Lipase-mediated stereoselective transformations of chiral organophosphorus P-boranes revisited: Revision of the absolute configuration of alkoxy(hydroxymethyl)phenylphosphine P-boranes

Article abstract of DOI:10.1016/j.tetasy.2011.08.024

The lipase-promoted kinetic resolution of a series of alkoxy(hydroxymethyl) phenylphosphine P-boranes proceeded with moderate stereoselectivity to give both the unreacted substrates and their O-acetyl derivatives. The absolute configurations of the produc

Full text of DOI:10.1016/j.tetasy.2011.08.024

Tetrahedron: Asymmetry 22 (2011) 1581–1590
Contents lists available at SciVerse ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Lipase-mediated stereoselective transformations of chiral organophosphorus
P-boranes revisited: revision of the absolute configuration of
alkoxy(hydroxymethyl)phenylphosphine P-boranes
⇑
´
´
Małgorzata Kwiatkowska, Grzegorz Krasinski, Marek Cypryk, Tomasz Cierpiał, Piotr Kiełbasinski
´
Centre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Sienkiewicza 112, 90-363 Łódz, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
The lipase-promoted kinetic resolution of a series of alkoxy(hydroxymethyl)phenylphosphine P-boranes
proceeded with moderate stereoselectivity to give both the unreacted substrates and their O-acetyl deriv-
atives. The absolute configurations of the products, which were earlier ascribed on the basis of the ster-
eoselective reduction of the corresponding phosphine oxides with borane and comparison with the
literature data concerning bicyclic phosphine oxides, were disputed by theoretical calculation. Some
additional studies were carried out, including theoretical calculations and more accurate chemical corre-
lation, which proved that the borane reduction of acyclic phosphine oxides proceeded with inversion of
configuration at the phosphorus center and, therefore, the former assignment of the absolute configura-
tions was incorrect. On this basis, the stereochemistry of the enzymatic reaction was ultimately deter-
mined. A mechanism of the borane reduction of acyclic phosphine oxides explaining inversion of
configuration at phosphorus is proposed.
Received 1 August 2011
Accepted 25 August 2011
Available online 6 October 2011
Ó 2011 Elsevier Ltd. All rights reserved.
BH3
P
BH3
P
1. Introduction
BH3
P
AcOCH=CH2
lipase
RO
Ph
+
Chiral, non-racemic organophosphorus compounds containing
a stereogenic phosphorus atom play an important role in various
areas of current research, such as asymmetric organic synthesis,
biochemistry and catalysis. For example, trivalent phosphorus
compounds, especially tertiary phosphines, are used as chiral li-
gands in transition metal catalysts. Recently, there has been grow-
ing interest in the synthesis and transformation of borane
complexes of trivalent phosphorus compounds.1,2 In contrast to
phosphines and other derivatives of trivalent phosphorus, which
are prone to oxidation and are usually difficult to handle, they
are stable compounds and can be easily converted into the corre-
sponding PIII compounds without racemization. Following this ten-
dency and taking into account the fact that many hydrolytic
enzymes proved to be capable of recognizing and stereoselectively
binding heteroatom stereogenic centers, among them those lo-
cated on the phosphorus,3 we4 and others5,6 applied biocatalytic
methods in the synthesis of optically active borane complexes of
chiral PIII compounds. Our investigations concerned the kinetic res-
olution of racemic P-stereogenic alkoxy(hydroxymethyl)phenyl-
phosphine P-boranes 1 via their enzymatic acetylation (Eq. 1).
OH
RO
Ph
Ph
RO
(+)-(R)-2
OAc
OH
rac-1
(-)-(S)-1
a, R=Me b, R=Et c, R=i-Pr
ð1Þ
The results showed that the P-boranes, in contrast to the anal-
ogous P-stereogenic alkoxy(hydroxymethyl)phenylphosphine oxi-
des
described previously (Eq. 2),7,8 were relatively poor
3
substrates for lipase-catalyzed transformations and underwent
similar reaction that was much slower and with low stereoselectiv-
ity (e.g., for the CAL-B-catalyzed acetylation of 3c the enantiomer
ratio E = 32, while for 1c E = 3).
AcOCH=CH2
lipase
O
P
O
P
O
P
RO
Ph
OH
OH
OAc
+
Ph
RO
(-)-(R)-3
RO
Ph
(+)-(S)-4
rac-3
a, R=Me b, R=Et c, R=Pri
ð2Þ
⇑
Corresponding author.
E-mail address: piokiel@cbmm.lodz.pl (P. Kiełbasin´ ski).
0957-4166/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2011.08.024

1582
M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
In contrast to the phosphine oxides 3, whose absolute configu-
ferent behavior of enzymes seemed very interesting. Attempting
to explain this phenomenon, we decided to perform molecular
modeling for the enzymatic reaction. The theoretical calculations
cast doubt on this result, predicting the same stereochemical
course of both reactions shown in Eqs. (1) and (2), which means
that the replacement of oxygen by a borane group at the phospho-
rus stereogenic center does not influence the stereorecognition by
the enzyme.11 Since there is no doubt about the absolute configu-
ration of the phosphine oxides 3 and 4,7,8 we have surmised that
the absolute configuration of P-boranes 1 and 2 might have been
wrongly ascribed4 and decided to reinvestigate it using both theo-
retical calculations and a more detailed chemical correlation.
rations were known,7,8 the direct determination of the absolute
configuration of 1 using X-ray analysis could not be performed
since no crystals could be obtained. Therefore, a simple chemical
correlation was performed, which was based on the reduction of
enantiomerically enriched hydroxymethyl(i-propoxy)phenylphos-
phine oxide (ꢀ)-(R)-3c and acetoxymethyl(i-propoxy)phenylphos-
phine oxide (+)-(S)-4c with borane.4 As a conclusion, the absolute
configurations (+)-(R) and (ꢀ)-(S) were ascribed to the resulting
1c, assuming, by analogy to the borane reduction of bicyclic phos-
phine oxides,9,10 that the reaction proceeded with retention of con-
figuration at the phosphorus center. On this basis, the
stereochemistry of the lipase-catalyzed acetylation of 1 was con-
sidered to be as shown in Eq. 1.4 Moreover, comparison of the ste-
reochemistry of the two analogous enzymatic reactions, the
acetylation of 1 (Eq. 1)4 and the acetylation of 3 (Eq. 2),7,8 indicated
that the same enzymes recognized and preferentially transformed
the enantiomers of 1 and 3 (and led to the corresponding acetates 2
and 4) which had opposite spatial arrangement around the
phosphorus.
