Synthesis of natural/13C-enriched D-tagatose from natural/13C-enriched D-fructose

Article abstract of DOI:10.1016/j.carres.2021.108377

A concise, easily scalable synthesis of a rare ketohexose, D-tagatose, was developed, that is compatible with the preparation of D-[UL-¹³C₆]tagatose. Epimerization of the widely available and inexpensive ketohexose D-fructose at the C-4 position via an oxidation/reduction (Dess-Martin periodinane/NaBH₄) was a key step in the synthesis. Overall, fully protected natural D-tagatose (3.21 g) was prepared from D-fructose (9 g) on a 50 mmol scale in 23% overall yield, after five steps and two chromatographic purifications. D-[UL-¹³C₆]Tagatose (92 mg) was prepared from D-[UL-¹³C₆]fructose (465 mg, 2.5 mmol) in 16% overall yield after six steps and four chromatographic purifications.

Full text of DOI:10.1016/j.carres.2021.108377

Journal Pre-proof
13
13
Synthesis of Natural/ C-Enriched D-Tagatose from Natural/ C-Enriched D-
Fructose
Mojmír Suchý, Thomas A. Charlton, Robert N. Ben, Adam J. Shuhendler
PII:
S0008-6215(21)00146-4
DOI:
https://doi.org/10.1016/j.carres.2021.108377
CAR 108377
Reference:
To appear in: Carbohydrate Research
Received Date: 29 April 2021
Revised Date: 8 June 2021
Accepted Date: 14 June 2021
Please cite this article as: M. Suchý, T.A. Charlton, R.N. Ben, A.J. Shuhendler, Synthesis of
13
13
Natural/ C-Enriched D-Tagatose from Natural/ C-Enriched D-Fructose, Carbohydrate Research,
https://doi.org/10.1016/j.carres.2021.108377.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2021 Elsevier Ltd. All rights reserved.

Synthesis of Natural/13C-Enriched D-Tagatose from Natural/13C-Enriched D-Fructose
Mojmír Suchý,1,2 Thomas A. Charlton,1 Robert N. Ben,1 Adam J. Shuhendler1,2*
1Department of Chemistry & Biomolecular Sciences, University of Ottawa, Ottawa, Ontario,
Canada
2University of Ottawa Heart Institute, Ottawa, Ontario, Canada
*Corresponding author
Abstract
A concise, easily scalable synthesis of a rare ketohexose, D-tagatose, was developed, that is
compatible with the preparation of D-[UL-13C6]tagatose. Epimerization of the widely available
and inexpensive ketohexose D-fructose at the C-4 position via an oxidation/reduction (Dess-
Martin periodinane/NaBH4) was a key step in the synthesis. Overall, fully protected natural D-
tagatose (3.21 g) was prepared from D-fructose (9 g) on a 50 mmol scale in 23% overall yield,
after five steps and two chromatographic purifications. D-[UL-13C6]Tagatose (92 mg) was
prepared from D-[UL-13C6]fructose (465 mg, 2.5 mmol) in 16% overall yield after six steps and
four chromatographic purifications.
Graphical abstract
Keywords
D-tagatose, D-fructose, D-[UL-13C6]tagatose, D-[UL-13C6]fructose, isotopic labelling, rare
carbohydrates
1

Introduction
D-Tagatose (1, Figure 1) is a naturally occurring rare ketohexose found in small quantities in
various dairy products. The sweetness of D-tagatose (92%) is on par with that of the widely used
sucrose, however with no cooling effect or aftertaste [1]. The use of D-tagatose is associated
with a decreased rate of tooth decay [2], and its low caloric value leads to weight loss at
medically desirable rates among healthy individuals [3]. D-Tagatose is now considered a safe
energy-reducing healthy lifestyle-compatible artificial sweetener [1].
The chemical structure of D-tagatose (1) and its relationship to the corresponding C-4 epimer D-
fructose (2) is depicted in Figure 1. As determined by detailed NMR analysis [4], D-tagatose (1)
predominantly (> 90%) occurs in pyranose form (79% α-D-tagatopyranose, 14% β-D-
tagatopyranose, Figure 1) in aqueous solution. Under identical conditions, the major anomer of
D-fructose (2) in solution is β-D-fructopyranose (72%), however a substantial proportion of β-D-
fructofuranose (20%) is also present [4]. To simplify the discussion, the most predominant
forms of D-tagatose (α-D-tagatopyranose 1, Figure 1) and D-fructose (β-D-fructopyranose 2,
Figure 1) will be used, and the corresponding carbohydrates will be referred to as D-tagatose and
D-fructose.
2

Figure 1. Isomeric forms of D-tagatose (1) and D-fructose (2) present in aqueous solution. The
most prevalent forms, α-D-tagatopyranose and β-D-fructopyranose (boxes), will be used to
represent 1 and 2 in all subsequent graphics associated with this work.
The favourable artificial sweetener properties associated with D-tagatose (1) have sparked
significant interest by the food industry to develop methodologies allowing for the economical
production of multi-ton quantities per year [1]. Biotransformation-based approaches, primarily
based on enzymatic isomerization of readily available and cheaper carbohydrate sources (e. g.
whey, milk) are now used to produce D-tagatose (1) on an industrial scale [1,5]. Small scale
chemical synthesis of D-tagatose (1) and its protected derivatives in a laboratory setting appears
to be significantly less developed, with no recent examples (since year 2000) describing enzyme-
free organic synthesis of D-tagatose (1) or its protected derivatives.
The synthetic approaches to D-tagatose (1) can be divided into two subgroups based on the
carbohydrates utilized as starting materials: D-galactose-based (Scheme 1, left) and D-fructose-
based (Scheme 1, right). The first example of the direct isomerisation of D-galactose (3) to D-
tagatose (1) is based on heating 3 in pyridine for an extended period, followed by a tedious
3

