Access to phosphacymantrene-2-carboxylic acid and 2-carboxaldehyde derivatives

Article abstract of DOI:10.1021/om9904596

Two new approaches to α-functional phosphacymantrenes are described. The reaction of [Mn₂(CO)₁₀] with a phosphole-2-carboxylate affords the corresponding phosphacymantrene2-carboxylate 2. Replacing one carbonyl by one triphenylphosphine on manganese enhances the reactivity of phosphacymantrenes toward electrophilic substitution. Thus, a Vilsmeier formylation leading to a phosphacymantrene-2-carboxaldehyde such as 5 becomes possible. The reduction of 5 affords the corresponding primary alcohol 6 and the methyl derivative 7.

Full text of DOI:10.1021/om9904596

5688
Organometallics 1999, 18, 5688-5690
Access to P h osp h a cym a n tr en e-2-ca r boxylic Acid a n d
2-Ca r boxa ld eh yd e Der iva tives
Bernard Deschamps, Louis Ricard, and Franc¸ois Mathey*
Laboratoire “He´te´roe´le´ments et Coordination”, UMR 7653 CNRS, DCPH, Ecole Polytechnique,
91128 Palaiseau Cedex, France
Received J une 14, 1999
Two new approaches to R-functional phosphacymantrenes are described. The reaction of
[Mn2(CO)10] with a phosphole-2-carboxylate affords the corresponding phosphacymantrene-
2-carboxylate 2. Replacing one carbonyl by one triphenylphosphine on manganese enhances
the reactivity of phosphacymantrenes toward electrophilic substitution. Thus, a Vilsmeier
formylation leading to a phosphacymantrene-2-carboxaldehyde such as 5 becomes possible.
The reduction of 5 affords the corresponding primary alcohol 6 and the methyl derivative 7.
In tr od u ction
optimization of the reaction conditions allowed us to
obtain the corresponding phosphacymantrene 2 in
50-70% yield (eq 1).
Enantiopure planar-chiral ferrocenylphosphines play
a central role in asymmetric catalysis.1 As a logical
extension of the same concepts, several promising
studies have been recently published on planar-chiral
phosphaferrocenes.2,3 From a synthetic standpoint, the
access to these molecules critically depends on the
straightforward preparation of phosphaferrocene-2-car-
boxaldehydes by a simple Vilsmeier formylation.4 Even
more recently, our attention was caught by a report of
Kudis and Helmchen describing the first successful
applications of enantiopure planar-chiral cymantre-
nylphosphines in asymmetric catalysis.5 Put together,
all these data led us to envisage the possible use of the
readily available phosphacymantrenes6 in the same
area. Unfortunately, only Friedel-Crafts acylations are
possible with these species,7 and both the 2-carboxylic
acid and 2-carboxaldehyde derivatives are presently
unknown. Thus, it appeared necessary to break this
synthetic bottleneck.
Two observations need further comments. First, it is
necessary to use 0.5 equiv of [Mn2(CO)10] to obtain the
best results. Any excess results in lower yields. This
observation is consistent with a radical mechanism
involving a mononuclear species such as [Mn(CO)4]•. The
intervention of such a species at 140 °C is conceivable
since both D(Mn-Mn) and D(Mn-CO) are low and
approximately equal at 36-38 kcal mol-1.9 After coor-
dination of [Mn(CO)4]• at P, a further loss of CO and
the expulsion of a phenyl radical would complete the
process. Of course, this mechanism is purely speculative
at the moment. Second, the functional phosphacyman-
trene 2 is extremely sensitive toward nucleophiles. It
is necessary to use a preacidified silica gel column to
obtain satisfactory results during the purification pro-
cess. Conversely, the resistance of 2 toward acidic media
is quite good. The acid hydrolysis of 2 at 100 °C indeed
affords the corresponding carboxylic acid 3 in good yield.
The sensitivity of 2 toward nucleophiles combined
with the relative inertness of the carboxylate function-
ality led us to investigate the synthesis of other func-
tional phosphacymantrenes. A transposition of the
synthesis of 2 with the 2-formylphosphole proved un-
successful. We reasoned that the failure of the Vilsmeier
formylation with phosphacymantrenes was due to the
Resu lts a n d Discu ssion
We first decided to investigate the reaction of
[Mn2(CO)10] with some 2-functional 1-phenylphospholes.
Our first choice was the readily available 1-phenyl-2-
ethoxycarbonyl-3,4-dimethylphosphole 1.8 A careful
(1) The main application of such phosphines is certainly the produc-
tion of the herbicide (S)-metolachlor in quantities of 10 000 tons per
year; see: Togni, A. Angew. Chem., Int. Ed. Engl. 1996, 35, 1475.
