Production of tagatose and talose through isomerization of galactose in a buffer solution under subcritical water conditions

Article abstract of DOI:10.1016/j.carres.2020.108031

Galactose was isomerized in pure water or in 10 mmol/L sodium phosphate buffer at 160 °C under pressurized conditions. The isomerization of galactose to tagatose and talose in phosphate buffer resulted in 14% and 1.4% yields, respectively, which were significantly higher than those obtained in subcritical pure water (0.6% and <0.1%, respectively). The effect of the temperature on isomerization was examined between 100 and 160 °C. The most remarkable isomerization was observed at 120 °C or higher. The effect of the buffer solution type was also examined. The pH drop of the treated solution was lesser in MOPS and PIPES buffers than in the phosphate buffer; however, the isomerization was less likely to occur in MOPS and PIPES buffers. The relationship between the pH drop during the reaction in phosphate buffer and the yields of tagatose and talose revealed that the isomerization proceeded only when the pH was >6.3. Our results indicate that the galactose isomerization by Lobry de Bruyn-Alberda-van Ekenstein transformation rarely occurred at low pH due to the formation of organic acids.

Full text of DOI:10.1016/j.carres.2020.108031

CarbohydrateResearch493(2020)108031
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Production of tagatose and talose through isomerization of galactose in a
buffer solution under subcritical water conditions
Yuichiro Onishia, Yuya Furushiroa, Yusuke Hirayamaa, Shuji Adachib, Takashi Kobayashia,∗
a Division of Food Science and Biotechnology, Graduate School of Agriculture, Kyoto University, Sakyo-ku, Kyoto, 606-8502, Japan
b Faculty of Bioenvironmental Science, Kyoto University of Advanced Science, Kameoka, Kyoto, 621-8555, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
Galactose was isomerized in pure water or in 10 mmol/L sodium phosphate buffer at 160 °C under pressurized
conditions. The isomerization of galactose to tagatose and talose in phosphate buffer resulted in 14% and 1.4%
yields, respectively, which were significantly higher than those obtained in subcritical pure water (0.6%
and < 0.1%, respectively). The effect of the temperature on isomerization was examined between 100 and
160 °C. The most remarkable isomerization was observed at 120 °C or higher. The effect of the buffer solution
type was also examined. The pH drop of the treated solution was lesser in MOPS and PIPES buffers than in the
phosphate buffer; however, the isomerization was less likely to occur in MOPS and PIPES buffers. The re-
lationship between the pH drop during the reaction in phosphate buffer and the yields of tagatose and talose
revealed that the isomerization proceeded only when the pH was > 6.3. Our results indicate that the galactose
isomerization by Lobry de Bruyn-Alberda-van Ekenstein transformation rarely occurred at low pH due to the
formation of organic acids.
Buffer
Galactose
Isomerization
Rare sugar
Subcritical water
1. Introduction
subcritical aqueous alcohol treatment due to the limited solubility of
galactose in an aqueous alcohol [16]. For example, tagatose and talose
Rare sugars are saccharides that are present in very small quantities
in nature. Some of them are low in calories and have a sweet taste
[1–3]. In addition, various beneficial physiological functions have been
reported for rare sugars, such as anti-tumor [4,5], anti-inflammatory
[4,6], anti-hypertensive effects [4]. Moreover, rare sugars are reported
to prevent obesity [5] and diabetes [1,7]. Therefore, they have at-
tracted much attention as functional food ingredients. However, prac-
tical methods for the mass production of most rare sugars have not yet
been developed. Several studies have been reported to produce rare
sugars, using enzymatic methods [1,3–5], alkaline isomerization
[8–10], and isomerization using subcritical fluids [11–15].
were produced at the concentration of 11 g/L and 2.6 g/L at most,
respectively, in aqueous ethanol [12]. In addition, alcohols are flam-
mable and expensive for industrial production. Therefore, to produce
rare sugars practically, there is a need to develop efficient methods
using subcritical water that can dissolve sugars at high concentrations.
