Efficient asymmetric syntheses of 1-phenyl-phosphindane, derivatives, and 2- or 3-oxa analogues: Mission accomplished

Article abstract of DOI:10.1021/ol500970x

A highly enantioselective synthesis of unsubstituted 1-phenyl-phosphindane and its P-borane and P-oxide derivatives was effectively established via a new fluoride-triggered desilylative carbocyclization strategy. Preparation of the oxygen atom-doped 1-phenyl-3-oxa-1-phosphindane-P-borane analogue was otherwise achieved via a tandem P-α-iodination-intra-O-alkylation.

Full text of DOI:10.1021/ol500970x

Letter
pubs.acs.org/OrgLett
Efficient Asymmetric Syntheses of 1‑Phenyl-phosphindane,
Derivatives, and 2- or 3‑Oxa Analogues: Mission Accomplished
Slavko Rast, Barbara Mohar,* and Michel Stephan*,†
National Institute of Chemistry, Hajdrihova 19, 1000 Ljubljana, Slovenia
S
* Supporting Information
ABSTRACT: A highly enantioselective synthesis of unsubstituted 1-
phenyl-phosphindane and its P-borane and P-oxide derivatives was
effectively established via a new fluoride-triggered desilylative carbocyc-
lization strategy. Preparation of the “oxygen atom-doped” 1-phenyl-3-oxa-
1-phosphindane-P-borane analogue was otherwise achieved via a tandem
P-α-iodination−intra-O-alkylation.
access. Prepared for the first time in 2007 in 70% ee by the
high level of development both academically and
Aindustrially has been attained especially in asymmetric Glueck group via Pd(Me-DuPHOS)(E-stilbene)-catalyzed
intramolecular asymmetric iodoarene phosphination,4a the
preparation of the P-oxide in 82% ee under asymmetric
hydrogenation of olefins catalyzed by transition-metal com-
plexes of enantiopure phosphorus-based ligands.1 P-Stereogenic
oxidative Appel conditions was recently published by the
Gilheany group.4b In the same period, we have reported the
formation of its 3-spiro derivatives 5TwistP·BH3 in >99.9% ee
(20% yield) via a radical intramolecular benzannulation
ligands backed by well-adapted processes for their preparation
have greatly impacted these advances.2 A multitude of chiral
ligand designs were proven to be quite efficient in catalysis,
among which phosphacyclic motif-incorporated ones displayed
notable performance. The 4−7-membered phosphacyclic
ligands are the most encountered in the literature of this last
category with phospholanes occupying center stage.2,3 Various
mono- and diphosphine designs have been invented wherein
the core skeleton is optionally decorated with arenes,
heteroatoms, and/or neighboring bulky groups (Figure 1).
The diverse plethora of chiral P-compounds excludes devising
one general access route to them all, but a best-route-per-design
approach is the most adequate.
following the Juge−
́
Stephan asymmetric route to phosphines.4c
Also, the closely related 1-o-anisyl-phosphindane-P-borane has
been obtained by Imamoto et al. in 65% ee relying on an l-
menthyl phosphinite-P-borane reduction/Pd(PPh3)4 cross-
coupling sequence.4d Nevertheless, providing a convenient
general solution for this long-standing challenge appeared to be
relevant to our ongoing asymmetric catalysis-related P-
research.5
We referred again to the particularly well-suited Juge−
́
Stephan asymmetric route which generally enables access to a
broad spectrum of P-stereogenic compounds.5d,6 Its prime
advantage is the chemospecific and highly stereoselective
stepwise displacement of the (+)- or (−)-ephedrine auxiliary
from an enantiopure 1,3,2-oxazaphospholidine-2-borane com-
plex (oxazaPB) toward either P-compound enantiomer.7
Aiming at the stepwise construction of the 1-phosphindane
cyclic core, we have contemplated a novel key ring-closure step
between a suitably α-functionalized o-tolyl group and a
functionalized methyl group following their successive
introduction onto the P-atom.
Clearly, synthetic hurdles still remain for some of the basic
sought-after structures preventing their application in full-
power catalysis. For instance, the chiral 1-aryl-phosphindane
framework (Figure 1) proved to be notoriously difficult to
́
Thus, applying an adapted Juge−Stephan approach to the
target molecule,8 (2R,4S,5R)-oxazaPB (derived from (−)-ephe-
drine) ring opening with lithium o-lithio-benzyloxide led to (o-
hydroxymethyl-phenyl)aminophosphine-P-borane 1 (86%
yield) as a single diastereomer after recrystallization of the
crude (Scheme 1).9 Noteworthy, diethyl ether as the solvent
was crucial for the reaction success instead of the
recommended THF in the standard route conditions. The
Received: April 2, 2014
Published: April 28, 2014
Figure 1. Selection of chiral (hetero)phospholanes.
