Unexpected highly diastereoconvergent Grignard additions to d-xylofuranose-derived t-butanesulfinyl aldimines

Article abstract of DOI:10.1016/j.tetlet.2011.11.002

Unexpected high levels of diastereoconvergence (dr >15:1) were observed in the addition of a series of Grignard reagents in THF to d-xylose-derived t-BS aldimines 2a,b affording (S_S,5R)- and (R_S,5R)-adducts. This anomaly was absent w

Full text of DOI:10.1016/j.tetlet.2011.11.002

Tetrahedron Letters 53 (2012) 119–122
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Unexpected highly diastereoconvergent Grignard additions to
D-xylofuranose-derived t-butanesulfinyl aldimines
Bing Wang a,b, , Pei-Pei Zhang a
⇑
a Department of Chemistry, Fudan University, 220 Handan Road, Shanghai 200433, China
b Institutes of Biomedical Sciences, Fudan University, 138 Yixueyuan Road, Shanghai 200032, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Unexpected high levels of diastereoconvergence (dr >15:1) were observed in the addition of a series of
Received 31 August 2011
Revised 11 October 2011
Accepted 18 October 2011
Available online 7 November 2011
Grignard reagents in THF to D-xylose-derived t-BS aldimines 2a,b affording (SS,5R)- and (RS,5R)-adducts.
This anomaly was absent when using ethereal solutions of organometallic reagents, revealing the subtle
solvent effects. This study illustrates the scope and limitations of N-t-BS imine chemistry in complex
systems.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
t-Butanesulfinyl imine
Grignard reagent
Addition
Diastereoconvergent
Carbohydrate
Recent development in the chemistry of t-butanesulfinyl imines
has greatly advanced the asymmetric synthesis of chiral
O
O
O
S
H
a
-
1) 2M HCl/THF
*
O
O
N
O
O
branched amines, amino alcohols, amino acids, vicinal diamines,
aziridines, etc.1 For simple substrates, the sense of asymmetric
induction was solely dictated by the sulfinyl auxiliary, whereas
in substrates with multiple stereogenic centers, the effect of this
chiral auxiliary also predominated.2 To date, only a few clear-cut
exceptions have been noted, in which a diastereomeric pair of N-
t-BS imines exhibited the same diastereofacial selectivity toward
the C@N bond regardless of the sulfinyl chirality.3 As a part of
our ongoing projects concerning alkaloids4 and nucleosides,5 we
planned to exploit the powerful and predictable chiral induction
of t-butanesulfinyl for stereodivergent synthesis based on carbohy-
drate scaffolds. However, an unexpected result was encountered,
and herein we report our preliminary findings.
2) NaIO4/MeOH
3) t-BuS(O)NH2
CuSO4/DCM
70-75%
O
O
Bn
O
Bn O
1
2a: (SS)-
2b: (RS)-
(from D-glucose)
Scheme 1. Preparation of D-xylofuranose-derived t-butanesulfinyl imines.
Initially, the addition of vinyl Grignard reagent to 2a and 2b was
examined.12 Both reactions appeared to be highly diastereoselec-
tive, affording N-t-BS allylic amines which were virtually diaste-
reopure (dr >50:1) in high yields, as judged by NMR of the crude
adducts. The absolute configuration of 3a, the adduct derived from
SS-sulfinimine 2a, was unambiguously established by single crystal
X-ray crystallography to be (SS,5R).13 The other adduct 3b, derived
from RS-sulfinimine 2b, was an oil. Under the assumption that the
chiral sulfinyl dictated the diastereoselectivity, we reasoned that
3b was of the (RS,5S) configuration. In order to prepare a crystalline
derivative for rigorous structural determination by X-ray analysis,
the chiral auxiliary was selectively removed using 2 M HCl without
affecting the 1,2-acetonide, and the amine was derivatized as the
corresponding 3,5-dinitrobenzamide 4b (Scheme 2). Unfortu-
nately, it was still an oil, and this prompted us to prepare the anal-
ogous N-3,5-dinitrobenzoylated 4a from 3a for a comparison of
their NMR spectra. Indirect assignment of the C-5 configuration
of 4b can thus be made. To our surprise, the conceived epimers
4a and 4b were identical in all respects.14 Thus we concluded that
the C-5 configuration of 3b was also R. More importantly, this
Compared to analogous nitrones6 and oximes,7 sugar-derived
N-t-BS imines have rarely been explored.8 The t-butanesulfinyl
imines in interest were easily prepared from D-glucose in five rou-
tine steps (Scheme 1). 3-O-Benzyl-1,2:5,6-diacetonide 19 was
selectively deprotected at the less hindered site,10 and the result-
ing 5,6-diol was oxidatively cleaved by NaIO4. The crude aldehyde
was condensed with RS- and SS-t-butanesulfinamides, respectively,
to afford a diastereomeric pair of Ellman’s imines 2a and 2b with a
carbohydrate backbone in good yields.11
⇑
Corresponding author. Tel./fax: +86 21 54237570.
