Formation of epoxides from pentacoordinated organoarsenic compounds with a β-hydroxyethyl group

Article abstract of DOI:10.1039/c0nj00035c

A series of pentacoordinated organoarsenic compounds (arsoranes) bearing a β-hydroxyethyl group (4a, 6a and 6b) was synthesized. The crystal structures were determined by single crystal X-ray analysis. Treatment of these arsoranes with KH almost quantitatively gave the corresponding epoxide. The reaction of 4a having an unsubstituted β-hydroxyethyl group with KH was monitored by ¹H and ¹⁹F NMR in CD₃CN, suggesting that a hexacoordinate arsenic anion was formed as the intermediate. However, a further stereochemical study of the epoxide formation suggested that the reaction proceeded in the S_N2 manner and not in the ligand coupling reaction (LCR) of the intermediate hexacoordinate arsenic anion.

Full text of DOI:10.1039/c0nj00035c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Formation of epoxides from pentacoordinated organoarsenic compounds
with a b-hydroxyethyl groupwzy
Xin-Dong Jiang,z Shiro Matsukawa,8 Yuta Fukuzaki and Yohsuke Yamamoto*
Received (in Montpellier, France) 18th January 2010, Accepted 18th March 2010
DOI: 10.1039/c0nj00035c
A series of pentacoordinated organoarsenic compounds (arsoranes) bearing a b-hydroxyethyl
group (4a, 6a and 6b) was synthesized. The crystal structures were determined by single crystal
X-ray analysis. Treatment of these arsoranes with KH almost quantitatively gave the
corresponding epoxide. The reaction of 4a having an unsubstituted b-hydroxyethyl group with
1
KH was monitored by H and 19F NMR in CD3CN, suggesting that a hexacoordinate arsenic
anion was formed as the intermediate. However, a further stereochemical study of the epoxide
formation suggested that the reaction proceeded in the SN2 manner and not in the ligand
coupling reaction (LCR) of the intermediate hexacoordinate arsenic anion.
Introduction
The reaction of a phosphonium ylide with a ketone or an
aldehyde producing an olefin, called the Wittig reaction, is one
of the most useful methods for carbon–carbon double bond
formation, and thus extensively used in organic synthesis.1 The
stereoselectivity of the produced olefin generally depends on
Scheme 1
the substituents of the ylide; phosphonium ylides bearing
electron-donating groups on the ylidic carbon (unstabilized
ylides) produce Z-olefins, whereas those having electron-
withdrawing groups (stabilized ylides) produce the thermo-
dynamically favorable E-olefins.
Unlike the phosphonium ylides, arsenic analogues (arsonium
ylides) are much less common.2 However, arsonium ylides are
essentially more reactive than phosphonium ylides due to the
higher nucleophilicity of the ylidic carbon of the former. In
addition, it is attractive that the product type varies depending
Scheme 2
on the nature of the substituent. For example, the reactions of
arsonium ylides with carbonyl compounds form olefins or
epoxides. Stabilized ylides, e.g., Ph3AsQCHCOPh, produce
olefins with a high E-selectivity under mild reaction conditions,
During the course of our study on hypervalent compounds
involving the group 15 elements,5 we have found that the
whereas the unstabilized ylides, e.g., Ph3AsQCHMe, form
a-anion of the 10-P-5 spirophosphorane having an ethoxy-
epoxides with a moderate to high E-selectivity.3 Semistabilized
carbonylmethyl group (1) works as a Wittig-type reagent.6
(benzylic and allylic) ylides generally give a mixture of olefins
The reactions of the a-anion with aromatic and aliphatic
aldehydes produce olefins with an excellent Z-selectivity
and epoxides. Hsi and Koreeda have shown that the product
distribution from the reaction of an allylic ylide with several
(up to Z/E = >98 : 2) (Scheme 2). Furthermore, the
aldehydes can be controlled by the cation of the base
(Scheme 1).4
phosphoranes with a cyano or an amide functionality also
form olefins.7 The cyano derivative showed a high Z-selectivity
only when reacted with aliphatic aldehydes, whereas the amide
Department of Chemistry, Graduate School of Science,
Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima,
to exclusively give the Z-olefins. The detailed stereochemical
739-8526, Japan. E-mail: yyama@sci.hiroshima-u.ac.jp;
derivatives reacted with both aliphatic and aromatic aldehydes
study of the decomposition reactions of the diastereomeric
phosphoranes having an a,b-diphenyl-b-hydroxyethyl group
(2) clarified that the rate of formation of the Z-olefin was
faster than that of the E-olefin (Scheme 3).8
Fax: +81-82-424-0723; Tel: +81-82-424-7430
w Dedicated to the memory of Pascal Le Floch.
z This article is part of a themed issue on Main Group chemistry.
y CCDC reference numbers 761610 (4a), 761611 (5a), 761613 (5b),
761612 (6a) and 761614 (6b). For crystallographic data in CIF or other
electronic format see DOI: 10.1039/c0nj00035c
z Current address: The Key Laboratory for Special Functional
Materials of Ministry of Education, Henan University, Kaifeng,
Henan 475-004, China.