© 2014 American Chemical Society
2688
dx.doi.org/10.1021/ol500970x | Org. Lett. 2014, 16, 2688−2691

Organic Letters
Letter
Scheme 1. Enantioselective Synthesis of 1-Phenyl-phosphindane (8) and Derivatives
subsequent H2SO4-mediated “methanolysis step’’ also accord-
ing to the general route furnished the unexpected 1-phenyl-2-
oxa-1-phosphindane-P-borane (2, 96% yield) instead of the
methyl phosphinite-P-borane.10 Interestingly, this crystalline
cyclic phosphinite-P-borane was formed in 97.5% ee, and
further ee upgrading to 99.8% was achieved by single
recrystallization from MeOH (75% total yield) (Figure
2).11,12 This phosphinite-P-borane is a valuable synthon for
exclusive attack on the P-atom even at very low temperatures.14
However, the enantioselectivity was dependent on the reaction
conditions, affording lower ee’s than usually obtained with
methyl phosphinite-P-boranes.5,6,15 Initial results in THF were
<40% ee. Hence, an optimization study was conducted with this
very reactive yet easily handled intermediate 2, varying
temperature, solvent, and the organolithium additive (see the
Supporting Information, Table S1).16 Of the conditions
examined following Route A, a maximum of 90.2% ee (RP)
was attained for 4 (80% yield) in the presence of (−)-sparteine
(1.1 equiv) in Et2O at −78 °C. Fortunately, the two-step Route
B via 3 was more selective leading to (RP)-3 (97% yield) in up
to 97.7% ee in cumene at −78 °C employing MeLi·LiBr.
Curiously, with MeLi in the presence of (R,R)- or (S,S)-trans-
N,N,N′,N′-tetramethyl-1,2-cyclohexanediamine (TMCDA, 2.2
equiv) in toluene at −78 °C, the ent-3 was obtained in 95.3% ee
(SP).17 Thus, the additive chirality does not seem to influence
the P-attack sense. However, the switch in the major attack
sense observed with MeLi vs TMSCH2Li must stem from their
relative steric bulk which influences the reagent aggregation18 in
the different media and the approach to 2 (Figure 3).
Subsequently, P-α-silylation of (RP)-3 led to (RP)-4 in high
yield (98%).19
Figure 2. ORTEP drawing of ent-2 (SP) derived from (+)-ephedrine at
the 50% probability level (one of the two molecules of the unit cell is
shown). Selected bond lengths (Å) and angles (deg): P1−B1
1.880(4), P1−O1 1.599(2), O1−C1 1.492(7), B1−P1−O1 115.3(2),
B1−P1−C11 114.9(2).
Further on, the hydroxyl group of (RP)-4 was mesylated21
into (RP)-5, but this unfortunately failed to react under the
desilylative conditions en route to cyclization.22 However, this
the preparation of a variety of P-chiral compounds due to its
minimal sterics, high reactivity of the cycle, and the possibility
to manipulate subsequently the benzylic position.
Conversion of the hydroxyl group of P-(o-hydroxymethyl-
phenyl) into a leaving group which would be intramolecularly
displaced by a generated P-α-anion was the contemplated key
ring-formation step. In general, α-deprotonation of (alkyl)-
phosphine-P-boranes is readily achieved with organolithiums,
but a compatible approach was required in this particular case
for the subsequent intramolecular cyclization onto the P-(α-
activated o-tolyl) group present in the same molecule. A P-
CH2TMS group as a masked P-α-anion appeared to be
appropriate allowing a F−-induced desilylative carbocycliza-
tion.13 Thus, a direct introduction of the P-CH2TMS group
(Route A) or a P-Me group followed by its silylation (Route B)
was pursued.
Figure 3. Organolithiums attack sense on oxazaPB,5,6,20a and cyclic (2)
or acyclic (methyl5,6 or l-menthyl20c) phosphinite-P-boranes.
