McMurry coupling of 2-acylphosphacymantrenes: E- vs Z-stereochemistry of the 1,2-bis(phosphacymantrenyl) alkenes

Article abstract of DOI:10.1021/om020144h

The McMurry coupling of a 2-acetylphosphacymantrene with the TiCl₄/Zn couple preferentially provides the Z-alkene as a mixture of the rac and meso diastereomers. The coupling of the 2-benzoylphosphacymantrene mainly gives the E-alkene together with the 2-benzyl reduction product. The alkene stereochemistries have been established by X-ray crystal structure analyses. A titanium-carbene intermediate is suggested in the second case.

Full text of DOI:10.1021/om020144h

Organometallics 2002, 21, 2635-2638
2635
McMu r r y Cou p lin g of 2-Acylp h osp h a cym a n tr en es: E- vs
Z-Ster eoch em istr y of th e 1,2-Bis(p h osp h a cym a n tr en yl)
Alk en es
Patrick Toullec, Louis Ricard, and Franc¸ois Mathey*
Laboratoire “He´te´roe´le´ments et Coordination”, UMR CNRS 7653, DCPH, Ecole Polytechnique,
91128 Palaiseau Cedex, France
Received February 20, 2002
The McMurry coupling of a 2-acetylphosphacymantrene with the TiCl4/Zn couple
preferentially provides the Z-alkene as a mixture of the rac and meso diastereomers. The
coupling of the 2-benzoylphosphacymantrene mainly gives the E-alkene together with the
2-benzyl reduction product. The alkene stereochemistries have been established by X-ray
crystal structure analyses. A titanium-carbene intermediate is suggested in the second case.
In tr od u ction
display the E-stereochemistry,6 so we decided to inves-
tigate the McMurry coupling of the easily accessible
2-acylphosphacymantrenes in which the absence of
P-substituents might favor the Z-stereochemistry.7
A lot of work has been devoted to the synthesis of
hetero analogues of porphyrins.1 However, the phos-
phorus analogues remain conspicuous by their absence.
There are at least two difficulties that hamper the
synthesis of phosphaporphyrins. The first one is purely
structural. The C-P-C intracyclic angle in a phosphole
unit typically lies around 90°,2 whereas the correspond-
ing C-N-C angle in a pyrrole unit is less acute at ca.
110°. Besides, the covalent radius of phosphorus is
substantially larger than the covalent radius of nitro-
gen: 1.10 vs 0.70 Å. As a consequence, the exact
P-analogue of a porphyrin probably does not exist, the
central hole being insufficient to accommodate the
bigger P atoms and the more acute CPC angles of the
phosphole units. The second difficulty lies in the lack
of aromatic character of the phosphole ring,3 which
prevents the use of the classical functionalization
techniques that are employed in most of the routes to
porphyrins. Since the structural features of the phos-
phole ring suggest that monatomic linkers such as the
sp2-C units of porphyrins are inappropriate for the
synthesis of phosphaporphyrins, it is logical to investi-
gate the use of the larger 1,2-alkenediyl units.4 Thus
we were led to investigate possible routes to Z-1,2-bis-
(2-phospholyl)alkenes. The McMurry coupling5 of 2-
acylphosphole derivatives was an obvious choice. A
preliminary report on the coupling of 2-formylphos-
pholes stated that the resulting alkenes essentially
Resu lts a n d Discu ssion
Our starting products were the readily available
2-acetyl (1) and 2-benzoyl (2) derivatives.8 The coupling
of the 2-acetylphosphacymantrene 1 was carried out in
refluxing THF with a TiCl4/Zn couple (eq 1).