In this context, we focused on the decomposition reaction of
the 10-As-5 spiroarsorane with a b-hydroxyethyl group.
We now report the synthesis and structure of 10-As-5
b-hydroxyethylspiroarsoranes, and the exclusive epoxide
formation from these arsoranes under basic conditions.
8 Current address: Department of Chemistry, Faculty of Science,
Toho University, 2-2-1 Miyama, Funabashi 274-8510, Japan.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1623–1629 | 1623

View Article Online
solvated dichloromethane. The ORTEP diagrams are shown
in Fig. 1, and the selected bond parameters are summarized in
Table 1. For all the structures, the arsenic atoms are penta-
coordinated, the geometries around the arsenic atoms are a
slightly distorted trigonal bipyramid (TBP), and the oxygen
atoms occupy the apical sites. As a whole, As–O and As–C
lengths for 6a and 6b are longer than those for the other
compounds due to steric repulsions originating from the bulky
monodentate ligand. The As–O1 lengths for 6a (1.937(3) and
1.938(3) A) and 6b (1.940(3) A) are longer than the As–O2
lengths by 0.02–0.04 A. The O1ꢁ ꢁ ꢁO3 lengths for 6a are 2.80 A
(1st molecule) and 2.73 A (2nd), and that of 6b is 2.78 A. This
suggests that, for 6a and 6b, the hydroxy proton forms a
hydrogen bond with the apical oxygen atom (O1), giving rise
to the elongation of the As–O1 lengths. On the other hand, the
hydroxy group of 4a forms intermolecular O–Hꢁ ꢁ ꢁF hydrogen
bonds, where the interatomic distance between O3 and F5
(3/2 ꢂ x, ꢂ1/2 + y, z) is 3.463(4) A, and that between O3 and
F12 (ꢂ1/2 + x, y, 1/2 ꢂ z) is 2.974(3) A. As for the bond
angles, the O1–As–O2 angles for the C2F5 derivatives 5b and
6b (167.7–170.21) are farther from the ideal 1801 than those for
the CF3 derivatives (173.1–177.51), and the C1–As–C2 angles
for the former (134.6–134.81) are farther from the ideal 1201
than those for the latter (120.6–127.71). These results indicate
that the geometries for the C2F5 derivatives are more distorted
from an ideal TBP than that for the CF3 derivatives.
Scheme 3
Results and discussion
At first, the syntheses of b-hydroxyethylarsoranes using two
kinds of bidentate ligands were examined. The Martin ligand
derived from the hexafluorocumyl alcohol9 and our ligand
derived from 1,1,1,2,2,4,4,5,5,5-decafluoro-3-phenyl-3-pentanol10
were used as the bidentate ligand. The reaction of the arsine
3a11 with 2-bromoethanol in the presence of NaH gave the
a,b-unsubstituted derivative 4a in 35% yield (Scheme 4). The
synthesis of the b,b-diphenyl derivative was next examined.
The treatment of 3a with NaH, followed by 2-chloro-1,1-
diphenylethanol (ClCH2CPh2OH), resulted in recovery of
the reagents. Thus, arsines 3a and 3b were treated with
phenacyl bromide (BrCH2COPh) to form the corresponding
b-keto derivatives 5a (20%) and 5b (20%), respectively. The
b-keto derivatives were then treated with PhLi to afford the
desired b,b-diphenyl derivatives 6a (74%) and 6b (67%)
(Scheme 5).
The decomposition reactions were then examined
(Scheme 6). We found that the treatment of compound 4a
with KH in CD3CN quantitatively gave the arsoranide 7a-K.
The reaction was thus monitored by 1H and 19F NMR at 25 1C
(Fig. 2). After 10 min, four signals at d = ꢂ73.3 (br s), ꢂ73.5
(br s), ꢂ73.9 (br m) and ꢂ74.2 (br s) ppm (A) were mainly
observed in the 19F NMR along with a pair of low-intensity
quartets corresponding to the arsoranide 7a-K (d = ꢂ73.3 and
ꢂ76.0 ppm). Additional signals (B) appeared at d = ꢂ74.3 (q),
ꢂ74.4 (m), ꢂ74.6 (m) and ꢂ74.9 (q) ppm during the reaction.
The A and B signals gradually decreased and the signals for
7a-K increased as the reaction proceeded, and finally only the
latter ones could be observed. After monitoring for 6.5 h at
25 1C and then heating for 4 h at 45 1C, the crude mixture was
separated to give the arsoranide 7a-K in 96% as an isolated
product. Although ethylene oxide (8) was detected at
Single crystals of 4a, 5a, 5b, 6a and 6b grown from hexane/
dichloromethane mixed solutions were subjected to single
crystal X-ray analysis. For 6a, two independent molecules
were found in the unit cell. The crystal of 6b contained
Scheme 4
1
d = 2.54 ppm in the H NMR (CD3CN), the highly volatile
8 (bp 10 1C) could not be isolated. Thus, similar reactions
using the b,b-diphenyl derivatives 6a and 6b were carried out
to almost quantitatively form the 1,1-diphenylethylene oxide
(9) (isolated yield 93% from 6a, 94% from 6b) together with
the corresponding arsoranide 7a-K and 7b-K, respectively. The
reactions of 6a and 6b were completed within 20 min at 30 1C.
