Effects of phosphorus substituents on reactions of α- alkoxyphosphonium salts with nucleophiles

Article abstract of DOI:10.1002/chem.201200480

The effects of phosphorus substituents on the reactivity of α-alkoxyphosphonium salts with nucleophiles has been explored. Reactions of α-alkoxyphosphonium salts, prepared from various acetals and tris(o-tolyl)phosphine, with a variety of nucleophiles proceeded efficiently. These processes represent the first examples of high-yielding nucleophilic substitution reactions of α-alkoxyphosphonium salts. The reactivity of these salts was determined by a balance between steric and electronic factors, respectively, represented by cone angles θ and CO stretching frequencies ν (steric and electronic parameters, respectively). In addition, a novel reaction of α-alkoxyphosphonium salts derived from Ph₃P with Grignard reagents was observed to take place in the presence of O₂ to afford alcohols in good yields. A radical mechanism is proposed for this process that has gained support from isotope-labeling and radical-inhibition experiments. A dramatic change in the reactivity of an α-alkoxyphosphonium salt toward nucleophiles is observed due to the steric and electronic nature of the phosphine substituents. By changing the type of phosphorus substituents, the reaction pathway can be controlled to proceed selectively by substitution or a new radical reaction (see scheme; OTf=trifluoromethansulfonate, TMS=trimethylsilyl, o-tol=tolyl). Copyright

Full text of DOI:10.1002/chem.201200480

FULL PAPER
DOI: 10.1002/chem.201200480
Effects of Phosphorus Substituents on Reactions of a-Alkoxyphosphonium
Salts with Nucleophiles
Akihiro Goto, Kazuki Otake, Ozora Kubo, Yoshinari Sawama, Tomohiro Maegawa, and
Hiromichi Fujioka*[a]
Abstract: The effects of phosphorus
substituents on the reactivity of a-al-
koxyphosphonium salts with nucleo-
philes has been explored. Reactions of
a-alkoxyphosphonium salts, prepared
salts. The reactivity of these salts was
determined by balance between
steric and electronic factors, respective-
ly, represented by cone angles q and
CO stretching frequencies n (steric and
electronic parameters, respectively). In
addition, a novel reaction of a-alkoxy-
phosphonium salts derived from Ph3P
with Grignard reagents was observed
to take place in the presence of O2 to
afford alcohols in good yields. A radi-
cal mechanism is proposed for this pro-
cess that has gained support from iso-
tope-labeling and radical-inhibition ex-
periments.
a
from various acetals and trisACTHUNRTGNE(GNU o-tolyl)-
phosphine, with a variety of nucleo-
philes proceeded efficiently. These pro-
cesses represent the first examples of
high-yielding nucleophilic substitution
reactions of a-alkoxyphosphonium
Keywords: phosphines · radicals ·
nucleophilic substitution
· super-
oxides · a-alkoxyphosphonium salts
Introduction
followed by H2O. In this process, a collidinium salt inter-
mediate is generated [Scheme 1, Eq. (1)], which reacts with
a variety of nucleophiles, such as H2O, alcohols, and the
azide ion.[4] In contrast, related salts, derived by the reac-
a-Alkoxyphosphonium salts serve as useful reactants in
Wittig reactions to generate vinyl ethers. This process has
been widely applied in natural-products total syntheses.[1,2]
In contrast to their utility in olefination reactions, a-alkoxy-
phosphonium salts have only rarely been explored as sub-
strates for other interesting reactions.[1c,d] In one earlier
study, Anders and co-workers probed nucleophilic substitu-
tion reactions of a-siloxyphosphonium salts[1d] prepared by
the reactions of Ph3P with various aldehydes in the presence
of TMSOTf or TMSI. The treatment of these salts with
carbon nucleophiles, such as Et4N+CNÀ and MeMgI, was
observed to take place to produce the corresponding substi-
tution products in low-to-moderate yields (23–48%). Al-
though interesting, the nucleophilic substitution reactions of
a-alkoxyphosphonium salts studied by Anders and co-work-
ers have not been investigated fully. One possible reason for
this lies in the fact that the salts derived from Ph3P have low
reactivity with nucleophiles because Ph3P does not serve as
a good leaving group.
Scheme 1. Reactivity of pyridinium cationic intermediates. TESOTf=
triethylsilyl trifluoromethanesulfonate.
tions of pyridine and 2-picoline with acetals, were observed
to be unreactive with H2O [Scheme 1, Eq. (2)]. We hypothe-
sized that the observed reactivity differences are a conse-
quence of steric factors that govern the stability/reactivity of
the salts.
In a previous study, we developed a novel method for
acetal deprotection,[3] which involves the treatment of these
substrates with TESOTf and 2,4,6-collidine or 2,6-lutidine
These earlier findings stimulated our interest in exploring
the reactivity of cationic intermediates, such as those that
arise from reactions of acetals with phosphines. In a recent
effort aimed at elucidating the reactivity profiles of a-alk-
oxyphosphonium salts, we observed that salts derived from
[a] A. Goto, K. Otake, O. Kubo, Dr. Y. Sawama, Dr. T. Maegawa,
Prof. Dr. H. Fujioka
Graduate School of Pharmaceutical Sciences
Osaka University, 1–6 Yamada-oka
Suita, Osaka, 565-0871 (Japan)
Fax : (+81)6-6879-8229
E-mail: fujioka@phs.osaka-u.ac.jp
acetals and trisACHTUNTRGNEUNG(o-tolyl)phosphine ((o-tol)3P) are highly reac-
tive substrates in nucleophilic substitution reactions
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201200480.
[Scheme 2, Eq. (1)].[5] In addition, we found that an unusual
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

tion, we observed that the salts derived from (m-tol)3P and
(p-tol)3P do not undergo hydrolysis to form 3a (Table 1, en-
tries 4 and 5). We also found that the a-alkoxyphosphonium
salt prepared from the electron-deficient phosphine (o-
FC6H4)3P and 1a reacts efficiently with H2O to afford 3a
and that the corresponding salt derived from the electron-
rich phosphine (o-MeOC6H4)3P is unreactive with H2O
(Table 1, entries 6 and 7). Finally, the sterically more bulky
phosphine (2,4,6-Me3C6H2)3P does not participate in salt for-
mation with 1a (Table 1, entry 8) and the salt derived from
nBu3P is unreactive with H2O (Table 1, entry 9).
Scheme 2. Substitution reaction of a-alkoxyphosphonium salts. Nu=nu-
cleophile.
reaction of a-alkoxyphosphonium salts derived from Ph3P
with Grignard reagents takes place efficiently to afford alco-
Detection of the phosphonium salt intermediate: The results
summarized above suggest that the reactivity of a-alkoxy-
phosphonium salts with water depends critically on both the
steric bulk and electronic properties of the phosphorus sub-
stituents. To verify that the reaction described above does
indeed proceed through the sequence shown in Table 1, we
attempted to isolate the proposed a-alkoxyphosphonium
salt intermediates 2. Although many reports exist that de-
scribe the generation and isolation of a-alkoxyphosphonium
salts derived from Ph3P, [1] those salts that arise from the
other phosphines such as (o-tol)3P have not been character-
ized to date. The results of an experiment aimed at isolating
the a-alkoxyphosphonium salt 2a, prepared from (o-tol)3P,
showed that this substance is unstable and gradually under-
goes decomposition during NMR spectroscopic measure-
ments made at room temperature. In contrast, the a-alkoxy-
phosphonium salt 2b, formed by reaction of acetal 1a and
(p-tol)3P, is readily isolated and stable (Figure 1). The
1H NMR spectrum of 2b contained characteristic resonances
that correspond to an a-alkoxyphosphonium salt (d=
5.55 ppm)[2d] and signals in the 13C, 19F, and 31P NMR spectra
of this salt were consistent with its structural assignment.[8]
hols through
a novel radical mechanism [Scheme 2,
Eq. (2)].[6] The results of this investigation along with a dis-
cussion of how steric and electronic effects contribute to
governing the reactivity of a-alkoxyphosphonium salts are
presented.