E-mail address: wangbing@fudan.edu.cn (B. Wang).
0040-4039/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2011.11.002

120
B. Wang, P.-P. Zhang / Tetrahedron Letters 53 (2012) 119–122
O
S
H
O
S
H
O
O
N
N
O
O
O
O
Bn
O
Bn
O
2a: (SS)-
2b: (RS)-
DCM, -78oC
DCM, -78oC
MgBr
MgBr
O2N
1) HCl, MeOH
NO2
1) HCl, MeOH
2) ArCOCl, Et3N
DCM
O
S
O
S
2) ArCOCl, Et3N
DCM
O
O
O
N
H
N
H
O
N
H
O
O
O
O
O
O
Bn
O
Bn
O
Bn O
X-ray
4a (= 4b)
3a
3b
Scheme 2. Vinyl Grignard addition to 2a,b and correlation of C-5 absolute configurations of 3a and 3b. ORTEP drawing for 3a.
means that the addition of vinylmagnesium bromide to 2a and 2b
proceeded in an unexpected highly diastereoconvergent fashion,
both from the Re face of the C@N double bond.
Stimulated by this unusual phenomenon, we then investigated
the diastereoselectivities for the additions of other carbanions to
determine whether this was general in scope (Table 1). Initially,
when using (for convenience) commercial 3.0 M ethereal solutions
of PhMgBr or EtMgBr, the stereochemical outcome diverged with
that of vinyl addition, especially for the SS-imine 2a. Phenyl
addition to 2a afforded predominantly (5S)-adduct 3d (entry 3),
whose structure was established by X-ray analysis (Fig. 1).15 The
analogous ethyl addition to 2a was almost non-selective (entry
5). Additions to 2b proceeded with lower drs (7.5–12:1) as com-
pared to the vinyl addition (entries 4 and 6). When using
TMSC„CMgBr in Et2O as a nucleophile, no desired adducts were
detected (entries 7 and 8). It occurred to us that the apparent irreg-
ular behaviors of different Grignard species might be attributed to
their respective co-solvents, although the volume of these ethereal
Grignard solutions was just a fraction of CH2Cl2 used as the solvent
for 2.
Indeed, after shifting the co-solvent to THF, all these additions
proceeded in a diastereoconvergent manner, yielding the 5R-ad-
ducts in high drs (>15:1) regardless of the sulfur chirality (Table
2). Notably, addition of TMSC„CMgBr prepared in THF afforded
the propargylic adducts 3k,l in good yields and high drs (entries
9 and 10), with the exception that these reactions were carried
out at ꢀ20 °C overnight. In view of the importance of co-solvent,
we also tested on substrate 2a using THF as the sole solvent, as
Figure 1. ORTEP drawing for 3d.
well as inverse order of addition. In the former case, no reaction
was observed, and 2a was recovered (entry 5). When a solution
of 2a in CH2Cl2 was added to 0.7 M PhMgBr in THF under ꢀ78 °C,
the high dr was maintained, albeit the conversion dropped to
ꢁ50% (entry 6). These results clearly indicated a pronounced effect
of the co-solvent: Grignard species dissolved in THF afforded high
levels of diastereoconvergence, while ethereal solutions of the
same nucleophiles produced widely varying and often unsatisfac-
tory drs. Such solvent effect has not been emphasized in previous
studies.16
The methods of stereochemistry determination for the adducts
3a–l were outlined in Table 3. Diimide reduction17 of the vinyl
group of 3a and 3b afforded ethyl analogs 3g and 3i, respectively.