However, the isomerization yield of sugars was reportedly low even
in subcritical water, since most of the sugars is mainly decomposed
[11]. When sugars decompose, release of some organic acids during the
reaction leads to the pH drop [17–21]. It is expected that a buffer so-
lution may counter the pH drop and improve yields of rare sugars.
Therefore, in this study, rare sugars were synthesized using galactose as
an inexpensive raw material, and by treating it in a buffer solution
under subcritical water conditions.
Subcritical water is the water that maintains liquid state at 100 °C or
higher under pressurized conditions. Reducing monosaccharides are
known to isomerize to other sugars in subcritical water [11]. When
heated under high pressure, aqueous alcohols also become subcritical
state. It has been reported that high yields of rare sugars can be ob-
tained from common reducing sugars in subcritical aqueous alcohol
than in subcritical water [12–15]. For example, galactose isomerizes to
rare sugars of tagatose and talose [12]. Isomerization by subcritical
aqueous alcohol treatment is simple and can shorten the reaction time.
However, concentrations of the rare sugars cannot be increased by the
2. Results and discussion
2.1. Isomerization in the subcritical buffer solution
The isomerization of galactose to tagatose and talose in a sodium
phosphate buffer was investigated under pressurized conditions. The
concentration of the buffer was adjusted at lower concentration
∗ Corresponding author.
E-mail address: kobayashi.takashi.6w@kyoto-u.ac.jp (T. Kobayashi).
https://doi.org/10.1016/j.carres.2020.108031
Received 26 March 2020; Received in revised form 28 April 2020; Accepted 7 May 2020
Availableonline11May2020
0008-6215/©2020ElsevierLtd.Allrightsreserved.

Y. Onishi, et al.
CarbohydrateResearch493(2020)108031
Fig. 1. Time courses for the conversion of galactose into tagatose and talose in
the phosphate buffer or in pure water under subcritical water conditions.
Fig. 2. Dependence of the yield of the rare sugars and isomerization selectivity
on the treatment temperature at 90 s.
(10 mmol/L, pH 7.0) because the downstream process (industrial se-
paration and purification of rare sugars) will be simplified, contributing
the low-cost production of rare sugars. When galactose (5 wt%) was
treated with phosphate buffer at 160 °C, the conversion of galactose
was significantly increased compared to the treatment with subcritical
pure water (Fig. 1). Some time courses for the formation of tagatose
showed downward convex at 30 s, showing the effect of heat transfer
between the heating medium (silicone oil) and reaction mixture. The
effect became, however, negligible at treatment times longer than 60 s.
The concentration of galactose in the buffer solution decreased by ca.
25% within 60 s. In contrast, when the treatment time was extended, no
further decrease was observed in the galactose content. After 60 s, 14%
tagatose and 1.4% talose were formed by the isomerization of ga-
lactose, and no further increase in tagatose yield was observed even
after extending the treatment time to 120 s. In fact, further prolonging
the treatment caused a slight decrease in the tagatose yield although the
yield of talose (ca. 1.6%) remained unchanged.
the change in yield was insignificant.
The selectivity of galactose isomerization to tagatose was rather low
below 110 °C. When the treatment temperature was increased to
140 °C, the selectivity increased to about 70%, but did not increase
further with the increase in the temperature. This could be due to the
decomposition of tagatose at higher temperatures.
In contrast, both, the yield and selectivity of galactose isomerization
to talose, gradually increased between 130 and 145 °C and remained
almost constant above 145 °C. We observed a difference in the tem-
perature dependence between isomerization to tagatose and talose. This
could reflect the difference in the activation energy for the isomeriza-
tion between these two saccharides. Based on these results, subsequent
studies were performed at 140 °C, unless otherwise specified.
It was reported that yields of tagatose and talose obtained by sub-
critical aqueous-ethanol treatment were 14% and 3%, respectively,
with tagatose selectivity of 58%, and were almost comparable to results
in this study [12]. However, the previous method required the treat-
ment temperature of 180 °C and time of 500 s. On the other hand, the
140 °C treatment gave almost the same yield within 120 s in phosphate
buffer. In addition, the solubility of galactose in 60% (v/v) ethanol in
water is ca. 10 wt%. On the other hand, the solubility in water is ca.