purification to afford a negligible yield (7%) of the desired D-tagatose (1, Scheme 1) [6].
Treatment of D-galactose (3) with dibenzylamine, followed by catalytic hydrogenation and
diazotization/hydrolysis afforded D-tagatose (1) in 16% overall yield (Scheme 1) [7]. More
recently, D-galactose (3) was converted to D-tagatose (1) upon alkaline isomerization in the
presence of phenylboronic acid. The yield was satisfactory (53%, Scheme 1) and the
methodology has also been extended to produce D-[1-13C]tagatose and D-[2-13C]tagatose [8]. In
addition to D-galactose (3), D-galacturonic acid (4) was converted to D-tagatose (1) over 4 steps
and 25% overall yield, with the reduction of the D-galacturonic acid methyl ester (LiAlH4) being
the key step in the synthesis (Scheme 1) [9]. A multistep process involving the formation of
bicyclic derivatives has also been utilized to generate D-tagatose (1) from 1-Bn-β-D-galactose (5,
Scheme 1) in seven steps and good overall yield (76%) [10].
To the best of our knowledge, there are only two examples wherein D-tagatose (1) or its
protected form, was synthesized using D-fructose (2) as a starting material. Conversion of 1,3-di-
O-isopropylidene-β-D-fructopyranoside (6, Scheme 1) to 1,2-di-O-Ac-3,4-di-O-ethylidene-β-D-
tagatopyranose (7, Scheme 1) in four steps was described in 1999 [11], however, the
deprotection to 1 was not carried out. Similarly, the relative stereochemical relationship of C4 in
D-tagatose (1) and D-fructose (2) was previously exploited to generate 1. In this approach,
tosylation of D-fructose (2; 1,6 positions) followed by the formation of acetonide (2,3 positions)
afforded 8 in 32% yield starting from 2 (Scheme 1) [12]. An oxidation/reduction (DMSO/Ac2O
then NaBH4) resulted in the formation of intermediate 9 bearing the C-4-tagato configuration
(Scheme 1). Two step deprotection (NaHg in MeOH/H2O then reflux in aqueous AcOH)
furnished the desired D-tagatose (1) in 6 steps and 14% overall yield [12]. The toxicity and
environmental concerns associated with the use of mercury, even as an amalgam, encouraged the
search for alternative methods to generate D-tagatose from D-fructose.
4

Scheme 1 Previous synthetic approaches to D-tagatose (1) based on D-galactose (3) or related
compounds 4 or 5 (left) and D-fructose (2, right) compared to the presented work (bottom).
Recently we have initiated research devoted to understanding the structure-activity relationship
of ketohexose analogs through metabolomic-based studies. During this work, we were required
to prepare functionalized ketohexose derivatives with 13C-enrichment in order to enable our
metabolomics approach. To satisfy this need, we have developed a concise and efficient
synthetic methodology allowing for the conversion of inexpensive D-fructose (2) to D-tagatose
(1). Herein we present a synthetic method that works consistently on a small (2.5 mmol) or large
5

(50 mmol) scales, and is fully compatible with the preparation of multi-milligram quantities of
rare and expensive D-[UL-13C6]tagatose (1a).
Results and discussion
While a complex mixture of pyranose and furanose forms [4] (Figure 1) exists in solution, it was
shown previously [13] that reaction with MeOH in the presence of a catalytic amount of acid
leads to the formation of methyl-D-fructofuranoside (anomeric mixture). Subsequent protection
of the 1,3-positions (acetonide) then leads to the formation of a single isomer, methyl-1,3-di-O-
isopropylidene-α-D-fructofuranoside [14] (10, Scheme 2). Among different reagents for the acid
catalyzed formation of methyl-D-fructofuranoside {H2SO4 [15], p-toluenesulfonic acid (p-TSA)
[14], ion exchange resin Amberlyst 15 [13]} we proceeded with Amberlyst 15. In addition to
excellent yields, the ion exchange resin can be filtered off, greatly simplifying workup and
purification. The resulting filtrate is concentrated, dissolved in MeOH and 2,2-
dimethoxypropane (acetone dimethyl acetal), followed by the addition of a catalytic amount of p-
TSA [14]. We have first attempted to obtain crude 10 by extraction, however the recovery of the
product is not satisfactory (< 40%) on a small scale (2.5 mmol), requiring isolation of 10 [16] by
flash column chromatography (FCC, see the Experimental and Supporting Information). On the
other hand, when working on 50 mmol scale, p-TSA was neutralized with Et3N, the reaction
mixture was concentrated and sufficiently pure crude methyl-1,3-di-O-isopropylidene-α-D-
fructofuranoside (10) was isolated upon partitioning the residue between brine and
CH2Cl2/MeOH (9:1). With intermediate 10 in hand, we proceeded with the selective protection
of the C-6-hydroxyl group. Since we intended to carry out one-step global deprotection at the
end of the synthesis, we have selected the methoxymethyl (MOM) protecting group. The MOM
group is readily introduced, is acid sensitive, and possesses reasonably low molecular mass (MW
45) towards atom economy. Treatment of methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside
(10) with chloromethyl methyl ether (MOM-Cl) afforded 6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-fructofuranoside (11, Scheme 2) in moderate yield. The reaction is
somewhat complicated by the presence of two hydroxy groups in 10, with the best results
achieved when toluene (CH2Cl2 giving ca. 5% lower yields, and unsatisfactorily slow progress
6

was observed in THF) was used as a solvent and N,N-diisopropylethylamine (DIPEA or
Hünnig’s base) was used as a base (see Experimental and Supporting information).
Having prepared 6-MOM-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (11), we
proceeded with the oxidation/reduction reaction sequence to invert C-4-fructo to C-4-tagato
stereochemistry. Treatment of 11 with Dess-Martin periodinane [17] in CH2Cl2 proceeded
smoothly with the formation of unstable ketone 12 (Scheme 2). Ketone 12 was isolated by
aqueous extraction, however, the attempts to obtain this intermediate in pure form led to
extensive decomposition [18]. Crude ketone 12 was dissolved in EtOH, and brief treatment with
NaBH4 at low temperature (0 °C) resulted in the formation of 6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-tagatofuranoside (13, Scheme 2). While a small amount of the starting
material (compound 11, C-4-fructo epimer) was obtained, the C-4-tagato inverted product 13 can
be consistently obtained in ca. 5:1 molar ratio. The two diastereomers are readily separated by
FCC. The stereoselectivity of the reduction is rationalized based on the steric hindrance
facilitating the hydride attack from the α-face. Similar outcomes were observed previously when
performing the C-4-fructo to C-4-tagato inversion of 8 to 9 (Scheme 1) [12]. The identity of 6-
MOM-methyl-1,3-di-O-isopropylidene-α-D-tagatofuranoside (13) was verified by detailed
1
spectroscopic analysis. H NMR spectra associated with 13 (red, Figure 2) and 11 (green, Figure
2; see the full view in the Supporting Information) were compared showing remarkable
differences. A similar pattern was also observed when 13C NMR spectra associated with 13 and
11 were compared (Supporting Information). As expected, high resolution mass spectra (HR-
MS) of 11 and 13 showed the same value, 301.1 [M + Na]+, further confirming the two
molecules are epimers. When performing the reaction sequence on 50 mmol scale, 3.2 g (11.5
mmol) of 6-MOM-methyl-1,3-di-O-isopropylidene-α-D-tagatofuranoside (13) was obtained from
9 g of D-fructose (23% overall yield) after five steps and two chromatographic purifications
(Scheme 2).
7