The formation of the four expected isomers (meso, rac;
Z, E) was monitored by 31P NMR of the crude reaction
mixture after removal of the inorganic byproducts: 3a :
δ
31P -22.2 (major); 3b: δ 31P -26.1 (minor); 3c: δ 31P
-28.2 (major); 3d : δ 31P -28.4 (minor). The two major
isomers were recovered as pure products by chroma-
tography. The first eluted species was 3c, isolated in
45% yield. In the mass spectrum (EI, 70 eV), a small
molecular peak appears at m/z 552, and the base peak
at m/z 329 corresponds to the loss of 6 CO and 1 Mn.
(1) Latos-Grazynski, L. In The Porphyrin Handbook; Kadish, K. M.,
Smith, K. H., Guilard, R., Eds.; Academic Press: San Diego, 2000; Vol.
2, pp 361-416.
(2) Phosphole, C. P.; McPhail, A. T. J . Chem. Soc., Dalton Trans.
1973, 1888. Phospholide, D. T.; Theopold, K. H. Angew. Chem., Int.
Ed. Engl. 1989, 28, 1367.
(3) Quin, L. D. In Phosphorus-Carbon Heterocyclic Chemistry: The
Rise of a New Domain; Mathey, F., Ed.; Pergamon: Oxford, 2001; pp
219-362. Mathey, F. Chem. Rev. 1988, 88, 429.
(4) Expanded tetraoxaporphyrins with 1,2-ethenediyl linkers have
been recently described: Vogel, E.; J ux, N.; Do¨rr, J .; Pelster, T.; Berg,
T.; Bo¨hm, H.-T.; Lex, J .; Bremm, D.; Hohlneicher, G. Angew. Chem.,
Int. Ed. 2000, 39, 1101.
(6) Niemi, T.-A.; Coe, P. L.; Till, S. J . J . Chem. Soc., Perkin Trans.
1 2000, 1519.
(7) The McMurry coupling of 2-acetylcymantrene is known to give
predominantly the Z-alkene, see: Pitter, S.; Huttner, G.; Walter, O.;
Zsolnai, L. J . Organomet. Chem. 1993, 454, 183. See also: Top, S.;
Kaloun, E. B.; Toppi, S.; Herrbach, A.; McGlinchey, M. J .; J aouen, G.
Organometallics 2001, 20, 4554.
(5) Reviews: McMurry, J . E. Chem. Rev. 1989, 89, 1513. Fu¨rstner,
A.; Bogdanovic, B. Angew. Chem., Int. Ed. Engl. 1996, 35, 2442.
(8) Mathey, F. Tetrahedron Lett. 1976, 4155. Mathey, F.; Mitschler,
A.; Weiss, R. J . Am. Chem. Soc. 1978, 100, 5748.
10.1021/om020144h CCC: $22.00 © 2002 American Chemical Society
Publication on Web 05/30/2002

2636 Organometallics, Vol. 21, No. 13, 2002
Toullec et al.
F igu r e 1. ORTEP drawing of one molecule of 3c. Selected
bond distances (Å) and angles (deg): Mn(1)-P(1) 2.3845-
(8), P(1)-C(1) 1.757(2), P(1)-C(4) 1.787(2), C(1)-C(2)
1.409(3), C(2)-C(3) 1.430(3), C(3)-C(4) 1.417(3), C(4)-C(5)
1.491(3), C(5)-C(6) 1.345(3), C(6)-C(7) 1.493(3); C(1)-
P(1)-C(4) 89.0(1), C(4)-C(5)-C(6) 119.7(2), C(5)-C(6)-
C(7) 120.3(2).
F igu r e 3. ORTEP drawing of one molecule of 4. Selected
bond distances (Å) and angles (deg): Mn(1)-P(2) 2.3745-
(6), P(2)-C(1) 1.758(2), P(2)-C(4) 1.787(2), C(1)-C(2)
1.406(3), C(2)-C(3) 1.425(2), C(3)-C(4) 1.424(2), C(4)-C(5)
1.497(2), C(5)-C(5a) 1.352(3); C(1)-P(2)-C(4) 88.90(8),
C(4)-C(5)-C(5a) 120.3(2).
ditions but a longer reaction time, the reaction led to a
major coupled isomer 4 plus a reduction product 5 (eq
2).