2. Results and discussion
2.1. Synthesis of (hydroxymethyl)phenylphosphine P-boranes
Enantiomerically pure alkoxy(hydroxymethyl)phenylphos-
phine boranes 1 were synthesized by reduction of the correspond-
ing enantiomerically pure phosphine oxides 3 or 4, obtained via
iterative enzymatic resolution (Eq. 2), with borane–THF complex
(Eq. 3, Table 1, entries 1–4). The reaction proceeded smoothly,
but unwanted by-products, a secondary phosphine borane 5 and
hydroxyphosphine borane 6 were always formed. They were the
products of a subsequent/simultaneous reduction of the alkoxy
group and their content in the reaction mixture was higher in
the case of lower alkyl substituents. For obvious reasons, no forma-
tion of these types of by-products was not observed in the reduc-
tion of t-butyl(hydroxymethyl)phenylphosphine 7 (Table 1, entry
5). It should be noted that 7 was synthesized according to the lit-
erature12 and was not enantiomerically pure. When acetates 4
were subjected to the reaction with borane, the acetyl group was
always removed.
BH3
O
P
RO
Ph
OAc
P
Ph
RO
(+)-(R)-2
OAc
(+)-(S)-4
This would mean that a simple replacement of the oxygen atom
with the borane moiety at the phosphorus atom would result in an
inversion of stereochemistry of the enzymatic reaction. Such dif-
Table 1
Borane reduction of (hydroxymethyl)phenylphosphine oxides
Entry
1
R1
R2
Phosphine oxide
Phosphine borane
Symbol
[a
]
D
Absol. config.
Symbol
Yield (%)
11
[a]
D
+80.2a
MeO
H
3a
ꢀ24.7a
ꢀ39.0b
+53.0a
ꢀ22.2a
ꢀ58.3b
+38.8a
ꢀ4.8a,c
(R)
1a
2
3
MeO
i-PrO
Ac
H
4a
3c
(S)
(R)
1a
1c
10
60
ꢀ80.2a
+101.0a
4
5
i-PrO
t-Bu
Ac
H
4c
7
(S)
(S)
1c
14
45
95
ꢀ102a
+3.9a,c
a
b
c
In chloroform, c 1.
In methanol, c 1.
ee = 45%.
O
P
O
P
O
P
TosCl, Et3N
92%
NaI. acetone
89%
I
OH
OTos
i-PrO
i-PrO
Ph
i-PrO
Ph
Ph
(+)-(S)-9
(+)-(S)-3c
(-)-(S)-8
BH3
P
O
BH3/THF
25%
Bu3Sn, AIBN
92%
P
i-PrO
Ph
Me
Ph
Me
i-PrO
(+)-(R)-10
(+)-11
Scheme 1. Synthesis of phosphine oxide 10 and phosphine borane 11.

M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
1583
X
½
aꢁD
¼
xi½aꢁiD
BH3
P
BH3
P
BH3
P
O
P
R1
R1
i
H
HO
Ph
BH3 / THF
OR2
OH
+
+
OH
OH
where xi is the fractional population of conformer i, whose specific
Ph
Ph
Ph
rotation is ½aꢁiD.
3, 4 or 7
6
5
1 or 14
The Boltzmann distribution of the conformers was calculated on
the basis of the Gibbs free energies of the conformers obtained at
the B3LYP/6-311++G(2d,2p)//B3LYP/6-31+G⁄ level (Tables 2 and 3).
The lowest energy conformers of 1c and 3c, considered in opti-
cal rotation calculations are shown in Figures 1 and 2, respectively,
The specific rotation for particular conformers can adopt very
different values and can even change the sign. This was particularly
apparent for the 1c conformers for which the specific rotation var-
ied from ꢀ300 to +190 (see Table 2). However, among the numer-
ous conformers, only a few have a determining effect on the
specific rotation. There are two such essential conformers for 3c
and three for 1c. Calculations predict that the enantiomers of 3c
and 1c having analogous spacial arrangement should show specific
ð3Þ
The phosphine oxide (+)-(R)-10 was synthesized according to the
procedure previously described by us7 and transformed into the
corresponding borane 11 (Scheme 1).
2.2. Determination of the absolute configuration of P-boranes 1
2.2.1. Theoretical calculations
Since in a previous paper,4 only the O-i-propyl derivatives 1c
and 2c were examined, we limited our calculations to these deriv-
atives. Although the absolute configuration of 3c was known, we
also performed calculations for this derivative, as this allowed us
to verify the reliability of the methodology applied.
In order to estimate the value of the specific rotation for 1c and
3c, we calculated the specific rotation of each conformer using the
B3LYP/6-311++G(2d,2p) method for the D line of Na (k = 589 nm).
The net specific rotation for the given compound was calculated
as the population-weighted average of the rotations of individual
conformations according to the equation shown below,13
rotations with the same sign, that is, calculated [a] for (R)-enan-
D
tiomers 3c and 1c in the gas phase are ꢀ54.3 and ꢀ67.4, respec-
tively. The predicted gas phase specific rotations show
appreciable deviations from the experimental data, 32 for 3c and
34 for 1c. However, the specific rotation of 1c calculated in CHCl3
(ꢀ85.2), is much closer to the experimental value (ꢀ102.0) in
terms of absolute numbers than the gas phase value. However,
due to the convergence problems with optimization of some
Figure 1. Seven lowest energy conformers of (R)-1c.