(2) (a) Ganter, C.; Brassat, L.; Glinsbo¨ckel, C.; Ganter, B. Organo-
metallics 1997, 16, 2862. (b) Ganter, C.; Brassat, L.; Ganter, B. Chem.
Ber./ Recl. 1997, 130, 1771. (c) Ganter, C.; Brassat, L.; Ganter, B.
Tetrahedron: Asymmetry 1997, 8, 2607. (d) Ganter, C.; Glinsbo¨ckel,
C.; Ganter, B. Eur. J . Inorg. Chem. 1998, 1163.
(3) (a) Garrett, C. E.; Fu, G. C. J . Org. Chem. 1997, 62, 4534. (b)
Qiao, S.; Hoic, D. A.; Fu, G. C. Organometallics 1998, 17, 773. (c) Qiao,
S.; Fu, G. C. J . Org. Chem. 1998, 63, 4168.
(4) de Lauzon, G.; Deschamps, B.; Fischer, J .; Mathey, F.; Mitschler,
A. J . Am. Chem. Soc. 1980, 102, 994.
(5) Kudis, S.; Helmchen, G. Angew. Chem., Int. Ed. Engl. 1998, 37,
3047.
(6) Mathey, F. Tetrahedron Lett. 1976, 46, 4155.
(7) Mathey, F.; Mitschler, A.; Weiss, R. J . Am. Chem. Soc. 1978,
100, 5748.
(8) Deschamps, E.; Mathey, F. Bull. Soc. Chim. Fr. 1992, 129, 486.
(9) Goodman, J . L.; Peters, K. S.; Vaida, V. Organometallics 1986,
5, 815.
10.1021/om9904596 CCC: $18.00 © 1999 American Chemical Society
Publication on Web 11/25/1999

Access to R-Functional Phosphacymantrenes
Organometallics, Vol. 18, No. 26, 1999 5689
13C, and 31P NMR spectra were recorded on a Bruker AC 200
SY spectrometer operating at 200.13, 50.32, and 81.01 MHz,
respectively. All chemical shifts are reported in ppm downfield
from internal TMS (1H and 13C) and external 85% H3PO4 (31P).
Mass spectra (EI) were obtained at 70 eV by the direct inlet
method.
2-Eth oxyca r bon yl-3,4-d im eth ylp h osp h a cym a n tr en e 2.
Phosphole 1 (1 g, 3.8 × 10-3 mol) and [Mn2(CO)10] (0.75 g, 1.9
× 10-3 mol) were heated at 140 °C for 1.5 h in 15 mL of xylene.
After cooling and evaporation of the solvent, the residue was
chromatographed on silica gel (Merck 60, 0.063-0.200 mm).
Before use, the silica gel was treated by a HCl solution in ether
in order to destroy the basic sites. Elution with hexane, then
with 70:30 hexane/dichloromethane, afforded 2 as a yellow oil
F igu r e 1. X-ray crystal structure of 5. Selected bond
lengths (Å) and angles (deg.): P(1)-C(1) 1.767(5),
P(1)-C(4) 1.792(5), C(1)-C(2) 1.382(7), C(2)-C(3)
1.443(6), C(3)-C(4) 1.420(6), C(4)-C(5) 1.456(7), O(1)-C(5)
1.190(7), Mn(1)-P(1) 2.420(1), Mn(1)-P(2) 2.254(1),
Mn(1)-CO 1.777(5) and 1.784(4); C(1)-P(1)-C(4) 86.8(2),
OC-Mn(1)-CO 91.2(2), P(2)-Mn(1)-CO 90.1(1) and
90.3(1), P(2)-Mn(1)-P(1) 97.61(5).
1
in 50-70% yields (0.6-0.85 g). 31P NMR (CDCl3): δ -23. H
NMR (CDCl3): δ 1.28 (t, 3H, CH3(Et)), 2.17 (s, 3H, Me), 2.46
2
(s, 3H, Me), 4.16 (m, 2H, CH2(Et)), 4.69 (d, 1H, J H-P ) 32.1
Hz, CH-P). 13C NMR (CDCl3): δ 13.56 (s, CH3(Et)), 14.03 (s,
CH3), 15.48 (s, CH3), 60.97 (s, CH2(Et)), 95.59 (d, 1J C-P ) 61.2
high electron-withdrawing power of the [Mn(CO)3] com-
plexing group. We supposed that the substitution of one
CO by a phosphine ligand would increase the sensitivity
of the phospholyl toward electrophilic attack. Indeed, a
Vilsmeier formylation was successfully performed with
the substituted phosphacymantrene 410 (eq 2).