F igu r e 2. ORTEP drawing of one molecule of 3a . Selected
bond distances (Å) and angles (deg): Mn(1)-P(1) 2.3859-
(8), P(1)-C(1) 1.763(2), P(1)-C(4) 1.795(2), C(1)-C(2)
1.402(3), C(2)-C(3) 1.438(2), C(3)-C(4) 1.422(2), C(4)-C(5)
1.491(2), C(5)-C(6) 1.346(2), C(6)-C(7) 1.490(2); C(1)-
P(1)-C(4) 88.75(8), C(4)-C(5)-C(6) 120.8(2), C(5)-C(6)-
C(7) 121.2 (1).
The 13C NMR spectrum shows the sp2-carbons of the
bridge as a pseudotriplet at 131.48 with ∑J CP ) 16.5
Hz. The stereochemistry was established by X-ray
crystal structure analysis (Figure 1). The CdC bridge
displays the desired Z-stereochemistry. The two phos-
phacymantrene rings lie almost parallel in a head-to-
tail disposition corresponding to the rac-diastereomer
with the two Mn(CO)3 moieties oriented away from each
other in order to minimize steric crowding. The CdC
double bond is well localized at 1.345(3) Å and shows
no distortion with all the C-CdC angles close to 120°.
The second major isomer, 3a , displays spectral charac-
teristics similar to 3c. In particular, the carbons of the
The mass spectrum of 4 (EI) displays a small (M + 2)
peak at m/z 678 and a base peak at m/z 454 correspond-
ing to the loss of 6 CO and 1 Mn. In the 13C spectrum,
the carbons of the bridge appear as a multiplet at
140.20. The X-ray crystal structure analysis (Figure 3)
demonstrates that 4 displays a center of symmetry and
a E-stereochemistry. No distortion of the CdC bond of
the bridge is visible. The structure of the reduction
1
product 5 was deduced from mass and H and 13C NMR
spectroscopy. The formation of the reduction product
and the change of stereochemistry suggest a change of
mechanism in the McMurry coupling of the benzoyl
derivative 2 when compared to 1. In line with recent
proposals on the coupling mechanism of hindered
ketones,9 we suggest a titanium-carbene intermediate
(TidC) whose hydrolysis would rationalize the forma-
tion of 5. The reasons behind the change of stereochem-
istry are less clear, but it must be recalled that the
preferential Z-stereochemistry of the coupling products
derived from 1 is consistent with those observed in the
coupling of acetylcymantrenes.7 Finally, it must be
bridge appear as a pseudotriplet at 130.76 with ∑J CP
)
18 Hz. Once again, the stereochemistry was established
by X-ray crystal structure analysis (Figure 2). The Cd
C bridge displays the Z-stereochemistry; the two phos-
phacymantrene rings are approximately parallel with
a head-to-head disposition corresponding to the meso-
diastereomer. The structural parameters are very simi-
lar to those of 3c.
The findings were entirely different with the 2-ben-
zoylphosphacymantrene 2. Under similar reaction con-
(9) Villiers, C.; Ephritikhine, M. Chem. Eur. J . 2001, 7, 3043.

1,2-Bis(phosphacymantrenyl) Alkenes
Organometallics, Vol. 21, No. 13, 2002 2637
carbon); 224.66 (s, CO). MS: m/z 524 (M+ - CO, 12%); 468
(M+ - 3CO, 18%); 440 (M+ - 4CO, 3%); 384 (M+ - 6CO, 52%);
329 (M+ - 6CO - Mn, 100%). Anal. Calcd for C22H20O6P2Mn2:
C, 47.83; H, 3.62. Found: C, 47.41; H, 3.94.
stressed that efficient decomplexation techniques are
available for the conversion of phosphacymantrenes into
phospholes,10 thus providing a solution to the synthetic
problems outlined in the Introduction. However, as
stated by one of the referees,11 the X-ray data of 3a and
3c suggest that, in a coplanar conformation, the two
phosphorus atoms of a Z-1,2-bis(phospholyl)alkene unit
would be only about 1.43 Å apart. This is too close for
accommodating a metal. Thus probably, other linkers
than the Z-1,2-alkenediyl are still to be found for the
successful synthesis of phosphaporphyrins.