Although the reaction rates have not yet been measured, the
decomposition reactions of the b,b-diphenyl derivatives 6 are
obviously faster than that of the unsubstituted 4a. This should
be due to the steric crowding of 6 compared to 4a, i.e., the
greater steric repulsion of 6 should be a driving force for the
decomposition.
Two mechanisms for the epoxide formation can be considered
(Scheme 7). One is the SN2 reaction, in which the alkoxide
anion attacks the a-carbon, and the other is the ligand coupling
reaction (LCR),12 in which hexacoordinate intermediates once
Scheme 5
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1624 | New J. Chem., 2010, 34, 1623–1629

View Article Online
Fig. 1 The ORTEP diagrams of 4a, 5a, 5b, 6a and 6b showing the thermal ellipsoids at the 30% probability level. For 6a, one of the two
independent molecules is shown. The hydrogen atoms other than those of the hydroxy group and the solvated molecules are omitted for clarity.
Table 1 Selected bond lengths (A) and angles (1) for 4a, 5a, 5b, 6a and 6b
4a
5a
5b
6aa
6b
Bond length/A
As–O1
As–O2
As–C1
As–C2
As–C3
1.912(2)
1.9119(19)
1.917(3)
1.925(2)
1.935(3)
1.915(2)
1.914(2)
1.917(3)
1.916(3)
1.931(3)
1.9138(16)
1.9009(16)
1.929(2)
1.917(3)
1.938(3)
1.937(3), 1.938(3)
1.914(3), 1.915(3)
1.931(4), 1.938(4)
1.938(4), 1.933(4)
1.951(4), 1.963(4)
1.940(3)
1.904(3)
1.933(4)
1.924(4)
1.946(4)
Bond angle/1
O1–As–O2
C1–As–C2
C1–As–C3
C2–As–C3
176.09(8)
120.64(11)
120.73(12)
118.62(12)
177.51(9)
121.73(13)
123.70(13)
114.53(14)
170.17(8)
134.60(11)
116.49(11)
108.89(11)
174.23(15), 173.07(13)
125.7(2), 127.66(18)
115.01(19), 108.96(18)
119.22(19), 123.36(18)
167.67(14)
134.84(17)
106.39(17)
118.76(17)
a
Bond lengths and angles for the two independent molecules are shown.
the four CF3 groups of such an anion are magnetically
unequivalent. However, there is no proof on the direct formation
of the epoxide from the hexacoordinate anions derived from 4
or 6. Although it is not yet clear, the B signals might
correspond to the isomeric hexacoordinate anions Int-20 or
Int-200 because we observed the two types of isomerizations in
the hexacoordinate phosphate anions of this type (10A to
10B,13b,c 11A to 11B10c) (Scheme 8).
The two mechanisms should differ in the stereochemistry of
the product, i.e., SN2 mechanism results in inversion of the
stereochemistry of the carbon a to the arsenic atom, whereas
the LCR mechanism retains it. To stereochemically clarify the
mechanism, some preliminary investigations were carried
out.14 We synthesized the diastereomeric a,b-dimethyl-b-
phenyl derivatives 12a, and one diastereomer 12a-SAsRR
Scheme 6
formed after deprotonation decompose to give the epoxide.13
The four broad signals A observed in the above-described
reaction (Fig. 2) could be those corresponding to a hexacoordinate
arsenic anion bearing an oxaarsetane ring, e.g. Int-2, because
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1623–1629 | 1625

View Article Online
Scheme 8
Scheme 9
Fig. 2 Time course of the reaction of 4a with KH in CD3CN at 25 1C.
(a) 10 min, (b) 3 h, (c) 6.5 h, and (d) after heating at 45 1C for 4 h
from (c).
In conclusion, we synthesized a series of pentacoordinated
organoarsenic compounds bearing a b-hydroxyethyl group,
and the structures were determined by single crystal X-ray
analysis. We found that these compounds decomposed upon
treatment with KH in CD3CN to almost quantitatively give
the corresponding epoxide. Further investigations on the
mechanism of the epoxide formation are in progress.
Experimental
General
The melting points were measured using a Yanaco micro-
1
melting point apparatus. The H NMR (400 MHz) and 19F
NMR (376 MHz) were recorded using JEOL EX-400 or JEOL
1
AL-400 spectrometers. The H NMR chemical shifts (d) are
given in ppm downfield from Me4Si, determined by residual
chloroform (d = 7.26 ppm) or acetonitrile (d = 1.90 ppm).
The 19F NMR chemical shifts (d) are given in ppm downfield
from the external CFCl3. The elemental analyses were
performed using a Perkin-Elmer 2400 CHN elemental analyzer.
All reactions were carried out under N2. Tetrahydrofuran
(THF) and diethyl ether (Et2O) were freshly distilled from
Na-benzophenone, n-hexane was distilled from Na, and the
other solvents were distilled from CaH2. Merck silica gel
60 was used for the column chromatography.