Scheme 2 Synthesis of D-tagatose and D-[UL-13C6]tagatose from D-fructose and D-[UL-
13C6]fructose. All [UL-13C6]-enriched derivatives bear the same number as natural carbohydrate
containing molecules containing an “a” suffix, e.g. D-tagatose (1), D-[UL-13C6]tagatose (1a).
To complete the synthesis of D-tagatose, we evaluated several different methods for the acid-
mediated hydrolysis of 13. Since heating in diluted HCl led to the darkening of the reaction
mixture and low yields of 1, we first investigated the use of Amberlyst 15 in dioxane/water.
While the acetal-based protecting groups (positions 1, 2 and 3) were completely removed after
stirring for 48 hours at room temperature, the MOM group resisted the removal even upon
extending the reaction time. The resulting 6-MOM-D-tagatofuranose (14, mixture of anomers)
was obtained in 69% yield (Scheme 3, Supporting Information) after FCC purification. When 6-
MOM-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (11) was treated under identical
conditions, the corresponding 6-MOM-D-fructofuranose (16, mixture of anomers) was isolated
as a major product (55%, 2:1 ratio, Scheme 2) along with 6-MOM-methyl-α-D-fructofuranoside
(15, 28%, Scheme 3). Compound 15 was still present even upon extending the reaction time to
96 h.
8

Scheme 3 Amberlyst 15-mediated partial deprotection of 11 and 13
When aqueous HCl (1 M) was used, complete deprotection of 13 took place (after 48 h) with the
formation of D-tagatose (1) in good yield after FCC purification and lyophilization (69%,
Scheme 2). The 1H and 13C NMR spectra of synthetic D-tagatose and corresponding
commercially sourced D-tagatose are in excellent agreement (Figure 3, panel A), and remarkable
differences are observed when the spectra associated with D-tagatose (1) are compared to those
associated with D-fructose (2, Figure 3, panel A).
The presented synthetic methodology offers several advantages, compared to a previously
described method utilizing an oxidation/reduction reaction sequence as a key step in converting
D-fructose (2) to D-tagatose (1) [12] (Scheme 1). The overall yield is slightly higher (16% over
six steps, compared to 14% over six steps described previously), it is significantly more atom
economical with 35% of the molecular mass lost upon global deprotection as compared to 66%
in the previous synthesis, and the use of standard FCC for the fully deprotected product isolation
simplifies the procedure and makes it more broadly accessible. Moreover, while we could
achieve global deprotection in one step under mild conditions (Amberlyst 15 in dioxane/1 M
HCl), two steps requiring global deprotection involving toxic reagents (NaHg) and harsh
conditions (reflux in aqueous AcOH) was utilized in the previous synthesis [12].
9

The newly developed methodology was then applied for the preparation of expensive and
scarcely available D-[UL-13C6]tagatose (1a) on a 2.5 mmol scale [19]. While the preparation of
D-[1-13C]tagatose and D-[2-13C]tagatose has been described previously [8], literature reports
describing the preparation of D-[UL-13C6]tagatose (1a) were not available. In order to
distinguish between the natural and 13C-enriched derivatives, all [UL-13C6]-enriched derivatives
bear an “a” suffix, e.g. D-tagatose (1), D-[UL-13C6]tagatose (1a). Treatment of D-[UL-
13C6]fructose (2a) with Amberlyst 15 in MeOH, followed by protecting the 1,3-positions as
acetonide resulted in the formation of [UL-13C6]-methyl-1,3-di-O-isopropylidene-α-D-
fructofuranoside (10a, Scheme 2) in reasonable yield. All 13C-enriched molecules were
characterized by multiple 1- and 2-D NMR techniques (1H NMR, 13C-decoupled 1H NMR, 13C
1
NMR, 13C-13C COSY) as well as HR-MS spectrometry. H NMR spectrum of 10a possessed a
complex pattern due to the presence of multiple short- and long-distance heteronuclear (1H-13C)
couplings. The 13C-decoupled 1H NMR spectrum associated with 10a overlayed with 1H NMR
spectrum of natural 10 is depicted in the Supporting Information, and excellent agreement
between the two spectra is observed although some hyperfine structure in 13C-decoupled 1H
NMR spectrum associated with 10a is lost due to the decoupling. The signals associated with
13C-enriched carbons (splitting pattern consistent with the presence of multiple 13C-13C short-
and long-distance couplings) were detected in the 13C NMR spectrum of 10a, along with signals
due to the presence of “natural” carbon (one signal due to the methyl group in position 2, three
signals due to the acetonide in positions 1 and 3, see the Supporting Information). Moreover, all
expected correlations were detected in the 13C-13C COSY NMR spectrum (Supporting
Information). A difference of 6 Da was observed in the HR-MS spectrum between the natural
10, 257.1 [M + Na]+ and 13C-enriched 10a, 263.1 [M + Na]+ further confirming the successful
preparation of 13C-enriched acetonide 10a. The MOM protecting group was introduced as
described previously for “natural” analog 11, furnishing [UL-13C6]-6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-fructofuranoside (11a) in moderate 44% yield (Scheme 2). All spectral data
resulting from 11a exhibited the features described above, with more detail provided in Figure 2
and the Supporting Information. C-4 epimerization of 11a (Dess-Martin periodinane/NaBH4)
proceeded in excellent yield (86%, Scheme 2) furnishing [UL-13C6]-6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-tagatofuranoside (13a). Overall, 465 mg (2.5 mmol) of D-[UL-13C6]fructose
(2a) was converted to 136 mg of 13a (0.48 mmol, 19% overall after five steps and three FCC
10

purifications). An overlay of 13C-decoupled 1H NMR spectra associated with 11a and 13a with
those associated with “natural” carbohydrate-containing derivatives 11 and 13 are depicted in
Figure 2.
Figure 2 13C-decoupled 1H NMR spectrum of [UL-13C6]-6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-tagatofuranoside (13a, black) overlaid with 1H NMR spectrum of 6-MOM-
methyl-1,3-di-O-isopropylidene-α-D-tagatofuranoside (13, red), 1H NMR spectrum of 6-MOM-
methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (11, green) and 13C-decoupled 1H NMR
spectrum of [UL-13C6]-6-MOM-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (11a,
purple). The spectra were acquired in CDCl3 at 300 MHz.
With the fully protected 13C-enriched D-tagatose derivative 13a in hand, the global acid-mediated
deprotection (Amberlyst 15 in dioxane/2 M HCl) was carried out (Scheme 2). D-[UL-
13C6]tagatose (1a) was purified by FCC and was obtained in good yield (83%) and significant
quantity (92 mg, 0.49 mmol from 169 mg, 0.6 mmol of 13a) as a crystalline solid upon
lyophilization. All spectral characteristics associated with the synthetic D-[UL-13C6]tagatose
(1a) are depicted in Figure 3, identical spectra acquired for D-fructose (2) and D-[UL-
13C6]fructose (2a) are shown for comparison. A commercial sample of D-[UL-13C6]tagatose (1a)
was also obtained [19], and the corresponding spectra were acquired and compared to those
associated with the synthetic D-tagatose (1) and D-[UL-13C6]tagatose (1a). Notably, 1H NMR
spectra of both, commercial and synthetic D-tagatose (1) and 13C-decoupled 1H NMR spectra
(both commercial and synthetic) of D-[UL-13C6]tagatose (1a) are in good agreement, although
some hyperfine structure is lost due to the decoupling. An agreement between commercial and
synthetic samples of D-[UL-13C6]tagatose (1a) was also remarkable, when their 13C NMR spectra
and 13C-13C COSY NMR spectra were compared (Figure 3). In HR-MS an expected increase by
6 Da was detected, yielding 209.1 [M + Na]+.
11