Similarly, removal of TMS (K2CO3/MeOH) in 3k and 3l followed by
saturation of the triple bond produced 3g and 3i, respectively. In
this manner, the C-5 configuration of 3g–l was established. Re-
moval of t-BS and N-benzoylation of 3d and 3f gave the same benz-
amide 4d to establish the C-5 configuration of 3c–f. Thus, with the
aid of X-ray and chemical correlation, the structures of all adducts
were unambiguously assigned.
Table 1
Diastereoselectivity of Grignard addition to 2a,b
O
S
H
O
S
R
5
*
*
*
O
O
RMgBr, DCM
N
N
H
O
O
3
3
O
O
Bn
O
Bn
O
2a, 2b
3a-j
Entry
Imine
R, concn, co-solvent
3 (%)a
drb
1
2
3
4
5
6
7
8
2a
2b
2a
2b
2a
2b
2a
2b
Vinyl, 0.7 M, THF
Vinyl, 0.7 M, THF
Ph, 3.0 M, Et2O
Ph, 3.0 M, Et2O
Et, 3.0 M, Et2O
Et, 3.0 M, Et2O
TMS-C„C, 0.7 M, Et2O
TMS-C„C, 0.7 M, Et2O
3a, 87
3b, 85
50:1c
50:1c
1:16c
7.5:1c
1.3:1d
12:1c
nd
3c, 3d, 92
3e, 3f, 85
3g, 3h, 93
3i, 3j, 87
Complex
Complex
nd
a
b
c
Combined isolated yields of both diastereomers.
Diastereomeric ratio of (5R)- to (5S)-.
Determined by NMR.
Although it has been reported that carbanion addition to the
analogous carbohydrate-derived nitrones6b and N-benzyl imines18
proceeded in moderate to high stereoselectivity, the uniformly
d
Determined by the isolated yields of each diastereomer.

B. Wang, P.-P. Zhang / Tetrahedron Letters 53 (2012) 119–122
121
Table 2
Acknowledgments
Diastereoconvergent Grignard additions to 2a,b
O
S
H
O
S
R
Financial support from the National Natural Science Foundation
of China (20602008, 20832005, and 21072032) is gratefully
acknowledged. We also thank the IBS (Fudan University) for
funding.
*
5
*
O
O
RMgBr, DCM
N
N
H
O
O
3
3
O
O
Bn
O
Bn
O
2a, 2b
Imine
3a-l
3 (%)a
Supplementary data
Entry
R, concn, co-solvent
drb
1
2
3
2a
2b
2a
2b
2a
2a
2a
2b
2a
2b
Vinyl, 0.7 M, THF
Vinyl, 0.7 M, THF
Ph, 0.7 M, THF
Ph, 0.7 M, THF
Ph, 0.7 M, THF
Ph, 0.7 M, THF
Et, 0.7 M, THF
Et, 0.7 M, THF
TMS-C„C, 0.7 M, THF
TMS-C„C, 0.7 M, THF
3a, 87
3b, 85
50:1
50:1
15:1
25:1
nd
15:1
20:1
20:1
18:1
30:1
Supplementary data (Experimental procedures and spectral
data.) associated with this article can be found, in the online ver-
sion, at doi:10.1016/j.tetlet.2011.11.002.
3c, 3d, 90
3e, 3f, 88
NR
3c, 3d, 42
3g, 3h, 83
3i, 3j, 75
3k, 65
4
5c
6d
7
References and notes
8
9e
10e
1. For leading reviews, see: (a) Robak, M. T.; Herbage, M. A.; Ellman, J. A. Chem.
Rev. 2010, 110, 3600; (b) Ferreira, F.; Botuha, C.; Chemla, F.; Pérez-Luna, A.
Chem. Soc. Rev. 2009, 38, 1162; (c) Lin, G.-Q.; Xu, M.-H.; Zhong, Y.-W.; Sun, X.-
W. Acc. Chem. Res. 2008, 41, 831.
3l, 80
a
b
c
Combined isolated yields of both diastereomers.
