2026 Volume 27 Issue 3 Pages 43-48
In the chemical isomerization of common sugars to rare sugars, browning of the reaction mixture due to byproduct formation remains a major challenge. In this study, two aldopentoses, two aldohexoses, and three aldo-disaccharides were utilized as substrates to investigate how the heating rate from room temperature to approximately 115°C at 0.754±0.011 and 0.401±0.007 °C/min (fast vs. slow heating) influences both ketose yield and the degree of browning, the latter of which was evaluated by the change in absorbance at 440 nm (ΔA440). For all tested sugars, ΔA440 was markedly lower under fast heating than under slow heating conditions during rare sugar production. This trend suggests that the activation energy for isomerization is higher than that for the concurrent browning pathways. Furthermore, the tendency for browning followed the order of pentoses > hexoses > disaccharides. These findings demonstrate that fast heating provides a practical strategy to suppress thermal degradation and browning while maintaining efficient rare sugar production.
In recent years, increasing attention has been directed toward the physiological functions and physicochemical properties of rare sugars, which occur only in limited quantities in nature [1]. Beyond serving as simple energy sources, rare sugars are now recognized for their diverse and distinctive biological activities, including the suppression of postprandial blood glucose elevation, anti-obesity effects, and antioxidant activity [2]. Consequently, intensive research and development efforts aimed at effectively utilizing rare sugars are underway across various fields, including the food and pharmaceutical industries. Production methods for converting common carbohydrates into rare sugars include biocatalytic approaches [1] and chemical approaches based on the Lobry de Bruyn-Alberda van Ekenstein (LBAE) transformation under alkaline conditions [3,4].
Biocatalytic production of rare sugars offers high selectivity due to the strict substrate specificity of the enzymes involved, thereby minimizing byproduct formation. However, this approach lacks versatility because each substrate requires a specific enzyme, and the costs for enzyme production and stabilization are substantial [5]. Conversely, chemical production is versatile and cost-effective, utilizing inexpensive catalysts or reagents across a broad spectrum of sugars [3]. Nonetheless, low reaction selectivity typically generates numerous byproducts, which complicates downstream separation and purification.
Because chemical strategies offer greater versatility than biochemical methods, we previously reported the production processes for rare sugars under subcritical water conditions [4]. These chemical approaches utilized a variety of reaction media, including a sodium phosphate buffer (hereafter referred to as phosphate buffer), an aqueous arginine solution, and eggshell and scallop shell powders as underutilized resources for valorization [4]. However, as noted above, major drawbacks of chemical methods include a drop in pH of the reaction mixture due to the formation of acidic byproducts, which typically leads to low yields, and significant browning of the final solution. These limitations not only diminish the yield of target rare sugars but also substantially increase the downstream purification burden, posing significant obstacles to practical application.
The activation energies for sugar isomerization and the subsequent formation of browning products can differ significantly. Previous studies investigating sugar isomerization in phosphate buffers have primarily been conducted under isothermal conditions, where the reactor is maintained at a predetermined temperature while varying the reaction time [6-11]. Consequently, the effect of the temperature history in reactor during external heating remains unexplored, although, for aqueous arginine systems, the effects of microwave-induced internal heating on both yield and browning at different applied powers were reported [12].
In this study, we investigated the effect of heating rate on the isomerization of sugars exclusively within phosphate buffer systems, utilizing two aldopentoses, two aldohexoses, and three aldo-disaccharides as substrates. The reaction temperature was increased linearly at two distinct rates to compare the resulting ketose (rare sugar) yields and degrees of browning. By examining a structurally diverse set of carbohydrates differing in carbon number and glycosidic linkage, this work elucidates how thermal history influences both isomerization efficiency and side-reaction-induced browning. Ultimately, our objective is to establish guidelines for designing optimized thermal profiles that simultaneously maximize target product yields and suppress undesirable browning during chemical rare-sugar production.
d-Ribose, d-xylose, d-glucose, d-galactose, maltose monohydrate (hereafter referred to as maltose), cellobiose, lactose monohydrate (hereafter referred to as lactose), disodium hydrogen phosphate, sodium dihydrogen phosphate, HPLC-grade acetonitrile, and HPLC-grade distilled water were purchased from Fujifilm Wako Pure Chemical Corp. (Osaka, Japan). Because all sugars used in this study belong to the d-series, the “d-” prefix denoting absolute configuration is omitted hereafter.