Scheme 7
(S configuration at As,15 R at the a and b carbon) was isolated.
The stereochemistry of 12a-SAsRR was determined by single
crystal X-ray analysis. The treatment of 12a-SAsRR with KH
in THF for 20 min exclusively gave (2R,3R)-2,3-dimethyl-2-
phenyloxirane (13-RR), indicating that the stereochemistry at
the a-carbon was inverted (Scheme 9). This result suggests that
the reaction proceeds in the SN2 manner, hence this should
be true for the reaction of 4a and 6. Thus, for the reaction
of 4a, the hexacoordinate anion Int-2 should exist as a
stable intermediate which does not form the epoxide
(Fig. 2).
Syntheses
4a. Under N2, a solution of 3a (151.2 mg, 0.269 mmol) in
CH3CN (3 mL) was added to a solution of NaH (70.0 mg,
1.75 mmol) in CH3CN (3 mL) at 0 1C, and the mixture was
stirred for 10 min at room temperature. The mixture was then
cooled to 0 1C, BrCH2CH2OH (0.08 mL, 1.13 mmol) was
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1626 | New J. Chem., 2010, 34, 1623–1629

View Article Online
2
2
added, and the mixture was stirred for 18 h at 35 1C. The
mixture was extracted with Et2O (2 ꢃ 30 mL), the organic
layer was washed with brine (2 ꢃ 20 mL) and dried over
anhydrous MgSO4. After removing the solvents by evaporation,
the resulting crude mixture was separated by preparative TLC
(d, JFF = 283.4 Hz, 1F), ꢂ117.2 (d, JFF = 283.4 Hz, 2F),
ꢂ119.8 (dm, 2JFF = 283.4 Hz, 1F), ꢂ120.1 ppm (dm, 2JFF
=
283.4 Hz, 1F); elemental analysis: calcd (%) for
C30H15F20O3As: C 41.02, H 1.72; found: C 40.98, H 1.51.
6a. Under N2, PhLi (1.14 M cyclohexane/Et2O solution,
0.14 mL, 0.159 mmol) was added to a solution of 5a (55.7 mg,
0.0821 mmol) in Et2O (3.0 mL) at ꢂ78 1C. The mixture was
then stirred for 2 h at ꢂ78 1C. The reaction was quenched with
saturated aqueous NH4Cl (20 mL) at ꢂ78 1C. The mixture was
extracted with Et2O (2 ꢃ 40 mL), and the organic layer was
washed with brine (2 ꢃ 30 mL) and dried over anhydrous
MgSO4. After removing the solvents by evaporation, the
resulting crude mixture was separated by preparative TLC
(n-hexane : CH2Cl2 = 4 : 1), followed by HPLC (ODS,
MeCN) to afford 6a (RT = 25.6 min: 46.2 mg, 0.0610 mmol,
74%) as a white solid. Colorless crystals of 6a suitable for the
X-ray analysis were obtained by recrystallization from
n-hexane/CH2Cl2. Mp: 57.0–58.0 1C (decomp.); 1H NMR
(n-hexane
: CH2Cl2 = 2 : 1) to afford 4a (57.8 mg,
0.0956 mmol, 35%) as a white solid. Colorless crystals of 4a
suitable for the X-ray analysis were obtained by recrystallization
from n-hexane/CH2Cl2. Mp: 149.3ꢂ150 1C; 1H NMR
(CDCl3): d = 7.31–8.35 (m, 2H), 7.80 (br s, 2H), 7.71–7.75
(m, 4H), 3.96–4.06 (m, 1H), 3.90–3.98 (m, 1H), 3.00–3.09 (m,
1H), 2.85–2.93 (m, 1H), 2.61 ppm (s, 1H); 19F NMR (CDCl3):
d = ꢂ75.4 (q, 4JFF = 8.6 Hz, 6F), ꢂ75.5 ppm (q, 4JFF = 8.6 Hz,
6F); elemental analysis: calcd (%) for C20H13F12O3As: C
39.76, H 2.17; found: C 39.56, H 2.06.
5a. Under N2, a solution of 3a (160.9 mg, 0.287 mmol) in
CH3CN (5 mL) was added to a solution of NaH (18.9 mg,
0.472 mmol) in CH3CN (5 mL) at 0 1C, and the mixture was
stirred for 10 min at room temperature. The mixture was then
cooled to 0 1C, BrCH2COPh (74.4 mg, 0.373 mmol) in CH3CN
(5 mL) was added, and the mixture was stirred for 19 h at
35 1C. The mixture was extracted with Et2O (2 ꢃ 40 mL), and
the organic layer was washed with brine (2 ꢃ 30 mL) and dried
over anhydrous MgSO4. After removing the solvents by
evaporation, the resulting crude mixture was separated by
preparative TLC (n-hexane : CH2Cl2 = 3 : 1) to afford 5a
(39.0 mg, 0.0574 mmol, 20%) as a white solid. Colorless
crystals of 5a suitable for the X-ray analysis were obtained
by recrystallization from n-hexane/CH2Cl2. Mp: 98.1ꢂ99.0 1C
3
(CDCl3): d = 7.93 (d, JHH = 8 Hz, 2H), 7.70 (br d,
3
3JHH = 8 Hz, 2H), 7.60 (t, JHH = 8 Hz, 2H), 7.49 (t,
3
3JHH = 8 Hz, 2H), 7.33–7.35 (m, 3H), 7.22 (t, JHH = 8 Hz,
3
2H), 7.18 (t, JHH = 8 Hz, 2H), 6.94–6.97 (m, 3H), 5.35 (s,
2
1H), 3.87 (d, JHH = 13.9 Hz, 1H), 3.80 ppm (d, JHH
2
=
=
4
13.9 Hz, 1H); 19F NMR (CDCl3): d = ꢂ74.3 (q, JFF
4
8.6 Hz, 6F), ꢂ75.1 ppm (q, JFF = 8.6 Hz, 6F); elemental
analysis: calcd (%) for C32H21F12O3AsꢁH2O: C 49.63, H 2.99;
found: C 49.94, H 2.64; MS(EI(+)): m/z = 756 [M]+, 757
[M + 1]+, 758 [M + 2]+, 559 [M ꢂ CH2CPh2OH]+.