1584
M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
Figure 2. Eight lowest energy conformers of (R)-3c.
Table 2
Boltzmann distribution of conformers and the specific rotation calculated for (R)-hydroxymethyl(i-propoxy)phenylphosphine P-borane 1c (in parentheses, specific rotations in
CHCl3 are given)
Conformer
[a
]
D
xi
Molecular specific rotation
(R)-1cA
(R)-1cB
(R)-1cC
(R)-1cD
(R)-1cE
(R)-1cF
(R)-1cG
ꢀ141.2 (ꢀ148.6)
+39.6 (+43.4)
0.470 (0.528)
0.43 (0.37)
0.0004 (0.0)
0.0004 (0.0014)
0.0007 (0.0011)
0.0004 (0.0005)
0.098 (0.1003)
Net specific rotation
ꢀ66.3 (ꢀ78.4)
+17.05 (+16.0)
+0.1 (0.0)
+189.4 (+196.1)
ꢀ4.7 (+23.1)
0.00 (+0.03)
ꢀ302.15 (ꢀ324.3)
ꢀ136.0 (ꢀ134.15)
ꢀ182.5 (ꢀ223.1)
ꢀ0.2 (ꢀ0.35)
ꢀ0.05 (ꢀ0.07)
ꢀ17.8 (ꢀ22.37)
ꢀ67.4 (ꢀ85.2)
conformer structures of 3c in CHCl3, the calculation of optical rota-
tion of 3c in solution could not be completed. Thus, we believe that
the discrepancy between the experimental and calculated specific
rotation of 3c is mainly due to neglecting the solvent effect in the
calculations.
with THF or borane with dimethyl sulfide (in both cases the results
were identical) to give (+)-1 of unknown absolute configuration. If
the reaction proceeded with retention of configuration, its absolute
configuration should be (R) (Route A), if with inversion, its config-
uration should be (S) (Route B). The subsequent reaction, removal
of the borane moiety by 1-octene followed by the addition of sul-
fur, leading to phosphine sulfide (+)-12, must proceed with reten-
tion of configuration.14 Hence, the absolute configuration of (+)-12
obtained was either (S) (Route A) or (R) (Route B). Finally, oxidation
of (+)-12 with iodoxybenzene gave the starting (ꢀ)-(R)-3. This
means that either all of the transformations proceeded with reten-
tion of configuration (Route A) or two of the three reactions,
2.2.2. Chemical correlation
The detailed chemical correlation was performed using the
phosphine oxides 3a and 3c and phosphine boranes 1a and 1c
(Scheme 2). Since in each case the stereochemical course was the
same, the transformations are presented using only the compound
number. Thus, (ꢀ)-(R)-3 was reacted with a complex of borane

M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
1585
Table 3
To sum up, the absolute configuration of 1 is (+)-(S)-1a,c and
(ꢀ)-(R)-1a,c, which is in full agreement with the theoretical calcu-
lations presented above and which is contrary to the previous
chemical correlation.4 In this context, the stereochemistry of the
enzymatic reaction shown in Eq. 1, must also be reversed and read
as follows (Eq. (4)).
Boltzmann distribution of conformers and the specific rotation calculated for (R)-
hydroxymethyl(i-propoxy)phenylphosphine oxide 3c
Conformer
[
a
]
xi
Molecular specific rotation
D
(R)-3cA
(R)-3cB
(R)-3cC
(R)-3cD
(R)-3cE
(R)-3cF
(R)-3cG
(R)-3cH
ꢀ86.3
ꢀ25.2
ꢀ23.2
0.43072
0.54204
0.00005
0.01182
0.00658
0.00003
0.00866
0.00011
ꢀ37.16
ꢀ13.66
0.00
ꢀ1.72
ꢀ1.22
0.00
ꢀ0.50
ꢀ0.02
ꢀ54.30
BH3
P
BH3
P
ꢀ145.8
ꢀ186.0
ꢀ27.2
BH3
P
AcOCH=CH2
lipase
RO
Ph
Ph
RO
OH
RO
Ph
OAc
OH
+
ꢀ58.1
ð4Þ
ꢀ195.5
rac-1
(-)-(R)-1
(+)-(S)-2
Net specific rotation
a, R=Me b, R=Et c, R=i-Pr
namely the reduction of 3 with borane and oxidation with iodoxy-
benzene, proceeded with inversion (Route B). As this sequence of
reactions could not give an answer, (ꢀ)-(R)-3 was independently
transformed into phosphine sulfide 12. The initial acetylation gave
acetate (ꢀ)-(R)-4, which was treated with Lawesson reagent to af-
ford the thiono derivative (ꢀ)-13. Removal of the acetyl group
using sodium methoxide led to (ꢀ)-12, thus the opposite enantio-
mer to that obtained in the previous reaction. Since the Lawesson
reagent is known to transform phosphine oxides into phosphine
sulfides with retention of configuration at phosphorus,15 the (S)-
configuration was tentatively ascribed to (ꢀ)-12. Moreover, its oxi-
dation with iodoxybenzene produced the opposite enantiomer of
the starting 3, that is, (+)-(S)-3 with a diminished enantiomeric ex-
cess, which undoubtedly proved that this reaction proceeded with
predominant inversion of the configuration at the phosphorus, the
result being contrary to the earlier literature reports.15–17 This
overall result allowed us to discard Route A and to ultimately con-
clude that the reduction of the phosphine oxide 3 with borane,
which leads to phosphine borane (+)-(S)-1, proceeds with inversion
of configuration.
2.3. Stereochemistry of the reaction of various phosphine
oxides with borane
Since it has been found that the borane reduction of the phos-
phine oxides 3, containing one C–O–P bond and the hydroxy-
methyl substituent, proceeds with different stereochemistry
(inversion at phosphorus) than the reduction of the bicyclic phos-
phine oxides, which have only C–P bonds and are deprived of the
hydroxymethyl substituent,9,10 we decided to investigate the im-
pact of the ‘extra’ oxygen atoms on the mechanism of the reduc-
tion. To this end, two sequences of transformations were
performed. In the first one, the stereochemistry of the borane
reduction of t-butyl(hydroxymethyl)phenylphosphine oxide
7
(thus having only direct C–P bonds and a pending hydroxymethyl
substituent) was studied (Scheme 3). In the second one, i-prop-
oxymethylphenylphosphine oxide 10 (thus having one C–O–P
bond and no hydroxymethyl substituent) was used as the subject
of analogous stereochemical investigations (Scheme 4).