1
Hz, dC-COOEt), 99.40 (d, J C-P ) 62.0 Hz, P-C-H), 118.80
(d, 2J C-P ) 7.9 Hz, C-CH3), 115.70 (d, 2J C-P ) 5.8 Hz, C-CH3),
167.10 (d, 2J C-P ) 18.5 Hz, COOEt), 222.54 (s, CO). Mass: m/z
323 (M+ + 1, 14), 322 (M+, 13), 266 (M+ - 2CO, 10), 238 (M+
- 3CO, 51.6), 194 (M+ + H - 3CO - OEt, 100). Anal. Calcd
for C12H12MnO5P: C, 44.72; H, 3.73. Found: C, 46.09; H, 3.73.
3,4-Dim eth ylp h osp h a cym a n tr en e-2-ca r boxylic Acid 3.
Ester 2 (1 g, 3.1 × 10-3 mol) was heated at 100 °C for 2 h in
a mixture of formic acid (8 mL) and sulfuric acid (2 mL). After
cooling, the crude mixture was extracted with chloroform. After
evaporation, the acid 3 was precipitated in toluene; 0.45-0.65
g was isolated (yield 50-70%). 31P NMR (CD3OD): δ -22.2.
1H NMR (CD3OD): δ 2.17 (s, 3H, CH3), 2.44 (s, 3H, CH3), 4.88
(d, 1H, 2J H-P ) 35.3 Hz, P-C-H). 13C NMR (CD3OD): δ 13.77
(s, CH3), 15.55 (s, CH3), 97.20 (d, 1J C-P ) 61.0 Hz, C-COOH),
1
2
100.57 (d, J C-P ) 61.5 Hz, P-C-H), 114.32 (d, J C-P ) 13.1
2
Hz, C-CH3), 117.27 (d, J C-P ) 5.9 Hz, C-CH3), 170.18 (d,
The crystal structure of 5 is shown in Figure 1.
The aldehyde group is almost coplanar with the phos-
pholyl ring (P(1)-C(4)-C(5)-O(1) dihedral angle ) 2.1°)
and the carbonyl is syn to the ring phosphorus. The
Ph3P-Mn(1)-centroid plane bisects the P(1)-C(1) un-
substituted ring bond in order to minimize steric repul-
sion. Otherwise, the structure is very similar to that
already published for a 2-benzoylphosphacymantrene.7
As expected, the aldehyde is an efficient starting point
for the synthesis of other functional derivatives. Its
reduction by a stoichiometric amount of LiAlH4 at low
temperature affords the alcohol 6, whereas an excess
leads to the methyl derivative 7 (eq 3).
2J C-P ) 22.8 Hz, COOH), 221.11 (s, CO). IR (decalin): ν max
2027.8, 1960.6, 1950.3 cm-1 (CO). Mass: m/z 294 (M+, 11.7),
238 (M+ - 2CO, 11.7), 237 (M+ -2CO -H), 8.5), 210 (M+
-3CO, 100). Anal. Calcd for C10H8MnO5P: C, 40.81; H, 2.72.
Found: C, 39.98; H, 2.72.
[(η5-2-F or m yl-3,4-d im eth ylp h osp h olyl)(tr ip h en ylp h os-
p h in e)]m a n ga n esed ica r bon yl 5. A mixture of 410 (1.0 g,
2 × 10-3 mol), freshly distilled POCl3 (0.4 mL), and N-methyl-
N-phenylformamide (0.5 g) in dichloromethane (10 mL) was
heated at 55 °C for 2 h. After hydrolysis and neutralization
with CO3Na2, the CH2Cl2 layer was dried over magnesium
sulfate and evaporated. The residue was chromatographed
on silica gel with hexane/CH2Cl2 (70:30) as the eluent; 0.75 g
of 5 was isolated as
a
orange powder (yield 70%). 31P
NMR (CDCl3): δ 86.0 (s, PPh3), -19.8 (s, cyclic P). 1H NMR
(CD2Cl2) δ 1.94 (s, 3H, CH3), 2.36 (s, 3H, CH3), 4.55 (dd, 1H,
The mechanism of formation of 7 is presently unclear.
With functional derivatives such as 2 and 5, we are now
able to develop the chemistry of phosphacymantrenes.
4
3
2J H-P )35.8 Hz, J H-P ) 2.4 Hz, dC-H), 8.84 (d, J H-P ) 4.2
Hz, CHO). 13C NMR (CD2Cl2) δ 13.80 (s, CH3), 15.36 (s, CH3),
1
2
101.52 (d, J C-P ) 59.9 Hz, C-P cycle), 108.82 (d, J C-P ) 4.6
Hz, C-CH3), 112.83 (d, 2J C-P )7.6 Hz, C-CH3), 192.82 (d, 2J C-P
) 26.70 Hz, CHO), 230.40 (m, CO). Mass: m/z 512 (M+, 2.5),
456 (M+ - 2CO, 100), 194 (M+ - Ph3P - 2CO, 24.4). Anal.
Calcd for C27H23MnO3P2: C, 63.29; H, 4.52. Found: C, 62.50;
H, 4.92.
