The Michaelis-Arbuzov rearrangement of anomeric thiocyanates: synthesis and application of S-glycosyl thiophosphates, thiophosphonates and thiophosphinates as glycosyl donors

Article abstract of DOI:10.1016/j.tetlet.2007.09.162

Reaction of anomeric thiocyanates with triethyl phosphite, dimethyl phenylphosphonite and methyl diphenylphosphinite afforded the corresponding S-glycosyl thiophosphates, thiophosphonates and thiophosphinates in good yields. These derivatives were applied

Full text of DOI:10.1016/j.tetlet.2007.09.162

Available online at www.sciencedirect.com
Tetrahedron Letters 48 (2007) 8482–8486
The Michaelis–Arbuzov rearrangement of anomeric thiocyanates:
synthesis and application of S-glycosyl thiophosphates,
thiophosphonates and thiophosphinates as glycosyl donors
Marta Piekutowska and Zbigniew Pakulski*
Institute of Organic Chemistry, Polish Academy of Sciences, Kasprzaka 44/52, 01-224 Warszawa, Poland
Received 27 July 2007; revised 17 September 2007; accepted 26 September 2007
Available online 29 September 2007
Abstract—Reaction of anomeric thiocyanates with triethyl phosphite, dimethyl phenylphosphonite and methyl diphenylphosphinite
afforded the corresponding S-glycosyl thiophosphates, thiophosphonates and thiophosphinates in good yields. These derivatives
were applied as glycosyl donors in the synthesis of benzyl glycosides and disaccharides with excellent stereoselectivity.
Ó 2007 Elsevier Ltd. All rights reserved.
The Michaelis–Arbuzov reaction (also known as the
Arbuzov reaction or Arbuzov rearrangement) is well
known and widely used in general organic chemistry.1
It results in pentavalent phosphorus derivatives with
the formation of a new carbon–phosphorus bond. In
most common form, the reaction usually requires pro-
longed heating at high temperature.
unsymmetrical O,O,S-trialkyl thiophosphates, however,
examples of the Michaelis–Arbuzov reaction of organic
thiocyanates are extremely rare.4
The effectiveness of glycosylation is a key problem in
oligosaccharide synthesis and strongly depends upon
the nature of the leaving group at the anomeric centre
and its method of activation. Although numerous versa-
tile leaving groups are known,5 no single, universally
applicable method for glycoside bond formation is avail-
able. The need for new, readily available, stable and
reactive glycosyl donors still persists. During our studies
directed towards the synthesis of oligosaccharides6 we
focused our attention on leaving groups containing a
phosphorus atom. In this regard, anomeric thiocya-
nates7,8 are particularly interesting starting materials
for the preparation of anomeric thiophosphorus
derivatives.
Due to their chemical properties,2 thiocyanate deriva-
tives (often classified as pseudohalogen derivatives)
may also be considered as starting materials for the
Michaelis–Arbuzov rearrangement. It has been demon-
strated that alkyl and aryl thiocyanates are highly reac-
tive reagents in the above reaction. The reaction
involves S–CN bond fission and gives alkyl cyanates
and O,O,S-trialkyl thiophosphates (Scheme 1).3 There-
fore, organic thiocyanates are potentially very promising
substrates for the highly regioselective preparation of
In this Letter, we report the preparation of S-glycosyl
thiophosphates, thiophosphonates and thiophosphi-
nates via Michaelis–Arbuzov reaction of glycosyl thio-
cyanates with simple phosphorus reagents. We have
found that all the thiophosphorus derivatives obtained
during this study act as excellent glycosyl donors form-
ing O-glycoside bonds with excellent stereoselectivity.
To our knowledge, thiophosphonates and thiophosphi-
nates have never been used as glycosyl donors.
R1R2P+ SR3
R1R2P OR
R3SCN
R1R2P SR3
O
RCN
+
+
O
R
CN-
Scheme 1.
Keywords: Thiocyanates; Thiophosphates; Thiophosphonates;
Thiophosphinates; Glycosyl donors; Glycosylation.