Figure 3 Detailed spectral characterization of synthetic D-tagatose (1) and D-[UL-13C6]tagatose
(1a). Panel A: 13C-decoupled 1H NMR spectrum of the commercial D-[UL-13C6]tagatose (1a,
black) overlayed with the 13C-decoupled 1H NMR spectrum of the synthetic D-[UL-13C6]tagatose
(1a, red), 1H NMR spectrum of the synthetic D-tagatose (1, green), 1H NMR spectrum of the
commercial D-tagatose (1, purple) and 1H NMR spectrum of the commercial D-fructose (2, blue).
Panel B: 13C NMR spectrum of the commercial D-tagatose (1, black) overlayed with 13C NMR
spectrum of the synthetic D-tagatose (1, red) and 13C NMR spectrum of the commercial D-
fructose (2, green). Panel C: 13C NMR spectrum of the commercial D-[UL-13C6]tagatose (1a,
black) overlayed with the 13C NMR spectrum of the synthetic D-[UL-13C6]tagatose (1a, red) and
13C NMR spectrum of the commercial D-[UL-13C6]fructose (2a, green). Panel D: 13C-13C COSY
NMR spectrum of the commercial D-[UL-13C6]tagatose (1a, bottom), 13C-13C COSY NMR
spectrum of the synthetic D-[UL-13C6]tagatose (1a, center) and 13C-13C COSY NMR spectrum of
12

the commercial D-[UL-13C6]fructose (2a, top). All spectra were acquired in D2O at 300 MHz
(1H NMR) or 75 MHz (13C NMR).
Conclusions
In conclusion we have developed a new synthetic methodology suitable for the preparation of the
rare ketohexose D-tagatose (1), along with the 13C-enriched counterpart D-[UL-13C6]tagatose
(1a). The main advantages associated with our synthetic approach are short time (the whole
synthetic sequence can be completed in one week), good reliability (similar yields are obtained
upon multiple repetitions of synthetic sequence), reasonable atom economy through the careful
selection of protecting groups, the use of standard FCC for the purification of intermediates and
final products as well as full compatibility with the preparation of expensive 13C-enriched
carbohydrates (similar yields are obtained when altering between “natural” and 13C-enriched
carbohydrates). The presented methodology allows for the preparation of scarce and very
expensive D-[UL-13C6]tagatose (1a) on multi-milligram quantities, providing scientists interested
in metabolomic studies with rare carbohydrates with rapid and simplified access to 1a. Although
beyond the scope of the presented work we also believe that the robust and reliable synthetic
approach described herein will be compatible with other 13C enrichment patterns and potentially
also radioactive isotopes (14C) present in D-fructose and D-tagatose.
Experimental
General information
All reagents were commercially available. In D-[UL-13C6]fructose (2), all six carbon atoms were
13C enriched (99% enrichment level). All solvents were HPLC grade and used as such, except
for water which was deionized (18.2 MΩ·cm-1) and dichloromethane which was dried over
Al2O3, in a solvent purification system. Organic extracts were dried with Na2SO4 and solvents
were removed under reduced pressure in a rotary evaporator. Aqueous solutions were
lyophilized. Flash column chromatography (FCC) was carried out using silica gel, mesh size
230 – 400 Å. Thin layer chromatography (TLC) was carried out on Al backed silica gel plates;
13

compounds were visualized by orcinol stain. Specific rotations []D were determined by
polarimeter at ambient temperature using a 2 mL, 1 cm path length cell; the units are 10-1 deg
cm2 g-1 and the concentrations are reported in g/100 mL. NMR spectra were recorded on 300
MHz spectrometer; for 1H (300 MHz),  values were referenced as follows CDCl3 (7.26 ppm);
CD3OD (3.31 ppm); D2O (4.79 ppm) for 13C (75 MHz) CDCl3 (77.0 ppm), CD3OD (49.0 ppm).
High resolution mass spectra (HR-MS) were obtained by electron spray ionization (ESI) time-of-
flight (TOF) method.
Methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (10)
An ion exchange resin, Amberlyst 15 (50 mg) was added to a stirred solution of D-fructose (2,
450 mg, 2.5 mmol) in MeOH (25 mL). The mixture was stirred for 18 hours (h) at room
temperature (rt), the resin was filtered off, the filter was washed with MeOH and the filtrate was
concentrated to leave colorless oil.
This material was dissolved in MeOH (2.5 mL), and 2,2-dimethoxypropane (acetone dimethyl
acetal, 2.5 mL), p-toluenesulfonic acid (p-TSA, 5 mg, 2.6  10-5 mol) was added and the mixture
was stirred for 1 h at rt. The reaction was quenched by the addition of saturated NaHCO3
solution (2 drops), the solvent was evaporated, and the residue was subjected to FCC on 50 g
SiO2, CH2Cl2/MeOH (19:1). Evaporation of the eluate afforded colorless oil, methyl-1,3-di-O-
isopropylidene-α-D-fructofuranoside (10, 346 mg, 59% over two steps); [α]D +29.2 (c 1.07,
CHCl3), lit. [16d] [α]D +34.5 (c 0.50, CHCl3), lit. [16c] [α]D +38.6 (c 2.00, CHCl3); RF 0.17
1
[CH2Cl2/MeOH (19:1)]. H NMR (CDCl3) δ 4.24 – 4.12 (m, 1H); 4.04 – 3.90 (m, 4H); 3.88 –
3.71 (m, 2H); 3.36 (s, 0.3H); 3.32 (s, 2.7H); 2.58 (br s, D2O exch., 2H); 1.52 (s, 0.3H); 1.47 (s,
2.7H); 1.38 (s, 2.7H); 1.35 (s, 0.3H). 13C NMR (CDCl3) δ 101.2, 98.7, 87.9, 79.4, 77.7, 62.8,
61.8, 48.7, 27.9, 19.1. HRMS (ESI) m/z found 257.1010 [M + Na]+ (257.1001 calcd for
C10H18O6Na). Spectral data matched those reported previously [14, 16c, 16d].
6-MOM-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (11), large scale procedure
An ion exchange resin, Amberlyst 15 (1 g) was added to a stirred solution of D-fructose (2, 9 g,
50 mmol) in MeOH (250 mL). The mixture was stirred for 18 h at rt, the resin was filtered off,
the filter was washed with MeOH and the filtrate was concentrated to leave colorless oil.
14