1
6b. Under N2, PhLi (1.14 M cyclohexane/Et2O solution,
0.08 mL, 0.0912 mmol) was added to a solution of 5b (50.7 mg,
0.0577 mmol) in Et2O (3.0 mL) at ꢂ78 1C. The mixture was
then stirred for 2 h at ꢂ78 1C. The reaction was quenched with
saturated aqueous NH4Cl (20 mL) at ꢂ78 1C. The mixture was
extracted with Et2O (2 ꢃ 40 mL), and the organic layer was
washed with brine (2 ꢃ 30 mL) and dried over anhydrous
MgSO4. After filtering the organic layer through SiO2 and
removing the solvents by evaporation, the resulting crude
mixture was separated by HPLC (ODS, MeCN) to afford 6b
(RT = 32.0 min: 37.0 mg, 0.0386 mmol, 67%) as a white solid.
Colorless crystals of 6b suitable for the X-ray analysis were
obtained by recrystallization from n-hexane/CH2Cl2. Mp:
(decomp.); H NMR (CDCl3): d = 8.34–8.36 (m, 2H), 7.81
(br s, 2H), 7.78 (d, 3JHH = 8 Hz, 2H), 7.71–7.74 (m, 4H), 7.55
(t, 3JHH = 8 Hz, 1H), 7.39 (t, 3JHH = 8 Hz, 2H), 4.40 (d, 2JHH
=
14.6 Hz, 1H), 4.25 ppm (d, JHH = 14.6 Hz, 1H); 19F NMR
2
4
(CDCl3): d = ꢂ75.2 (q, JFF = 8.6 Hz, 6F), ꢂ75.6 ppm
4
(q, JFF = 8.6 Hz, 6F); elemental analysis: calcd (%) for
C26H15F12O3As: C 46.04, H 2.23; found: C 46.18, H 2.07.
5b. Under N2, a solution of 3b (367.6 mg, 0.469 mmol) in
CH3CN (5 mL) was added to a solution of NaH (17.8 mg,
0.445 mmol) in CH3CN (5 mL) at 0 1C, and the mixture was
stirred for 10 min at room temperature. The mixture was then
cooled to 0 1C, BrCH2COPh (70.0 mg, 0.351 mmol) in CH3CN
(5 mL) was added, and the mixture was stirred for 21 h at
35 1C. The mixture was extracted with Et2O (2 ꢃ 40 mL), and
the organic layer was washed with brine (2 ꢃ 40 mL) and dried
over anhydrous MgSO4. After removing the solvents by
evaporation, the resulting crude mixture was separated by
preparative TLC (n-hexane : CH2Cl2 = 3 : 1) to afford 5b
(83.6 mg, 0.0951 mmol, 20%) as a white solid. Colorless
crystals of 5b suitable for the X-ray analysis were obtained
by recrystallization from n-hexane/CH2Cl2. Mp: 148.7ꢂ149.5 1C
(decomp.); 1H NMR (CDCl3): d = 8.57 (d, 3JHH = 8 Hz, 2H),
1
139.0ꢂ140.0 1C (decomp.); H NMR (CDCl3): d = 8.21 (d,
3
3JHH = 8 Hz, 2H), 7.71 (br d, JHH = 8 Hz, 2H), 7.75–7.62
3
(m, 4H), 7.32 (d, JHH = 8 Hz, 2H), 7.12–7.23 (m, 6H),
6.91–6.98 (m, 2H), 4.85 (s, 1H), 3.76 ppm (s, 2H); 19F NMR
3
5
(CDCl3): d = ꢂ78.2 (t, JFF = 8.6 Hz, JFF = 8.6 Hz, 6F),
3
ꢂ78.4 (d, JFF = 8.6 Hz, 6F), ꢂ114.6 (dm, 2JFF = 287.1 Hz,
2
2
2F), ꢂ115.2 (dm, JFF = 287.1 Hz, 2F), ꢂ116.3 (d, JFF
=
2
287.1 Hz, 2F), ꢂ118.1 ppm (dm, JFF = 287.1 Hz, 2F);
elemental analysis: calcd (%) for C36H21F20O3As: C 45.21,
H 2.21; found: C 45.42, H 2.00; MS(EI(+)): m/z = 956 [M]+,
957 [M + 1]+, 958 [M + 2]+, 759 [M ꢂ CH2CPh2OH]+.