*
Glycosyl thiocyanates 1–5 (Fig. 1) were readily obtained
Corresponding author. Tel.: +48 22 343 20 05; fax: +48 22 632 66
81; e-mail: pakul@icho.edu.pl
by treatment of the corresponding glycosyl bromides
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.09.162

M. Piekutowska, Z. Pakulski / Tetrahedron Letters 48 (2007) 8482–8486
8483
in our hands, 1-butyl-3-methylimidazolium tetrafluoro-
borate ([bmim]BF4) as solvent for the rearrangement
of glycosyl thiocyanate 2 drastically reduced the yield
of this reaction. It also promoted decomposition of the
glycosyl thiocyanate, which resulted in contamination
of the desired product with large amounts of inseparable
sugar byproducts. A similar reaction was also performed
with tributyl phosphite to afford thiophosphate 8. The
same reactions were performed for galactopyranosyl
thiocyanate 3 and b-mannopyranosyl thiocyanate 4.
The expected glycosyl thiosphosphates 9–12 were
obtained in moderate yields and the results are summa-
rized in Table 1.
BzO
BzO
OR
OBz
O
BzO
BzO
O
O
SCN
SCN
R
RO
RO
BzO
BzO
RO
BzO
R'
1: R = Ac
2: R = Bz
3
4: R = SCN, R' = H
5: R = H, R' = SCN
Figure 1.
with potassium thiocyanate in the presence of 18-crown-
6 as described earlier.7
An initial Michaelis–Arbuzov reaction with tetra-O-
acetyl glucose derived thiocyanate 1 was carried out in
neat triethyl phosphite at 60 °C and afforded the
expected thiophosphate 6 (Fig. 2) in low yield (20%).
Acetylated glucosyl thiocyanate 1 was unstable at the
high temperature required for reaction, hence we
decided to use the more stable perbenzoylated thiocya-
nates 2–5 for further studies.9
According to the 31P NMR data the purity of thiophos-
phates 7, 9 and 11 fluctuated between 77% and 99% and
the products were contaminated with H2P(O)(OEt),
HP(O)(OEt)2 and HOP(O)(OEt)2. The purities of
compounds 8, 10 and 12 were only 30–46%, byprod-
ucts including HP(O)(OBu)2 and HOP(O)(OBu)2 were
formed.
Reaction of thiocyanate 2 with triethyl phosphite was
performed at various temperatures. The highest yield
of thiophosphate 7 (51%) was obtained at 120 °C. For
some unknown reason, addition of a small amount of
toluene as cosolvent practically stopped the reaction.
Recent literature data10 suggests that ionic liquids influ-
ence the Michaelis–Arbuzov reaction by increasing the
yield and lowering the reaction temperature. However,
The unique chemical properties of a-mannopyranosyl
thiocyanate 5 should also be mentioned here. As we
have observed previously, a-mannopyranosyl thio-
cyanate 5 was completely unreactive towards Grignard
reagents.7 This compound also remained stable during
Michaelis–Arbuzov rearrangement, and we did not
observe any reaction with triethyl phosphite, dimethyl
BzO
BzO
BzO
OR
OBz
O
O
O
SP(OR')2
O
SP(OR')2
O
SP(OR')2
O
BzO
BzO
RO
RO
BzO
RO
BzO
6: R = Ac, R' = Et
7: R = Bz, R' = Et
8: R = Bz, R' = Bu
9: R' = Et
10: R' = Bu
11: R' = Et
12: R' = Bu
Figure 2.
Table 1. The Michaelis–Arbuzov rearrangement of glucosyl thiocyanates 2–4
Entry
RSCN
P(OR0)3
Solvent or ionic liquid
Product
Temp. (°C)a
Time
Yieldb (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
1
2
2
2
2
2
3
3
3
3
4
2
2
3
3
4
4
P(OEt)3
P(OEt)3
P(OEt)3
P(OEt)3
P(OEt)3
P(OEt)3
P(OEt)3
P(OEt)3
P(OEt)3
P(OEt)3
P(OEt)3
P(OBu)3
P(OBu)3
P(OBu)3
P(OBu)3
P(OBu)3
P(OBu)3
—
—
—
—
6
60
20
80
120
110
20
20
80
120
60
80
80
120
60
80
80
7 h
24 h
3 h
30 min
18 h
24 h
24 h
2 h
30 min
4 h
3.5 h
24 h
30 min
8 h
8 h
4.5 h
30 min
20
26
33
51
Traces
<10
26
30
25
Traces
24
22
33
25
30
30
40
7
7
7
Toluene
[bmim]BF4
—
—
7
7
9
9
—
9
[bmim]BF4
9
—
—
—
—
—
—
—
11
8
8
10
10
12
12
120
a Bath temperature.
b Purity was determined by 31P NMR (161.9 MHz, CDCl3) spectroscopy and the yield was calculated for pure product.