Ratio of (5R)- to (5S)-, determined by 1H NMR of the crude adducts.
THF as the solvent for substrate.
2. For examples of predominant stereo-induction of chiral t-BS over that of
a-
substituents, see: (a) Voituriez, A.; Pérez-Luna, A.; Ferreira, F.; Botuha, C.;
Chemla, F. Org. Lett. 2009, 11, 931; (b) Hjelmgaard, T.; Faure, S.; Lemoine, P.;
Viossat, B.; Aitken, D. J. Org. Lett. 2008, 10, 841; (c) Evans, J. W.; Ellman, J. A. J.
Org. Chem. 2003, 68, 9948; (d) Liu, J.; Li, Y.; Hu, J. J. Org. Chem. 2007, 72, 3119;
(e) Prakash, G. K. S.; Mandal, M.; Olah, G. A. J. Am. Chem. Soc. 2002, 124, 6538.
3. (a) McMahon, J. P.; Ellman, J. A. Org. Lett. 2004, 6, 1645; (b) Risseeuw, M. D. P.;
Mazurek, J.; van Langenvelde, A.; van der Marel, G. A.; Overkleeft, H. S.;
Overhand, M. Org. Biomol. Chem. 2007, 5, 2311; (c) Wang, B. J. Org. Chem. 2010,
75, 6012. This is not to be confused with reversal of stereoselectivity, which is
relatively common for this chiral auxiliary upon variation of reaction
conditions (solvent, countercation, additive).
4. (a) Chen, B.-L.; Wang, B.; Lin, G.-Q. J. Org. Chem. 2010, 75, 941; (b) Wang, B.; Lin,
G.-Q. Eur. J. Org. Chem. 2009, 5038; (c) Wang, B.; Wang, Y.-J. Org. Lett. 2009, 11,
3410; (d) Wang, B.; Zhong, Z.; Lin, G.-Q. Org. Lett. 2009, 11, 2011; (e) Wang, B.;
Liu, R.-H. Eur. J. Org. Chem. 2009, 2845; (f) Liu, R.-H.; Fang, K.; Wang, B.; Xu, M.-
H.; Lin, G.-Q. J. Org. Chem. 2008, 73, 3307; (g) Wang, B.; Fang, K.; Lin, G.-Q.
Tetrahedron Lett. 2003, 44, 7981; (h) Wang, B.; Yu, X.-M.; Lin, G.-Q. Synlett 2001,
904; (i) Liu, D.-G.; Wang, B.; Lin, G.-Q. J. Org. Chem. 2000, 65, 9114.
5. Sun, Z.-H.; Wang, B. J. Org. Chem. 2008, 73, 2462.
d
e
Inverse addition of 2a in CH2Cl2 to PhMgBr-THF, 50% conversion.
At ꢀ20 °C, 12 h.
Table 3
Summary of C-5 configuration determination
Correlation
1) K2CO3/MeOH
2) HN=NH
HN=NH
method A:
vinyl
Et
TMS-C≡ C
adducts
adducts
adducts
same t-BS config.
same C-5 config.
Correlation
method B:
1) HCl/MeOH
2) ArCOCl
adducts
same benzamide
different t-BS config.
same C-5 config.
Adduct
t-BS
R
C-5
Determination method
6. For leading references, see: (a) Racine, E.; Bello, C.; Gerber-Lemaire, S.; Vogel,
P.; Py, S. J. Org. Chem. 2009, 74, 1766; (b) Karanjule, N. S.; Markad, S. D.; Shinde,
V. S.; Dhavale, D. D. J. Org. Chem. 2006, 71, 4667; (c) Karanjule, N. S.; Markad, S.
D.; Dhavale, D. D. J. Org. Chem. 2006, 71, 6273.
7. Masson, G.; Philouze, C.; Py, S. Org. Biomol. Chem. 2005, 3, 2067.
8. For pyranose-derived sulfinimines, see: (a) Raunkjær, M.; Oualid, F. E.; van der
Marel, G. A.; Overkleeft, H. S.; Overhand, M. Org. Lett. 2004, 6, 3167; For a
ribose-derived t-BS imine, see: (b) Luo, Y.-C.; Zhang, H.-H.; Xu, P.-F. Synlett
2009, 833. No anomalous result was reported.