3
7.75–7.82 (m, 6H), 7.71 (d, JHH = 8 Hz, 2H), 7.54
3
(t, JHH = 8 Hz, 1H), 7.39 (t, JHH = 8 Hz, 2H), 4.30 (d,
3
Reactions with KH
2JHH = 16.1 Hz, 1H), 4.08 ppm (d, 2JHH = 16.1 Hz, 1H); 19
F
NMR (CDCl3): d = ꢂ78.1 (s, 6F), ꢂ79.2(dd, 3JFF = 17.2 Hz,
4a. A CD3CN (0.5 mL) solution of 4a (26.2 mg,
5JFF = 4.9 Hz, 6F), ꢂ115.6 (dq, JFF = 283.4 Hz, JFF
=
0.0433 mmol) was added to a CD3CN (0.3 mL) suspension
2
3
2
17.2 Hz, 2F), ꢂ116.1 (d, JFF = 283.4 Hz, 1F), ꢂ116.2
of KH (excess), then the mixture was stirred for 10 min at 0 1C.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1623–1629 | 1627

View Article Online
The supernatant was transferred to an NMR tube under N2,
and the reaction was monitored by 1H and 19F NMR at 25 1C.
During the reaction, the H NMR (CD3CN) showed a signal
MeCN) to afford 9 (RT = 16.8 min: 6.0 mg, 0.031 mmol,
93%) as a colorless liquid and 7a-K (RT = 14 min: 18.6 mg,
0.031 mmol, 95%) as a white solid. 9: 1H NMR (CDCl3): d =
7.30–7.37 (m, 10H), 3.29 ppm (s, 2H); MS(EI(+)): m/z = 196
[M]+, 197 [M + 1]+, 198[M + 2]+, 166 [M ꢂ CH2O]+. The
spectral data of 7a-K were consistent with those of the same
product described above.
1
for 8 (d = 2.54 ppm), however, the signal disappeared
probably due to the low boiling point of 8 (10 1C). After
monitoring for 6.5 h at 25 1C, the mixture was heated at 45 1C
for 4 h to complete the reaction. The mixture was quenched
with aqueous NH4Cl (10 mL). The mixture was extracted with
Et2O (2 ꢃ 20 mL), the organic layer was washed with water
(2 ꢃ 20 mL) and dried over anhydrous MgSO4. After removing
the solvents by evaporation, the resulting crude mixture was
separated by preparative TLC (n-hexane : CH2Cl2 = 3 : 1),
followed by HPLC (ODS, MeCN) to afford 7a-K (RT =
15.6 min: 25.1 mg, 0.0419 mmol, 96%) as a white solid. 7a-K:
1H NMR (CDCl3): d = 8.05 (d, 3JHH = 7.6 Hz, 2H), 7.55 (br
6b. A CD3CN (0.5 mL) solution of 6b (28.4 mg, 0.0296 mmol)
was added to a CD3CN (0.3 mL) suspension of KH
(excess), then the mixture was stirred for 10 min at 0 1C.
The supernatant was transferred to an NMR tube under N2,
and the reaction was monitored by 1H and 19F NMR at 30 1C.
1
The H NMR (CD3CN) showed a characteristic signal for 9
3
3
4
(d = 3.29 ppm). The reaction was completed within 20 min.
The mixture was quenched with aqueous NH4Cl (10 mL). The
mixture was then extracted with Et2O (2 ꢃ 30 mL), the organic
layer was washed with water (2 ꢃ 20 mL) and dried over
anhydrous MgSO4. After removing the solvents by evaporation,
the resulting crude mixture was separated by preparative TLC
(n-hexane : CH2Cl2 = 4 : 1), followed by HPLC (ODS,
MeCN) to afford 9 (RT = 16.8 min: 5.5 mg, 0.031 mmol,
94%) as a colorless liquid and 7b-K (RT = 14 min: 22.4 mg,
0.0281 mmol, 95%) as a white solid. 7b-K: 1H NMR (CDCl3):
d = 7.99 (d, 3JHH = 7.8 Hz, 2H), 7.63 (br d, 3JHH = 7.8 Hz,
d, JHH = 7.6 Hz, 2H), 7.37 (td, JHH = 7.6 Hz, JHH = 1.2
3
4
Hz, 2H), 7.30 ppm (td, JHH = 7.6 Hz, JHH = 1.2 Hz, 2H);
19F NMR (CDCl3): d = ꢂ74.0 (q, JFF = 8.6 Hz, 6F),
4
4
ꢂ76.6 ppm (q, JFF = 8.6 Hz, 6F).
6a.
A CD3CN (0.5 mL) solution of 6a (24.7 mg,
0.0326 mmol) was added to a CD3CN (0.3 mL) suspension
of KH (excess), then the mixture was stirred for 10 min at 0 1C.
