugar alcohol‐containing solutions (Allan et al., 2018; Woodbury et al., 2022). Similarly, compared to sucrose, the monosaccharides (glucose and fructose) and the equimolar glc:fru mixture (similar to invert sugar) in this study did not elevate the T _gel as much, had higher paste viscosity, and promoted both amylose and amylopectin retrogradation. In sucrose solutions, where glucose and fructose are linked together via a glycosidic linkage, the T _gel was elevated to a greater extent, paste viscosity was diminished, and retrogradation was inhibited as concentration was increased most likely due to sucrose having a greater number of equatorial or effective hydroxyl groups (N _OH,s), increased solution viscosity, and a lower Φ _w,eff than unbound monosaccharides. The properties of the sugar solutions that best correlated to their effects on starch thermal properties were related to both sugar type and concentration and included viscosity, Φ _w,eff, and to a lesser extent a _w and whether or not phase separation occurred.

HFCS, similar in composition to a 50:50 glc:fru mixture or invert sugar but with varying fructose contents (e.g., 42% and 55%), is commonly co‐formulated with sucrose in commercial baked goods to lower product cost while ensuring desirable product quality attributes are met related to texture, appearance, flavor, and sweetness (Zargaraan et al., 2016). The effects of the glc:fru mixture solutions on various wheat starch thermal properties were similar to glucose and fructose at low concentrations (15% to 30%) regardless of the measurement method. However, at higher concentrations (45% and 60%), phase separation is likely to occur which explained why the glc:fru mixture had effects either in between those of glucose and fructose, or similar to fructose, depending on the static or dynamic nature of the measurement. For example, under static conditions (i.e., conditions where agitation was not applied like in the DSC) phase separation was permitted due to fructose more easily diffusing to interact with starch chains resulting in T _gel and ΔH _ret,7 values more similar to fructose than glucose. Under dynamic conditions, like those in the RVA and subsequent oscillatory rheometry experiments, phase separation was limited by the mixing action of the RVA, thus resulting in the glc:fru mixture having an effect on pasting intermediate between that of pure glucose and fructose solutions at the 45% concentration and lower pasting and Kʹ_ret,7 than fructose at the 60% concentration. The effects of a 50:50 glc:fru mixture on baked good systems are expected to be relatively similar to those of HFCS, resulting in products with increased firmness/hardness with time, decreased volume and spread, and darker coloration compared to sucrose (Coleman & Harbers, 1983; Curley & Hoseney, 1984).

4. CONCLUSION

The gelatinization, pasting, and retrogradation properties of wheat starch in water were significantly altered by the addition of sugars to the solutions dependent upon sugar type, concentration, and associated physicochemical properties. Increasing the concentration of a sugar in solution results in less water entering the starch granule, due to both the presence of the sugar and the reduced a _w of the solution, and a concomitant concentration‐dependent increase in the T _gel of starch relative to the T _gel in water. The extent of T _gel elevation was further dependent on solution viscosity, volumetric density of hydrogen bonds (an inverse relationship), and potential for stabilizing intermolecular hydrogen bonds between the sugar and the starch in the amorphous regions, all of which varied by sugar type. The highest T _gel and PT values were found in the sucrose solutions. After gelatinization, starch paste viscosity parameters (PV and FV) increased more in monosaccharide solutions than sucrose because monosaccharides promoted amylose leaching and granule swelling during pasting. Solutions of fructose and the glc:fru mixture at higher concentrations tended to promote amylopectin retrogradation, whereas retrogradation in sucrose and glucose solutions was inhibited. The glc:fru mixture generally exhibited effects on starch properties similar to those of pure glucose and fructose at low concentrations (15% and 30%). At higher concentrations, the effects of the glc:fru mixture relative to those of the individual monosaccharides on starch properties varied depending on whether or not experimental conditions supported glc:fru phase separation. When phase separation occurred, the glc:fru mixture had effects more similar to those of fructose than glucose. These findings underscore the importance of sugar physicochemical properties and solution dynamics in how sugars alter starch thermal properties.

AUTHOR CONTRIBUTIONS

Travest J. Woodbury: Formal analysis; writing—original draft. Sarah L. Pitts: Data curation; writing—review and editing. Adrianna M. Pilch: Data curation; writing—review and editing. Paige Smith: Visualization; Writing—review and editing. Lisa J. Mauer: Conceptualization; methodology; supervision; writing—review and editing.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Woodbury, T. J. , Pitts, S. L. , Pilch, A. M. , Smith, P. , & Mauer, L. J. (2023). Mechanisms of the different effects of sucrose, glucose, fructose, and a glucose–fructose mixture on wheat starch gelatinization, pasting, and retrogradation. Journal of Food Science, 88, 293–314. 10.1111/1750-3841.16414