This material was dissolved in MeOH (50 mL), and 2,2-dimethoxypropane (acetone dimethyl
acetal, 50 mL), p-TSA (95 mg, 0.5 mmol) was added, and the mixture was stirred for 1 h at rt.
The reaction was quenched by the addition of Et3N (70 μL, 0.5 mmol), the solvent was
evaporated. The residue was partitioned between brine (50 mL) and CH2Cl2/MeOH (9:1, 2 
100 + 3  50 mL). Combined organic extract was dried and was concentrated; the residue was
co-evaporated with toluene (100 mL) to leave methyl-1,3-di-O-isopropylidene-α-D-
fructofuranoside (10, 9.27 g) as a colorless oil of sufficient purity to be used for the next step.
Methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (10, 9.27 g, 39.57 mmol) was dissolved in
toluene (200 mL), diisopropylethylamine (DIPEA, 8.62 mL, 49.47 mmol) was added and the
mixture was cooled to 0 °C. Freshly distilled chloromethyl methyl ether (MOM-Cl, 3.76 mL,
49.47 mmol) was added drop wise (over 1 min period), the mixture was stirred for 1 h at 0 °C
and for 18 h at rt. The reaction was quenched by the addition of saturated NH4Cl solution and
brine (50 mL each), the mixture was stirred vigorously for 10 min at rt, was transferred into a
separatory funnel and the organic phase was collected. The aqueous phase was extracted with
EtOAc (100 + 2  50 mL), combined organic extract was dried and was concentrated. The
residue was subjected to FCC on 350 g SiO2, hexanes/EtOAc (1:1). Evaporation of the eluate
afforded pale yellow oil, 6-MOM-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside [11,
4.01 g, 29%, based on D-fructose (2)]; [α]D +21.9 (c 0.96, MeOH); RF 0.42 [hexanes/EtOAc
(1:1)], RF 0.21 [petroleum ether/EtOAc (7:3)], the values were determined upon two
1
developments of the same TLC plate. H NMR (CDCl3) δ 4.68 (s, 2H); 4.21 (ddd, J = 5.5, 2.5,
1.5 Hz, 1H); 4.02 – 3.87 (m, 4H); 3.75 (dd, J = 10.5, 7.0, 1H); 3.69 (dd, J = 9.5, 5.5, 1H); 3.38
(s, 3H); 3.32 (s, 3H); 2.52 (br s, D2O exch., 1H); 1.44 (s, 3H); 1.36 (s, 3H). 13C NMR (CDCl3) δ
101.9, 98.5, 96.6, 86.2, 79.4, 78.4, 68.4, 61.7, 55.2, 48.8, 27.7, 19.5. HRMS (ESI) m/z found
301.1259 [M + Na]+ (301.1263 calcd for C12H22O7Na).
6-MOM-methyl-1,3-di-O-isopropylidene-α-D-tagatofuranoside (13)
Dess-Martin periodinane (6.7 g, 15.81 mmol) was added to a stirred solution of 6-MOM-methyl-
1,3-di-O-isopropylidene-α-D-fructofuranoside (11, 4 g, 14.37 mmol) in CH2Cl2 (200 mL). The
mixture was stirred for 90 min at rt, the reaction was quenched by the addition of Na2S2O3 ·
5H2O solution (10%) and saturated NaHCO3 solution (150 mL each). The mixture was stirred
15

vigorously for 10 min at rt, was transferred into a separatory funnel and the organic phase was
collected. The aqueous phase was extracted with CH2Cl2 (2  60 mL), combined organic extract
was dried and was concentrated to leave crude ketone 12 used for the next step without further
purification.
The residue was dissolved in EtOH (450 mL) and the resulting solution was cooled to 0 °C.
NaBH4 (652 mg, 17.25 mmol) was added, the mixture was stirred for 15 min at 0 °C, the solvent
was evaporated, and the residue was partitioned between brine (250 mL) and CH2Cl2 (120 + 2 
60 mL). Combined organic extract was dried and was concentrated; the residue was subjected to
FCC on 250 g SiO2, petroleum ether/EtOAc (7:3). Evaporation of the eluate afforded pale
yellow oil, 6-MOM-methyl-1,3-di-O-isopropylidene-α-D-tagatofuranoside (13, 3.21 g, 80%);
[α]D +24.6 (c 1.04, MeOH); RF 0.32 [petroleum ether/EtOAc (7:3)], the value was determined
1
upon two developments of the same TLC plate. H NMR (CDCl3) δ 4.68 (dd, J = 9.5, 6.5 Hz,
2H); 4.60 (dd, J = 6.5, 5.5 Hz, 1H); 4.22 (m, 1H); 3.98 (d, J = 5.5 Hz, 1H); 3.88 (dd, J = 15.5,
12.5 Hz, 2H); 3.84 (d, J = 3.0 Hz, 1H); 3.72 (dd, J = 11.0, 7.0 Hz, 2H); 3.36 (s, 3H); 3.22 (s,
3H); 2.75 (br s, D2O exch., 1H); 1.45 (s, 3H); 1.41 (s, 3H). 13C NMR (CDCl3) δ 100.3, 99.2,
96.7, 79.5, 73.2, 71.8, 66.6, 62.1, 55.2, 48.7, 27.3, 20.4. HRMS (ESI) m/z found 301.1257 [M +
Na]+ (301.1263 calcd for C12H22O7Na).
6-MOM-D-tagatofuranose (14)
Amberlyst 15 (30 mg) was added to a stirred solution of 6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-tagatofuranoside (13, 140 mg, 0.5 mmol) in dioxane and water (1 mL each).
The mixture was stirred for 48 h at rt, the resin was filtered off (Pasteur pipette with a cotton
plug), the filter was washed with MeOH and the filtrate was concentrated. The residue was
subjected to FCC on 10 g SiO2, CH2Cl2/MeOH (9:1). Evaporation of the eluate afforded
colorless oil, 6-MOM-D-tagatofuranose (14, 77 mg, 69%); [α]D -7.0 (c 1.32, MeOH); RF 0.11
1
[CH2Cl2/MeOH (9:1)]. H NMR (CD3OD) δ 4.63 (m, 2H); 4.59 – 4.06 (m, 3H); 3.88 – 3.44 (m,
4H); 3.37 (m, 3H). 13C NMR (CD3OD) δ 106.5, 104.2, 97.8, 98.7, 80.2, 79.4, 79.3, 73.2, 73.1,
72.1, 70.2, 68.5, 67.8, 64.2, 64.0, 55.5. HRMS (ESI) m/z found 247.0797 [M + Na]+ (247.0794
calcd for C8H16O7Na).
6-MOM-D-fructofuranose (16)
16