Route A
Route B
BH3
BH3
P
O
P
BH3/THF
P
RO
Ph
OH
Ph
RO
OH
Ph
RO
OH
or
inversion
retention
(+)-(R)-1
(-)-(R)-3
(+)-(S)-1
1-octene
S8, THF
retention
O
S
S
PhIO2
P
P
P
RO
OH
Ph
OH
Ph
RO
(+)-(S)-12
OH
or
A) retention
B) inversion
Ph
RO
(+)-(R)-12
(-)-(R)-3
AcCl, Et3N
S
P
S
P
O
P
O
P
PhIO2
MeONa
RO
Ph
Lawesson
retention
OH
Ph
RO
(-)-(S)-13
OAc
Ph
OH
Ph
RO
OAc
inversion
RO
(-)-(S)-12
(+)-(S)-3
(-)-(R)-4
Scheme 2. Chemical correlation of the absolute configurations of 1 and 3.

1586
M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
BH3
O
P
BH3/THF
inversion
P
Ph
t-Bu
t-Bu
Ph
(+)-(S)-14
OH
OH
(-)-(S)-7
1-octene
S8, THF
retention
O
S
P
PhIO2
P
Ph
t-Bu
(-)-(S)-7
OH
t-Bu
Ph
OH
inversion
(-)-(R)-15
AcCl, Et3N
S
P
S
P
O
MeONa
Lawesson
retention
Ph
t-Bu
(+)-(S)-17
OAc
Ph
t-Bu
(+)-(S)-15
OH
P
Ph
t-Bu
(+)-(S)-16
OAc
Scheme 3. Chemical correlation of the absolute configurations of 7 and 14.
BH3
P
BH3
O
BH3/THF
P
Ph
i-PrO
Ph
or
Me
Me
retention
P
i-PrO
Ph
Me
i-PrO
inversion
(+)-(R)-11
(+)-(S)-11
(+)-(R)-10
1-octene
S8, THF
retention
S
S
O
PhIO2
P
P
P
Ph
i-PrO
i-PrO
Ph
or
Me
Me
Me
A) retention
B) inversion
i-PrO
(-)-(R)-18
Ph
(+)-(R)-10
(-)-(S)-18
Lawesson
O
S
P
PhIO2
P
Ph
i-PrO
Ph
Me
Me
inversion
i-PrO
10
(-)-(S)-
(+)-(S)-18
Scheme 4. Chemical correlation of the absolute configurations of 10 and 11.
Scheme 3 clearly shows that also in the case of t-butyl(hydroxy-
methyl)-phenylphosphine oxide 7, the borane reduction proceeds
with inversion of configuration. This can be undoubtedly con-
cluded on the basis of the absolute configurations of 7 and 14,
which are known from the literature.5,12,18 Nevertheless, some
additional transformations, analogous to those presented in
Scheme 2, have been performed and proved that their stereochem-
ical course was identical as for 1 and 3.

M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
1587
All of the discussions presented above for the correlation be-
tween 3 and 1 (Scheme 2) and 7 and 14 (Scheme 3) also proved va-
lid for the correlation between 10 and 11 (Scheme 4).
phosphine oxides9,10 does not apply to this transformation. A pro-
posal of a new mechanism is shown in Scheme 5 for the reduction
of 3 to 1, although it may be extended to the reduction of the
remaining types of acyclic phosphine oxides. It seems reasonable
to assume that instead of a hydride shift from boron to phosphorus
in the initially formed O-boron-phosphine oxide 19, which would
lead to a pentacoordinate phosphorane 20,9,10 an attack of the hy-
dride anion on the oxygen atom takes place. This causes breaking
of the phosphorus–oxygen bond and an attack of the electron pair
of this bond on the next borane molecule, which ultimately results
in the inversion of configuration at the phosphorus (Scheme 5).
Such a difference between the mechanisms of the reduction of
bicyclic and acyclic phosphine oxides may be explained in terms
of both the higher ability of cyclic phosphorus derivatives to form
hypervalent pentacoordinate intermediates, and steric hindrance
around phosphorus in bicyclic phosphine oxides.
It is noteworthy to add that the reaction sequence: transforma-
tion of a phosphine oxide into a phosphine sulfide by the Lawesson
reagent (proceeding with retention of configuration) and oxidation
of the resulting phosphine sulfide to the parent phosphine oxide
with iodoxybenzene (proceeding with inversion of configuration)
constitutes an example of a Walden cycle. The only drawback is
the lack of full stereoselectivity during the oxidation step, which
proceeds with a partial racemization and causes a loss of steric
integrity of the compounds involved.
2.4. Mechanism of the reaction of phosphine oxides with
borane
All the results presented above undoubtedly prove that the bor-
ane reduction of acyclic phosphine oxides proceeds with inversion
of configuration, irrespective of the substituents at phosphorus.
Therefore, the mechanism proposed by Keglevich et al. for bicyclic
For the sake of completeness, another possible explanation of
the stereochemical course of the reaction investigated was taken
into account (Scheme 6).19 It assumes a simple nucleophilic attack
of a hydride anion on the phosphonium derivative 19, based on the
H2
B
BH3
H
O
P
O
P
RO
Ph
BH3
- H2BOH
OH
OH
P
Ph
RO
H3B
OH
Ph
RO
P
RO
Ph
OH
BH3
(+)-(S)-1
19
(-)-(R)-3
H2B
Ph
O
H
P
OH
OR
20
Scheme 5. Proposed mechanism of the reduction of acyclic phosphine oxides with borane.