Results and Discussion
Effect of the phosphine component of a-alkoxyphosphoni-
um salts: In the initial phase of the current investigation
into reactivity issues, we employed a-alkoxyphosphonium
salts 2, obtained by reactions of n-dodecanal dimethyl acetal
(1a) with different types of phosphines (Table 1). The reac-
tions of these salts with water were observed to generate al-
dehydes 3a.
Table 1. Effect of the phosphane component on the reactions of a-alk-
oxyphosphonium salts with water.[a]
R
Time A [h]
Time B [h]
Yield [%]
1
2
Ph
o-tol
o-tol
m-tol
0.5
0.5
0.5
0.5
0.5
0.5
0.5
5
96
1
1
47
78
95
0
3[b]
4
Figure 1. Structures of a-alkoxyphosphonium salts.
24
24
24
24
5
6
7
8
p-tol
0
o-FC6H4
o-MeOC6H4
2,4,6-Me3C6H2
nBu
74
0
Correlation of the reactivity of the phosphonium salts and
the steric parameter q of the phosphine group: One of the
major driving forces for nucleophilic substitution reactions
of a-alkoxyphosphonium salts is relief of the steric hin-
drance offered by groups surrounding the phosphorus atom.
Owing to this feature, our at-
[c]
[c]
–
–
9
1.5
24
0
[a] Reaction procedure: PR3 (3.0 equiv) and TESOTf (2.0 equiv) were
added to acetal 1a (1.0 equiv) in CH2Cl2 (0.1m) at 08C. The resulting
mixtures were stirred for time A and then H2O was added with further
stirring at room temperature for time B. [b] Temperature=À58C. [c] No
a-alkoxyphosphonium salt was formed. tol=tolyl.
tention focused on a steric pa-
rameter that corresponds to the
phosphine moiety. The cone
In contrast to the a-alkoxyphosphonium salt formed from
1a and Ph3P, which undergoes slow (4 days) hydrolysis to
give aldehyde 3a in only moderate yield (Table 1, entry 1),
the analogue derived from (o-tol)3P is highly reactive with
H2O, rapidly (1 h) forming 3a at room temperature (78%),
and even at À58C (95%) (Table 1, entries 2 and 3). In addi-
angle q (Figure 2), introduced
by Tolman in the 1970s,[7] is
a
well-known parameter fre-
quently used to evaluate the
steric bulk of phosphines that
participate in transition-metal- metal center.
Figure 2. Cone angle q. M=
&
2
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Reactivity of a-Alkoxyphosphonium Salts with Nucleophiles
FULL PAPER
catalyzed reactions. Because q value is an indicator of steric
hindrance (as q increases the steric bulk of the phosphine
increases), we have used it as a parameter to correlate the
hydrolytic reactivities of a-alkoxyphosphonium salts.
As the data displayed in Table 2 demonstrate, the salt
formed from the reaction of 1a with the less-bulky phos-
phine nBu3P (q=1328) does not participate in a hydrolysis
reaction (Table 2, entry 1). Although the salt derived from
Figure 3. Electronic parameter n.
Studies were carried out to determine if the hydrolytic re-
activities of a-alkoxyphosphonium salts correlate with phos-
phine electronic parameters n (Table 3). The results of inves-
tigations with salts that contain phosphines with identical
Table 2. Correlation between reactivity and phosphane steric effects.[a]
Table 3. Correlation between reactivity and phosphane electronic
effects.[a]
PR3
q [8]
Result
Time [h]
Yield [%]
1[b]
2[b]
3
4
5
nBu3P
Ph3P
ACHTUNGTRENNUNG
132
145
161
178
194
212
205
24
96
24
4
0
47
79
87
95
ACHTUNGTRENNUNG
R
1
[c]
[c]
PR3
n [cmÀ1
]
q [8]
Result
6
(2,4,6-Me3C6H2)3P
(o-MeOC6H4)3P
–
–
7[b]
24
0[d]
Time [h]
Yield [%]
1[b]
2[b]
3
C
2066.7
2068.9
2071.3
2074.1
2090.0
2058.3
2066.6
145
145
145
145
184
205
194
24
96
24
1
–
24
1
0
47
71
93
–
0
95
[a] Reaction procedure: PR3 (3.0 equiv) and TESOTf (2.0 equiv) were
added to acetal 1a (1.0 equiv) in CH2Cl2 (0.1m) at À58C. The resulting
mixtures were stirred for 0.5 h and then H2O was added with further stir-
ring at room temperature. [b] Temperature=08C. [c] No a-alkoxyphos-
phonium salt was formed. [d] a-Alkoxyphosphonium salt was formed.
4
5
[c]
[c]
6[b]
7
[a] Reaction procedure: PR3 (3.0 equiv) and TESOTf (2.0 equiv) were
added to acetal 1a (1.0 equiv) in CH2Cl2 (0.1m) at À58C. The resulting
mixtures were stirred for 0.5 h and then H2O was added with further stir-
ring at room temperature. [b] Temperature= 08C. [c] a-Alkoxyphospho-
nium salt was quickly decomposed.
Ph3P (q=1458) is only moderately reactive (47% after 96 h;
Table 2, entry 2), those salts that arise from (o-tol)Ph2P (q=
1618) and (o-tol)2PhP (q=1788) undergo more rapid hydrol-
ysis reactions (Table 2, entries 3 and 4), and the salt contain-
ing the (o-tol)3P moiety (q=1948) is the most reactive in the
series studied (Table 2, entry 5). Interestingly, the data clear-
ly show that as the cone angle of the phosphine moiety in-
creases, the rates and efficiencies of the hydrolysis reactions
of the corresponding a-alkoxyphosphonium salt increase.
cone angles show that the electronic nature of the phosphine
moieties greatly affect the reactivity of a-alkoxyphosphoni-
um salts (Table 3, entries 1–4). For example, the salt derived
from (p-tol)3P (n=2066.7 cmÀ1), which contains a more elec-
tron-rich phosphine than Ph3P (n=2068.9 cmÀ1), displays
low reactivity toward H2O, whereas the salt that arises from
(p-FC6H4)3P (n=2071.3 cmÀ1) is highly reactive with water.
The salt derived from (m-FC6H4)3P (n=2074.1 cmÀ1), which
is more electron poor than (p-FC6H4)3P, displays excellent
hydrolytic reactivity (93%). In addition, we observed that
the a-alkoxyphosphonium salt containing the highly elec-
tron-withdrawing (C6F5)3P moiety (n=2090.0 cmÀ1) is so un-
stable that it undergoes decomposition during its formation
(Table 3, entry 5) and that the salt from the most electron-
rich phosphine (o-MeOC6H4)3P (n=2058.3 cmÀ1) does not
react with H2O (Table 3, entry 6). Although the bulky phos-
phines (o-tol)3P (n=2066.6 cmÀ1) and (p-tol)3P have compa-
rable electronic parameters, the reactivity of their corre-
sponding salts are different as a consequence of the strong
steric effect displayed by the former phosphine (Table 3,
entry 7). The data summarized above show that the relative
reactivities of a-alkoxyphosphonium salts toward nucleo-
philes can be readily predicted by considering the balance
Correlation of the reactivity of the phosphonium salts and
the electronic parameter n of the phosphine groups: Al-
though (o-MeOC6H4)3P has a cone angle of 2058, the salt
formed from this phosphine does not undergo a hydrolysis
reaction (Table 2, entry 7). This observation suggests that
the nucleophilic substitution reactivities of a-alkoxyphos-
phonium salts are not only governed by steric factors. In ad-
dition, the a-alkoxyphosphonium salt derived from (o-
FC6H4)3P, which contains electron-withdrawing aryl substitu-
ents, displays a high reactivity toward H2O (74%, 24 h;
Table 1, entry 6). These results support the hypothesis that
the reactivities of a-alkoxyphosphonium salts should also be
governed by electronic factors. The parameter n, introduced
by Tolman,[7] corresponds to the electronic effects of phos-
phines on the CO stretching frequencies in nickel tricarbon-
yl phosphine complexes Ni(CO)3–PR3 (Figure 3). A larger
value of n indicates that the phosphine at the nickel center
is more electron withdrawing.
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

H. Fujioka et al.
Table 4. Reactivity of various a-alkoxyphosphonium salts with H2O.[a]
between the steric and electronic effects exerted by
the phosphine moieties.
Substitution reactions of a-alkoxyphosphonium
salts: From the investigations summarized above,
we found that two a-alkoxyphosphonium salts, de-
rived by reactions of acetal 1a with (m-FC6H4)3P
and (o-tol)3P, participate in efficient nucleophilic
substitution reactions with water. In studies aimed
at exploring the substrate and nucleophilic scope of
substitution reactions of a-alkoxyphosphonium
salts, (o-tol)3P was selected as the phosphine com-
ponent owing to its low cost, commercial availabili-
ty, and ease of handling.[5] In the initial phase of the
investigation, we also probed the use of other inex-
pensive Lewis acids (i.e., BF3·OEt2, AlCl3, and
TiCl4) in place of TESOTf in the salt-forming reac-
tion. However, we observed that these Lewis acids
are ineffective at promoting this process.[8] There-
fore, the combination of TESOTf and (o-tol)3P was
used to induce reactions to form a-alkoxyphospho-
nium salts from different acetals.
Various a-alkoxyphosphonium salts 4 (Table 4),
prepared by reactions of O,O-acetals 1 with (o-
tol)3P, were subjected to hydrolysis reactions. The
results demonstrate that hydrolysis reactions of the
salts derived from aliphatic, aromatic, and unsatu-
rated acetals take place to form aldehydes 3 in
good yields (Table 4, entries 1–4). In addition, the
reactions of a phosphonium salt that contains an
acid-labile functional group (34%; Table 4, entry 5)
and a salt derived from the cyclic acetal 1 f (42%;
Table 4, entry 6) form aldehydes in moderate yields.
Since the pKa values of phosphine products formed
in this process are low, the reactions produce acids,
which might account for the lower than expected
yields of the aldehydes. Consistent with this propos-
al is the observation that the addition of the bulky
base 2,6-di-tert-butylpyridine to the reaction mix-
tures leads to improved yields (Table 4, entries 5
and 6). In the case of the bulky acetal 1g, reaction
with (o-tol)3P in the presence of TESOTf smoothly
affords a salt that reacts efficiently with H2O
(Table 4, entry 7). Finally, an acetal-selective hy-
drolysis reaction was accomplished by using a salt
derived from the mixed ketal/acetal 1h (Table 4,
entry 8).
Substrate
Product
Time [h]
1
Yield [%]
95
1
2
2
72
89
87
3
0.5
4
1
97
5[b]
1.5
(34)[d]
95
6[b]
1.5
0.5
(42)[d]
7
72
79
8[b,c]
1.5
[a] Reaction procedure: (o-tol)3P (3.0 equiv) and TESOTf (2.0 equiv) were added to
acetal 1 (1.0 equiv) in CH2Cl2 (0.1m) at À58C. The resulting mixtures were stirred for
0.5 h and then H2O was added with further stirring at room temperature. [b] 2,6-Di-
tert-butylpyridine (3 equiv) followed by TESOTf. [c] Reaction was performed at a con-
centration of 0.2m. [d] The reaction was conducted in the absence of 2,6-di-tert-butyl-
pyridine.
Table 5. Reactivity of various a-alkoxyphosphonium salts with PhSLi.[a]
Substrate
Product
Time [h]
4
Yield [%]
97
1
2
2
93
86
3[b]
0.5
4[b]
5[c]
0.5
3
74
72
In studies that probed reactions with other nu-
cleophiles, a-alkoxyphosphonium salts 4, prepared
from aliphatic O,O-acetals 1a and 1b and (o-tol)3P,
were treated with PhSLi. In both cases, an efficient
reaction occurs to form the respective mixed acetals
5 in excellent yields (Table 5, entries 1 and 2). Re-
actions of aromatic 1c and a,b-unsaturated 1d
acetal-derived phosphonium salts with PhSLi also
[a] Reaction procedure: (o-tol)3P (3.0 equiv) and TESOTf (2.0 equiv) were added to
acetal 1 (1.0 equiv) in CH2Cl2 (0.1m) at À58C. The resulting mixtures were stirred for
0.5 h and then PhSLi (1.2 equiv) was added with further stirring at room temperature.
[b] iPr2NEt (3 equiv) followed by PhSLi. [c] PhSLi (2 equiv) was used.
yielded dithioacetal products. Interestingly, the incorpora-
blocking overreaction, thus enabling production of the de-
tion of iPr2NEt into these reaction mixtures was effective in
sired O,S-acetals 5c and 5d in high yields (Table 5, entries 3
&
4
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Reactivity of a-Alkoxyphosphonium Salts with Nucleophiles
Table 6. Reactions of a-alkoxyphosphonium salt with C-nucleophiles.[a]
FULL PAPER
acetal 10 by reaction of n-decanal with CH318OH
and transformed it into the corresponding Ph3P-
containing salt. Analysis showed that the alcohol
product from the reaction of this salt with PhMgBr
shows that it does not contain an 18O label
(Scheme 4).[8]
Product (time, yield)
TMSCN
PhMgBr
MeMgBr
Owing to the unprecedented nature of the alco-
hol-forming reaction, we carried out mechanistic
studies that used the a-methoxyphosphonium salt
12a, which was prepared from benzaldehyde di-
methyl acetal (1i)[9] and Ph3P. We observed that
treatment of 12a with five equivalents of PhMgBr
results in the formation of major products, the
nature and yield of which vary with the reaction
temperature. Accordingly, the benzhydrol methyl
ether 7 f is generated as the major product when
the reaction is conducted at room temperature,
whereas benzhydrol 9b predominates in the reac-
tion at 08C (Scheme 5).
Moreover, 9b was obtained in yields that vary
between 30 and 65% when multiple reactions were
performed under identical conditions at 08C. Be-
cause phenol and triphenylphosphine oxide (Ph3P=
O) are produced as byproducts in these processes,
we assumed that the N2 gas employed contained O2
as a contaminant[10] and that the contaminant was
responsible for the alcohol-forming reaction and
the formation of PhOH and Ph3P=O through the
respective oxidations of PhMgBr[11] and Ph3P. Fur-
thermore, we felt that the erratic yields of the pre-
(4 h, 95%)
(2 h, 97%)
(1 h, 90%)
(6 h, 99%)
(0.5 h, 95%)
(1 h, 87%)
(5 h, 95%)
(5 h, 90%)
(1 h, 79%)[b]
(1 h, 70%)
(1.5 h, 52%)[c]
(0.5 h, 79%)
(5 h, 83%)
(1 h, 79%)[b,d]
(1 h, 52%)[b]
[a] Reaction procedure: (o-tol)3P (3.0 equiv) and TESOTf (2.0 equiv) were added to
acetal 1 (1.0 equiv) in CH2Cl2 (0.1m) at À58C. The resulting mixtures were stirred for
0.5 h and then TMSCN (3 equiv) or PhMgBr (1.2 equiv) or MeMgBr (1.2 equiv) were
added with stirring at room temperature. [b] 2,6-Di-tert-butylpyridine (3 equiv) was
added followed by TESOTf. [c] PhMgBr (2 equiv) was used. [d] Reaction was per-
formed at a concentration of 0.2m. TMSCN=trimethylsilyl cyanide.
and 4). A salt prepared from the bulky acetal 1g also reacts
with PhSLi in a good yield (Table 5, entry 5).
Reactions of a-alkoxyphosphonium salts 4 with a variety
of C-nucleophiles were also examined (Table 6). The results
Scheme 3. Reaction of the a-alkoxyphosphonium salt derived from 1a
and Ph3P with PhMgBr.
show that TMSCN reacts with these salts to afford the cor-
responding nitriles 6a–e in excellent yields. Moreover, these
salts react with MeMgBr and PhMgBr to generate the corre-
sponding adducts 7a–e and 8a–e in good yields. The acetal-
selective transformation of mixed ketal/acetal compounds
1h to 7e in good yield is noteworthy.
An unusual reaction of a-alkoxyphosphonium salts with
Grignard reagents and the reaction mechanism: During
studies that focused on the reactivity of a-alkoxyphosphoni-
um salts, we uncovered an unusual reaction that occurs
when Grignard reagents are used as nucleophiles. Specifical-
ly, we observed that the hydrolytically less-reactive a-meth-
oxyphosphonium salt 2c, prepared from Ph3P and acetal 1a,
participates in a reaction with excess PhMgBr to afford the
unexpected alcohol 9a (45%) as a major product along with
a lesser amount of the expected substitution product 7a
(20%) (Scheme 3). We initially believed that formation of
9a was a consequence of a secondary reaction of 7a. To test
this hypothesis, we synthesized the 18O-labeled dimethyl
Scheme 4. An isotopic labeling experiment that used the 18O-labeled di-
methyl acetal 10.
Scheme 5. Temperature-dependent reaction of a-methoxyphosphonium
salt 12a with PhMgBr.
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

H. Fujioka et al.
Table 7. Reaction of 12a with PhMgBr under various conditions.[a]
Table 8. Reactions of 12a with a variety of Grignard reagents.[a]
PhMgBr [equiv]
Atmosphere
T [8C]
Yield [%]
R
Product
Yield [%]
1
2
3
4
5
6
5
5
5
5
1
3
O2
dry air
Ar
dry air
dry air
dry air
0
0
0
40
81
1
2
3
4
5
6
Ph
p-tol
p-MeOC6H4
p-FC6H4
isopropenyl
Et
9b
9c
9d
9e
9 f
9g
81
86
62
57
trace
ND
ND[b]
57
À40
0
trace
49
0
[a] Reaction procedure: Ph3P (3.0 equiv) and TESOTf (2.0 equiv) to
acetal 1i (1.0 equiv) in CH2Cl2 (0.2m) under O2/Ar or dry air at 08C. The
resulting mixtures were stirred for 0.5 h and then PhMgBr (5.0 equiv)
was added quickly with further stirring at 08C for 2–4 h. [b] Not detect-
ed.
[a] Reaction procedure: Ph3P (3.0 equiv) and TESOTf (2.0 equiv) were
added to acetal 1i (1.0 equiv) in CH2Cl2 (0.2m) under dry air at 08C. The
resulting mixtures were stirred for 0.5 h and then RMgBr (5.0 equiv) was
added quickly with stirring at 08C for 2–4 h.
Table 9. Reactions of various a-alkoxyphosphonium salts with PhMgBr
under an O2 atmosphere.[a]
sumably identical reactions that generated 9b were a conse-
quence of different levels of contaminating O2 in the N2. On
the basis of this hypothesis, we carried out reactions of a-
methoxyphosphonium salt 12a with PhMgBr at 08C in an
O2 or dry-air atmosphere (Table 7, entries 1 and 2, respec-
tively). Under these conditions, 9b was consistently formed
in the same yield. In addition, 9b was not generated when
an argon atmosphere was employed (Table 7, entry 3).
Therefore, O2 is essential for the production of the unex-
pected product. Also, although requiring a longer time, reac-
tion at À408C occurs to form 9b in the same yield as at 08C
(Table 7, entry 4). The use of one and three equivalents of
PhMgBr led to decreased yields of 9b (Table 7, entries 5
and 6). Finally, based on the assumption that O2 promotes
the formation of 9b, the reaction was conducted using 18O2
gas (Scheme 6). Analysis of the alcohol product 9b showed
that it contained an 18O-labeled hydroxy group.
Substrate
Product
Temp. [8C]
Yield [%]
1
2
R=Me (1j)
R=OMe (1c)
R=Br (1k)
R=CF3 (1l)
9c
9d
9h
9i
0
À40
0
À40
81
83
78
63
3
4[b,c]
5
9j
À40
66
6
9b
9b
0
83
60
7[c]
À40
8
9
9a
9k
0
0
56
55
10
9l
0
70
Scheme 6. Isotopic labeling experiment with 18O2 gas.
11[c]
12
9m
9n
À40
56
63
Reaction of a-alkoxyphosphonium salts with Grignard re-
agents under dry air: We next examined the reaction of a-
alkoxyphosphonium salt 12a with various Grignard reagents
under optimal conditions that employed a dry-air atmos-
phere (Table 8). A variety of aromatic Grignard reagents
possessing electron-donating or -withdrawing groups were
observed to react with 12a at 08C to generate the corre-
sponding benzhydrol derivatives 9b–e (Table 8, entries 1–4).
In contrast, alkenyl and alkyl Grignard reagents were not ef-
fective substrates for this reaction (Table 8, entries 5 and 6).
The results of additional studies demonstrated that several
aromatic and heteroaromatic a-alkoxyphosphonium salts 12,
generated from the corresponding O,O-acetals 1c,j–m, also
participate in the reaction to form diarylmethanols 9c,d,h–j
0
[a] Reaction procedure: Ph3P (3.0 equiv) and TESOTf (2.0 equiv) were
added to acetal 1 (1.0 equiv) in CH2Cl2 (0.2m) in dry air at 08C. The re-
sulting mixtures were stirred for 0.5 h and then PhMgBr (5.0 equiv) was
added quickly with stirring at 08C for 2–4 h. [b] The reaction of 1 with
Ph3P and TESOTf was performed at room temperature. [c] TMSOTf was
used instead of TESOTf. TMSOTf=trimethylsilyl trifluoromethanesulfo-
nate.
in good yields (Table 9, entries 1–5). The cyclic acetal 1n
and diisopropyl acetal 1o are also transformed to benz-
hydrols 9b in good yields (Table 9, entries 6 and 7). In addi-
tion, aliphatic a-alkoxyphosphonium salts derived from
&
6
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Reactivity of a-Alkoxyphosphonium Salts with Nucleophiles
FULL PAPER
Table 10. Correlation between the reactivity and structural effect of
phosphane.[a]
O,O-acetals 1a,b,p–r also react with PhMgBr to give the cor-
responding alcohols 9a,k–n in moderate-to-good yields
(Table 9, entries 8–12). In some cases, the alcohol-forming
reaction proceeds at À408C, whereas the nucleophilic substi-
tution process takes place competitively at 08C.
On the basis of the observations summarized above, the
novel radical pathway depicted in Scheme 7 is proposed for
the formation of the alcohol product. In this route, superox-
PR3
q [8]
n [cmÀ1
]
Yield [%]
9a
7a
1
Ph3P
ACHTUNGTRENNUNG
145
194
145
145
2068.9
2066.6
2066.1
2074.1
56
18
12
57
0
73
0
2[b]
3
4[b]
G
29
[a] Reaction procedure: PR3 (3.0 equiv) and TESOTf (2.0 equiv) were
added to acetal 1a (1.0 equiv) in CH2Cl2 (0.2m) under dry air at 08C.
The resulting mixtures were stirred for 0.5 h and then PhMgBr
(5.0 equiv) was added quickly with stirring at 08C for 2 h. [b] Tempera-
ture=À58C.
(q=1458) with PhMgBr under dry air was observed to
afford only 9a, whereas the reaction of the salt derived from
the more bulky (o-tol)3P (q=1948) afforded predominantly
the substitution product 7a (Table 10, entries 1 and 2). Thus,
it appears that salts containing a bulky phosphine moiety
with a good leaving-group ability, undergo the substitution
reaction in preference to the radical process. The influence
of phosphine electronic factors was probed by studying reac-
tions of salts derived from electron-rich and electron-poor
phosphines. The salt containing (p-MeOC6H4)3P (n=
2066.1 cmÀ1) reacted only very slowly with PhMgBr under
dry air (Table 10, entry 3). On the other hand, the salt ob-
tained from the more electron-poor phosphine (m-FC6H4)3P
(n=2074.1 cmÀ1) reacted efficiently to yield 9a as a major
product with 7a (Table 10, entry 4). Although the salts
formed from (o-tol)3P and (m-FC6H4)3P are highly reactive
with H2O (Table 3), they reacted with PhMgBr to form dif-
ferent major products. These results suggest that electronic
and steric factors have the same effects on both reaction
types. This phenomenon is consistent with the mechanism
for alcohol formation depicted in Scheme 7 because both
pathways involve the nucleophilic addition of water or a su-
peroxide anion to the positively charged phosphorus center
of the salt as key steps.
Scheme 7. Plausible mechanism for alcohol formation in the reaction of
a-alkoxyphosphonium salts with Grignard reagents.
ide radical anion A and the aryl radical are formed by reac-
tion of the Grignard reagent with O2.[12] The highly nucleo-
philic[13] superoxide radical anion A then adds to the posi-
tively charged phosphorous center in the salt 12 followed by
oxidative elimination via an oxaphosphetane-type inter-
mediate. The aldehyde intermediate 13 produced in this
manner quickly reacts with the Grignard reagent to afford
alcohol 9.
Support for this mechanistic proposal was gained in
a study of the reaction carried out in the presence of the
radical trap 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO).
Significantly, TEMPO was observed to inhibit alcohol for-
mation. In addition, the fact that triphenylphosphine oxide
is generated in the reaction supports the operation of the
oxidative elimination process. Finally, the observation that
aliphatic Grignard reagents react to produce only trace
amounts of product alcohols 9 is consistent with the fact
that these reagents are more reactive with O2 than their aryl
counterparts[11c] and, as a result, they are fully consumed
prior to their reactions with the intermediate aldehyde.
Conclusion
In the investigation described above, we have examined the
reactivity of a-alkoxyphosphonium salts with nucleophiles.
The results show that both the steric and electronic proper-
ties of substituents on the phosphine moieties of these salts
affect the efficiencies of reactions that form hydrolysis prod-
ucts. In addition, efficient substitution reactions of various
types of nucleophile with a-alkoxyphosphonium salts pre-
pared from various acetals and (o-tol)3P were developed. Fi-
nally, we uncovered an unusual reaction of the a-alkoxy-
phosphonium salts with aryl Grignard reagents in the pres-
ence of O2, which leads to the formation of alcohol products
through a novel radical pathway.
Phosphine structure/salt reactivity relationships in the alco-
hol-forming reaction: In a continuing effort, we determined
how the nature of the phosphine moiety governs the effi-
ciency of the alcohol-forming reaction of a-alkoxyphospho-
nium salts. Reactions were carried out with salts containing
phosphine moieties with different steric q and electronic fac-
tors n (Table 10).
In an exploration of the steric effect of the phosphine
moiety, the reaction of the salt obtained from 1a and Ph3P
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

H. Fujioka et al.
trans-2-Methoxy-4-phenylbut-3-enenitrile (6d):[20] Colorless oil; 1H NMR
(300 MHz, CDCl3): d=3.52 (3H, s), 4.84 (1H, d, J=5.9 Hz), 6.18 (1H,
dd, J=15.9, 5.9 Hz), 6.91 (1H, d, J=15.9 Hz), 7.32–7.43 ppm (5H, m).
Experimental Section
General procedure for substitution reactions of a-alkoxyphosphonium
salts with nucleophiles (Tables 4–6): In a flame-dried Schrenk tube filled
with nitrogen, TESOTf (2 equiv) was added slowly to a stirred a solution
of an O,O-acetal 1 (1 equiv) and (o-tol)3P (3 equiv) in dry CH2Cl2 (0.1m)
at À58C. The resulting solution was stirred for 0.5 h at 58C. After the dis-
appearance of 1 and the formation of the a-alkoxyphosphonium salt (as
monitored by TLC analysis), H2O (10 mLmmolÀ1) or a nucleophile (1.2–
3 equiv) was added to the reaction mixture. The solution was then stirred
at room temperature. After disappearance of the a-alkoxyphosphonium
salt (TLC), saturated aqueous NaHCO3 was added to the mixture. The
resulting solution was extracted with CH2Cl2 to give an organic layer that
was dried over anhydrous Na2SO4. Concentration in vacuo gave a residue
that was subjected to column chromatography on silica gel to give prod-
ucts 3 and 5–8. Compounds 3a–d are commercially available, and com-
pounds 3e,f, 5b–d, 6b–d, 7b, and 8b–d are known compounds.
2-Isopropoxytridecanonitrile (6e): Colorless oil; IR (KBr): 3156, 2255,
1794, 1722, 1377, 1329, 1119 cmÀ1 1H NMR (300 MHz, CDCl3): d=0.88
;
(3H, t, J=6.7 Hz), 1.17 (3H, d, J=6.2 Hz), 1.22–1.39 (19H, m), 1.42–1.50
(2H, m), 1.77–1.84 (2H, m), 3.86 (1H, septet, J=6.2 Hz), 4.16 ppm (1H,
t, J=6.7 Hz); 13C NMR (100 MHz, CDCl3): d=14.1, 20.9, 22.7, 24.8, 29.0,
29.31, 29.33, 29.5, 29.6, 31.9, 34.0, 66.2, 72.1, 119.3 ppm; HRMS (FAB):
m/z calcd for C16H31NONa: 276.2303 [M+Na]+; found: 276.2278.
1-Methoxydodecylbenzene (7a): Colorless oil; IR (KBr): 2926, 2853,
1454, 1263, 1099 cmÀ1 1H NMR (300 MHz, CDCl3): d=0.87 (3H, t, J=
;
6.8 Hz), 1.24–1.41 (18H, m), 1.59–1.65 (1H, m), 1.76–1.83 (1H, m), 3.20
(3H, s), 4.07 (1H, t, J=6.6 Hz), 7.24–7.37 ppm (5H, m); 13C NMR
(75 MHz, CDCl3): d=14.1, 22.7, 25.8, 29.3, 29.5, 29.56, 29.58, 29.60, 29.62,
31.9, 38.2, 56.6, 84.2, 126.7, 127.4, 128.3, 142.5 ppm; HRMS (EI): m/z
calcd for C19H32O: 276.2445; found: 276.2453.
1-Methoxy-4-[methoxyACHTNUTGRNEUNG
(phenyl)methyl]benzene (7b):[21] Colorless oil;
12-tert-Butyldimethylsilyloxy-1-dodecanal (3e):[14] Colorless oil; 1H NMR
(400 MHz, CDCl3): d=0.04 (6H, s), 0.89 (9H, s), 1.24–1.35 (14H, m),
1.47–1.51 (2H, m), 1.60–1.64 (2H, m), 2.39–2.43 (2H, m), 3.59 (2H, t, J=
1H NMR (300 MHz, CDCl3): d=3.28 (3H, s), 3.69 (3H, s), 5.12 (1H, s),
6.77 (2H, d, J=8.6 Hz), 7.17 (2H, d, J=8.6 Hz), 7.21–7.25 ppm (5H, m).
1-Methoxy-(12-tert-butyldimethylsilyloxy)dodecylbenzene (7c): Colorless
6.6 Hz), 9.76 ppm (1H, s).
oil; IR (KBr): 3051, 2305, 1389, 1360, 1098 cmÀ1 1H NMR (400 MHz,
;
1
3-(1,4-Dioxaspiro
[4.5]decan-6-yl)propanal (3 f):[15] Colorless oil; H NMR
(400 MHz, CDCl3): d=1.21–1.81 (10H, m), 1.94–2.02 (1H, m), 2.36–2.55
(2H, m), 3.91–4.00 (4H, m), 9.77 ppm (1H, t, J=1.8 Hz).
CDCl3): d=0.05 (6H, s), 0.89 (9H, s), 1.19–1.40 (16H, m), 1.47–1.52
(2H, m), 1.60–1.65 (1H, m), 1.77–1.82 (1H, m), 3.20 (3H, s), 3.60 (2H, t,
J=6.6 Hz), 4.08 (1H, t, J=6.6 Hz), 7.26–7.29 (3H, m), 7.33–7.37 ppm
(2H, m); 13C NMR (100 MHz, CDCl3): d=À5.3, 18.4, 25.78, 25.82, 26.0,
29.4, 29.5, 29.56, 29.59, 32.9, 38.2, 56.6, 63.3, 84.1, 126.7, 127.4, 128.3,
142.5 ppm; elemental analysis (%) calcd for C25H46OSi: C 73.83, H 11.40;
found: C 73.67, H 11.26.
(1-Methoxydodecyl)phenylsulfide (5a): Colorless oil; IR (KBr): 2926,
2855, 2253, 1468, 1379, 1098 cmÀ1 1H NMR (300 MHz, CDCl3): d=0.88
;
(3H, t, J=6.7 Hz), 1.24–1.32 (16H, m), 1.41–1.46 (2H, m), 1.69–1.78
(2H, m), 3.47 (3H, s), 4.62 (1H, t, J=6.6 Hz), 7.26–7.33 (3H, m), 7.46–
7.49 ppm (2H, m); 13C NMR (75 MHz, CDCl3): d=14.1, 22.7, 26.2, 29.1,
29.3, 29.48, 29.52, 29.6, 31.9, 35.6, 55.3, 91.0, 127.4, 128.7, 133.4,
133.5 ppm; HRMS (EI): m/z calcd for C19H32OS: 308.2174; found:
308.2193.
(1-Methoxy-3-phenylpropyl)phenylsulfide (5b):[16] Colorless oil; 1H NMR
(300 MHz, CDCl3): d=1.98–2.12 (2H, m), 2.76 (2H, t, J=7.6 Hz), 3.48
(3H, s), 4.54 (1H, t, J=6.6 Hz), 7.13–7.31 (8H, m), 7.45–7.48 ppm (2H,
m).
1-Isopropoxydodecylbenzene (7d): Colorless oil; IR (KBr): 2926, 1452,
1
1375, 1119 cmÀ1; H NMR (400 MHz, CDCl3): d=0.87 (3H, t, J=6.2 Hz),
1.07 (3H, d, J=6.2 Hz), 1.14 (3H, d, J=6.2 Hz), 1.24–1.31 (17H, m),
1.41–1.43 (1H, m), 1.53–1.60 (1H, m), 1.70–1.76 (1H, m), 3.45 (1H,
septet, J=6.2 Hz), 4.29 (1H, t, J=6.2 Hz), 7.22–7.34 ppm (5H, m);
13C NMR (100 MHz, CDCl3): d=14.1, 21.2, 22.7, 23.5, 26.1, 29.3, 29.54,
29.56, 29.60, 29.63, 31.9, 38.9, 68.7, 79.3, 126.5, 127.1, 128.1, 144.1 ppm;
HRMS (FAB): m/z calcd for C21H36O: 304.2766 [M]+; found: 304.2779.
ACHTUNGTRENNUNG
[Methoxy(4-methoxyphenyl)methyl]phenylsulfide (5c):[17] Colorless oil;
1-(Triethylsilyloxyethoxy)-3-(1,4-dioxaspiro
(diastereo mixture) (7e): Colorless oil; IR (KBr): 3053, 2875, 1454, 1261,
1094 cmÀ1 1H NMR (400 MHz, CDCl3): d=0.59 (6H, q, J=7.8 Hz), 0.94
ACHTUNGTRNE[NUNG 4.5]decan-6-yl)propylbenzene
1H NMR (300 MHz, CDCl3): d=3.48 (3H, s), 3.78 (3H, s), 5.69 (1H, s),
;
6.78–6.83 (2H, m), 7.20–7.26 (5H, m), 7.31–7.36 ppm (2H, m).
(9H, t, J=7.8 Hz), 1.17–1.36 (4H, m), 1.42–1.61 (5H, m), 1.69–1.91 (4H,
m), 3.31–3.41 (2H, m), 3.68–3.74 (2H, m), 3.79–3.93 (4H, m), 4.20 (1H, t,
J=6.4 Hz), 7.22–7.33 ppm (5H, m); 13C NMR (100 MHz, CDCl3): d=4.4,
6.8, 23.9, 24.3, 24.4, 24.7, 29.1, 29.2, 34.7, 36.2, 36.7, 44.5, 44.7, 62.3, 64.6,
64.7, 69.9, 70.1, 83.1, 110.88, 110.90, 126.6, 126.8, 127.2, 127.3, 128.2,
142.8, 143.1 ppm; HRMS (EI): m/z calcd for C25H42O4Si: 434.2852;
found: 434.2849.
trans-(1-Methoxy-3-phenylallyl)phenylsulfide (5d):[18] Colorless oil;
1H NMR (300 MHz, CDCl3): d=3.56 (3H, s), 5.30 (1H, dd, J=5.7,
1.3 Hz), 6.20 (1H, dd, J=15.9, 5.7 Hz), 6.50 (1H, dd, J=15.9, 1.3 Hz),
7.19–7.36 (8H, m), 7.44–7.49 ppm (2H, m).
(1-Isopropoxydodecyl)phenylsulfide (5e): Colorless oil; IR (KBr): 3154,
1
3061, 1439, 1377, 1300 cmÀ1; H NMR (300 MHz, CDCl3): d=0.88 (3H, t,
J=6.7 Hz), 1.12 (3H, d, J=5.8 Hz), 1.18–1.32 (19H, m), 1.41–1.42 (2H,
m), 1.71 (2H, m), 4.12 (1H, septet, J=5.8 Hz), 4.71 (1H, t, J=6.5 Hz),
7.26–7.32 (3H, m), 7.47–7.51 ppm (2H, m); 13C NMR (75 MHz, CDCl3):
d=14.1, 20.8, 22.7, 23.1, 26.3, 29.1, 29.3, 29.49, 29.53, 29.6, 31.9, 36.7, 69.0,
86.7, 127.4, 128.6, 133.1, 134.1 ppm; HRMS (EI): m/z calcd for C21H36OS:
336.2487; found: 336.2513.
2-Methoxytridecane (8a): Colorless oil; IR (KBr): 2959, 2924, 2855, 2820,
1
1464, 1373, 1261, 1142, 1123, 1092 cmÀ1; H NMR (300 MHz, CDCl3): d=
0.88 (3H, t, J=6.6 Hz), 1.12 (3H, d, J=6.0 Hz), 1.26–1.38 (20H, m),
3.25–3.31 (1H, m), 3.31 ppm (3H, s); 13C NMR (75 MHz, CDCl3): d=
14.1, 19.0, 22.7, 25.4, 29.3, 29.6, 29.7, 29.8, 31.9, 36.3, 55.9, 76.9 ppm;
HRMS (FAB): m/z calcd for C14H30OLi: 221.2457 [M+Li]+; found:
221.2474.
(3-Methoxybutyl)benzene (8b):[22] Colorless oil; 1H NMR (300 MHz,
CDCl3): d=1.16 (3H, d, J=6.2 Hz), 1.64–1.90 (2H, m), 2.59–2.77 (2H,
m), 3.25–3.35 (1H, m), 3.33 (3H, s), 7.15–7.30 ppm (5H, m).
4-Methoxy-(1-methoxyethyl)benzene (8c):[23] Colorless oil; 1H NMR
(300 MHz, CDCl3): d=1.42 (3H, d, J=6.4 Hz), 3.19 (3H, s), 3.81 (3H,
s), 4.25 (2H, q, J=6.4 Hz), 6.89 (2H, d, J=8.5 Hz), 7.23 ppm (2H, d, J=
8.5 Hz).
trans-1,3-Diphenyl-3-methoxyprop-1-ene (8d):[24] Colorless oil; 1H NMR
(300 MHz, CDCl3): d=1.33 (3H, d, J=6.2 Hz), 3.32 (3H, s), 3.85–3.93
(1H, m), 6.09 (1H, dd, J=16.0, 7.6 Hz), 6.53 (1H, d, J=16.0 Hz), 7.22–
7.41 ppm (5H, m).
2-Methoxytridecanonitrile (6a): Colorless oil; IR (KBr): 2928, 2855,
2255, 1462, 1109 cmÀ1 1H NMR (300 MHz, CDCl3): d=0.88 (3H, t, J=
;
6.7 Hz), 1.26–1.29 (16H, m), 1.43–1.50 (2H, m), 1.79–1.87 (2H, m), 3.49
(3H, s), 4.04 ppm (1H, t, J=6.6 Hz); 13C NMR (75 MHz, CDCl3): d=
14.1, 22.7, 24.7, 29.0, 29.3, 29.4, 29.6, 31.9, 33.3, 57.9, 70.7, 118.2 ppm;
HRMS (FAB): m/z calcd for C14H27NONa: 248.1990 [M+Na]+; found:
248.1968.
2-Methoxy-4-phenylbutyronitrile (6b):[19] Colorless oil; 1H NMR
(300 MHz, CDCl3): d=2.05–2.25 (2H, m), 2.82 (2H, t, J=7.6 Hz), 3.48
(3H, s), 3.96 (1H, t, J=6.7 Hz), 7.17–7.26 (3H, m), 7.29–7.34 ppm (2H,
m).
2-Methoxy-2-(4-methoxyphenyl)acetonitrile (6c):[19]
Colorless
oil;
1H NMR (300 MHz, CDCl3): d=3.42 (3H, s), 3.75 (3H, s), 5.06 (1H, s),
6.86 (2H, d, J=8.5 Hz), 7.33 ppm (2H, d, J=8.5 Hz).
2-(Triethylsilyloxyethoxy)tridecane (8e): Colorless oil; IR (KBr): 2928,
1464, 1377, 1240 cmÀ1, 1092; 1H NMR (400 MHz, CDCl3): d=0.61 (6H,
&
8
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Reactivity of a-Alkoxyphosphonium Salts with Nucleophiles
FULL PAPER
q, J=8.0 Hz), 0.88 (3H, t, J=6.8 Hz), 0.96 (9H, t, J=8.0 Hz), 1.12 (3H,
d, J=6.4 Hz), 1.26–1.37 (19H, m), 1.50–1.56 (1H, m), 3.39–3.47 (2H, m),
3.52–3.58 (1H, m), 3.73 ppm (2H, t, J=6.0 Hz); 13C NMR (100 MHz,
CDCl3): d=4.4, 6.7, 14.1, 19.7, 22.7, 25.5, 29.4, 29.64, 29.66, 29.68, 29.8,
31.9, 36.6, 62.5, 69.6, 75.9 ppm; HRMS (FAB): m/z calcd for C21H47O2Si:
359.3345 [M+H]+; found: 359.3376.
38.9, 81.2, 127.2, 127.5, 127.8, 142.2 ppm; HRMS (FAB): m/z calcd for
C18H30ONa: 285.2194 [M+Na]+; found: 285.2206.
5-Bromo-1-phenylpentan-1-ol (9m):[27] Colorless oil; 1H NMR (500 MHz,
CDCl3): d=1.40–1.46 (1H, m), 1.55–1.61 (1H, m), 1.68–1.75 (1H, m),
1.78–1.84 (1H, m), 1.88 (2H, quintet, J=7.0 Hz), 1.94 (1H, s), 3.38 (2H,
t, J=6.7 Hz), 4.66 (1H, t, J=6.7 Hz), 7.25–7.37 ppm (5H, m).
General procedure for reactions of a-alkoxyphosphonium salts with
Grignard reagents under dry air (Tables 8 and 9): In a flame-dried
Schrenk tube filled with dry air, TESOTf (2 equiv) was added slowly to
a stirred solution of 1 (1 equiv) and Ph3P (3 equiv) in dry CH2Cl2 (0.2m)
at 08C and the resulting solution was stirred for 0.5 h at this temperature.
After the disappearance of 1 and the formation of the a-alkoxyphospho-
nium salt (as monitored by TLC analysis), the Grignard reagent
(5.0 equiv) was added quickly to the reaction mixture, and the resulting
solution was stirred at 08C. After disappearance of the a-methoxyphos-
phonium salt (TLC), saturated aqueous NH4Cl was added to the mixture.
The resulting solution was extracted with CH2Cl2 and the organic layer
was dried over anhydrous Na2SO4. Concentration in vacuo gave a residue
that was subjected to column chromatography on silica gel to give prod-
ucts 9. Compound 9b is commercially available, and compounds 9c–e,
9g–j, and 9m are known compounds.
Acknowledgements
This study was financially supported by the Hoansha Foundation, the
Uehara Memorial Foundation, and Grant-in-Aid for scientific research
from the Ministry of Education, Culture, Sports, Science and Technology
of Japan.
[1] a) S. V. Ley, B. Lygo, H. M. Organ, A. Wonnacott, Tetrahedron 1985,
41, 3825; b) W. Tꢁckmantel, K. Oshima, K. Utimoto, Tetrahedron
Lett. 1986, 27, 5617; c) S. Kim, Y. C. Kim, Synlett 1990, 115; d) E.
Anders, K. Hertlein, A. Stankowiak, E. Irmer, Synthesis 1992, 577.
[2] For recent applications in total synthesis, see: a) W. T. Lambert,
S. D. Burke, Org. Lett. 2003, 5, 515; b) G. A. Molander, K. M.
George, L. G. Monovich, J. Org. Chem. 2003, 68, 9533; c) E.
Quꢂron, R. Lett, Tetrahedron Lett. 2004, 45, 4527; d) W. T. Lambert,
G. H. Hanson, F. Benayoud, S. D. Burke, J. Org. Chem. 2005, 70,
9382; e) M. E. Jung, J. Cordova, M. Murakami, Org. Lett. 2009, 11,
3882; f) M. T. Crimmins, M. C. Mans, A. D. Rodrꢃguez, Org. Lett.
2010, 12, 5028.
1-Phenyldodecan-1-ol (9a): White solid; IR (KBr): 3605, 3383, 1815,
1
1454 cmÀ1; H NMR (400 MHz, CDCl3): d=0.88 (3H, t, J=6.9 Hz), 1.24–
1.43 (18H, m), 1.66–1.83 (2H, m), 1.89 (1H, s), 4.64 (1H, t, J=6.6 Hz),
7.24–7.36 ppm (5H, m); 13C NMR (100 MHz, CDCl3): d=14.1, 22.7, 25.8,
29.3, 29.52, 29.54, 29.58, 29.60, 29.63, 31.9, 39.1, 74.7, 125.9, 127.5, 128.4,
144.9 ppm; HRMS (FAB): m/z calcd for C18H30ONa: 285.2194 [M+Na]+
; found: 285.2167.
(p-tolyl)methanol (9c):[25] Colorless oil; 1H NMR (400 MHz,
PhenylACHTUNGTRENNUNG
CDCl3): d=2.31 (3H, s), 2.31 (1H, s), 5.75 (1H, s), 7.12 (2H, d, J=
7.8 Hz), 7.22–7.25 (3H, m), 7.29 (2H, d, J=7.8 Hz), 7.32–7.36 ppm (2H,
[3] a) H. Fujioka, Y. Sawama, N. Murata, T. Okitsu, O. Kubo, S. Matsu-
da, Y. Kita, J. Am. Chem. Soc. 2004, 126, 11800; b) H. Fujioka, T.
Okitsu, Y. Sawama, N. Murata, R. Li, Y. Kita, J. Am. Chem. Soc.
2006, 128, 5930.
m).
(p-Methoxyphenyl)phenylmethanol (9d):[25] Colorless oil; 1H NMR
(400 MHz, CDCl3): d=2.22 (1H, d, J=2.8 Hz), 3.78 (3H, s), 5.79 (1H, d,
J=2.8 Hz), 6.84–6.87 (2H, m), 7.23–7.38 ppm (7H, m).
[4] a) H. Fujioka, T. Okitsu, Y. Sawama, T. Ohnaka, Y. Kita, Synlett
2006, 3077; b) H. Fujioka, T. Ohnaka, T. Okitsu, O. Kubo, K. Oka-
moto, Y. Sawama, Y. Kita, Heterocycles 2007, 72, 529; c) H. Fujioka,
T. Okitsu, T. Ohnaka, R. Li, O. Kubo, K. Okamoto, Y. Sawama, Y.
Kita, J. Org. Chem. 2007, 72, 7898; d) H. Fujioka, O. Kubo, K.
Senami, K. Okamoto, T. Okitsu, Y. Kita, Heterocycles 2009, 79,
1113.
[5] H. Fujioka, A. Goto, K. Otake, O. Kubo, K. Yahata, Y. Sawama, T.
Maegawa, Chem. Commun. 2010, 46, 3976.
[6] H. Fujioka, A. Goto, K. Otake, O. Kubo, Y. Sawama, T. Maegawa,
Chem. Commun. 2011, 47, 9894.
(p-Fluorophenyl)phenylmethanol (9e):[25] Colorless oil; 1H NMR
(400 MHz, CDCl3): d=2.30 (1H, s), 5.80 (1H, s), 6.98–7.03 (2H, m),
7.24–7.34 ppm (7H, m).
(p-Bromophenyl)phenylmethanol (9g):[25] Colorless oil; 1H NMR
(400 MHz, CDCl3): d=2.43 (1H, s), 5.74 (1H, s), 7.19–7.36 (7H, m),
7.42–7.44 ppm (2H, m).
(p-Trifluoromethylphenyl)phenylmethanol
(9h):[25]
Colorless
oil;
1H NMR (400 MHz, CDCl3): d=2.40 (1H, s), 5.85 (1H, s), 7.27–7.35
(5H, m), 7.49 (2H, d, J=8.2 Hz), 7.58 ppm (2H, d, J=8.2 Hz).
1
Phenyl(2-thienyl)methanol (9i):[25] Pale-yellow solid; H NMR (500 MHz,
[7] C. A. Tolman, Chem. Rev. 1977, 77, 313.
[8] See the Supporting Information for details.
CDCl3): d=2.44 (1H, d, J=3.4 Hz), 6.05 (1H, d, J=3.4 Hz), 6.88 (1H, d,
J=3.1 Hz), 6.94 (1H, dd, J=4.9, 3.1 Hz), 7.26 (1H, dd, J=4.9, 1.2 Hz),
7.29–7.32 (1H, m), 7.37 (2H, t, J=7.3 Hz), 7.45 ppm (2H, dd, J=7.3,
1.8 Hz).
1,3-Diphenylpropan-1-ol (9j):[26] White solid; 1H NMR (400 MHz,
CDCl3): d=1.92 (1H, d, J=3.2 Hz), 2.01–2.13 (2H, m), 2.65–2.74 (2H,
m), 4.65–4.68 (1H, m), 7.16–7.38 ppm (10H, m).
[9] The a-alkoxyphosphonium salts prepared from Ph3P were stable
and easily isolated, but they are highly polar compounds and cannot
be easily handled; thus, we used the salt generated from the acetal
and Ph3P in situ and the reactions afforded the products in similar
yields whether the salt that was isolated or not.
[10] We used N2 gas from a nitrogen tank through a N2 gas line, in which
a small amount of O2 from the air may have been present.
[11] a) C. W. Porter, C. Steel, J. Am. Chem. Soc. 1920, 42, 2650; b) H.
Gilman, A. Wood, J. Am. Chem. Soc. 1926, 48, 806; c) M. T. Goebel,
C. S. Marvel, J. Am. Chem. Soc. 1933, 55, 1693; d) M. S. Kharasch,
W. B. Reynolds, J. Am. Chem. Soc. 1943, 65, 501; e) C. Walling, S. A.
Buckler, J. Am. Chem. Soc. 1955, 77, 6032; f) J. F. Garst, C. D.
Smith, A. C. Farrar, J. Am. Chem. Soc. 1972, 94, 7707. For a review,
see: g) G. Sosnovsky, J. H. Brown, Chem. Rev. 1966, 66, 529.
[12] The reaction of O2 and Grignard reagents are known to produce
radical anion species; see: C. Walling, A. Cioffari, J. Am. Chem.
Soc. 1970, 92, 6609; the aryl radical species coupled to itself to
afford a biaryl compound, which was detected as a byproduct.
[13] a) E. J. Corey, K. C. Nicolaou, M. Shibasaki, Y. Machida, C. S.
Shiner, Tetrahedron Lett. 1975, 16, 3183; b) D. T. Sawyer, J. S. Valen-
tine, Acc. Chem. Res. 1981, 14, 393, and references sited therein.
2-Methyl-1-phenylundecan-1-ol (diastereo mixture) (9k): White solid; IR
(KBr): 3605, 3445, 2930, 2249, 1815, 1493, 1454, 1377, 1015 cmÀ1
;
1H NMR (400 MHz, CDCl3): d=0.74 (1.2H, d, J=6.9 Hz), 0.87 (1.8H,
dd, J=5.5, 1.4 Hz), 0.90 (3H, t, J=6.6 Hz), 1.04–1.41 (16H, m), 1.75–1.87
(2H, m), 4.42 (0.4H, d, J=6.9 Hz), 4.52 (0.6H, d, J=5.5 Hz), 7.23–
7.35 ppm (5H, m); 13C NMR (100 MHz, CDCl3): d=14.1, 14.3, 15.6, 22.7,
27.0, 27.2, 29.31, 29.33, 29.58, 29.60, 29.7, 29.8, 29.9, 31.88, 31.90, 32.2,
33.1, 40.09, 40.14, 78.2, 79.1, 126.3, 126.7, 127.2, 127.4, 128.12, 128.13,
143.6, 143.9 ppm; HRMS (FAB): m/z calcd for C18H30ONa: 285.2194
[M+Na]+; found: 285.2195.
2,2-Dimethyl-1-phenyldecan-1-ol (9l): Pale-yellow oil; IR (KBr): 3609,
1794, 1468, 1385, 1096 cmÀ1 1H NMR (400 MHz, CDCl3): d=0.81 (3H,
;
s), 0.88 (3H, t, J=6.6 Hz), 0.89 (3H, s), 1.17–1.38 (14H, m), 1.81 (1H, d,
J=2.3 Hz), 4.45 (1H, d, J=2.3 Hz), 7.24–7.33 ppm (5H, m); 13C NMR
(100 MHz, CDCl3): d=14.1, 22.6, 23.1, 23.9, 29.4, 29.7, 30.6, 31.9, 38.1,
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

H. Fujioka et al.
[14] W. He, C. E. Soll, S. S. Chavadi, G. Zhang, J. D. Warren, L. E. N.
Quadri, J. Am. Chem. Soc. 2009, 131, 16744.
[23] P. Mꢁller, P. Nury, G. Bernardinelli, Eur. J. Org. Chem. 2001, 4137.
ˇ
ˇ
[24] A. V. Malkov, I. R. Baxendale, D. Dvorꢄk, D. J. Mansfield, P. Kocov-
´
[15] G. Pandey, S. Hajra, M. K. Ghorai, K. R. Kumar, J. Org. Chem.
1997, 62, 5966.
sky, J. Org. Chem. 1999, 64, 2737.
[25] J. Karthikeyan, M. Jeganmohan, C.-H. Cheng, Chem. Eur. J. 2010,
16, 8989.
[26] A. F. Trindade, P. M. P. Gois, L. F. Veiros, V. Andre, M. T. Duarte,
C. A. M. Afonso, S. Caddick, F. G. N. Cloke, J. Org. Chem. 2008, 73,
4076.
[16] A. Mendoza, D. S. Matteson, J. Organomet. Chem. 1978, 156, 149.
[17] F. Ono, R. Negoro, T. Sato, Synlett 2001, 1581.
[18] T. Miura, Y. Masaki, Tetrahedron 1995, 51, 10477.
[19] T. Soga, H. Takenoshita, M. Yamada, T. Mukaiyama, Bull. Chem.
Soc. Jpn. 1990, 63, 3122.
[27] W. S. Trahanovsky, N. S. Fox, J. Am. Chem. Soc. 1974, 96, 7968.
[20] S. Hꢁnig, H. Reichelt, Chem. Ber. 1986, 119, 1772.
[21] D. E. Laycock, H. Alper, J. Org. Chem. 1981, 46, 289.
[22] X. Li, L. Li, Y. Tang, L. Zhong, L. Cun, J. Zhu, J. Liao, J. Deng, J.
Org. Chem. 2010, 75, 2981.
Received: February 14, 2012
Revised: July 5, 2012
Published online: && &&, 0000
&
10
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Reactivity of a-Alkoxyphosphonium Salts with Nucleophiles
FULL PAPER
Phosphorus Chemistry
A. Goto, K. Otake, O. Kubo,
Y. Sawama, T. Maegawa,
H. Fujioka* ................... &&&& — &&&&
A dramatic change in the reactivity of
an a-alkoxyphosphonium salt toward
nucleophiles is observed due to the
steric and electronic nature of the
phosphine substituents. By changing
the type of phosphorus substituents,
the reaction pathway can be controlled
to proceed selectively by substitution
a new radical reaction (see scheme;
OTf=trifluoromethansulfonate,
TMS=trimethylsilyl, o-tol=tolyl).
Effects of Phosphorus Substituents on
Reactions of a-Alkoxyphosphonium
Salts with Nucleophiles
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!