32 wt% [22]. Therefore, to improve the productivity (concentration) of
rare sugars, it is essential to use water as the reaction medium.
Delidovicha et al. reported the efficient combination of conversion
of glucose to fructose in a phosphate buffer (0.3–0.7 mol/L, initial pH
7.3–8.5) and its extractive recovery using 1-octanol and o-hydro-
xymethyl phenylboronic acids (final yield = 51%), indicating the ef-
fectiveness of phosphate [23]. Meanwhile, although the yield was
lower, isomerization proceeded even at neutral pH of 7.0 in a phos-
phate buffer with much lower concentration (10 mmol/L) without any
other additives in this study.
In contrast, the tagatose yield was only 0.6% in subcritical pure
water at 60 s, and prolonged treatment did not increase the efficiency of
its formation. After the subcritical pure water treatment, talose was
present only in trace amounts (yield < 0.1%). These results suggest that
the isomerization of galactose to the rare sugars is enhanced in the
phosphate buffer even at low concentrations.
2.2. Effects of temperature on the yield and selectivity of rare sugars
The yield of tagatose gradually decreased at treatment times longer
than 120 s at 160 °C in phosphate buffer (Fig. 1). This decrease could be
ascribed to the gradual thermal decomposition of tagatose at high
temperatures. Therefore, the reaction temperature in the phosphate
buffer was lowered. Hence, the isomerization of galactose was eval-
uated in the range of 100–160 °C. Our results showed that the progress
of the reaction was significantly low at 120 °C or lower, and more than
90% of galactose remained in solution even after 300 s or longer
treatment. The remaining fraction of galactose gradually decreased at
140 °C; it reached about 75% after 180 s but did not decrease after-
wards. The yields of tagatose and talose at 140 °C after 180 s were 13%
and 1.3%, respectively, which were comparable to those obtained after
the 60-s treatment at 160 °C.
2.3. Effect of buffer type on the isomerization
The use of other neutral pH buffers was investigated to further
improve the yield of the rare sugars. We investigated PIPES (piperazine-
N,N′-bis(2-ethanesulfonic acid), pKa = 6.80) and MOPS (3-(N-mor-
pholino)propanesulfonic acid, pKa = 7.20) buffers (10 mmol/L) at
140 °C because they have the pKa values of ca. 7 and would further
suppress the side reactions and improve the selectivity of rare sugars
(Fig. 3). However, compared with the phosphate buffer, PIPES and
MOPS buffers suppressed the isomerization of galactose; moreover the
suppression was more remarkable for the MOPS buffer. The yields of
The yield of the rare sugars and isomerization selectivity were
evaluated at 90 s (Fig. 2). The selectivity was defined as the ratio of the
amount of the formed rare sugar to the amount of the consumed ga-
lactose. At temperatures below 120 °C, the yield of tagatose was low
(< 1.0%), but at temperatures between 120 and 150 °C, the yield
sharply increased to about 13%. At temperatures higher than 150 °C,
2

Y. Onishi, et al.
CarbohydrateResearch493(2020)108031
comparable to that in the pure water treatment, indicating that the
buffer solution was too dilute to buffer optimally. We speculate that
organic acids formed even in the presence of the buffer, thereby re-
sulting in the loss of the buffering capacity.
We tried to lower the reaction temperature to suppress the forma-
tion of organic acids. Our results showed that the buffering capacity
was maintained at 140 °C during the first 60 s (pH was maintained in
the range from 6.7 to 6.8 during the first 60 s); however, the formation
of organic acids resulted in a pH drop during prolonged treatments. At
300 s or longer, the pH dropped to 5.2, and the buffering capacity was
completely lost. The magnitude of pH drop was, however, smaller than
observed at the 160 °C treatment. Further, the decrease of pH was much
lesser in the PIPES and MOPS buffers at 140 °C, and at lower tem-
peratures in the phosphate buffer (pH > 6.5 at 300 s). Moreover,
under these conditions, the decrease in galactose concentration was
also comparatively small (Figs. 1 and 3). Collectively, these results
show that suppression of organic acids formation facilitated the main-
tenance of the buffering capacity.