BH3
BH3
O
BH3
BH3
O
-
O
P
O
P
O
BH3
OH
OH
O
P
OH
Ph
RO
Ph
Ph
RO
H
RO
H2B
19
(-)-(R)-3
BH3
P
RO
Ph
RO
Ph
RO
Ph
BH3
P
H
OH
OH
OH
P
P
BH3
RO
Ph
OH
- BH2OH
OBH2
(+)-(S)-1
21
22
Scheme 6. An alternative mechanism of the reduction of acyclic phosphine oxides with borane.

1588
M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
fact that the borane–THF complex is sufficiently activated to gen-
erate the hydride. The nucleophilic substitution with the hydride
proceeds with inversion of the configuration at the phosphorus
atom and produces the protonated phosphine 21, with a simulta-
neous regeneration of the borane–THF complex. Abstraction of
the proton from the phosphorus atom in 21 leads to phosphine
22, which ultimately reacts with the next borane molecule to give
the phosphine P-borane 1 with inversion of configuration. Never-
theless, at this stage, neither mechanism can be proven.
(10 mL) and the solution was refluxed for 2 weeks (31P NMR con-
trol). The solvent was evaporated, and to the residue water was
added and the mixture was extracted with CH2Cl2. The organic
solution was washed with an aqueous solution of Na2S2O3 and
dried over MgSO4. After evaporation of the solvent, the crude mix-
ture was purified by column chromatography using CH2Cl2 as elu-
ent. Yield: 0.3022 g (89%). [a]D = + 26.2 (c 1.12, CHCl3). Ee = 100%
[HPLC Chiralpak OD; Hexane: (i-PrOH/EtOH 4:1) 98:2; fl. 0.5 mL/
min; tR = 39.09 min]. 31P NMR (CDCl3): d 32.9. 1H NMR (CDCl3): d
1.24 (1.24, d, J = 6.15 Hz, 3H), 1.45 (d, J = 6.15 Hz, 3H), 3.05–3.25
(m, 2H), 4.63–4.77 (m, 1H), 7.43–7.89 (m, 5H). MS: m/z 325 (M+H).
3. Conclusions
4.2.3. Synthesis of isopropoxy(methyl)phenylphosphine oxide
(+)-(R)-10
A reinvestigation of the sterical course of the borane reduction
of acyclic phosphine oxides was performed using both theoretical
calculations and a detailed chemical correlation. It allowed us to
prove that the reduction proceeds with full inversion of configura-
tion at the phosphorus and to correct the former wrongly ascribed
absolute configurations of the resulting phosphine P-boranes. On
this basis, the formerly established stereochemistry of the lipase-
Compound (+)-(S)-9 (0.2883 g, 0.8898 mmol) and Bu3SnH
(0.3096 g, 1.0678 mmol) and AIBN (a few mg) were dissolved in
benzene (20 mL) and refluxed for one week. The next sample of
Bu3SnH (0.060 g) and AIBN were added and the refluxing was con-
tinued for another day. After completion of the reaction (31P NMR
control), the solvent was evaporated and the residue was purified
by column chromatography using CH2Cl2 as eluent. Yield: 0.616 g
promoted kinetic resolution of
a series of alkoxy(hydroxy-
methyl)phenylphosphine P-boranes was also corrected and ulti-
mately determined. An interesting example of a Walden cycle
was achieved via the reaction sequence: transformation of a phos-
phine oxide into a phosphine sulfide by the Lawesson reagent (pro-
ceeding with retention of configuration) and oxidation of the
resulting phosphine sulfide to the parent phosphine oxide with
iodoxybenzene (proceeding with a predominant inversion of con-
figuration). A mechanism of the borane reduction of acyclic phos-
phine oxides explaining inversion of configuration at the
phosphorus center has also been proposed.
(92%). [a]D = +40.0 (c 1.06, CHCl3), ee = 100%. [HPLC, Chiralpak
OD; Hexane: (i-PrOH/EtOH 4:1) 98:2; fl. 0.5 mL/min;
tR = 20.3 min]. 31P NMR (CDCl3): d 40.7. 1H NMR (CDCl3): d 1.14
(d, J = 6.15 Hz, 3H), 1.37 (d, J = 6.15 Hz, 3H), 1.64 (d, J = 14.58 Hz,
3H), 4.43–4.59 (m, 1H), 7.41–7.87 (m, 5H). HRMS: calcd for
C10H15O2P 198.080969, found 198.08101.
4.3. Synthesis of phosphine boranes—general procedure
A phosphine oxide 3, 7, or 10 was dissolved in a solution of BH3
in THF (1 M, 6 equiv) and the mixture was stirred at room temper-
ature. The reaction was monitored by 31P NMR and was stopped
when the substrate was consumed. The solvent was removed un-
der vacuum. The residue was separated by column chromatogra-
phy using dichloromethane as solvent to give 1, 14, or 11,
respectively.
4. Experimental
4.1. General
NMR spectra were recorded on a Bruker instrument at 200 MHz
for 1H and 81 MHz for 31P with CDCl3 as solvent. Optical rotations
were measured on a Perkin–Elmer 241 MC polarimeter at 20 °C.
Column chromatography was carried out using Merck 60 silica
gel. TLC was performed on Merck 60 F254 silica gel plates. The enan-
tiomeric excess (ee) values were determined by chiral HPLC (Varian
Pro Star 210, Chiralpak OD). Enantiomerically pure 3a and 3c were
obtained according to the literature,7,8 respectively. Enantiomeri-
cally enriched 7 was also prepared according to the literature.12
4.3.1. For compound (+)-(S)-14 see the literature5
4.3.1.1. Compound (+)-(S)-1a.
Yield: 11%. [a]D = +80.2 (c 1.15,
CHCl3), ee = 100% [HPLC: Chiralpak OD; Hexane: (i-PrOH/EtOH 4:1)
98:2; fl. 0.5 mL/min; tR = 31.6 min]. 31P NMR (CDCl3): d 112.85 (q,
JP-B = 63.04 Hz). 1H NMR (CDCl3): d 0.75 (br q, JB-H = 94.71 Hz, 3H,
BH3), 1.90 (br s, 1H, OH), 3.72 (d, 2H, J = 11.67 Hz), 4.12 (s, 2H,
CH2OH), 7.46–7.92 (m, 5H, Ph). MS (CI): m/z 183 (MꢀH).