8484
M. Piekutowska, Z. Pakulski / Tetrahedron Letters 48 (2007) 8482–8486
BzO
BzO
OBz
O
OBz
O
BzO
BzO
O
SP(OMe)Ph
O
SP(OMe)Ph
O
SP(OMe)Ph
O
BzO
BzO
BzO
BzO
BzO
BzO
13
14
15
Figure 3.
Table 2. The Michaelis–Arbuzov rearrangement of glycosyl thiocyanates 2–4 with dimethyl phenylphosphonite PhP(OMe)2
Entry
RSCN
Solvent or ionic liquid
Product
Temp. (°C)a
Time
Yieldb (%)
1
2
3
4
2
3
3
4
—
—
13
14
14
15
40
40
20
50
30 min
30 min
1 h
51
53
47
41
[bmim]NTf2
—
1 h
a Bath temperature.
b Purity was determined by 31P NMR (161.9 MHz, CDCl3) spectroscopy and the yield was calculated for pure product.
phenylphosphonite or methyl diphenylphosphinite. The
reason for the unreactivity of a-mannopyranosyl thiocy-
anate 5 remains unknown, but we suspect that steric hin-
drance plays a role.
case, due to the instability of the final products at the
higher temperatures, the reactions were conducted
below 80 °C. Due to the high polarity, chromatographic
separation of S-glycosyl thiophosphinates 16–18 was
relatively simple and the purities of the products were
usually higher than 95%. As side products,
Ph2P(O)(OMe), Ph2P(OH) and H2P(O)(OMe) were
detected with the aid of 31P NMR. Detailed results are
presented in Table 3.
When dimethyl phenylphosphonite was used instead of
triethyl phosphite in the reaction with glycosyl thiocya-
nates 2–4 under similar conditions, the expected S-gly-
cosyl thiophosphonates 13–15 (Fig. 3) were obtained
as an inseparable mixture of diastereoisomers in an
approximate ratio of 1.2:1.0 in good yields. In this case,
the purity of thiophosphonates 13–15 varied between
70% and 99%; HP(O)(Ph)(OMe) and HP(O)(Ph)(OH)
were detected as contaminants by 31P NMR. Here,
[bmim]NTf2 proved to be a suitable solvent for the
above reaction (Table 2, entry 3). However, despite the
relatively high reaction yield, it promoted a series of
transformations of dimethyl phenylphosphonite afford-
ing a complex mixture of side products. The results
are collected in Table 2.
It should be emphasized that in all the cases studied the
configuration at the stereogenic centre connected to the
sulfur atom (anomeric position) was fully preserved.
Reactions were also regioselective affording S-glycosyl
derivatives as the sole products.
Next, we focused on the use of thiophosphorus deriva-
tives 7–17 as glycosyl donors. Prior work with S-glyco-
syl thiophosphates (very few analogues of such
derivatives had been previously applied to carbohydrate
chemistry11) suggested that thiophilic silver salts may
serve as activators for such donors.12 Unexpectedly,
only traces of the desired products were detected
upon reaction of 7 and benzyl alcohol (19) in the pres-
ence of silver triflate. We found that trimethylsilyl
triflate was a convenient activating agent for all the
donors tested. Preliminary results on glycosylations
involving the use of donors 7–17 are outlined in Table
4. In all cases, this glycosylation method furnished
exclusively the b-linked glycosides (for the gluco and
galacto series) and the a-linked mannosides (Fig. 5). It
is very important to note, that organophosphorus
Extension of these reaction conditions to methyl diph-
enylphosphinite as a starting material led to S-glycosyl
thiophosphinates 16–18 (Fig. 4) in low yields. In this
BzO
BzO
OBz
O
OBz
O
BzO
BzO
O
SPPh2
O
SPPh2
O
SPPh2
O
BzO
BzO
BzO
BzO
BzO
BzO
16
17
18
Figure 4.
Table 3. The Michaelis–Arbuzov rearrangement of glycosyl thiocyanates 2–4 with methyl diphenylphosphinite Ph2P(OMe)
Entry
RSCN
Solvent or ionic liquid
Product
Temp. (°C)a
Time
Yieldb (%)
1
2
3
4
5
6
2
2
3
3
3
4
—
—
16
16
17
17
17
18
40
80
rt
40
20
50
30 min
10 min
24 h
30 min
30 min
3 h
20
25
20
25
Traces
Traces
—
[bmim]NTf2
—
a Bath temperature.
b Purity was determined by 31P NMR (161.9 MHz, CDCl3) spectroscopy and the yield was calculated for pure product.