McMu r r y Cou p lin g of 2-Ben zoyl-3,4-d im eth ylp h os-
p h a cym a n tr en e. A 1.17 g sample of zinc dust (18 mmol) was
placed in a 150 mL Schlenk tube under nitrogen atmosphere,
and 50 mL of distilled THF was added. The Schlenk tube was
cooled to 0 °C, and 1 mL of titanium tetrachloride (9 mmol)
was added slowly. The resulting solution was refluxed for 2
h, then 0.9 g (2.55 mmol) of 2-acetyl-3,4-dimethylphosphacy-
mantrene dissolved in 20 mL of distilled THF was added. The
solution was refluxed for 48 h. The solvents were removed
under reduced pressure, and the crude mixture was redis-
solved in 50 mL of diethyl ether. The solution was treated with
2 × 50 mL of saturated sodium hydrogen carbonate aqueous
solution and washed with 50 mL of water. The organic layer
was collected, dried with magnesium sulfate, and filtered, and
the solvents were removed under vacuum to give an orange
oil that partly crystallized. Chromatography over silica gel
with a 97:3 hexane/ether mixture as the eluant first gave 250
mg of 5, then 350 mg of 4.
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All reactions were performed
under an inert atmosphere with dry, deoxygenated solvents
by using vacuum line and Schlenk tube techniques. 2-Acetyl-
3,4-dimethylphosphacymantrene and 2-benzoyl-3,4-dimeth-
ylphosphacymantrene were prepared according to known
procedures.8 All other reagents are commercial grade and used
as received. NMR spectra were measured on a Bruker 300
MHz multinuclear spectrometer. Chemical shifts are expressed
in ppm from internal TMS (1H and 13C) or external 85% H3-
PO4 (31P); couplings constants are expressed in Hz. Mass
spectra (electron impact) were measured at 70 eV by the direct
inlet method. Elemental analyses were performed at the
Service de Microanalyse du CNRS, Gif-sur-Yvette, France.
McMu r r y Cou plin g of 2-Acetyl-3,4-dim eth ylph osph acy-
m a n tr en e. A 1.57 g sample of zinc dust (24 mmol) was placed
in a 150 mL Schlenk tube under nitrogen atmosphere, and 50
mL of distilled THF was added. The Schlenk tube was cooled
to 0 °C, and 1.3 mL of titanium tetrachloride (12 mmol) was
added slowly. The resulting solution was refluxed for 2 h, then
1.16 g (4 mmol) of 2-acetyl-3,4-dimethylphosphacymantrene
dissolved in 20 mL of distilled THF was added. The solution
was refluxed for 18 h. The solvents were removed under
reduced pressure, and the crude mixture was redissolved in
70 mL of diethyl ether. The solution was treated with 2 × 50
mL of saturated sodium hydrogen carbonate aqueous solution
and washed with 50 mL of water. The organic layer was
collected, dried with magnesium sulfate, and filtered, and the
solvents were removed under vacuum to give 1.03 g of a yellow
crystalline solid as a mixture of the four isomers of the final
products. Chromatography over silica gel with a 97:3 hexane/
ether mixture as the eluent was carried out, yielding 490 mg
of 3c first, then 300 mg of 3a . Total yield of the two Z-isomers:
72.5%.