Amberlyst 15 (30 mg) was added to a stirred solution of 6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-fructofuranoside (11, 140 mg, 0.5 mmol) in dioxane and water (1 mL each).
The mixture was stirred for 96 h at rt, the resin was filtered off (Pasteur pipette with a cotton
plug), the filter was washed with MeOH and the filtrate was concentrated. The residue was
subjected to FCC on 10 g SiO2, CH2Cl2/MeOH (9:1). Two fractions were obtained upon
evaporation of eluate. Less polar fraction, colorless oil, 6-MOM-methyl-α-D-fructofuranoside
1
(15, 33 mg, 28%), [α]D +63.7 (c 0.60, MeOH); RF 0.26 [CH2Cl2/MeOH (9:1)]. H NMR
(CD3OD) δ 4.66 (s, 2H); 4.02 (d, J = 4.0 Hz, 1H); 3.93 (m, 1H); 3.84 (dd, J = 6.5, 4.0 Hz, 1H);
3.76 - 3.59 (m, 4H); 3.31 (s, 3H); 3.31 (s, 3H). 13C NMR (CD3OD) δ 109.3, 97.7, 83.1, 82.7,
79.7, 68.9, 60.2, 55.6, 48.8. HRMS (ESI) m/z found 261.0950 [M + Na]+ (261.0950 calcd for
C9H18O7Na). More polar fraction, colorless oil, 6-MOM-D-fructofuranose (16, 62 mg, 55%);
1
[α]D -1.8 (c 0.88, MeOH); RF 0.11 [CH2Cl2/MeOH (9:1)]. H NMR (CD3OD) δ 4.67 (s, 1.7H);
4.66 (s, 0.3H); 4.08 - 3.99 (m, 2H); 3.92 - 3.81 (m, 1H); 3.77 - 3.47 (m, 4H); 3.67 (m, 3H). 13C
NMR (CD3OD) δ 106.0, 103.3, 97.7 (2  C), 84.1, 81.8, 81.5, 78.8, 77.5, 77.2, 70.5, 68.9, 65.1,
64.5, 55.6 (2  C). HRMS (ESI) m/z found 247.0789 [M + Na]+ (247.0794 calcd for
C8H16O7Na).
D-Tagatose (1)
Amberlyst 15 (160 mg) was added to a stirred solution of 6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-tagatofuranoside (13, 756 mg, 2.72 mmol) in dioxane and 1 M HCl (5 mL
each). The mixture was stirred for 48 h at rt, the resin was filtered off (Pasteur pipette with a
cotton plug), the filter was washed with water, the filtrate was cooled to 0 °C. The pH was
adjusted to ~ 6.5 – 7 (20% NaOH solution), the liquids were evaporated, and the residue was
subjected to FCC on 40 g SiO2, CH2Cl2/MeOH (4:1). Evaporation of the eluate afforded
colorless oil, this was dissolved in small amount of water (~ 4 ml), was transferred into a Falcon
tube and was lyophilized to afford colorless crystalline solid, D-tagatose (1, 338 mg, 69%); [α]D -
1
5.8 (c 0.83, water), lit. [20] [α]D -5.0 (c 5.00, water); RF 0.07 [CH2Cl2/MeOH (4:1)]. H NMR
(D2O) δ 4.50 – 3.54 (m, 7H). 13C NMR (D2O) δ 102.5, 98.3, 98.1, 80.0, 70.9, 70.6, 69.8, 69.2,
66.3, 63.9, 63.7, 63.5, 62.6, 62.2, 60.9, 60.1. Spectral data matched those reported previously [4]
and those associated with the commercial sample of D-tagatose (1).
[UL-13C6]-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (10a)
17

Amberlyst 15 (50 mg) was added to a stirred solution of D-[UL-13C6]fructose (2a, 465 mg, 2.5
mmol) in MeOH (25 mL). The mixture was stirred for 18 h at rt, the resin was filtered off, the
filter was washed with MeOH and the filtrate was concentrated to leave colorless oil.
This material was dissolved in MeOH (2.5 mL), and 2,2-dimethoxypropane (acetone dimethyl
acetal, 2.5 mL), p-TSA (5 mg, 2.6  10-5 mol) was added and the mixture was stirred for 1 h at rt.
The reaction was quenched by the addition of saturated NaHCO3 solution (2 drops), the solvent
was evaporated, and the residue was subjected to FCC on 50 g SiO2, CH2Cl2/MeOH (95:5).
Evaporation of the eluate afforded colorless oil, [UL-13C6]-methyl-1,3-di-O-isopropylidene-α-D-
1
fructofuranoside (10a, 308 mg, 51% over two steps); RF 0.17 [CH2Cl2/MeOH (19:1)]. H NMR
(CDCl3) δ 4.48 – 3.48 (m, 7H); 3.37 (d, J = 3.5 Hz, 0.3H); 3.33 (d, J = 3.5 Hz, 2.7H); 2.30 (br s,
1
D2O exch., 2H); 1.53 (s, 0.3H); 1.48 (s, 2.7H); 1.39 (s, 2.7H); 1.36 (s, 0.3H). H NMR{13C}
(CDCl3) δ 4.23 – 3.72 (m, 7H); 3.37 (s, 0.3H); 3.33 (s, 2.7H); 2.30 (br s, D2O exch., 2H); 1.53 (s,
0.3H); 1.48 (s, 2.7H); 1.39 (s, 2.7H); 1.36 (s, 0.3H). 13C NMR (CDCl3) δ 101.9 – 100.6 (m);
98.7 – 97.4 (m); 88.3 – 87.3 (m); 80.1 – 79.5 (m); 78.1 – 75.8 (m); 73.5 – 71.7 (m); 63.5 – 59.9
(m); 48.7 (q, J = 2.0 Hz); 27.8 (t, J = 2.0 Hz); 19.2 (t, J = 1.5 Hz). HRMS (ESI) m/z found
13
263.1213 [M + Na]+ (263.1202 calcd for 12C4 C6H18O6Na).
[UL-13C6]-6-MOM-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (11a)
[UL-13C6]-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (10a, 308 mg, 1.28 mmol) was
dissolved in toluene (13 mL), DIPEA (280 μL, 1.6 mmol) was added and the mixture was cooled
to 0 °C. Freshly distilled MOM-Cl (120 μL, 1.6 mmol) was added, the mixture was stirred for 1
h at 0 °C and for 1 h at rt. The reaction was quenched by the addition of saturated NH4Cl
solution and brine (7 mL each), the mixture was stirred vigorously for 10 min at rt, was
transferred into a separatory funnel and the organic phase was collected. The aqueous phase was
extracted with EtOAc (3  20 mL), combined organic extract was dried and was concentrated.
The residue was subjected to FCC on 60 g SiO2, hexanes/EtOAc (1:1). Evaporation of the eluate
afforded pale yellow oil, [UL-13C6]-6-MOM-methyl-1,3-di-O-isopropylidene-α-D-
fructofuranoside (11a, 161 mg, 44%); RF 0.42 [hexanes/EtOAc (1:1)], RF 0.21 [petroleum
ether/EtOAc (7:3)], the values were determined upon two developments of the same TLC plate.
1H NMR (CDCl3) δ 4.69 (d, J = 6.0 Hz, 2H); 4.49 – 4.12 (m, 3H); 4.00 – 3.92 (m, 1H); 3.78 –
3.43 (m, 3H); 3.39 (d, J = 0.5 Hz, 3H); 3.32 (d, J = 4.0 Hz, 3H); 2.32 (br s, D2O exch., 1H); 1.45
18