M. Piekutowska, Z. Pakulski / Tetrahedron Letters 48 (2007) 8482–8486
8485
Table 4. Glycosylation of benzyl alcohol (C6H5CH2OH, 19) and
methyl 2,3-O-isopropylidene-b-D-ribofuranoside (20) by donors 7–17
References and notes
1. Bhattacharya, A. K.; Thyagarajan, G. Chem. Rev. 1981,
81, 415–430.
2. Erian, A. W.; Sherif, S. M. Tetrahedron 1999, 55, 7957–
8024.
3. (a) Michalski, J.; Wieczorkowski, J. Bull. Acad. Pol. Sci.
Cl. III 1956, 4, 279–282; (b) Sheppard, W. A. J. Org.
Chem. 1961, 26, 1460–1462; (c) Pilgram, K.; Phillips, D. D.
J. Org. Chem. 1965, 30, 2388–2392.
4. (a) Ivanova, Z. M.; Kim, T. V.; Boldeskul, I. E.;
Gololobov, Y. G. Zh. Obsh. Khim. 1979, 49, 1464–1470;
(b) Cooper, D.; Trippett, S.; White, C. J. Chem. Res. (S)
1983, 234–235; (c) Yarmukhamedova, D. K.; Shtyrlina, A.
A.; Dubinina, N. P. Zh. Obsh. Khim. 1983, 53, 1755–1757;
(d) Yarmukhamedova, D. K.; Gilmanova, G. K.; Buchin,
P. I.; Cheplanova, I. V. Khim. Pharm. Zh. 1986, 20, 571–
573; (e) Nuretdinova, O. N.; Novikova, V. G.; Troitskaja,
L. B. Izv. Akad. Nauk, Ser. Khim. 1992, 2673–2675; (f)
Timperley, C. M.; Saunders, S. A.; Szpalek, J.; Waters, M.
J. J. Fluorine Chem. 2003, 119, 161–171.
Entry
Donor
Acceptor
Product
Yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
7
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
21
21
21
21
22
22
22
22
23
23
24
24
25
25
25
26
80
40a
80
57
51
31a
70
53
67a
69
54
85
44
50
63
52
8
13
16
9
10
14
17
12
15
7
13
9
14
17
15
a Contaminated by organophosphorus byproducts.
5. (a) Toshima, K.; Tatsuta, K. Chem. Rev. 1993, 93, 1503–
1531; (b) Nicolaou, K. C.; Mitchell, H. J. Angew. Chem.,
Int. Ed. 2001, 40, 1576–1624.
6. (a) Nepogodev, S. A.; Pakulski, Z.; Zamojski, A.; Holst,
O.; Brade, H. Carbohydr. Res. 1992, 232, 33–45; (b)
Pakulski, Z. Synthesis 2003, 2074–2078; (c) Pakulski, Z.
Polish J. Chem. 2005, 79, 361–367; (d) Blattner, R.;
Furneaux, R. H.; Pakulski, Z. Carbohydr. Res. 2006, 341,
2115–2125.
R1
HO
OMe
BzO
OBz
O
O
BzO
O
OBn
R
BzO
BzO
BzO
BzO
O
CMe2
20
O
OBn
21: R = OBz, R1 = H
22: R = H, R1 = OBz
23
_ ´
7. Pakulski, Z.; Pierozynski, D.; Zamojski, A. Tetrahedron
BzO
1994, 50, 2975–2992.
R1
BzO
BzO
OBz
O
O
8. (a) Kochetkov, N. K.; Klimov, E. M.; Malysheva, N. N.;
Demchenko, A. V. Tetrahedron Lett. 1989, 30, 5459–5462;
(b) Kochetkov, N. K.; Klimov, E. M.; Malysheva, N. N.;
Demchenko, A. V. Bioorg. Khim. 1990, 16, 701–710; (c)
Kochetkov, N. K.; Klimov, E. M.; Malysheva, N. N.;
Demchenko, A. V. Dokl. Akad. Nauk SSSR 1991, 321,
1203–1207; (d) Kochetkov, N. K.; Malysheva, N. N.;
Klimov, E. M.; Demchenko, A. V. Dokl. Akad. Nauk
SSSR 1991, 320, 634–637; (e) Klimov, E. M.; Demchenko,
A. V.; Malysheva, N. N.; Kochetkov, N. K. Bioorg. Khim.
1991, 17, 1660–1664; (f) Kochetkov, N. K.; Klimov, E.