The ring opening of the cyclic phosphinite-P-borane 2 with
the organolithiums proceeded well regioselectively with
2689
dx.doi.org/10.1021/ol500970x | Org. Lett. 2014, 16, 2688−2691

Organic Letters
Letter
product allowed advantageously ee upgrading to >99.9% by
recrystallization from MeOH at −15 °C.15 Substitution of the
MsO group with a Br ((RP)-6, 95%) overcame the reactivity
issue, and the targeted transformation to (RP)-7 proceeded
smoothly in 78% yield using CsF/4 Å molecular sieves in
DMA.23 Conversion under nonracemizing mild conditions of
(RP)-7 into its BH3-free phosphine (RP)-8 (98%) and the
corresponding P-oxide (SP)-9 (95%) was high-yielding with full
preservation of the enantiomeric integrity.4a,15,24
ACKNOWLEDGMENTS
■
This work was supported by the Ministry of Higher Education,
Science, and Technology of the Republic of Slovenia (Grant
No. P1-242). We thank Barbara Modec, University of
Ljubljana, for X-ray crystal structures determination on the
Supernova diffractometer at the EN-FIST Centre of Excellence,
Slovenia.
REFERENCES
■
Finally, additional interest arose to prepare the yet unknown
3-oxa analogue 11 (Scheme 2). The BH3-free P-Ph-substituted
(1) Handbook of Homogeneous Hydrogenation; de Vries, J. G., Elsevier,
C. J., Eds.; Wiley-VCH: Weinheim, 2006; Vols. 1−3.
(2) For recent literature surveys, see: (a) Grabulosa, A. P-Stereogenic
Ligands in Enantioselective Catalysis, 1st ed.; RSC Publishing:
Cambridge, U.K., 2011. (b) Phosphorous(III) Ligands in Homogeneous
Catalysis: Design and Synthesis, Kamer, P. C. J.; van Leeuwen, P. W. N.
M., Eds.; John Wiley & Sons, Ltd.: West Sussex, U.K., 2012. (c)
Scheme 2. Enantioselective Synthesis of Unsubstituted 1-
Phenyl-3-oxa-1-phosphindane-P-borane ((RP)-11) from an
SMS-Phos Precursor
Phosphorus Ligands in Asymmetric Catalysis; Borner, A., Ed.; Wiley-
̈
VCH: Weinheim, Germany, 2008; Vols. 1−3. (d) Holz, J.; Genson,
M.-N.; Zayas, O.; Borner, A. Curr. Org. Chem. 2007, 11, 61−106.
̈
(3) For R-DuPHOS and R-BPE (R = Me, Et, Pr, i-Pr, Ph), see:
(a) Burk, M. J.; Feaster, J. E.; Nugent, W. A.; Harlow, R. L. J. Am.
Chem. Soc. 1993, 115, 10125−10138. (b) Pilkington, C. J.; Zanotti-
Gerosa, A. Org. Lett. 2003, 5, 1273−1275. For CatASium M, see:
(c) Holz, J.; Zayas, O.; Jiao, H.; Baumann, W.; Spannenberg, A.;
Monsees, A.; Riermeier, T. H.; Almena, J.; Kadyrov, R.; Borner, A.
̈
cycle is the aromatic counterpart of the resolved P-tBu-
substituted component unit of BIBOP.3j,k Starting from our
previously described enantiopure (o-hydroxyphenyl)(methyl)-
(phenyl)phosphine-P-borane ((RP)-10, 85% overall yield from
oxazaPB),5d a tandem iodination−intra-O-alkylation of its P-α-
anion resulted in the cyclized product (RP)-11 in 82% yield.15
Chem.Eur. J. 2006, 12, 5001−5013. For P-chiral 2-Bn-substituted
1,2-bisphospholano-ethane or -benzene, see: (d) Hoge, G. J. Am.
Chem. Soc. 2003, 125, 10219−10227. (e) Hoge, G. J. Am. Chem. Soc.
2004, 126, 9920−9921. P-chiral i-Pr-BeePHOS (absolute P- and C-
configurations unknown) was prepared via double bisnucleophilic
attacks by dilithium o-phenylenediphosphide on enantiopure α-i-Pr-o-
fluorophenethyl mesylate. For this, see: (f) Shimizu, H.; Saito, T.;
́
In conclusion, although we had the Juge−Stephan route as a
Kumobayashi, H. Adv. Synth. Catal. 2003, 345, 185−189.
For
template for our enantioselective synthesis of 1-phenyl-
phosphindane (8), the surmounted challenges encountered
along the proceeding steps reassert the difficulty entailed in the
preparation of such a structure and bear testimony to the
versatility of the followed route. The basic cyclic structure 8 and
its P-BH3 7 and P-oxide 9 derivatives were prepared in >99.9%
ee and 32−34% overall yields via a new facile F−-triggered
desilylative carbocyclization strategy. Moreover, the highly
enantiodivergent ring opening of the cyclic precursor 2 offers a
stereocomplementary advantage to the opposite enantiomer
access by switching the ephedrine chirality. Finally, 1-phenyl-3-
oxa-1-phosphindane-P-borane (11) an analogue of 7·BH3, was
also efficiently prepared (>99.9% ee, 70% overall yield).