1
(s, 3H); 1.37 (s, 3H). H NMR{13C} (CDCl3) δ 4.69 (s, 2H); 4.23 – 4.19 (m, 1H); 4.02 – 3.87
(m, 4H); 3.78 – 3.67 (m, 2H); 3.39 (s, 3H); 3.32 (s, 3H); 2.33 (br s, D2O exch., 1H); 1.45 (s, 3H);
1.37 (s, 3H). 13C NMR (CDCl3) δ 102.5 – 101.3 (m); 86.9 – 85.8 (m); 80.1 – 77.8 (m); 68.7 –
68.1 (m); 62.1 – 61.4 (m); 61.8 (d, J = 2.5 Hz); 55.3 (s); 48.8 (d, J = 3.0 Hz); 27.8 (t, J = 2.0 Hz);
13
19.5 (s). HRMS (ESI) m/z found 307.1474 [M + Na]+ (307.1465 calcd for 12C6 C6H22O7Na).
[UL-13C6]-6-MOM-methyl-1,3-di-O-isopropylidene-α-D-tagatofuranoside (13a)
Dess-Martin periodinane (259 mg, 0.61 mmol) was added to a stirred solution of [UL-13C6]-6-
MOM-methyl-1,3-di-O-isopropylidene-α-D-fructofuranoside (11a, 158 mg, 0.56 mmol) in
CH2Cl2 (9 mL). The mixture was stirred for 90 min at rt, the reaction was quenched by the
addition of Na2S2O3 · 5H2O solution (10%) and saturated NaHCO3 solution (10 mL each). The
mixture was stirred vigorously for 10 min at rt, was transferred into a separatory funnel and the
organic phase was collected. The aqueous phase was extracted with CH2Cl2 (20 + 10 mL),
combined organic extract was dried and was concentrated to leave crude ketone 12a used for the
next step without further purification.
The residue was dissolved in EtOH (32 mL) and the resulting solution was cooled to 0 °C.
NaBH4 (25 mg, 0.67 mmol) was added, the mixture was stirred for 15 min at 0 °C, the solvent
was evaporated, and the residue was partitioned between brine (30 mL) and CH2Cl2 (2  20 + 10
mL). Combined organic extract was dried and was concentrated; the residue was subjected to
FCC on 45 g SiO2, petroleum ether/EtOAc (7:3). Evaporation of the eluate afforded pale yellow
oil, [UL-13C6]-6-MOM-methyl-1,3-di-O-isopropylidene-α-D-tagatofuranoside (13a, 136 mg,
86%); RF 0.32 [petroleum ether/EtOAc (7:3)], the value was determined upon two developments
1
of the same TLC plate. H NMR (CDCl3) δ 4.89 – 4.67 (m, 2H); 4.50 – 3.90 (m, 5H); 3.75 –
3.47 (m, 2H); 3.39 (s, 3H); 3.25 (d, J = 4.0 Hz, 3H); 2.82 (br d, J = 6.0 Hz, D2O exch., 1H); 1.48
1
(s, 3H); 1.44 (s, 3H). H NMR{13C} (CDCl3) δ 4.71 (dd, J = 9.5, 6.5 Hz, 2H); 4.66 – 4.60 (m,
1H); 4.28 – 4.22 (m, 1H); 4.02 – 3.86 (m, 4H); 3.78 – 3.66 (m, 1H); 3.39 (s, 3H); 3.25 (s, 3H);
2.81 (br d, J = 8.5 Hz, D2O exch., 1H); 1.48 (s, 3H); 1.43 (s, 3H). 13C NMR (CDCl3) δ 101.0 –
99.8 (m); 80.2 – 79.0 (m); 73.8 – 71.3 (m); 67.0 – 66.3 (m); 62.5 – 61.8 (m); 60.5 (q, J = 2.5 Hz);
59.9 (m); 59.2 (d, J = 2.5 Hz); 55.3 (s); 48.8 (q, J = 2.0 Hz); 27.4 (t, J = 1.5 Hz); 20.5 (s). HRMS
13
(ESI) m/z found 307.1455 [M + Na]+ (307.1465 calcd for 12C6 C6H22O7Na).
19

D-[UL-13C6]tagatose (1a)
Amberlyst 15 (35 mg) was added to a stirred solution of [UL-13C6]-6-MOM-methyl-1,3-di-O-
isopropylidene-α-D-tagatofuranoside (13a, 169 mg, 0.6 mmol) in dioxane and 2 M HCl (1.1 mL
each). The mixture was stirred for 24 h at rt, the resin was filtered off (Pasteur pipette with a
cotton plug), the filter was washed with water, the filtrate was cooled to 0 °C. The pH was
adjusted to ~ 6.5 – 7 (20% NaOH solution), the liquids were evaporated, and the residue was
subjected to FCC on 20 g SiO2, CH2Cl2/MeOH (4:1). Evaporation of the eluate afforded
colorless oil; this was dissolved in small amount of water (~ 4 ml), was transferred into a Falcon
tube and was lyophilized to afford colorless crystalline solid, D-[UL-13C6]tagatose (1a, 92 mg,
1
83%); [α]D -2.9 (c 0.52, water); RF 0.07 [CH2Cl2/MeOH (4:1)] . H NMR (D2O) δ 4.58 - 3.25
1
(m, 7H). H NMR{13C} (D2O) δ 4.35 - 3.37 (m, 7H). 13C NMR (D2O) δ 98.8 –97.5 (m); 71.4 –
68.7 (m); 66.8 – 65.8 (m); 64.2 – 61.9 (m); 61.2 – 59.9 (m). HRMS (ESI) m/z found 209.0743
[M + Na]+ (209.0733 calcd for 13C6H12O6Na).
Declaration of competing interests
The authors declare no competing financial interest.
Acknowledgements
This work was supported by NSERC Discovery Grant RGPIN 2015-05796 and the Canada
Research Chairs Program 950-230754 (A.J.S.), and funding from NSERC and CIHR (R.N.B.).
We thank Dr. Glenn Facey (University of Ottawa) for developing NMR experiments to perform
the detailed spectroscopic analysis of D-[UL-13C6]tagatose (1a) and 13C-enriched derivatives 10a,
11a and 13a. We thank Prof. A. Beauchemin and Mr. W. Gagné-Monfette for the permission
and assistance with the use of polarimeter.
Appendix A. Supplementary data
20