M.; Malysheva, N. N.; Demchenko, A. V. Carbohydr. Res.
1991, 212, 77–91; (g) Kochetkov, N. K.; Malysheva, N.
N.; Klimov, E. M.; Demchenko, A. V. Tetrahedron Lett.
1992, 33, 381–384; (h) Kochetkov, N. K.; Klimov, E. M.;
Malysheva, N. N.; Demchenko, A. V. Carbohydr. Res.
1992, 232, C1–C5; (i) Kochetkov, N. K.; Klimov, E. M.;
Malysheva, N. N.; Demchenko, A. V. Carbohydr. Res.
1993, 242, C7–C10.
O
OMe
BzO
R
O
BzO
O
OMe
BzO
O
O
O
24: R = OBz, R1 = H
25: R = H, R1 = OBz
O
O
CMe2
26
CMe2
Figure 5.
impurities present in some donor samples did not influ-
ence the glycosylation process, and simple chromato-
graphic separation afforded pure glycosides. The only
exceptions were reactions involving dibutyl thiophos-
phates 8, 10 and 12, which gave the desired glycosides
slightly contaminated with inseparable organophospho-
rus species.
9. General procedure for the Michaelis–Arbuzov rearrange-
ment: A mixture of P(OEt)3 (2–5 equiv) and thiocyanate 2
(1 equiv) was heated under an argon atmosphere in a
screw cap tube (for product details see Tables 1–3).
Column chromatography of the residue (hexane–ethyl
In conclusion, we have developed an efficient synthesis
of S-glycosyl thiophosphates, thiophosphonates and
thiophosphinates from anomeric thiocyanates. The
resulting S-glycosyl thiophosphates, thiophosphonates
and thiophosphinates act as powerful glycosylation
agents which, in the presence of TMSOTf as activator,
install b-glucosidic, b-galactosidic and a-mannosidic
linkages stereoselectively.
acetate, 7:3) gave S-glycosyl phosphate 7. 1H, 13C and 31
P
NMR spectra were consistent with those expected for the
structures assigned to the respective compounds. All
compounds gave satisfactory high resolution mass spectra.
31P NMR (161.9 MHz, CDCl3): d 22.4 (6), 22.6 (7), 22.8
(8), 22.9 (9), 23.1 (10), 23.2 (11), 23.4 (12), 42.9 and 41.1
(13), 42.7 and 41.3 (14), 43.1 and 41.1 (15), 42.4 (16), 42.4
(17).
Acknowledgement
10. Matveeva, E. V.; Odinets, I. L.; Kozlov, V. A.; Shaplov,
A. S.; Mastryukova, T. A. Tetrahedron Lett. 2006, 47,
7645–7648.
11. (a) Michalska, M.; Michalski, J.; Orlich, I. Bull. Acad. Pol.
Sci. Ser. Chim. 1974, 22, 1053–1057; (b) Michalska, M.;
This work was supported by a Grant No. 3 TO9A 110
29 from the Ministry of Science and Higher Education,
Poland.

8486
M. Piekutowska, Z. Pakulski / Tetrahedron Letters 48 (2007) 8482–8486
Michalski, J.; Orlich, I. Tetrahedron 1978, 34, 617–622; (c)
A. Helv. Chim. Acta 2004, 87, 3106–3118; (i) Cipolla, L.;
Redaelli, C.; Faria, I.; Nicotra, F. J. Carbohydr. Chem.
2006, 25, 163–171; (j) Knapp, S.; Ajayi, K. Tetrahedron
Lett. 2007, 48, 1945–1949.
Chmielewski, M.; BeMiller, J. N. Carbohydr. Res. 1981,
96, 73–78; (d) Lipka, P.; Michalska, M. Carbohydr. Res.
1983, 113, 317–320; (e) Kudelska, W. Pol. J. Chem. 1997,
71, 1548–1552; (f) Kudelska, W. Z. Naturforsch. 1998, 53b,
1277–1280; (g) Kudelska, W. Carbohydr. Res. 2000, 329,
687–691; (h) Kulesza, A.; Frank, C. G.; Aebi, M.; Vasella,
12. (a) Borowiecka, J. Liebigs Ann. Recueil 1997, 2147–2150;
´
(b) Borowiecka, J.; Stanczak, A. Il Farmaco 2001, 56, 257–
262.