The template synthesis of triaryl functionalised 1,5,9-triphosphacyclododecane on molybdenum using organocopper reagents

Article abstract of DOI:10.1039/a900509i

Free triphenyl substituted 1,5,9-triphosphacyclododecane has been prepared via the templated reaction of 1,5,9-trichloro-1,5,9-triphosphacyclododecane on molybdenum with phenylcopper or diphenylcuprate.

Full text of DOI:10.1039/a900509i

View Article Online / Journal Homepage / Table of Contents for this issue
The template synthesis of triaryl functionalised 1,5,9-
triphosphacyclododecane on molybdenum using organocopper
reagents†
David J. Jones,a Peter G. Edwards,*a Robert P. Toozeb and Thomas Albersa
a Department of Chemistry, Cardiff University, Cardiff, PO Box 912, UK CF1 3TB.
E-mail: edwardspg@cardiff.ac.uk
b ICI Acrylics PLC, PO Box 90, Wilton, Middlesbrough, Cleveland, UK TS90 8JE
Received 19th January 1999, Accepted 15th February 1999
Free triphenyl substituted 1,5,9-triphosphacyclododecane
has been prepared via the templated reaction of 1,5,9-
trichloro-1,5,9-triphosphacyclododecane on molybdenum
with phenylcopper or diphenylcuprate.
with CI4 is rapid at RT but a stable product could not be
isolated.
Reaction of 2 with PhLi or PhMgBr leads, in general, to the
formation of an insoluble brown material, presumably poly-
meric. In one case, however, reacting 2 with PhLi in diethyl
ether at 0 ЊC led to a 20% yield of the required product
[([12]aneP3Ph3)Mo(CO)3] 4 while reaction of 2 with PhMgBr in
THF at Ϫ40 ЊC led to a small yield of an unstable product
which appeared to be fully arylated but had three inequivalent
phosphorus atoms as indicated by 31P and 1H NMR. Reaction
of 2 with Ph2Zn in refluxing THF gave no reaction after 24
hours, however, reaction of 2 with PhCu or Ph2CuLi in THF
overnight at 30 ЊC or 1 hour in refluxing THF did lead to the
required product 4 in high yield, >70%. These reactions are
readily monitored by 31P NMR spectroscopy where a new
singlet at δ 10.9 due to 4 increases at the expense of the singlet
at δ 134 due to 2.
Previous routes to the symmetrical 12-membered tritertiary
triphosphamacrocycles ([12]aneP3R3)1 do not lead to the suc-
cessful formation of aryl functionalised macrocycles and the
yields for bulky alkyl or perfluoroalkyl substituents are typically
poor.2 In view of the interest in more relatively π-acidic aryl
phosphines and their importance in catalysis and other appli-
cations, we have investigated alternative synthetic routes to the
formation of aryl functionalised P3 macrocycles based upon the
[12]aneP3 core and report preliminary observations herein.
One approach to the formation of aryl substituted phos-
phines is via transmetallation of the chlorophosphine. Reac-
tions to form the required P–Cl from secondary phosphines
generally require forcing reaction conditions and reagents
such as phosgene or phosgene equivalents.3 In addition these
reactions would not be expected to work on coordinated
phosphines due to the unavailability of the lone pair and
coordination sites in the transition state. The chlorination of
free secondary phosphines with CCl4 in the presence of Et3N
has been reported; a 5-coordinate transition state was pro-
posed.4 Surprisingly we have found that [([12]aneP3X3)-
Mo(CO)3] (where X = Cl 2 or Br 3) can be readily formed in
high yield, >80%, from the reaction of [([12]aneP3H3)Mo(CO)3]
1 with the corresponding CX4 in the presence of Et3N (Scheme
1). This facile reaction is complete within 2 hours at ambient
Oxidation of 4 with I2 and decomplexation in ethanolic
NaOH followed by extraction into toluene following our
previously reported method2 results in isolation of the free
triphenyl-macrocycle, [12]aneP3Ph3 5, according to Scheme 2.
P
P
Ph
P
Cl
P
P
P
Ph
Cl
Ph
Cl
(i) or (ii)
Mo
Mo
OC
OC
CO
CO
OC
OC
2
4
(iii)
P
P
X
H
P
P
P
P
H
X
H
CX4 / Et3N
X
Mo
Mo
Ph
P
P
P
OC
OC
Ph
Ph
CO
CO
OC
OC
5
2 X = Cl
3 X = Br
1
Scheme 2 (i) PhLi or PhMgBr–THF Ϫ40 ЊC to RT; (ii) 6PhCu or
Ph2CuLi, THF, RT for 2 hours; (iii) I2, refluxing 1,1,2-trichloroethane,
excess NaOH in MeOH, diethyl ether–H2O extraction.
Scheme 1 Halogenation of the templated trisecondarymacrocycle by
reaction with an excess of CX4.
temperature and, as the 5-coordinate transition state is not
available for the kinetically inert Mo(0) complex, this suggests
that a free radical mechanism is involved. These rates are
significantly faster than those for the free phosphines reported
in the literature where reaction times of days or weeks were
required. Bromination is faster than chlorination while reaction
Fig. 1 shows the crystal structure of 4 with molybdenum in a
distorted octahedral environment.‡ Complex 4 crystallises with
two independent molecules in the asymmetric unit differing
only in the orientation of one phenyl ring; the phenyl ring is
approximately parallel to the C3v mirror plane through the
phosphorus to which it is attached in one form and orthogonal
in the other (Fig. 1 4b).5 There is no obvious structural reason
for this orientation and it may simply be due to crystal packing
forces. Five of the six β-methylene groups of the macrocycles
are disordered over the two possible positions to give the
pseudo-chair and -boat configurations for the adjacent P–P
chelate rings with the molybdenum, e.g. P1C4C5C6P2Mo, all
† Supplementary data available: experimental details for complexes 2–5.
For direct electronic access see http://www.rsc.org/suppdata/dt/1999/
1045/, otherwise available from BLDSC (No. SUP 57501, 6 pp.) or the
RSC Library. See Instructions for Authors, 1999, Issue 1 (http://
www.rsc.org/dalton).
J. Chem. Soc., Dalton Trans., 1999, 1045–1046
1045

