Reactivity of organozinc derivatives of phosphinines

Article abstract of DOI:10.1021/om9506936

The reactions of the 2-organozinc derivatives 1, 2, and 3 of 4,5-dimethylphosphinine with electrophiles of phosphorus, arsenic, tin, copper, silver, and mercury are discussed. The reactions of 2 and 3 with arsenic trichloride furnished the monosubstitution products 4 and 5, respectively, which were not isolable in pure form. Reactions of 2 and 3 with phosphorus electrophiles were partially successful, but the monosubstitution products 8 and 11, respectively, were not formed selectively; among the byproducts were 7, 9, and 10. Transmetalation reactions with tin electrophiles gave 13 (37%) and 16 (7%), which were isolated in pure form. Transmetalation of 2 with a copper salt furnished copper derivative 17, which was trapped below -80 °C with a Michael acceptor and chlorotrimethylsilane to give the 1,4-addition product 19. This compound was obtained in nearly pure form in a yield of 20% relative to the 2-iodophosphinine 10. Transmetalation of 3 with a copper salt furnished the copper derivative 20, which decomposed within 6 h at room temperature. This compound was also prepared via direct insertion of reactive Cu(0) into 12. Transmetalations with silver salts afforded the organosilver derivatives 21-21″. These silver species turned out to be rather unstable at room temperature. Transmetalation reactions with mercury salts were successful with 1, 2, and 3. The organomercury derivative 22 obtained via 1 and 2, respectively, decomposed at room temperature, but was characterized by ¹H, ¹³C, and ³¹P NMR spectroscopies. The relatively stable organomercury derivative 23, obtained from 3, was isolated in a yield of 50% relative to 10.