The supernatant was transferred to an NMR tube under N2,
and the reaction was monitored by 1H and 19F NMR at 30 1C.
1
The H NMR (CD3CN) showed a characteristic signal for 9
3
4
(d = 3.29 ppm). The reaction was completed within 20 min.
The mixture was quenched with aqueous NH4Cl (10 mL). The
mixture was then extracted with Et2O (2 ꢃ 20 mL), the organic
layer was washed with water (2 ꢃ 20 mL) and dried over
anhydrous MgSO4. After removing the solvents by evaporation,
the resulting crude mixture was separated by preparative TLC
(n-hexane : CH2Cl2 = 3 : 1), followed by HPLC (ODS,
2H), 7.45 (t, JHH = 7.8 Hz, JHH = 1.2 Hz, 2H), 7.37 ppm
(td, 3JHH = 7.8 Hz, 4JHH = 1.2 Hz, 2H); 19F NMR (CDCl3):
d = ꢂ78.0 (dd, 3JFF = 16.0 Hz, 5JFF = 3.8 Hz, 6F), ꢂ78.6 (s,
2
6F), ꢂ115.2 (s, 4F), ꢂ117.3 (dm, JFF = 284.6 Hz, 2F),
2
ꢂ119.5 ppm (dm, JFF = 284.6 Hz, 2F). The spectral data
of 9 were consistent with those of the same product
described above.
Table 2 Crystallographic data for 4a, 5a, 5b, 6a and 6b
Compound
4a
5a
5b
6a
6bꢁCH2Cl2
Formula
Molecular weight
Crystal system
Space group
Color
Habit
C20H13AsF12O3
604.22
Orthorhombic
Pbca
Colorless
Plate
C26H15AsF12O3
678.30
Triclinic
P1
Colorless
Plate
C30H15AsF20O3
878.34
Triclinic
P1
Colorless
Plate
C32H21AsF12O3
756.41
Monoclinic
C2/c
Colorless
Plate
C37H23AsCl2F20O3
1041.37
Monoclinic
C2/c
Colorless
Plate
ꢀ
ꢀ
Crystal dimensions/mm
a/A
b/A
c/A
a/1
b/1
0.60 ꢃ 0.40 ꢃ 0.40
9.8110(1)
17.1930(3)
24.8070(5)
90
0.60 ꢃ 0.40 ꢃ 0.30
9.2340(1)
10.9850(2)
14.1180(4)
85.094(1)
80.621(1)
66.995(1)
1300.18(5)
2
1.733
1.422
672
Mo-Ka, 0.71073
293
0.60 ꢃ 0.40 ꢃ 0.40
9.9960(1)
12.3680(2)
14.9450(3)
105.926(1)
105.845(1)
103.150(1)
1615.31(4)
2
1.806
1.201
864
Mo-Ka, 0.71073
293
0.30 ꢃ 0.20 ꢃ 0.20
40.3950(8)
12.0720(2)
26.7000(6)
90
93.526(1)
90
12995.6(4)
16
1.546
0.60 ꢃ 0.40 ꢃ 0.30
39.9674(1)
9.5830(5)
26.4609(6)
90
127.3997(10)
90
8051.2(5)
8
1.718
90
90
g/1
V/A3
4184.46(12)
8
1.918
1.754
2384
Z
Dcalcd/g cmꢂ3
Abs. coeff./mmꢂ1
F(000)
1.147
6048
Mo-Ka, 0.71073
173
1.107
4128
Mo-Ka, 0.71073
293
Radiation, l/A
T/K
Mo-Ka, 0.71073
173
Data, collected
Data/restrains/parameters
R1[I > 2s(I)]
wR2 (all data)
GOF
+h, +k, +l
4676/1/329
0.0398
0.1305
1.123
n-Hexane/CH2Cl2
+h, ꢄk, ꢄl
5599/0/379
0.0505
0.1446
1.178
n-Hexane/CH2Cl2
+h, ꢄk, ꢄl
7112/0/487
0.0381
0.1215
1.145
n-Hexane/CH2Cl2
+h, +k, ꢄl
14609/0/912
0.0640
0.2135
1.112
n-Hexane/CH2Cl2
ꢄh, +k, ꢄl
15491/0/568
0.0658
0.2053
1.162
n-Hexane/CH2Cl2
Solvent for crystallization
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1628 | New J. Chem., 2010, 34, 1623–1629

View Article Online
Single crystal X-ray analyses
2 (a) D. Lloyd, I. Gosney and R. A. Ormiston, Chem. Soc. Rev.,
1987, 16, 45–74; (b) I. Gosney, T. J. Lillie and D. Lloyd, Angew.
Chem., Int. Ed. Engl., 1977, 16, 487–488; (c) H. S. He, C. W.
Y. Chung, T. Y. S. But and P. H. Toy, Tetrahedron, 2005, 61,
1385–1405.
3 W. C. Still and V. J. Novack, J. Am. Chem. Soc., 1981, 103,
1283–1285.
4 J. D. Hsi and M. Koreeda, J. Org. Chem., 1989, 54, 3229–3231.
5 K.-y. Akiba, Chemistry of Hypervalent Compounds, Wiley-VCH,
New York, 1999.