2.2 Isomerization of aldose to corresponding ketoseEach sugar solution (5 mL, 5% (w/v) in 10 mmol/L phosphate buffer, pH 7.0) was dispensed into amber screw-cap test tubes (10 mmφ×105 mm, Maruemu Corp., Osaka). The tubes containing the substrate solutions were placed in a digital dry bath heater (model HDB-1N, AS ONE Corp., Osaka) and heated linearly from room temperature to approximately 115°C over a period of either 120 min (0.754±0.011 °C/min) or 240 min (0.401±0.007 °C/min). For the 120 min heating profile, tubes were removed at 0, 30, 60, 80, 85, 90, 95, 100, 105, 110, 115, and 120 min. For the 240 min heating profile, tubes were removed at 0, 60, 120, 160, 170, 180, 190, 200, 210, 220, 230, and 240 min. Thermal reactions were quenched immediately upon removal by immersing the tubes in an ice-water bath. To monitor the internal temperature profile during heating, a reference screw-cap tube containing 5 mL of pure water was fitted with a silicone stopper featuring a center-drilled hole for a thermometer (model AD-5605H, A&D, Tokyo, Japan); the stopper assembly was hermetically sealed using an adhesive (Super X, Cemedine, Tokyo).
Because temperature, ketose yield, and the absorbance increase at 440 nm, ΔA440, were measured at multiple reaction times, only a single experiment was performed for each experimental condition.
2.3 AnalysesThe concentrations of the remaining substrate (aldose) and the resulting rare sugar (ketose) in the reaction mixture were determined via high-performance liquid chromatography (HPLC). The HPLC system comprised a Shimadzu LC-10AD pump (Kyoto, Japan), a Resonac RI-101 refractive index detector (Tokyo), and a Shimadzu Chromatopac CR-8A data processor. Separation was achieved using a Cosmosil® Sugar-D column (4.6 mmφ × 250 mm, Nacalai Tesque, Kyoto). The mobile phase consisted of acetonitrile in water at a concentration of 80% (v/v) for pentose and hexose substrates, and 75% (v/v) for disaccharides, maintained at a flow rate of 1.0 mL/min. The sample injection volume was 5 μL.
The color intensity of the reaction mixture was quantified by measuring its absorbance at 440 nm using a UV-visible spectrophotometer (UV-1280, Shimadzu Corp.).
Figure 1 illustrates the time courses of temperature, maltulose yield, and the increase in absorbance at 440 nm (ΔA440), used as a browning index, during the thermal treatment of a 5% (w/v) maltose solution at two distinct heating rates (fast and slow). Under both conditions, negligible reaction occurred in the low-temperature region during the initial heating stage. Once the reaction mixture reached approximately 80°C (~80 min for fast heating and ~150 min for slow heating), maltulose formation initiated abruptly, followed by a noticeable increase in ΔA440. The maltulose yields at the maximum temperature (~115°C) were approximately 0.15 for fast heating and 0.16 for slow heating, indicating no substantial difference attributable to the heating rate. In contrast, the final ΔA440 reached approximately 0.05 under fast heating but doubled to about 0.10 under slow heating, demonstrating markedly enhanced browning at the lower heating rate. Although the maltulose yield at approximately 115°C was independent of the heating rate, the pronounced browning observed under slow heating is likely attributable to differences in the intrinsic kinetics of the two pathways and the residence time at elevated temperatures. The duration for which the reaction mixture remained above 80°C was more than twice as long under slow heating (~90 min) than under fast heating (~40 min). The side reactions, such as thermal degradation accompanied by browning, proceed irreversibly [6]. Thus, the accumulation of browning products is presumed to increase in proportion to the reaction time at high temperatures. Furthermore, the pH of the reaction mixture decreases over time because of the formation of organic acids (e.g., glycolic, formic, and acetic acids) generated via aldose degradation under subcritical water conditions [6]. In a 10 mmol/L phosphate buffer, the pH falls below 6 once the buffering capacity is exceeded, at which point the base-catalyzed LBAE transformation becomes essentially suppressed [6,7]. Consequently, under both slow and fast heating, the ketose yield levels off once the pH of the reaction mixture drops to 6 or below, whereas acid-catalyzed dehydration and polymerization reactions leading to browning (i.e., 5-hydroxymethylfurfural (HMF) formation followed by humin or caramel production) [8] continue to progress. Therefore, the discrepancy in thermal histories between the two heating profiles manifests as a pronounced increase in browning at comparable maltulose yields. These results suggest that fast heating is advantageous for suppressing browning side reactions while efficiently producing rare sugars.