Exp er im en ta l Section
All reactions were performed under nitrogen; the solvents
were purified, dried, and degassed by standard techniques. 1H,
(10) Bre`que, A.; Mathey, F., J . Organomet. Chem. 1978, 144, C9.

5690 Organometallics, Vol. 18, No. 26, 1999
Deschamps et al.
[(η5-2-Hyd r oxym eth yl-3,4-d im eth ylp h osp h olyl)(tr ip h -
en ylp h osp h in e)]m a n ga n esed ica r bon yl 6. To a solution of
5 (1.0 g, 1.95 × 10-3 mol) in THF (10 mL) cooled at -50 °C
was added with a syringe 0.3 mL (0.5 × 10-3 mol) of a 1 M
solution of LiAlH4 in THF. The reaction mixture was allowed
to warm to room temperature and hydrolyzed. After extraction
with CH2Cl2 and chromatography on silica gel with 90:10
CH2Cl2/AcOEt, 0.6-0.8 g of 6 was isolated as a yellow powder
(yield 80%). 31P NMR (CH2Cl2): δ 89.9 (s, PPh3), -43.3 (s, cyclic
Ph). 13C NMR (CDCl3): δ 11.91 (s, CH3), 14.61 (d, 2J C-P )25.9
1
Hz, CH3), 15.20 (s, CH3), 95.39 (d, J C-P ) 59.4 Hz, H-C-P
cycle), 102.93 (d, 2J C-P ) 4.0 Hz, C-CH3), 109.96 (d, 2J C-P )6.1
1
Hz, C-CH3), 112.08 (d, J C-P ) 55.3 Hz, P-C-CH3), 231.0
(m, CO). Mass: m/z 498 (M+, 3), 442 (M+ - 2CO, 100).
Cr ysta llogr a p h ic Da ta for 5. C27H23MnO3P2: M ) 512.33
g/mol; monoclinic; space group P21/n; a ) 10.1400(6) Å, b )
16.0040(8) Å, c ) 14.6200(7) Å, â ) 97.532(3)°, V ) 2352.1(2)
Å3; Z ) 4; D ) 1.447 g cm-3; m ) 0.725 cm-1; F(000) ) 1056.
Crystal dimensions 0.18 × 0.14 × 0.10. Total reflections
collected 17480 and 4215 with I > 2σ(I). Goodness of fit on F2
1.094; R(I > 2σ(I)) ) 0.0665, wR2 ) 0.1678 (all data);
maximum/minimum residual density 2.219(0.095)/-0.947-
(0.095) e Å-3. Data were collected on a KappaCCD diffracto-
meter at 150(1) K with Mo KR radiation (λ ) 0.71073 Å). Full
details of the crystallographic analysis are described in the
Supporting Information.
1
P). H NMR δ 1.90 (s, 3H, CH3), 2.36 (s, 3H, CH3), 3.49 (dd,
2
4
1H, J H-P ) 35.1 Hz, J H-P ) 4.4 Hz, P-C-H), 4.25 (m, 2H,
CH2OH). 13C NMR δ 12.22 (s, CH3), 15.12 (s, CH3), 60.67 (d,
2J C-P ) 20.9 Hz, CH2OH), 97.51 (d, J C-P ) 59.7 Hz, H-C-P
2
cycle), 103.72 (d, 2J C-P ) 4.9 Hz, C-CH3), 111.16 (d, 2J C-P )7.0
Hz, C-CH3), 113.82 (d, 1J C-P )56.6 Hz, C-CH2OH), 231.0 (m,
CO). Mass: m/z 514 (M+, 1.0%), 456 (M+ - 2CO, 29.3). Anal.
Calcd for C27H25MnO3P2: C, 63.05; H, 4.90. Found: C, 62.68;
H, 4.76.
[(η5-2,3,4-Tr im eth ylp h osp h olyl)(tr ip h en ylp h osp h in e)]-
m a n ga n esed ica r bon yl 7. The same reaction was run with
4-fold excess of LiAlH4 . The product was eluted with hexane/
CH2Cl2 (90:10) (yield 70%). 31P NMR (CH2Cl2): δ +89.0 (PPh3),
Su p p or tin g In for m a tion Ava ila ble: For 5, tables of
additional crystal data collection and refinement parameters,
atomic coordinates, thermal parameters, and complete tables
of distances and angles are provided. This material is available
free of charge via the Internet at http://pubs.acs.org.
1
3
-45.0 (cyclic P). H NMR (CH2Cl2) δ 1.80 (d, 3H, J H-P ) 9.5
Hz, P-C-CH3), 1.91 (s, 3H, CH3), 2.00 (s, 3H, CH3), 3.23 (dd,
2
4
1H, J H-P ) 34.1 Hz, J H-P ) 4.1 Hz, P-C-H), 7.36-7.52 (m,
OM9904596