As discussed above, the pH change was negligible in PIPES or MOPS
buffer. In contrast, the phosphate buffer showed almost no buffering
capacity. Nevertheless, the yields of the rare sugars were lower in the
PIPES and MOPS buffers than in the phosphate buffer. These results
demonstrate that the phosphate buffer not only acts as a weak buffering
solution, but also as a catalyst. Therefore, unless otherwise noted,
subsequent studies were performed in the phosphate buffer.
Fig. 3. Time courses for the isomerization of galactose in various buffers at
140 °C.
tagatose and talose at 300 s in PIPES buffer were 9.6% and 0.6%, re-
spectively, which were lower than observed in the phosphate buffer.
The yields of the rare sugars were even lower in MOPS (yield of taga-
tose = 3.6% at 300 s; yield of talose = 0.3% at 300 s), although they
were higher than by the subcritical pure water treatment. Therefore, it
was suggested that the type of buffer influenced the efficiency of ga-
lactose conversion to rare sugars. Although the role of the buffer so-
lution on the reaction mechanism is not clear, our results show that the
presence of the buffer solution even at low concentrations affects the
efficiency of galactose isomerization.
2.5. Relationship between pH of the treated solution and the yields of the
rare sugars
Formation of the rare sugars in the phosphate buffer proceeded
rapidly at the early stage of the reaction, and continued at a decreased
rate in the latter half of the treatment (Fig. 1). This was presumably due
to the pH drop caused by the loss of buffering capacity. Therefore, we
next focused on elucidating the relationship between the degree of re-
action progress and pH.
2.4. Changes in pH during the treatment
Fig. 4 shows the pH change during the treatment under subcritical
water conditions. When galactose was treated with subcritical pure
water at 160 °C, the pH dropped sharply at the initial period of the
reaction (0–60 s, pH 4.7 at 60 s), while after 60 s, the drop in pH was
less rapid. This could be explained by the immediate formation of or-
ganic acids by galactose decomposition. The buffering capacity of
phosphate buffer was limited at 160 °C, i.e. the drop in pH was
In the phosphate buffer, the relationship between the remaining
fraction of galactose and pH could be expressed by a single curve in-
dependent of the temperature (Fig. 5, Fig. S1). The reaction started at
pH ~7. The fraction of galactose and pH decreased with the reaction
progress, while the yield of the rare sugars increased. The tagatose yield
increased sharply and reached about 10% until the pH dropped from
7.0 to approximately 6.5. However, with the further drop in pH, the
yield of tagatose was slowed down, and it stabilized when pH dropped
Fig. 5. Dependence of the remaining fraction of galactose and the yields of
Fig. 4. Change in pH of the treated solution obtained at 100–160 °C.
tagatose and talose on the pH of the solutions at 100–160 °C.
3

Y. Onishi, et al.
CarbohydrateResearch493(2020)108031
below 6.2. The formation of talose showed a similar trend, although the
talose yield did not exceed 1.5%. In contrast, the remaining fraction of
galactose decreased slightly even at the pH below 6.2. This indicates
that galactose could be directly decomposed into organic acids, or that
its isomerization to the rare sugars and the formation of organic acids
by decomposition of the produced rare sugars were balanced.
As described above, the relationship between the remaining fraction
of galactose or the yields of the rare sugars and the pH of the reaction
mixture, did not depend on the temperature. These results suggest that
pH drop, due to the formation of organic acids, rather than the tem-
perature of the treatment was the crucial factor in determining the
overall reaction efficiency. This also indicates that the isomerization of
galactose rarely occurred if organic acids were present at a certain
level. It has been proposed that the isomerization in subcritical water
occurs through Lobry de Bruyn-Alberda-van Ekenstein (LBAE) trans-
formation due to the formation of hydroxide ions [8,9,20,24,25]. In this
study, the fact that the isomerization was suppressed by the acidity of
the phosphate buffer clearly indicates that the isomerization occurred
through LBAE transformation, and that LBAE transformation occurred
even in neutral or weakly acidic pH range (6.3–7.0). This could be due
to the high ion product and increase in hydroxide ion concentration
under the subcritical state of the buffer solution, which showed a
slightly acidic pH of 6.3 at room temperature. Taken together, these
results indicate that maintaining a pH above 6.3 in phosphate buffer is
effective for the production of rare sugars.
10 mmol/L sodium hydroxide solution to adjust pH at 7.0. MOPS buffer
(pH 7.0) was prepared by mixing 10 mmol/L MOPS aqueous solution
and 10 mmol/L sodium hydroxide solution. Nitrogen was then suffi-
ciently blown into the starting mixture to remove dissolved oxygen, and
its reservoir was connected to a nitrogen gas bag to prevent redissolu-
tion of oxygen.
The treatment of galactose was performed using a tubular reactor
(reactor volume = 1.9 mL), which was similar to those reported in
previous studies [26,27]. The reactor consisted of a stainless-reinforced
PEEK tube (0.8 mm I.D.) immersed in a silicone oil bath and back-
pressure regulator (P-880, Upchurch Scientific, Oak Harbor, WA, USA).
The temperature of the reactor varied between 100 and 160 °C. The
starting mixture was delivered into the reactor using an HPLC pump
(LC-10ADVP, Shimadzu, Kyoto, Japan). The flow rate was adjusted to
provide a residence time (reaction time) of 30–300 s. The pressure in-
side the reactor was maintained at approximately 5 MPa. The outlet
side of the reactor was immersed in a water bath to terminate the re-
action. The effluent coming out of the reactor was collected and ana-
lyzed by HPLC. All experiments were done in triplicate.
4.3. Analysis
The contents of the effluent were analyzed by HPLC. Galactose,
tagatose, and talose were quantified by an HPLC system equipped with
the RID-20A refractive index detector (Shimadzu) and the LC-20AD
HPLC pump (Shimadzu) connected to the COSMOSIL Sugar-D column
(3 mm I.D. × 250 mm, Nacalai Tesque). The eluent was 80% acet-
onitrile (v/v) at a flow rate of 0.4 mL/min. The products (rare sugars)
were confirmed by comparing the retention times of the present study
and previous report [12].
In contrast, in subcritical pure water, a slight decrease in the ga-
lactose concentration (conversion < 3%) caused a sharp drop in pH,
and the yields of tagatose and talose remained very low. These results
differed from those obtained in the phosphate buffer treatment, sug-
gesting that the treatment of galactose with subcritical pure water re-
sults mainly in the decomposition of sugars into organic acids, rather
than isomerization.
pH of the effluent was measured using a D-71 pH meter (HORIBA,
Kyoto, Japan) by dipping a pH electrode (9680-10D, HORIBA) into ca.
1 mL of the effluent.
3. Conclusion
Declaration of competing interest
Our results showed that when galactose was treated with a buffer
solution under subcritical water conditions, tagatose and talose were
produced through isomerization at higher yields than under a treatment
with subcritical pure water. The isomerization of galactose in sodium
phosphate buffer at 140 °C resulted in high yields of tagatose and talose
(13% and 1.3%, respectively), although the buffering capacity of the
solution was low. On the other hand, when MOPS and PIPES buffers
were used, the yields of the rare sugars were not so high in spite of the
lower pH drop in comparison to the pH drop in the phosphate buffer.
Therefore, the subcritical phosphate buffer treatment was effective for
the efficient production of the rare sugars. Moreover, maintaining the
pH of the reaction mixture at 6.3 or higher was essential for the iso-
merization in our study.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgment
This study was financially supported by JSPS KAKENHI (No.
17K07814; T. K.).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.carres.2020.108031.
4. Experimental
References
4.1. Materials
[1] S. Roy, J. Chikkerur, S.C. Roy, A. Dhali, A.P. Kolte, M. Sridhar, A.K. Samanta, J.
Food Sci. 83 (2018) 2699–2709.
D-Galactose, D-tagatose, and D-talose were purchased from FUJIFILM
Wako Pure Chemical (Osaka, Japan). Other chemicals (PIPES, MOPS,
sodium hydroxide, and acetonitrile) were purchased from FUJIFILM
Wako, Dojindo (Kumamoto, Japan) or Nacalai Tesque (Kyoto, Japan).