Detailed spectroscopic characterization associated with compounds 1, 1a, 10, 10a, 11, 11a, 13,
13a, 14-16.
References and notes
1. D.K. Oh, Tagatose: Properties, Applications, and Biotechnological Processes, Appl.
Microbiol. Biotechnol.76 (2007) 1-8, and references cited therein.
2. D. Wong, Sweetener Determined Safe in Drugs, Mouthwashes, and Toothpastes, Dent. Today
19 (2000) 34-35.
3. G.V. Levin, Tagatose, the New GRAS Sweetener and Health Product, J. Med. Food 5 (2002)
23-36.
4. S. Köpper, S. Freimund, The Composition of Keto Aldoses in Aqueous Solution as
Determined by NMR Spectroscopy, Helv. Chim. Acta 86 (2003) 827-843.
5. For selected examples see:
(a) P.S.J Cheetham, A.N. Wootton, Bioconversion of D-Galactose into D-Tagatose, Enzyme
Microb. Technol. 15 (1993) 105-108.
(b) H.J. Oh, H.J. Kim, D.K. Oh, Increase in D-Tagatose Production Rate by Site-Directed
Mutagenesis of L-Arabinose Isomerase from Geobacillus thermodenitrificans, Biotechnol. Lett.
28 (2006) 145-149.
(c) Z. Xu, S. Li, X. Feng, J. Liang, H. Xu, L-Arabinose Isomerase and Its Use for
Biotechnological Production of Rare Sugars, Appl. Microbiol. Biotechnol. 98 (2014) 8869-8878.
(d) Y. Liu, S. Li, H. Xu, L. Wu, Z. Xu, J. Liu, X. Feng, Efficient Production of D-Tagatose Using
a Food-Grade Surface Display System, J. Agric. Food Chem. 62 (2014) 6756-6762.
(e) L. Wen, K. Huang, M. Wei, J. Meisner, Y. Liu, K. Garner, L. Zang, X. Wang, X. Li, J. Fang,
H. Zhang, P.G. Wang, Facile Enzymatic Synthesis of Ketoses, Angew. Chem. Int. Ed. 54 (2015)
12654-12658.
21

(f) S.H. Lee, S.H. Hong, K.R. Kim, D.K. Oh, High-Yield Production of Pure Tagatose from
Fructose by a Three-Step Enzymatic Cascade Reaction, Biotechnol. Lett. 39 (2017) 1141-1148.
(g) Z. Zheng, J. Xie, P. Liu, X. Li, J. Ouyang, Elegant and Efficient Biotransformation for Dual
Production of D-Tagatose and Bioethanol from Cheesy Whey Powder, J. Agric. Food Chem. 67
(2019) 829-835.
(h) J.R. Bober, N.U. Nair, Galactose to Tagatose Isomerisation at Moderate Temperatures with
High Conversion and Productivity, Nat. Commun. 10 (2019) article 4548.
6. J.V. Karabinos, Psicose, Sorbose and Tagatose, Adv. Carbohydrate Res. 7 (1952) 99-136, and
references cited therein.
7. R. Grünnagel, H.D. Haas, Darstellung von D-Tagatose, Liebigs Ann. Chem, 721 (1969) 234-
235.
8. B. Bose, S. Zhao, R. Stenutz, F. Cloran, P.B. Bondo, G. Bondo, B. Hertz, I. Carmichael, A.S.
Serianni, Three Bond C-O-C-C Spin-Coupling Constants in Carbohydrates: Development of a
Karplus Relationship, J. Am. Chem. Soc. 120 (1998) 11158-11173.
9. P.A.J. Gorin, J.K.N. Jones, W.W. Reid, A Synthesis of D-Tagatose from D-Galacturonic Acid,
Can. J. Chem. 33 (1955) 1116-1118.
10. P.L. Barili, G. Berti, G. Catelani, F. D’Andrea, L. Miarelli, New Syntheses of D-Tagatose
and of 1,5-Anhydro-D-tagatose from D-Galactose Derivatives, Carbohydrate Res. 274 (1995)
197-208.
11. M. Frank, R. Mietchen, D. Degenring, D-Tagatose Derivatives from D-Fructose by a Facile
Epimerisation Procedure, Carbohydrate Res. 318 (1999) 167-170.
12. A.A.H. Al-Jobore, R.D. Guthrie, R.D. Wells, A New Route for the Synthesis of D-Tagatose,
Carbohydrate Res. 16 (1971) 474-476.
13. R.H. Furneaux, P.C. Tyler, L.A. Whitehouse, A Short Practical Synthesis of
Deoxymannojirimycin from D-Fructose, Tetrahedron Lett. 34 (1993) 3613-3616; no
experimental procedure is provided in the original reference.
22

14. K. Yu, X. Zhao, W. Wu, Z. Hong, An Efficient Procedure for Synthesis of Fructose
Derivatives, Tetrahedron Lett. 54 (2013) 2788-2790.
15. B.J. Trayner, T.N. Grant, F.G. West, C.I. Cheeseman, Synthesis and Characterization of 6-
Deoxy-6-fluoro-D-fructose as a Potential Compound for Imaging Breast Cancer with PET,
Bioorg. Med. Chem. 17 (2009) 5488-5495.
16. Intermediate 10 has been prepared previously by other routes, see
(a) R. Cortes-Garcia, L. Hough, A.C. Richardson, Acetalation of Sucrose by Acetal Exchange
with Concomitant Fission of the Glycosididc Bond. Some New Acetals of D-Glucose and Methyl
α-D-Fructofuranoside, J. Chem. Soc., Perkin Trans. 1 1981 (1981) 3176-3181.
(b) T. Hanaya, N. Sato, H. Yamamoto, An Efficient Synthesis of Methyl 1,3-O-Isopropylidene-
α-D-Fructofuranoside and 2,3:5,6-Di-O-isopropylidene-D-glucose dimethyl acetal Derivatives
from Sucrose, Carbohydrate Res. 340 (2005) 2494-2501.
(c) A. Lauritsen, R. Madsen, Synthesis of Naturally Occurring Iminosugars from D-Fructose by
the Use of Zinc-Mediated Fragmentation Reaction, Org. Biomol. Chem. 4 (2006) 2898-2905.
(d) P. Abayakoon, R. Epa, M. Petricevic, C. Bengt, J.W.Y. Mui, P.L. van der Peet, Y. Zhang,
J.P. Lingford, J.M. White, E.D. Goddard-Borger, S.J. Williams, Comprehensive Synthesis of
Substrates, Intermediates, and Products of the Sulfoglycolytic Embden-Meyerhoff-Parnas
Pathway, J. Org. Chem. 84 (2019) 2901-2910.
17. D.B. Dess, J.C. Martin, Readily Accessible 12-I-5 Oxidant for the Conversion of Primary
and Secondary Alcohols to Aldehydes and Ketones, J. Org. Chem. 48 (1983) 4155-4156.
18. Decomposition of intermediate ketone when converting 8 to 9 (Scheme 1), like that observed
by us (ketone 12, Scheme 2) was described in ref. 12.
19. While multigram quantities of D-[UL-13C6]fructose (2a) are available from multiple
commercial sources, we have been able to find only two commercial sources of D-[UL-
13C6]tagatose (1a). We were interested in purchasing up to 10 mg of 1a as a spectroscopic
standard, the quote requested from one of the commercial suppliers informed us that 10 mg will
be US$ 300 and it can be delivered within one month.
23

20. E.L. Totton, H.A. Lardy, The Synthesis of D-Tagatose by Biochemical Oxidation and by an
Improved Chemical Method, J. Am. Chem. Soc. 71 (1949) 3076-3078.
24

Declaration of interests
☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests: