Organic Letters
Letter
complete and clean deprotection of 2d (Table 1, entry 5).
Under the same reaction conditions, the related MOM
protected 2e, afforded only 90% conversion (Table 1, entry
6). We hypothesized that the slightly more electron-with-
drawing BOM group favored the deprotection process.
Scheme 2. Initial Experiments with Methyl Phosphinite
Borane 2a
To demonstrate the capacity of BOM-phosphinite 2d as a P-
stereogenic transfer reagent, we next tested the one-pot
deprotection/substitution reaction with PhLi using 2d as
starting material (Scheme 3).17 After deprotection of 2d in
boranes provide clean substitution reaction with organolithium
reagents.10 Reaction of 2a with PhLi was strongly solvent
dependent; no reaction was observed with ethereal solvents,
while in toluene 79% of the desired product was obtained but
with complete loss of the optical purity.7 The differential
Scheme 3. One-Pot Deprotection/Substitution Reaction
with BOM-Phosphinite 2d
́
behavior compared with the results reported by Juge is most
likely due to the presence of a bulky tert-butyl group attached
to phosphorus in 2a.11 At this stage, we envisioned that the
reaction on a borane-free trivalent phosphorus atom could take
̈
place through a stereospecific pathway. Borner and Bayardon
have demonstrated that trivalent phosphinites are more
reactive when compared to borane-protected ones.12 To this
end, deprotection of 2a with 1,4-diazabicyclo[2.2.2]octane
(DABCO) was examined (Scheme 2). As determined by 31P
NMR analysis, low conversion to the desired free phosphinite
3 was observed. In addition to this, due to its volatility any
efforts to isolate 3 were unsuccessful. Given these observations,
2a was deemed an inappropriate substrate for this strategy.
At this point, a phosphinite derivative that could be
efficiently deprotected under basic conditions was needed.
To this end, ten new phosphinites were prepared from a stable
section for the details of the synthesis and a complete list of
phosphinite derivatives. With the new phosphinites in hand,
we tested them for deprotection with DABCO in degassed
C6D6 and analyzed the conversion by 31P NMR. Selected
results for phosphinites 2b−e are shown in Table 1.14 Using 3
toluene at 80 °C, the reaction was cooled and 2 equiv of
phenyllithium were added. Finally, the resulting tertiary
phosphine was protected again and analyzed by chiral
HPLC. While addition of PhLi (1.9 M in Bu2O) at room
temperature produced the tert-butylmethylphenyl phosphine
3a with suboptimal optical purity, reducing the temperature to
−20 °C afforded 3a in 99% ee and excellent yield. Optical
rotation of 3a and comparison with literature data18 revealed
that the substitution took place with inversion of configuration
at the phosphorus center. We believe that chelation of lithium
through the BOM oxygen atoms facilitates an efficient
substitution process at phosphorus.
To demonstrate the utility of 2d in the synthesis of P-
stereogenic tertiary phosphines, we tested the scope of the
reaction with several organometallic reagents (Scheme 4). The
use of phenylmagnesium bromide in the substitution reaction
did not allow it to reach completion, which confirmed that
organolithium reagents are more efficient in this process. To
achieve the highest stereoselectivity in the substitution process,
the organolithium reagent was added at −78 °C and the
reaction was allowed to warm up to room temperature
overnight. In this way, reaction with commercial butyllithium
provided 3b in a highly stereospecific manner. By comparing
the sense of rotation of 3b with the one reported in the
literature,19 it was again confirmed that the substitution
reaction with alkyllithiums occurs with inversion of config-
uration at phosphorus center.
Next, we tested the reaction with noncommercial organo-
lithiums (Scheme 4). Aryllithiums were generated from the
corresponding aryl bromides by metalation with BuLi at low
temperature.20 We observed that in certain instances the
choice of solvent (Et2O, THF, or a mixture of both) in which
the organolithium was prepared was crucial in the reaction
performance. We believe that this solvent effect is not related
with the generation of the ArLi, but rather has a strong
influence in the P-substitution step. After careful optimization
for each case, (aryl)-tert-butylmethyl phosphines boranes 3c−i
were prepared in moderate to good yields and very high optical
purity, as determined by chiral HPLC. The method proved
useful for the formation of intermediates bearing heterocycles
(3h) or bicyclic rings (3i). Phosphine 3j, bearing an acetylide
moiety, was also be prepared efficiently. In contrast with most
methods, modification of the last step also allowed the
preparation of phosphine oxides. Addition of tert-butyl
Table 1. Highlighted Deprotection Assays with Different
Phosphinites
a
a
b
entry
R
substrate
base
conditions conversion
1
2
3
4
5
6
PhCH2
2b
2c
2c
2d
2d
2e
DABCO
DABCO
DABCO
DABCO
Quinuclidine 80 °C, 4 h
Quinuclidine 80 °C, 4 h
60 °C, 4 h
60 °C, 4 h
80 °C, 4 h
80 °C, 4 h
20%
45%
Decomp.
85%
98%
C6F5CH2
C6F5CH2
BnOCH2
BnOCH2
MeOCH2
90%
a
NMR tube experiments, concentration: 2 (0.39 M) and base (1.2
b
M). Conversion was determined by 31P NMR analysis.
equiv of DABCO, the benzyl phosphinite was deprotected with
only 20% conversion (Table 1, entry 1). The conversion
increased slightly when pentafluorobenzyl phosphinite 2c was
used; however, the starting material and product showed
decomposition when the temperature was increased to 80 °C
(Table 1, entries 2 and 3). Under the same reaction conditions
BOM derivative 2d provided 85% conversion and a cleaner
reaction profile (Table 1, entry 4).15 Finally, switching to
quinuclidine, which is a stronger base,16 gave a virtually
4803
Org. Lett. 2021, 23, 4802−4806
Rojo, Pep