2-Ben zyl-3,4-d im eth ylp h osp h a cym a n tr en e (5). 31P{1H}
NMR (CH2Cl2): δ -38.6. 1H NMR (CDCl3): δ 2.06 (s, 3H); 2.14
(s, 3H); 3.39 (d, J ) 1.6 Hz, 1H); 3.44 (s, 1H); 4.30 (d, J ) 35
3
Hz, 1H); 7.2-7.3 (m, 5H). C NMR (CDCl3): δ 13.30 (s); 16.55
(s); 36.10 (d, J ) 18.9 Hz, CH2); 94.07 (d, J ) 58.9 Hz, C5′);
110.00 (d, J ) 6.8 Hz, C4′); 113.51 (d, J ) 7.5 Hz, C3′); 120.71
(d, J ) 60.4 Hz, C2′); 127.32 (s, Cpara); 129.12 (s, Cortho); 129.27
(s, Cmeta); 140.50 (d, J ) 1.5 Hz, Cipso); 224.87 (s, CO). MS: m/z
339 (M+ - 1, 9%); 283 (M+ - 1 - 2CO, 4%); 255 (M+ - 1 -
3CO, 100%); 201 (M+ - 3CO - Mn, 14%).
r a c-(E)-1,2-Bis[3′,4′-d im eth ylp h osp h a cym a n tr en -2′-yl]-
1,2-d ip h en yleth en e (4). 31P{1H} NMR (CH2Cl2): δ -19.3. 1H
NMR (CDCl3): δ 1.89 (s, 12H); 4.13 (d, J ) 35.4 Hz, 2H); 7.05-
7.25 (m, 10H). 3C NMR (CDCl3): δ 15.05 (s); 16.18 (s); 93.80
(d, J ) 63.4 Hz, C5′); 108.05 (d, J ) 6.8 Hz, C4′); 115.00 (d, J
) 3 Hz, C3′); 119.14 (d, J ) 57.4 Hz, C2′); 127.78 (s, Cortho);
128.45 (s, Cmeta); 130.25 (s, Cpara); 140.20 (m, ethylenic carbon);
144.62 (s, Cipso); 224.12 (s, CO). MS: m/z 678 (M+ + 2H, 2%);
650 (M+ + 2H - CO, 10%); 592 (M+ - 3CO, 70%); 509 (M+
-
6CO + H, 55%); 454 (M+ - 6CO - Mn, 100%). Anal. Calcd for
C
32H24O6P2Mn2: C, 56.80; H, 3.55. Found: C, 56.68; H, 3.77.
X-r a y Cr ysta l Str u ctu r es. All data were collected on a
KappaCCD diffractometer at 150.0(1) K with Mo KR radiation
(λ ) 0.71073 Å). Full details of the crystallographic analysis
are described in the Supporting Information.
Crystallographic data for C22H20Mn2O6P2, 3a : M ) 552.20
g/mol; triclinic; space group P1h; a ) 7.558(5) Å, b ) 12.384(5)
Å, c ) 13.786(5) Å, R ) 108.110(5)°, â ) 98.760(5)°, γ )
104.430(5)°, V ) 1150.1(10) Å3; Z ) 2; D ) 1.595 g cm-3; µ )
1.273 cm-1; F(000) ) 560. Crystal dimensions 0.20 × 0.14 ×
0.10 mm. Total reflections collected 9356 and 5650 with I >
2σ(I). Goodness of fit on F2 1.018; R(I > 2σ(I)) ) 0.0326, wR2
) 0.0914 (all data); maximum/minimum residual density
r a c-(Z)-2,3-Bis[3′,4′-d im eth ylp h osp h a cym a n tr en -2′-yl]-
bu t-2-en e (3c). 31P{1H} NMR (CH2Cl2): δ -28.2. 1H NMR
(CDCl3): δ 1.82 (s, 6H, ethylenic CH3); 2.10 (s, 6H, CH3-C4′);
2.17 (s, 6H, CH3-C3′); 4.67 (dm, J ) 34.5 Hz; 2H). 3C NMR
(CDCl3): δ 13.77 (s, CH3-C4′); 16.57 (s, CH3-C3′); 26.16 (s,
ethylenic CH3); 92.25 (dm, J ) 68.7 Hz, C5′); 106.87 (s, C4′);
116.45 (s, C3′); 121.37 (dm, J ) 52.8 Hz, C2′); 131.48 (pseudo-
triplet, J ) 16.5 Hz; ethylenic carbon); 224.92 (s, CO). MS:
m/z 552 (M+, 1%); 524 (M+ - CO, 16%); 468 (M+ - 3CO, 18%);
440 (M+ - 4CO, 13%); 384 (M+ - 6CO, 75%); 329 (M+ - 6CO
- Mn, 100%). Anal. Calcd for C22H20O6P2Mn2: C, 47.83; H, 3.62.