Time-course profiles of temperature (lines), maltulose yield (circles), and absorbance increase at 440 nm (ΔA440) (triangles) during fast heating (closed symbols, solid lines) and slow heating (open symbols, dashed lines).
Figures 2(a)-(c) illustrate the relationship between the yields of corresponding ketoses and the ΔA440 for two aldopentoses (ribose and xylose), two aldohexoses (glucose and galactose), and three aldo-disaccharides (maltose, cellobiose, and lactose) under fast heating (solid lines, closed symbols) and slow heating (dashed lines, open symbols). The specific ketoses generated from the isomerization of ribose, xylose, glucose, galactose, maltose, cellobiose, and lactose are ribulose, xylulose, fructose, tagatose, maltulose, cellobiulose, and lactulose, respectively. Across all studied sugars, the ΔA440 at any given ketose yield was consistently higher under slow heating than under fast heating. Furthermore, structural variations among the sugars, specifically carbon number and degree of oligomerization (monosaccharide vs. disaccharide), led to distinct behaviors. For the aldopentoses (Fig. 2(a)), ΔA440 increased sharply even at low ketose yields (approximately 0.04-0.08 mol/mol), with xylose under slow heating conditions exhibiting the most pronounced browning. In contrast, the aldohexoses (Fig. 2(b)) exhibited a more gradual increase in absorbance at comparable yields relative to the pentoses. For the aldo-disaccharides (Fig. 2(c)), ΔA440 remained relatively low even as the ketose yield reached a higher range of 0.12-0.16 mol/mol. This indicates that for disaccharides, isomerization proceeds more readily with concomitant suppression of the browning pathways compared to their monosaccharide counterparts.
Relationship between ketose yield and absorbance increase at 440 nm (ΔA440) during the isomerization of aldoses to ketoses. Substrates include: (a) ribose (circles) and xylose (triangles); (b) glucose (circles) and galactose (triangles); and (c) maltose (circles), cellobiose (triangles), and lactose (squares). Open symbols with dashed lines and closed symbols with solid lines indicate slow and fast heating, respectively.
These results can be interpreted in terms of the activation energies (E) governing isomerization and browning. The isomerization of aldoses to ketoses and the browning reactions, comprising thermal degradation and polymerization, proceed as parallel, competing pathways [6]. The observation that fast heating suppresses browning per unit ketose yield suggests that the ratio of the isomerization rate constant to the browning rate constant (kiso/kbrown) increases with temperature. According to the Arrhenius equation, a temperature-dependent increase in this rate-constant ratio implies that the activation energy for isomerization (Eiso) is greater than that for browning (Ebrown), i.e., Eiso > Ebrown. Reported Eiso values for monosaccharide isomerization under subcritical water conditions generally range from 100 to 150 kJ/mol [13,14], whereas Ebrown values of approximately 50-100 kJ/mol are typically reported for the caramelization browning of sugars [15,16]. Thus, literature data strongly support the relationship Eiso > Ebrown. Under slow heating conditions, the reaction mixture spends an extended duration in the low- to mid-temperature regime (approximately 80-100 °C), where the kiso/kbrown ratio is lower and browning is kinetically favored. Consequently, browning products accumulate before substantial isomerization can occur. In contrast, fast heating quickly bypasses this low-selectivity temperature window, shifting the reaction kinetics into the high-temperature regime where isomerization selectivity is significantly enhanced. Fast heating therefore effectively suppresses browning.
Next, the influence of the structural characteristics of the sugars was considered. Thermal reactions of carbohydrates proceed via ring opening to form linear enediol intermediates [6]. The open chain contents (hereafter referred to as OCC) have been reported to be 0.092%, 0.072%, 0.0097%, and 0.054% for ribose, xylose, glucose, and galactose, respectively [17]; and 0.0020% and 0.0015% for maltose and cellobiose, respectively [18]. These values for the monosaccharides and disaccharides were chemically derived via rapid and selective reaction with the trace open-chain aldehyde forms present in aqueous solution. Although the OCC of lactose has not been reported, it is expected to be similarly low to those of maltose and cellobiose. Because aldopentoses exhibit higher OCC in solution than aldohexoses or disaccharides, their subsequent dehydration to furfural derivatives occurs more rapidly. Consequently, the browning rate relative to the isomerization yield is markedly higher for pentoses [19]. Under subcritical-water treatment in a phosphate buffer, pentoses (ribose and xylose) generate substantially larger amounts of glycolic acid and dihydroxyacetone as byproducts than hexoses, which is accompanied by a more pronounced decrease in pH [6]. These organic acids autocatalytically accelerate the pathway from ketoses to furfurals and subsequently to humins or caramelized products [8], explaining the pronounced browning observed for pentoses at a given ketose yield. Xylose, in particular, is well known for its high browning tendency during food processing [20], and the sharp increase in ΔA440 observed under slow heating in this study is consistent with this behavior.