4.2. Synthesis of phosphine oxide 10
4.3.1.2. Compound (+)-(S)-1c.
Yield: 60%. [a]D = +101 (c 1.15,
CHCl3), ee = 100%. For analytical data see the literature.4
4.2.1. Synthesis of
isopropoxy(phenyl)tosyloxymethylphosphine oxide (ꢀ)-(S)-8
Enantiomerically pure (+)-(S)-3c (0.2576 g, 1.204 mmol), TosCl
(0.235 g, 1.23 mmol) and Et3N (0.1246 g, 1.23 mmol) were dis-
solved in CH2Cl2 (10 mL) and the solution was stirred at rt. After
4 days (31P NMR control), the solvent was evaporated and the
crude mixture was separated by column chromatography using
4.3.1.3. Compound (+)-(S)-11.
Yield 25%, [a]D = +68.4 (c 2.25,
CHCl3), ee = 100% (HPLC: Chiralpak OD; Hexan: (i-PrOH/EtOH 4:1)
98:2; fl. 0.5 mL/min; tR = 8.59 min); 31P NMR (CDCl3): d 107.33 (q,
JP-B = 67.22 Hz); 1H NMR (CDCl3): d 0.80 (br q, JB-H = 94.67 Hz, 3H,
BH3), 1.12 (d, J = 6.14 Hz, 3H, (CH3)2CHO), 1.31 (d, J = 6.14 Hz, 3H,
(CH3)2CHO), 1.67 (d, J = 9.18 Hz, 3H, CH3), 4.41–4.48 (m, 1H,
(CH3)2CHO), 7.45–7.82 (m, 5H, Ph). MS (CI): m/z 195 (M+H).
CH2Cl2/MeOH 100:1 as eluent. Yield 0.4078 g (92%). [
a
]
D = ꢀ6.4 (c
1.46, CHCl3). Ee = 100% [HPLC Chiralpak OD; Hexane: (i-PrOH/EtOH
4:1) 98:2; fl. 0.5 mL/min; tR = 52.08 min]. 31P NMR (CDCl3): d 29.6.
4.3.1.4. Compound 5.
Compound 5 was isolated from the
1H NMR (CDCl3):
d 1.28 (1.24, d, J = 6.14 Hz, 3H), 1.38 (d,
reaction of (ꢀ)-(R)-3c with borane (Scheme 3). Yield: 15%.
[a] d 0.1 (q, JP-
D = +6.4 (c 1.01, CHCl3), 31P NMR (CDCl3):
J = 6.14 Hz, 3H), 2.43 (s, 3H), 4.10–4.39 (d AB, 2H), 4.59–4.76 (m,
1H), 7.24–7.82 (m, 9H). MS (CI): m/z 369 (M+H).
B = 41.53 Hz); 1H NMR (CDCl3): d 0.8 (br q, J = 92.27 Hz, 3H, BH3),
1.85 (br s, 1H), 4.27 (s, 2H), 5.57 (double m, JP-H = 376 Hz), 7.35–
7.84). MS (CI) m/z 155 (M+H).
4.2.2. Synthesis of iodomethyl(isopropoxy)phenylphosphine
oxide (+)-(S)-9
4.3.1.5. Compound 6.
(CDCl3): d 16.5 (q, JP-B = 48.62 Hz); 1H NMR (CDCl3): d 0.64 (br q,
Isolated as above. Yield 10%. 1P NMR
Compound (ꢀ)-(S)-8 (0.3852 g, 1.0467 mmol) and NaI
(0.6280 g, 4 equiv, 4.187 mmol) were dissolved in acetone

M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
1589
J = 94.97 Hz, 3H, BH3), 1.56 (br s, 1H), 2.19 (br s, 1H), 4.38 (s, 2H),
7.36–7.91 (m, 5H). MS (CI): m/z 171 (M+H).
4.4.4. Compound (ꢀ)-(R)-18
Yield 40%, [
a]
D = ꢀ25.2 (c 1, CHCl3), 100% ee 31P NMR (CDCl3): d
84.73. 1H NMR (CDCl3): d 1.08 (d, J = 6.18 Hz, 3H, (CH3)2CHO), 1.36
(d, J = 6.18 Hz, 3H, (CH3)2CHO), 1.98 (d, J = 13.63 Hz, 3H, CH3), 4.61–
4.79 (m, 1H, (CH3)2CHO), 7.41–8.01 (m, 5H, Ph). MS (CI): m/z 215
(M+H).
4.4. Transformation of phosphine boranes into phosphine
sulfides; general procedure
To a suspension of sulfur (67 mg, 2.094 mmol) in THF (3 mL)
under argon was added 1-octene (235 mg, 2.094 mmol) and a solu-
tion of 1a, 1c, 14, or 11 (0.524 mmol) in THF (3 mL). The mixture
was refluxed until the substrate was consumed (31P NMR control).
The solvent was removed under vacuum and the residue was puri-
fied by column chromatography in dichloromethane to give phos-
phine sulfides 12a, 12c, 15, or 18, respectively.
4.5. Oxidation of phosphine sulfides with iodoxybenzene—
general procedure
To a solution of 12, 15, or 18 (0.226 mmol) in chloroform (5 mL),
iodoxybenzene (53 mg, 1.2 equiv = 0.271 mmol) and a few milli-
grams of Montmorillonite K-10 as an activator were added. The
suspension was stirred at room temperature. The reaction was
monitored by TLC (CH2Cl2/MeOH 20:1) or 31P NMR. After filtration
of the solid material, the solvent was removed under vacuum and
the residue was purified by column chromatography using dichlo-
romethane/methanol in gradient as solvent. Analytical data for the
products were identical with those for the starting substrates. The
yields, specific rotations and ee values are given in Table 4.