Found: C, 47.29; H, 3.91.
0.927(0.068)/-0.525(0.068) e Å-3
.
Crystallographic data for C22H20Mn2O6P2, 3c: M ) 552.20
g/mol; monoclinic; space group P21/n; a ) 7.946(5) Å, b )
12.668(5) Å, c ) 23.622(5) Å, â ) 92.610(5)°, V ) 2375.3(18)
Å3; Z ) 4; D ) 1.544 g cm-3; µ ) 1.233 cm-1; F(000) ) 1120.
Crystal dimensions 0.20 × 0.20 × 0.20 mm. Total reflections
collected 16 064 and 6196 with I > 2σ(I). Goodness of fit on F2
1.032; R(I > 2σ(I)) ) 0.0255, wR2 ) 0.0733 (all data);
maximum/minimum residual density 0.400(0.074)/-0.456-
m eso-(Z)-2,3-Bis[3′,4′-d im et h ylp h osp h a cym a n t r en -2′-
1
yl]bu t-2-en e (3a ). 31P{1H} NMR (CH2Cl2): δ -22.2. H NMR
(CDCl3): δ 1.78 (s, 6H); 1.79 (s, 6H); 2.06 (s, 6H, ethylenic CH3);
4.46 (dm, J ) 35.0 Hz, 2H). 3C NMR (CDCl3): δ 14.57 (s, CH3-
C4′); 16.46 (s, CH3-C3′); 27.27 (s, ethylenic CH3); 95.95 (dm, J
) 64.1 Hz, C5′); 108.74 (s, C4′); 114.04 (s, C3′); 121.89 (dm, J )
57.4 Hz, C2′); 130.76 (pseudotriplet, J ) 18.2 Hz, ethylenic
(0.074) e Å-3
.
Crystallographic data for C32H24Mn2O6P2, 4: M ) 676.33
g/mol; monoclinic; space group P21/n; a ) 11.1367(2) Å, b )
12.6396(4) Å, c ) 11.4394(3) Å, â ) 114.2360(10)°, V ) 1468.32-
(7) Å3; Z ) 2; D ) 1.530 g cm-3; µ ) 1.013 cm-1; F(000) ) 688.
Crystal dimensions 0.16 × 0.16 × 0.16 mm. Total reflections
collected 7036 and 3272 with I > 2σ(I). Goodness of fit on F2
(10) Deschamps, B.; Toullec, P.; Ricard, L.; Mathey, F. J . Organomet.
Chem. 2001, 634, 131.
(11) We are grateful to Prof. Michael McGlinchey for this remark.

2638 Organometallics, Vol. 21, No. 13, 2002
Toullec et al.
1.059; R(I > 2σ(I)) ) 0.0384, wR2 ) 0.1063 (all data);
maximum/minimum residual density 0.980(0.077)/-0.519-
Su p p or tin g In for m a tion Ava ila ble: Listings of atomic
coordinates, including equivalent isotropic displacement pa-
rameters, bond lengths, and bond angles of 3a , 3c, and 4. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(0.077) e Å-3
.
Ack n ow led gm en t. The authors thank BASF for the
financial support of Patrick Toullec and of this research
program.
OM020144H