View Article Online
Acknowledgements
We would like to thank ICI Acrylics PLC for financial support
of this project (Strategic Research Fund).
Notes and references
‡ Crystal data for complex 4, C30H33MoO3P3, M = 630.41, T = 20 ЊC,
monoclinic, space group P21/n (no. 14), a = 9.399(4), b = 17.59(3),
c = 34.22(2) Å, β = 90.15(3)Њ, V = 5656(10) Å3, Z = 8, D = 1.481 g cmϪ3
,
µ(Mo-Kα) = 0.664 mmϪ1
, independent reflections = 14926 (Rint =
0.0968), 2θ range for data collection 3.6 < 2θ < 39.6Њ, R1 (I > 2σ) =
0.0484, wR2 (all data) = 0.0924. Equivalent 1,2- and 1,3-distances
involving the disordered atoms were restrained to be equal. Rigid-bond
and similarity restraints were used as well. Pseudo-orthorhombic twin-
ning was encountered (aЈ = Ϫa, bЈ = b, cЈ = c), contribution of minor
component 0.1954(15). CCDC reference number 186/1351. See http://
www.rsc.org/suppdata/dt/1999/1045/ for crystallographic files in .cif
format.
Fig. 1 ZORTEP plot of 4 (50% probability ellipsoids), showing the
two crystallographically independent molecules 4a and 4b. Selected
average bond distances (Å) and angles (Њ): P–C (ring) 1.817(13), Mo–P
2.493(4), Mo–C 1.933(13), C–O 1.174(13), non-bonded P–P 3.536(6);
P–Mo–P 90.37(14), C–Mo–C 90.27(5), C–Mo–P 177.38(4), R–P–Mo
116.3(4).
§ See supplementary data for experimental details: Selected data for
complex 2: δP 134.0; δH PCH2 2.33 and 2.14, PCH2CH2 2.22 and 1.57;
ν(CO) cmϪ1 1970 and 1879. Complex 3: δP 119.0; δH PCH2 2.55 and
2.27, PCH2CH2 2.21 and 1.74; ν(CO) cmϪ1 1961 and 1876. Complex 4:
δP 10.9; δH PCH2 2.06 and 1.90, PCH2CH2 2.05; δC 128.26 (dd, m-Ph,
have been drawn in the chair configuration with greatest occu-
pancy. The exception is the methylene (C38) lying in the C3v
plane opposite P4 which is found only in the chair configur-
ation. The metrical parameters are otherwise comparable to the
2-propyl analogue, [([12]aneP3Pri3)Mo(CO)3].1d
5
3JPC = 6, JPC = 3 Hz); ν(CO) cmϪ1 1917 and 1809. Compound 5,
δP Ϫ34.6; δH PCH2 2.16 and 1.93, PCH2CH2 1.73 and 1.57. Analytical
data: 2, found: C, 28.74; H, 3.41. Calc. for C12H18O3P3Cl3Mo: C, 28.51;
H, 3.59. 3, found: C, 23.10; H, 2.78. Calc. for C12H18O3P3Br3Mo: C,
22.56; H, 2.84. 4, found: C, 56.98; H, 5.32. Calc. for C30H33O3P3Mo: C,
57.15; H, 5.25. 5, found: C, 71.99; H, 7.38. Calc. for C27H33P3: C, 71.99;
H, 7.33%. Satisfactory mass spectroscopic data were obtained.
The methylene protons on both the α- and β-ring carbons are
diastereotopic and could be expected to have distinct environ-
ments while structural evidence indicates they could exist in
either the boat or chair conformation. In solution, a rapid equi-
librium between these conformations would be expected to
broaden their NMR resonances, the disorder of the β-ring
carbons in the solid state structure for 4 presumably reflects a
low energy difference between the two conformations. This