4.4.1. Compound (+)-(R)-12a
Yield 32%, [a]D = +23.4 (c 1.17, CHCl3), ee = 100% (HPLC: Chir-
alpak OD; Hexane : (i-PrOH/EtOH 4:1) 98:2; fl. 0.5 mL/min;
tR = 31.7 min (major), tR = 35.3 min (minor). 31P NMR (CDCl3): d
91.56. 1H NMR (CDCl3):
d 1.96 (br s, 1H, OH), 3.55 (d,
J = 13.21 Hz, 3H, CH3), 3.93 (AB, 2H, CH2OH), 7.45–7.95 (m, 5H,
Ph); MS (CI): m/z 203 (M+H); HRMS: calcd for C8H12 O2PS:
203.029566, found: 203.02954.
4.6. Sulfuration of (+)-(R)-10 with Lawesson reagent
4.4.2. Compound (+)-(R)-12c
To a solution of (+)-(R)-10 {[a]D = +40 (c 1.06, CHCl3), ee = 100%
Yield 92%, [
a
]
D = +26.1 (c 1.57, CHCl3), ee = 100% (HPLC: Chir-
(51 mg, 0.260 mmol)} in benzene (5 mL), Lawesson’s reagent
(105 mg, 0.260 mmol) was added. The reaction mixture was re-
fluxed until the substrate was consumed (31P NMR control). The
solvent was removed under vacuum and the residue was purified
by column chromatography using dichloromethane as solvent to
alpak OD; Hexane: (i-PrOH/EtOH 4:1) 98:2; fl. 0.5 mL/min;
tR = 16 min). 31P NMR (CDCl3): d 86.68. 1H NMR (CDCl3): d = 1.20
(d, J = 6.2 Hz, 3H, CH3), 1.39 (d, J = 6.2 Hz, 3H, CH3), 2.45 (br s, 1H,
OH), 3.96 (AB, 2H, CH2OH), 4.74–4.86 (m, 1H, (CH3)2CHO), 7.35–
7.99 (m, 5H, Ph); MS (CI): m/z = 231 (M+H); HRMS: calcd for
C10H15O2PS: 230.053011, found: 230.05345.
give pure product (+)-(S)-18. Yield 58%, [a]D = +25.2 (c 0.81, CHCl3),
ee = 100% (HPLC: Chiralpak OD; Hexane: (i-PrOH/EtOH 4:1) 98:2;
fl. 0.5 mL/min; tR = 7.96 min). For analytical data see (ꢀ)-(R)-18.
4.4.3. Compound (ꢀ)-(R)-15
Yield 92%, [
a
]
D = ꢀ5.7 (c 1.02, CHCl3), ee = 45% (HPLC: Chiralpak
4.7. Transformation of hydroxymethylphosphine oxides into
hydroxymethylphosphine sulfides with Lawesson’s reagent
OD; Hexane: (i-PrOH/EtOH 4:1) 98:2; fl. 0.5 mL/min; tR = 24.4 min
for (ꢀ)-(R), tR = 28.4 min for (+)-(S); 31P NMR (CDCl3): d 64.63; 1H
NMR (CDCl3): d 1.20 (d, J = 16.47 Hz, 9H, t-Bu), 3.06 (br s, 1H,
OH), 4.27 (AB, 2H, CH2OH), 7.44–7.85 (m, 5H, Ph); MS (CI): m/z
229 (M+H); HRMS: calcd for C11H18PSO 229.081602 found
229.08156.
The hydroxymethylphosphine oxides 3a, 3c, and 7 were first
acetylated using acetyl chloride–triethylamine under standard
conditions to obtain 4a, 4c, and 16. The sulfuration of the latter
was performed as for (+)-(R)-10 (vide supra). The resulting acet-
oxymethylphosphine sulfides 13a, 13c, and 17, respectively, were
then treated with sodium methoxide in methanol to give 12a,
12c, and 15, respectively (see Schemes 2 and 3). The results are
shown in Table 5.
Table 4
Oxidation of phosphine sulfides with iodoxybenzene
4.8. Computational details
Entry
Substrate
Product
a
a
Symbol
[
a
]
D
ee (%) Symbol
Yield (%)
[
a
]
D
ee (%)
Full conformation analysis of both hydroxymethyl(i-pro-
poxy)phenylphosphine P-borane 1c and hydroxymethyl(i-pro-
poxy)-phenylphosphine oxide 3c was performed using the
Spartan ‘08 program and the PM3 semiempirical method.20 Sixteen
and ten conformers were found for 1c and 3c, respectively. All con-
formers were further optimized at the B3LYP/6-31+G⁄ level with
1
2
3
4
5
(+)-(S)-12a +18.0 77
(+)-(S)-12c +26.1 >99
(ꢀ)-(R)-12c ꢀ22.2 85
(ꢀ)-(R)-3a
(ꢀ)-(R)-3c
(+)-(S)-3c
(ꢀ)-(S)-7
(+)-(R)-10
32 ꢀ10.7 45
68 ꢀ13.5 66
50 +10.2 60
(ꢀ)-(R)-15
(ꢀ)-(R)-18
ꢀ5.7
45
30 ꢀ1.9
17
ꢀ25.2 100
30 ꢀ29.8 70
a
In chloroform, c 1.