Crystals of 4a, 5a, 5b, 6a and 6b suitable for the X-ray
structural determination were mounted on a Mac Science
DIP2030 imaging plate diffractometer and irradiated with
graphite monochromated Mo-Ka radiation (l = 0.71073 A)
for the data collection. The unit cell parameters were
determined by separately autoindexing several images in each
data set using the DENZO program (MAC Science).16 For
each data set, the rotation images were collected in 3 degree
increments with a total rotation of 1801 about the f axis. The
data were processed using SCALEPACK. The structure was
solved by a direct method using the SHELX-97 program.17
Refinement on F2 was carried out using the full-matrix
least-squares by the SHELX-97 program.17 All non-hydrogen
atoms were refined using the anisotropic thermal parameters
except for the disordered fluorine atoms. The hydrogen atoms
were included in the refinement along with the isotropic
thermal parameters. The crystallographic data are summarized
in Table 2.y
6 S. Kojima, R. Takagi and K.-y. Akiba, J. Am. Chem. Soc., 1997,
119, 5970–5971.
7 S. Kojima, K. Kawaguchi, S. Matsukawa, K. Uchida and
K.-y. Akiba, Chem. Lett., 2002, 170–171.
8 (a) S. Kojima, K. Kawaguchi and K.-y. Akiba, Tetrahedron Lett.,
1997, 38, 7753–7756; (b) S. Kojima, K. Kawaguchi, S. Matsukawa
and K.-y. Akiba, Tetrahedron, 2003, 59, 255–265.
9 J. C. Martin, Science (Washington, D. C.), 1983, 221, 509–514, and
references therein.
10 (a) X.-D. Jiang, K.-i. Kakuda, S. Matsukawa, H. Yamamichi,
S. Kojima and Y. Yamamoto, Asian J. Chem., 2007, 2, 314–323;
(b) X.-D. Jiang, S. Matsukawa, H. Yamamichi and Y. Yamamoto,
Inorg. Chem., 2007, 46, 5480–5482; (c) X.-D. Jiang, S. Matsukawa,
H. Yamamichi and Y. Yamamoto, Heterocycles, 2007, 73,
805–824; (d) X.-D. Jiang, S. Matsukawa and Y. Yamamoto,
Dalton Trans., 2008, 3678–3687.
11 Y. Yamamoto, K. Toyata, Y. Wakisaka and K.-y. Akiba, Heteroat.
Chem., 2000, 11, 42–47.
12 G. Schroder, T. Okinaka, Y. Mimura, M. Watanabe,
¨
Acknowledgements
The authors are grateful to Central Glass Co., Ltd., for the
generous gift of the hexafluorocumyl alcohol. This study was
supported by a Grant-in-Aid for Scientific Research (No.
17350021) from the Ministry of Education, Culture, Sports,
Science and Technology, Japan.
T. Matsuzaki, A. Hasuoka, Y. Yamamoto, S. Matsukawa and
K.-y. Akiba, Chem.–Eur. J., 2007, 13, 2517–2529, and references
therein.
13 For the hexacoordinate phosphates bearing an oxaphosphetane
ring, see: (a) S. Matsukawa, S. Kojima, K. Kajiyama,
Y. Yamamoto, K.-y. Akiba, S. Re and S. Nagase, J. Am. Chem.
Soc., 2002, 124, 13154–13170; (b) S. Kojima and K.-y. Akiba,
Tetrahedron Lett., 1997, 38, 547–550; (c) T. Kawashima,
K. Watanabe and R. Okazaki, Tetrahedron Lett., 1997, 38,
551–554.
14 Details of the stereochemical study will be published elsewhere.
15 For the proposed nomenclature for the pentacoordinated atom,
see: J. C. Martin and T. M. Balthazor, J. Am. Chem. Soc., 1977, 99,
152–162.
References
1 For leading references, see: (a) W. S. Wadsworth, Jr., Org. React.,
1977, 25, 73–253; (b) J. I. G. Cadogan, Organophosphorus Reagents
in Organic Synthesis, Academic Press, New York, 1979;
(c) B. E. Maryanoff and A. B. Reitz, Chem. Rev., 1989, 89,
863–927; (d) S. E. Kelly, in Comprehensive Organic Synthesis, ed.
B. M. Trost and I. Fleming, Pergamon Press, Oxford, 1991, vol. 1,
pp. 730–817; (e) A. W. Johnson, Ylides and Imines of Phosphorus,
Wiley-Interscience, New York, 1993; (f) E. Vedejs and
M. J. Peterson, Top. Stereochem., 1994, 21, 1–157; (g) E. Vedejs
and M. J. Peterson, Adv. Carbanion Chem., 1996, 2, 1–85.
16 Z. Otwinowski and W. Minor, Macromolecular Crystallography, in
Methods in Enzymology, ed. C. W. Carter, Jr. and R. M.
Sweet, Academic Press, New York, 1997, vol. 276, part A,
pp. 307–326.
17 G. M. Sheldrick, Acta Crystallogr., 2008, A64, 112–122.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1623–1629 | 1629