Application of these compounds and the developed key steps
are manifold.
TangPhos, DuanPhos, and ZhangPhos, see: (g) Tang, W.; Zhang, X.
Angew. Chem., Int. Ed. 2002, 41, 1612−1614. (h) Liu, D.; Zhang, X.
Eur. J. Org. Chem. 2005, 646−649. (i) Zhang, X.; Huang, K.; Hou, G.;
Cao, B.; Zhang, X. Angew. Chem., Int. Ed. 2010, 49, 6421−6424. For
BIBOPs, see: (j) Tang, W.; Qu, B.; Capacci, A. G.; Rodriguez, S.; Wei,
X.; Haddad, N.; Narayanan, B.; Ma, S.; Grinberg, N.; Yee, N. Y.;
Krishnamurthy, D.; Senanayake, C. H. Org. Lett. 2010, 12, 176−179.
(k) Liu, G.; Liu, X.; Cai, Z.; Jiao, G.; Xu, G.; Tang, W. Angew. Chem.,
Int. Ed. 2013, 52, 4235−4238. For diazaphospholanes, see: (l) Clark,
T. P.; Landis, C. R. J. Am. Chem. Soc. 2003, 125, 11792−11793.
(4) (a) Brunker, T. J.; Anderson, B. J.; Blank, N. F.; Glueck, D. S.;
Rheingold, A. L. Org. Lett. 2007, 9, 1109−1112. (b) Carr, D. J.;
Kudavalli, J. S.; Dunne, K. S.; Muller-Bunz, H.; Gilheany, D. G. J. Org.
̈
̌
Chem. 2013, 78, 10500−10505. (c) Mohar, B.; Cusak, A.; Modec, B.;
Stephan, M. J. Org. Chem. 2013, 78, 4665−4673. (d) Imamoto, T.;
Yoshizawa, T.; Hirose, K.; Wada, Y.; Masuda, H.; Yamaguchi, K.; Seki,
H. Heteroatom Chem. 1995, 6, 99−104. Therein, by contrast 1-alkyl-
phosphindane-P-boranes were prepared in 99% ee via reduction/
alkylation.
ASSOCIATED CONTENT
* Supporting Information
■
S
̌
(5) (a) Stephan, M.; Sterk, D.; Modec, B.; Mohar, B. J. Org. Chem.
2007, 72, 8010−8018. (b) Zupancic, B.; Mohar, B.; Stephan, M. Adv.
Synth. Catal. 2008, 350, 2024−2032. (c) Stephan, M.; Sterk, D.;
̌
Zupancic, B.; Mohar, B. Org. Biomol. Chem. 2011, 9, 5266−5271.
Experimental procedures and characterization data. This
material is available free of charge via the Internet at http://
pubs.acs.org.
̌
̌
̌
̌
(d) Stephan, M.; Modec, B.; Mohar, B. Tetrahedron Lett. 2011, 52,
1086−1089.
AUTHOR INFORMATION
Corresponding Authors
■
(6) Juge,
1990, 31, 6357−6360.
́ ̂
S.; Stephan, M.; Laffitte, J. A.; Genet, J.-P. Tetrahedron Lett.
*E-mail: mstephan@phosphoenix.com.
*E-mail: barbara.mohar@ki.si.
Present Address
†PhosPhoenix SARL, 115, rue de l’Abbe
́
Groult, 75015 Paris,
France.
Notes
(7) This is contrary to the reproach for this route as presented in:
Han, Z. S.; Goyal, N.; Herbage, M. A.; Sieber, J. D.; Qu, B.; Xu, Y.; Li,
Z.; Reeves, J. T.; Desrosiers, J.-N.; Ma, S.; Grinberg, N.; Lee, H.;
Mangunuru, H. P. R.; Zhang, Y.; Krishnamurthy, D.; Lu, B. Z.; Song,
J.; Wang, G.; Senanayake, C. H. J. Am. Chem. Soc. 2013, 135, 2474−
2477.
(8) For the preparation of the lower homologue intermediates
The authors declare no competing financial interest.
towards SMS-Phos ligands’ precursor, see ref 5d.
2690
dx.doi.org/10.1021/ol500970x | Org. Lett. 2014, 16, 2688−2691

Organic Letters
Letter
(9) X-ray crystal structure analysis of ent-1 prepared from (2S,4R,5S)-
oxazaPB (derived from (+)-ephedrine) confirmed the retention of the
P-configuration. For this, see the Supporting Information.