Although browning was generally greater for pentoses, which exhibited higher OCC values, than for hexoses with lower OCC values, a higher OCC did not necessarily correspond to more pronounced browning within either the pentose or hexose group. Therefore, although OCC is an indicator of browning potential, it alone is insufficient to predict the susceptibility of a sugar to browning. For example, the OCC of glucose is approximately one-fifth that of galactose, yet the ΔA440 values after 120 or 240 min of reaction are nearly identical, and under fast-heating conditions the ΔA440 of glucose even exceeds that of galactose. This difference can be explained by the distinct reaction pathways of the two hexoses. Under high-temperature conditions, glucose is readily converted through enolization and dehydration reactions into chromophore precursors such as HMF, which subsequently undergo polymerization to form brown pigments [21]. In contrast, although galactose has a higher ring-opening propensity, the reaction intermediates it generates and their subsequent reaction pathways differ from those of glucose [22], resulting in a less efficient formation of colored products.
In contrast, the higher selectivity for isomerization of aldo-disaccharides over browning may be attributed to their larger molecular size and steric hindrance, which are presumed to suppress ring opening and subsequent condensation reactions [23]. Additionally, both maltose and lactose possess a reducing-end glucose residue whose C4 position is substituted through an α-glucosidic or β-galactosidic linkage, respectively. This structural feature suppresses the formation of C3 byproducts, such as dihydroxyacetone, which typically arise from the retro-aldol cleavage of glucose, and instead biases the byproduct distribution toward glycolic and formic acids [6]. In contrast, disaccharides exhibited substantially lower OCC values than either pentoses or hexoses, whereas their browning, although reduced, was not diminished to the same extent. This discrepancy is likely attributable to the partial hydrolysis of disaccharides under high-temperature conditions, producing hexoses that subsequently undergo browning relatively rapidly. These findings further indicate that OCC alone cannot adequately account for the browning behavior of sugars. Limiting the variety of these byproducts simplifies the multistep polymerization pathway from HMF to humins or caramel relative to monosaccharides, thereby suppressing the accumulation of browning products at comparable ketose yields. Taken together, these findings indicate that, across a wide range of aldoses, the application of fast heating, leveraging the intrinsic relationship Eiso > Ebrown, constitutes a broadly applicable strategy for efficiently producing isomerized sugars while minimizing thermal degradation and browning.
In this study, we investigated the effects of linear heating rates on the isomerization yields and browning intensities of seven structurally diverse aldoses. Our primary objective was to establish thermal history guidelines that optimize the chemical production of rare sugars while minimizing concurrent browning. For all sugars examined, browning was markedly lower under fast heating than under slow heating conditions during ketose production. This behavior is attributed to the activation energy of the isomerization reaction exceeding that of the browning reaction (Eiso > Ebrown). Fast heating allows the reaction mixture to pass quickly through the low- to mid-temperature regimes where browning is kinetically favored, thereby concentrating the reaction in the high-temperature regime where isomerization selectivity is significantly enhanced.
The structural characteristics of the sugars strongly influenced their browning behavior, with the tendency for browning decreasing in the order: aldopentoses > aldohexoses > aldo-disaccharides. This trend correlates with the higher equilibrium OCC in monosaccharides, which facilitates rapid thermal degradation, dehydration, and condensation via highly reactive intermediates. Collectively, these findings demonstrate that applying fast heating, to exploit the temperature dependence of reaction selectivity, provides a rational thermal-control strategy across a wide range of aldoses to efficiently produce rare sugars while suppressing competitive browning pathways. Furthermore, our results suggest that this heating-profile design is broadly applicable not only to phosphate-buffer systems but also to other chemical methods utilizing basic amino acid solutions or eggshell/shell-derived catalysts [6,24,25], provided that the condition Eiso > Ebrown is satisfied.
This work was financially supported by JSPS KAKENHI (Grant Number 24K08807 to T. K.). We also thank Mr. H. Ohta for his technical assistance.