9. (a) Drueckhammer, D. G.; Wong, C.-H. J. Org. Chem. 1985, 50, 5912; (b) Fleet, G.
W. J.; Smith, P. W. Tetrahedron 1987, 43, 971.
10. Nacro, K.; Lee, J.; Barchi, J. J.; Lewin, N. E.; Blumberg, P. M.; Marquez, V. E.
Tetrahedron 2002, 58, 5335.
3a
3b
3c
3d
3e
3f
3g
3h
3i
SS-
RS-
SS-
SS-
RS-
RS-
SS-
SS-
RS-
RS-
SS-
RS-
Vinyl
Vinyl
Ph
Ph
Ph
Ph
Et
Et
R
R
R
S
R
S
R
S
R
S
X-ray
Correlate with 3a (B)
Infer from 3d
X-ray
Infer from 3f
Correlate with 3d (B)
Correlate with 3a (A)
Infer from 3g
Correlate with 3b (A)
Infer from 3i
Et
Et
3j
3k
3l
TMS-C„C
TMS-C„C
R
R
Correlate with 3g (A)
Correlate with 3i (A)
11. Liu, G.; Cogan, D. A.; Owens, T. D.; Tang, T. P.; Ellman, J. A. J. Org. Chem. 1999, 64,
1278.
12. General procedures: To a cooled (ꢀ78 °C) solution of 2a (353 mg, 0.93 mmol) in
CH2Cl2 (5 mL) under Ar was added dropwise vinylmagnesium bromide (3.3 mL,
0.7 M in THF, 2.31 mmol), and the solution was stirred at the same
temperature for 1 h. The reaction was quenched by satd aq NH4Cl, diluted
with ether (50 mL), the organic layer was washed with brine, dried (Na2SO4),
and concentrated under reduced pressure. The residue was purified by silica
gel flash column chromatography eluated with EtOAc/Hexane. Compound 3a:
high drs observed for Grignard additions to 2a,b are still remark-
able and synthetically useful. In addition, compared to N-benzyl,
the N-t-BS group also enjoyed the benefit of easy removal. The ori-
gin of this unexpected diastereoconvergence is yet to be rational-
ized; nevertheless, it cannot be attributed solely to the
carbohydrate moiety, for the stereo-induction of the latter cannot
overcome that of the chiral sulfinyl completely. The present case
suggested subtle interplay between the two chiral auxiliaries.
To summarize, we report a rare example of the diastereoconver-
½ ꢂ
a 2D3 +6.1 (c 1.13, CHCl3); 1H NMR (500 MHz, CDCl3) d 7.39–7.26 (m, 5H), 6.00
(ddd, 1H, J = 17.2, 10.5, 5.4 Hz), 5.93 (d, 1H, J = 3.8 Hz), 5.49 (d, 1H, J = 17.2 Hz),
5.29 (d, 1H, J = 10.5 Hz), 4.66–4.45 (AB, 2H, JAB = 11.0 Hz), 4.64 (m, 1H), 4.32–
4.25 (m, 1H), 4.24 (dd, 1H, J = 7.5, 3.2 Hz), 4.11 (d, 1H, J = 3.2 Hz), 3.92 (d, 1H,
J = 8.2 Hz), 1.50 (s, 3H), 1.32 (s, 3H), 1.09 (s, 9H). 13C NMR (125 MHz, CDCl3) d
136.7, 136.3, 128.5, 128.1, 127.9, 117.7, 111.6, 105.0, 82.7, 81.4, 81.3, 71.9,
57.7, 56.0, 26.7, 26.2, 22.5. HR-ESI-MS m/z Calcd for C21H31NO5SNa (M+Na+)
gent addition of Grignard reagents to a pair of
D-xylofuranose-
432.1821. Found 432.1813. Compound 3b: ½a D23
ꢂ
ꢀ46.4 (c 2.23, CHCl3); 1H NMR
based t-butanesulfinyl aldimines 2a and 2b. Both the RS- and SS-
imines afforded predominantly (5R)-adducts. The use of THF as
the co-solvent for the Grignard reagents is essential to achieve high
and consistent diastereoselectivity. On the other hand, precaution
should be paid in assigning the stereochemistry of nucleophilic
additions to polysubstituted N-t-BS imines. In view of the versatile
synthetic potentials of the vinyl and alkynyl groups, the adducts
can be exploited in the asymmetric synthesis of azasugars or re-
lated chiral scaffolds. Work along this line is currently in progress
in this laboratory.