(10) (a) Conversion of (SP)-1 in ∼96% was reached within 1.5 h
compared to the usual 12−15 h for other (o-substituted aryl)(N-
ephedrino)(phenyl)phosphine-P-boranes.5,6 (b) Exploration of this
step using BF3·Et2O (1 equiv) Lewis acid5d at rt in Et2O, toluene, or
THF led to (RP)-2 in 95%, 95%, and 92% ee, respectively.
(11) The ee of enriched 2 collected from the recrystallization filtrate
was determined by HPLC on a Daicel Chiralcel OD column (25 cm)
conjugated with a Daicel Chiralcel OD-H column (15 cm), referenced
with ent-2 prepared from (2S,4R,5S)-oxazaPB (derived from
(+)-ephedrine). For details, see the Supporting Information.
(12) Full X-ray crystal structure analysis of ent-2 prepared from
(2S,4R,5S)-oxazaPB (derived from (+)-ephedrine) is contained in the
Supporting Information.
(13) To the best of our knowledge, the closest work in the literature
is Ph2PCH2TMS addition to carbonyls promoted by CsF (10 mol %)
in DMF at 50 °C. For this, see: Kawashima, T.; Mitsuda, N.; Inamoto,
N. Bull. Chem. Soc. Jpn. 1991, 64, 708−710. Therein, use of 4 Å
molecular sieves as a drying agent caused a decrease in yields.
(14) No RLi attack at the benzylic position was observed.
(15) (a) Ee was determined by HPLC analysis on chiral columns.
For details, see the Supporting Information. (b) The P-configurations
of compounds 3−9 were attributed based on the chiral HPLC order of
elution of the P-oxide 9 and comparison with the literature;4b therein,
the X-ray structure of (RP)-9 was also reported. (c) Phosphine-P-
boranes decomplexation in Et2NH and phosphines P-oxidation with
H2O2 occur usually with retention of the P-configuration;20c however,
the absolute P-configuration can change according to the stereo-
chemistry CIP priority rules.
(16) Commercial TMSCH2Li (1 M in pentane), MeLi (1.6 M in
Et2O), or MeLi·LiBr (1.5 M in Et2O) were used in the study.
(17) A reverse P-attack sense was also noted in THF or using
TMEDA or (−)-sparteine with MeLi but not in the TMSCH2Li case.
(18) For alkyllithium aggregation studies, see: (a) Su, C.; Hopson, R.;
Williard, P. G. J. Org. Chem. 2013, 78, 11733−11746. (b) Reich, H. J.
Chem. Rev. 2013, 113, 7130−7178.
(19) NH4Cl aq. treatment hydrolyzed the O-TMS-protected product
into 4.
(20) (a) Rippert, A. J.; Linden, A.; Hansen, H.-J. Helv. Chim. Acta
2000, 83, 311−321. Therein, a retention of the P-configuration was
also observed with the (S,S,S)-oxazaPB derived from (1S,2S)-
́
(+)-pseudoephedrine. (b) Leon, T.; Riera, A.; Verdaguer, X. J. Am.
Chem. Soc. 2011, 133, 5740−5743. Therein, an inversion of the P-
configuration was observed for the ring opening of (2R,4S,5R)-2-t-Bu-
(1,3,2)-oxazaphospholidine-P-borane derived from (1S,2R)-1-amino-
2-indanol. (c) Imamoto, T.; Oshiki, T.; Onozawa, T.; Kusumoto, T.;
Sato, K. J. Am. Chem. Soc. 1990, 112, 5244−5252.
(21) Mesylation of (2-hydroxymethyl-phenyl)(methyl)(phenyl)phos-
phine-P-oxide using MsCl/Et3N led to the P-(2-chloromethyl-phenyl)
́ ́
derivative. For this, see: Kaczmarczyk, S.; Madalinska, L.; Kiełbasinski,
P. Phosphorus, Sulfur Silicon 2013, 188, 249−253.
(22) According to 1H NMR a straight desilylation of 5 ensued
without cyclization.
(23) Under the adopted unoptimized reaction conditions, ∼3 equiv
of CsF were used, and addition of 4 Å molecular sieves proved to be
beneficial (on the contrary for the transformation in ref 13). Also,
byproducts were formed for the reaction in DMF.
(24) Decomplexation of 7 was carried out at rt in order to avoid
partial racemization of 8, as heating at 50 °C for 2 h resulted in a 2% ee
loss. This was confirmed by P-recomplexation of 8 with BH3 using
BMS and chiral HPLC analysis.
2691
dx.doi.org/10.1021/ol500970x | Org. Lett. 2014, 16, 2688−2691