Table 5
Reactions of hydroxymethylphosphine oxides with Lawesson’s reagent
Substrate, phosphine oxide
Product, phosphine sulfide
CH2OH
CH2OAc
CH2OAc
CH2OH
a
a
a
a
Symbol
[a
]
D
(ee [%])
Symbol
[a
]
D
(ee [%])
Symbol
Yield (%)
[a
]
D
(ee [%])
Symbol
Yield (%)
[
a
]
(ee [%])
D
(ꢀ)-(R)-3a
(ꢀ)-(R)-3c
(ꢀ)-(S)-7
ꢀ21.0 (85)
ꢀ17.5 (79)
ꢀ4.8 (43)
(ꢀ)-(R)-4a
(ꢀ)-(R)-4c
(+)-(S)-16
ꢀ45.2 (85)
ꢀ30.5 (79)
+13.2
(ꢀ)-(S)-13a
(ꢀ)-(S)-13c
(+)-(S)-17
83
70
75
ꢀ31.4 (85)
ꢀ18.2 (79)
+27.8
(ꢀ)-(S)-12a
(ꢀ)-(S)-12c
(+)-(S)-15
100
100
93
ꢀ18.0 (77)b
ꢀ9.3 (80)
+5.2 (45)
a
In chloroform, c 1.
b
Partial racemization due to the MeO–MeO exchange at the phosphorus atom upon treatment with MeONa/MeOH.

1590
M. Kwiatkowska et al. / Tetrahedron: Asymmetry 22 (2011) 1581–1590
tight convergence criteria using the GAUSSIAN 03 program,21 yielding
only seven and eight different structures for 1c and 3c, respec-
tively, since some individual conformations found at the PM3 level
converged to the same conformation at the DFT level. The station-
ary point nature of all of the obtained structures was verified by
the vibrational analysis. Vibrational analysis also provided thermal
corrections to calculated Gibbs free energies of conformers at
298 K. SCRF calculations of conformer distribution and optical rota-
tion have been performed using the CPCM method22 with UAKS
atomic radii assuming CHCl3 as solvent. The specific rotation was
calculated using the B3LYP/6-311++G(2d,2p) method for D line of
Na (k = 589 nm) as the population-weighted average of the rota-
tions of individual conformations according to Boltzmann distribu-
tion at the B3LYP/6-311++G(2d,2p)//B3LYP/6-31+G⁄ level.
8. Kiełbasin´ ski, P.; Albrycht, M.; Łuczak, J.; Mikołajczyk, M. Tetrahedron:
Asymmetry 2002, 13, 735–738.
9. Keglevich, G.; Chuluunbaatar, T.; Ludanyi, K.; Toke, L. Tetrahedron 2000, 56, 1–
}
6.
}
10. Keglevich, G.; Fekete, M.; Chuluunbaatar, T.; Dobo, A.; Harmal, V.; Toke, L. J.
Chem. Soc., Perkin Trans. 1 2000, 4451–4455.
´
´
11. Krasinski, G.; Cypryk, M.; Kwiatkowska, M.; Mikołajczyk, M.; Kiełbasinski, P.,
submitted for publication.
12. Shioji, K.; Ueno, Y.; Karauchi, Y.; Okuma, K. Tetrahedron Lett. 2001, 42, 6569–
6572.
13. Stephens, P. J.; Devlin, F. J.; Cheeseman, J. R.; Frisch, M. J.; Bortolini, O.; Besse, P.
Chirality 2003, 15, S57–S64.
14. Uziel, J.; Darcel, C.; Moulin, D.; Bauduin, C.; Juge, S. Tetrahedron: Asymmetry
2001, 12, 1441–1449.
15. Mikołajczyk, M.; Łuczak, J.; Kiełbasin´ ski, P.; Colonna, S. Tetrahedron: Asymmetry
2009, 20, 1948–1951.
16. Interestingly, the iodoxybenzene oxidation of O,S-dimethyl-O-p-nitrophenyl
phosphorodithioate15 and cyclic phosphine sulfides17 proceeds with full
retention of configuration at phosphorus.
17. Mielniczak, G.; Łopusin´ ski, A. Synlett 2001, 505–508.
18. It should be noted, that the absolute configuration of 14 was reported5 to be
(+)-(R). This is, however, an obvious error, which can be concluded from the
accompanying discussion in this reference. Thus, the absolute configuration of
14 is (+)-(S).
19. We thank a Referee for this suggestion
20. SPARTAN 08; Wavefunction Inc.: Irvine, CA.
References
_
1. Kiełbasin´ ski, P.; Zurawin´ ski, R.; Bałczewski, P.; Mikołajczyk, M.
Alkylphosphorus Compounds In Comprehensive Organic Functional Group
Transformations, 2 (COFGT II); Katritzky, A. R., Taylor, R. J. K., Eds.; Elsevier,
2005; Vol. 2, pp 383–461.
21. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken,
V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A.
J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G.
A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;.
Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe,
M.;. Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J.A.
GAUSSIAN 03, Revision D. 01; Gaussian, Inc.: Wallingford CT, 2004.
22. Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003, 24, 669–681.
2. E.g.: (a) Imamoto, T. Pure Appl. Chem. 2001, 73, 373–376; (b) Stankevic, M.;
Pietrusiewicz, K. M. J. Org. Chem. 2007, 72, 816–822; (c) Moraleda, D.; Gatineau,
D.; Martin, D.; Giordano, L.; Buono, G. Chem. Commun. 2008, 3031–3033; (d)
Julienne, D.; Delacroix, O.; Gaumont, A.-C. Compt. Rend. Chim. 2010, 13, 1099–
1103; (e) Gatineau, D.; Giordano, L.; Buono, G. J. Am. Chem. Soc. 2011, 133,
10728–10731.
3. Kiełbasin´ ski, P.; Mikołajczyk, M. Heteroatom-Containing Compounds. In Future
Directions in Biocatalysis; Matsuda, T., Ed.; Elsevier, 2007; pp 159–203.
_
4. Kiełbasin´ ski, P.; Albrycht, M.; Zurawin´ ski, R.; Mikołajczyk, M. J. Mol. Catal. B:
Enzym. 2006, 39, 45–49.
5. Shioji, K.; Karauchi, Y.; Okuma, K. Bull. Chem. Soc. Jpn. 2003, 76, 833–843.
6. Wiktelius, D.; Johansson, M. J.; Luthmann, K.; Kann, N. Org. Lett. 2005, 7, 4991–
4994.
7. Kiełbasin´ ski, P.; Omelan´ czuk, J.; Mikołajczyk, M. Tetrahedron: Asymmetry 1998,
9, 3283–3287.