(500 MHz, CDCl3) d 7.39–7.26 (m, 5H), 5.98 (d, 1H, J = 3.5 Hz), 5.91 (ddd, 1H,
J = 17.2, 10.4, 6.3 Hz), 5.35 (d, 1H, J = 17.2 Hz), 5.24 (d, 1H, J = 10.4 Hz), 4.68–
4.61 (AB, 2H, JAB = 11.2 Hz), 4.64 (m, 1H), 4.32–4.26 (m, 1H), 4.16 (d, 1H,
J = 3.3 Hz), 4.13 (dd, 1H, J = 7.6, 3.3 Hz), 3.66 (d, 1H, J = 7.5 Hz), 1.48 (s, 3H), 1.31
(s, 3H), 1.11 (s, 9H). 13C NMR (125 MHz, CDCl3) d 137.1, 136.2, 128.4, 128.0,
127.9, 117.2, 111.5, 105.0, 81.9, 81.8, 81.3, 71.6, 57.0, 55.7, 26.7, 26.1, 22.4. HR-
ESI-MS m/z Calcd for C21H31NO5SNa (M+Na+) 432.1821. Found 432.1814.
13. CCDC 720014 contains the crystallographic data for 3a. These data can be
obtained free of charge from the Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
14. Compound 4a: ½a D21
ꢂ
+23.1 (c 0.72, CHCl3); 1H NMR (500 MHz, CDCl3) d 9.03 (t,
1H, J = 2.0 Hz), 8.61 (d, 2H, J = 2.1 Hz), 7.92 (d, 1H, J = 8.1 Hz), 7.30–7.26 (m,

122
B. Wang, P.-P. Zhang / Tetrahedron Letters 53 (2012) 119–122
2H), 7.21–7.16 (m, 3H), 6.03 (d, 1H, J = 3.8 Hz), 5.84 (ddd, 1H, J = 17.2, 10.4,
16. (a) Liu, G.; Cogan, D. A.; Ellman, J. A. J. Am. Chem. Soc. 1997, 119, 9913; (b)
Cogan, D. A.; Liu, G.; Ellman, J. A. Tetrahedron 1999, 55, 8883. THF was found to
be deleterious to dr in the latter paper.
17. Hart, D. J.; Hong, W. P.; Hsu, L. Y. J. Org. Chem. 1987, 52, 4665.
18. For selected papers, see: (a) Compain, P.; Martin, O. R.; Boucheron, C.; Godin,
G.; Yu, L.; Ikeda, K.; Asano, N. ChemBioChem 2006, 7, 1356; (b) Bordier, A.;
Compain, P.; Martin, O. R.; Ikeda, K.; Asano, N. Tetrahedron: Asymmetry 2003,
6.0 Hz), 5.36 (d, 1H, J = 17.2 Hz), 5.32 (d, 1H, J = 10.4 Hz), 5.24 (m, 1H), 4.76–
4.46 (AB, 2H, JAB = 10.4 Hz), 4.75 (m, 1H), 4.36 (m, 1H), 4.21 (d, 1H, J = 3.2 Hz),
1.52 (s, 3H), 1.36 (s, 3H). 13C NMR (125 MHz, CDCl3) d 162.1, 148.4, 138.2,
135.8, 133.2, 128.7, 128.6, 128.5, 126.8, 120.5, 117.8, 111.9, 104.8, 84.0, 81.1,
78.8, 72.8, 52.6, 26.6, 26.0. HR-ESI-MS m/z Calcd for
C
24H26N3O9 (M+H+)
500.1669. Found 500.1676.
15. CCDC 730904 contains the crystallographic data for 3d. These data can be
obtained free of charge from the Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
14, 47. The carbohydrate moiety reported (
in our work, however, the relative sense of chiral induction was the same.
L-xylo-) was antipodal with that of 2