expected behaviour is seen for the new complexes and for the
free ligand where small differences in δ (NMR) are observed
for the diastereotopic α-methylene protons in 4§ and the α- and
β-methylene protons appear as broad multiplets. These reson-
ances are shifted significantly upfield in the free macrocycle 5 as
compared to 4 especially the β-methylene peaks δH 2.05 to 1.73
and 1.57.
1 (a) P. G. Edwards, J. S. Fleming and S. S. Liyanage, J. Chem. Soc.,
Dalton Trans., 1997, 193; (b) P. G. Edwards, J. S. Fleming, S. S.
Liyanage, S. J. Coles and M. B. Hursthouse, J. Chem. Soc., Dalton
Trans., 1996, 1801; (c) S. J. Coles, P. G. Edwards, J. S. Fleming and
M. B. Hursthouse, J. Chem. Soc., Dalton Trans., 1995, 4091; (d) S. J.
Coles, P. G. Edwards, J. S. Fleming and M. B. Hursthouse, J. Chem.
Soc., Dalton Trans., 1995, 1139.
2 P. G. Edwards, J. S. Fleming and S. S. Liyanage, Inorg. Chem., 1996,
35, 4563.
3 R. Rabinowitz and J. Pellon, J. Org. Chem., 1961, 26, 4623; W. A.
Henderson and S. A. Buckler, J. Am. Chem. Soc., 1960, 82, 5794;
E. Steininger, Chem. Ber., 1963, 96, 3184; A. N. Pudovik, G. V.
Romanov and V. M. Pozhidaev, IZV. Akad. Nauk. SSSR, Ser. Khim.,
1977, 9, 2172.
It is interesting to note that long range phosphorus coupling
is seen for the meta-carbon on the phenyl ring for 4 where a
doublet of doublets is seen, this is absent for the free ligand 5
where only a doublet is observed.
In conclusion we have developed a new, facile route to P–Cl
bond formation for coordinated secondary phosphines. This
simple transformation allows for alternative routes to function-
alised macrocycles. One of these routes, notably aryl/alkylation
using organocopper reagents, offers the possibility of intro-
ducing a broad range of substituents due to the extensive range
of organocopper reagents available and including options for
which alternative methods have failed. The chemistry of these
new ligands is currently being examined and this new class of
triaryl-substituted 1,5,9-triphosphacyclododecanes should
also allow a more extensive examination of structure/reactivity
relationships in their metal complexes. This study is currently
under investigation.
4 Y. A. Veits, E. G. Neganova, M. V. Filippov, A. A. Borisenko and
V. L. Foss, J. Gen. Chem. USSR (Engl. Transl.), 1961, 61, 114;
Y. A. Veits, E. G. Neganova, M. V. Filippova, A. A. Borisenko and
V. L. Foss, Zh. Obshch. Khim., 1991, 61, 130.
5 G. M. Sheldrick, SHELXS-94, Acta Crystallogr., Sect. A, 1990, 46,
467; G. M. Sheldrick, SHELXL-97, Program for the refinement of
Crystal Structures, Universität Göttingen, 1997; L. Zsolnai and
G. Huttner, ZORTEP, Universität Heidelberg, 1994; L. Guoguang,
PATTERN, Program for Drawing Diffraction Patterns, Karolinska
Institute, 1994.
Communication 9/00509I
1046
J. Chem. Soc., Dalton Trans., 1999, 1045–1046