[2] S.H. Bhuiyan, Y. Itami, Y. Rokui, T. Katayama, K. Izumori, J. Ferment. Bioeng. 85
(1998) 539–541.
[3] K. Beerens, T. Desmet, W. Soetaert, J. Ind. Microbiol. Biotechnol. 39 (2012)
823–834.
[4] Y.-R. Lim, D.-K. Oh, Appl. Microbiol. Biotechnol. 91 (2011) 229–235.
[5] W. Zhang, T. Zhang, B. Jiang, W. Mu, Biotechnol. Adv. 35 (2017) 267–274.
[6] D. Gao, N. Kawai, T. Nakamura, F. Lu, Z. Fei, T. Tamiya, Neurol. Med.-Chir. 53
(2013) 365–374.
4.2. Treatment of galactose with a buffer solution under subcritical water
conditions
[7] Y. Lu, G. Levin, T.W. Donner, Diabetes Obes. Metabol. 10 (2008) 109–134.
[8] J. De Bruijn, A. Kieboom, H. Van Bekkum, Rec. Trav. Chim. Pays-Bas. 105 (1986)
176–183.
Briefly, galactose was dissolved at a concentration of 5 wt% in
10 mmol/L sodium phosphate buffer, PIPES buffer, or MOPS buffer (pH
7.0) to prepare the starting mixture. PIPES buffer was prepared as
follows. PIPES (3.02 g, 10 mmol) was dissolved in 1 L of 10 mmol/L of
sodium hydroxide aqueous solution. To the solution, was added
[9] H. Isbell, H. Frush, C. Wade, C. Hunter, Carbohydr. Res. 9 (1969) 163–175.
[10] T. Nagasawa, K. Sato, T. Kasumi, J. Appl. Glycosci. 64 (2017) 27–32.
[11] C. Usuki, Y. Kimura, S. Adachi, Food Sci. Technol. Res. 13 (2007) 205–209.
[12] D.M. Gao, T. Kobayashi, S. Adachi, Food Chem. 175 (2015) 465–470.
4

Y. Onishi, et al.
CarbohydrateResearch493(2020)108031
[13] D.M. Gao, T. Kobayashi, S. Adachi, Biosci. Biotechnol. Biochem. 80 (2016)
998–1005.
[20] A. Sınağ, S. Gülbay, B. Uskan, M. Canel, Energy Convers. Manag. 51 (2010)
612–620.
[14] N. Soisangwan, D.-M. Gao, T. Kobayashi, P. Khuwijitjaru, S. Adachi, J. Food
Process. Eng. 40 (2017).
[21] F. Jin, J. Yun, G. Li, A. Kishita, K. Tohji, H. Enomoto, Green Chem. 10 (2008)
612–615.
[15] P. Khuwijitjaru, N. Milasing, S. Adachi, Sci. Eng. Health Stud. 12 (2018) 159–167.
[16] D.-M. Gao, T. Kobayashi, S. Adachi, Food Sci. Technol. Res. 21 (2015) 801–803.
[17] B.M. Kabyemela, T. Adschiri, R.M. Malaluan, K. Arai, Ind. Eng. Chem. Res. 38
(1999) 2888–2895.
[22] S.P. Gould, J. Dairy Sci. 23 (1940) 227.
[23] I. Delidovich, R. Palkovits, Green Chem. 18 (2016) 5822–5830.
[24] S.S. Toor, L. Rosendahl, A. Rudolf, Energy 36 (2011) 2328–2342.
[25] I. Delidovich, M.S. Gyngazova, N. Sánchez-Bastardo, J.P. Wohland, C. Hoppe,
P. Drabo, Green Chem. 20 (2018) 724–734.
[18] H. Kishida, F. Jin, X. Yan, T. Moriya, H. Enomoto, Carbohydr. Res. 341 (2006)
2619–2623.
[26] T. Kobayashi, P. Khuwijitjaru, S. Adachi, J. Appl. Glycosci. 63 (2016) 99–104.
[27] Y. Furushiro, T. Kobayashi, J. Appl. Glycosci. 67 (2020) 11–15.
[19] A. Kruse, A. Gawlik, Ind. Eng. Chem. Res. 42 (2003) 267–279.
5