Full text of DOI:10.1021/om9506936

802
Organometallics 1996, 15, 802-808
Rea ctivity of Or ga n ozin c Der iva tives of P h osp h in in es
Herman T. Teunissen and Friedrich Bickelhaupt*
Scheikundig Laboratorium, Vrije Universiteit, De Boelelaan 1083,
1081 HV Amsterdam, The Netherlands
Received September 1, 1995X
The reactions of the 2-organozinc derivatives 1, 2, and 3 of 4,5-dimethylphosphinine with
electrophiles of phosphorus, arsenic, tin, copper, silver, and mercury are discussed. The
reactions of 2 and 3 with arsenic trichloride furnished the monosubstitution products 4 and
5, respectively, which were not isolable in pure form. Reactions of 2 and 3 with phosphorus
electrophiles were partially successful, but the monosubstitution products 8 and 11,
respectively, were not formed selectively; among the byproducts were 7, 9, and 10.
Transmetalation reactions with tin electrophiles gave 13 (37%) and 16 (7%), which were
isolated in pure form. Transmetalation of 2 with a copper salt furnished copper derivative
17, which was trapped below -80 °C with a Michael acceptor and chlorotrimethylsilane to
give the 1,4-addition product 19. This compound was obtained in nearly pure form in a
yield of 20% relative to the 2-iodophosphinine 10. Transmetalation of 3 with a copper salt
furnished the copper derivative 20, which decomposed within 6 h at room temperature. This
compound was also prepared via direct insertion of reactive Cu(0) into 12. Transmetalations
with silver salts afforded the organosilver derivatives 21-21′′. These silver species turned
out to be rather unstable at room temperature. Transmetalation reactions with mercury
salts were successful with 1, 2, and 3. The organomercury derivative 22 obtained via 1 and
1
2, respectively, decomposed at room temperature, but was characterized by H, 13C, and 31P
NMR spectroscopies. The relatively stable organomercury derivative 23, obtained from 3,
was isolated in a yield of 50% relative to 10.
Ch a r t 1
In tr od u ction
The synthesis of phosphinine derivatives, which was
first reported by Ma¨rkl in 1966,1 is usually achieved by
a multistep procedure in which the phosphinine is
prepared from nonphosphaaromatic precursors.2 Dur-
ing recent years, however, synthetic routes have been
developed in which a new phosphinine derivative is
prepared by direct functionalization of a halogen-
substituted phosphinine. Mathey et al. investigated the
synthesis and reactivity of bromo- and chlorophosphin-
ines,3 whereas our research was focused on the synthe-
sis and reactivity of iodophosphinines.4 In a previous
paper,4e we reported the synthesis of 2-iodophosphinines
and their pentacarbonyltungsten complexes; further-
more, the synthesis of the organozinc derivatives of
phosphinines 1, 2, and 3 (Chart 1) was described. These
interesting organometallic compounds were prepared
either via insertion of zinc into the carbon-iodine bond
of 2-iodo-4,5-dimethylphosphinine in DMF (1) or THF/
TMEDA (2) or, in the case of the pentacarbonyltungsten
complex 3, by reaction of 2 with (acetonitrile)pentacar-
bonyltungsten (15).
In this paper we report the reactions of the organozinc
species 1, 2, and 3 with PCl3 and AsCl3 as it is well
established that conversion of organozinc derivatives
with phosphorus5 or arsenic6 electrophiles leads to the
corresponding nucleophilic substitution products. In
addition, some transmetalations of 1, 2, and 3 have been
explored; in line with the electronegativity scale of
Allred,7 organometallic derivatives of zinc (x ) 1.6) can
in principle be used for the preparation of organo-
metallic derivatives of tin (x ) 1.96), mercury (x ) 1.9),
copper (x ) 1.9), and silver (x ) 1.9).
X Abstract published in Advance ACS Abstracts, December 15, 1995.
(1) Ma¨rkl, G. Angew. Chem. 1966, 78, 907.
(2) Ma¨rkl, G. In Multiple Bonds and Low Coordination in Phospho-
rus Chemistry; Regitz, M., Scherer, O. J ., Eds.; Thieme Verlag:
Stuttgart, 1990; pp 220-257.
(3) (a) Le Floch, P.; Mathey, F. Tetrahedron Lett. 1989, 30, 817. (b)
Le Floch, P.; Ricard, L.; Mathey, F. Polyhedron 1990, 9, 991. (c) Le
Floch, P.; Carmichael, D.; Mathey, F. Organometallics 1991, 10, 2432.
(d) Le Floch, P.; Carmichael, D.; Mathey, F. Bull. Soc. Chim. Fr. 1992,
129, 291. (e) Le Floch, P.; Carmichael, D.; Mathey, F. Phosphorus
Sulfur 1993, 76, 33. (f) Le Floch, P.; Ricard, L.; Mathey, F. J . Chem.
Soc., Chem. Commun. 1993, 789. (g) Le Floch, P.; Carmichael, D.;
Ricard, L.; Mathey, F. J . Am. Chem. Soc. 1993, 115, 10665. (h) Le
Floch, P.; Kolb, A.; Mathey, F. J . Chem. Soc., Chem. Commun. 1994,
2065. (i) Trauner, H; Le Floch, P.; Lefour, J . M.; Ricard, L.; Mathey,
F. Synthesis 1995, 717.
Resu lts a n d Discu ssion
Rea ction s w ith P h osp h or u s a n d Ar sen ic Elec-
tr op h iles. The reactions of organozinc derivatives of
phosphinines with phosphorus and arsenic electrophiles
(4) (a) Teunissen, H. T.; Bickelhaupt, F. Tetrahedron Lett. 1992, 33,
3537. (b) Teunissen, H. T.; Bickelhaupt, F. Bull. Soc. Chim. Belg. 1992,
101, 609. (c) Bickelhaupt, F. Pure Appl. Chem. 1993, 65, 621. (d)
Teunissen, H. T.; Bickelhaupt, F. Phosphorus Sulfur 1993, 76, 75. (e)
Teunissen, H. T.; Bickelhaupt, F. Organometallics 1996, 15, 794.
(5) Sasse, K. In Methoden der Organischen Chemie (Houben-Weyl);
Thieme Verlag: Stuttgart, 1963; Vol. 12, No. 1, p 308.
(6) Samaan, S. In Methoden der Organischen Chemie (Houben-Weyl);
Thieme Verlag: Stuttgart, 1978; Vol. 13, No. 8, p 59.
(7) Allred, A. L. J . Inorg. Nucl. Chem. 1961, 17, 215.
0276-7333/96/2315-0802$12.00/0 © 1996 American Chemical Society

Organozinc Derivatives of Phosphinines
Organometallics, Vol. 15, No. 2, 1996 803
Sch em e 1
Sch em e 3
Sch em e 2
of 4:1:1.5, which means that the initial purity of 2
(97.6%) had been reduced to 61.5% in 8. Mainly because
of the presence of 10 (23%), the purification of 8 was
rather difficult. Crystallizations from pentane fur-
nished yellow crystals which contained substantial
amounts of 10. The stability of 8 at room temperature
was also a problem as an NMR solution of 8 showed
slow decomposition.
The reaction of 3 with PCl3 was similarly disappoint-
ing as selective formation of a monosubstitution product
did not occur. Before the reaction of 3 with PCl3, 3, 12
1
(δ(31P) 184.6 ppm, J (PW) ) 280.4 Hz), 9, and 7 were
present in a ratio of 21.8:1.6:1.2:1; afterward, the ratio
of 11 (δ(31P) 174.2 (P(Ar)), 158.4 (P(Cl2)) 2J (PP) ) 278.1
1
were limited to 2 and 3 and carried out in THF; the
inevitable presence of DMF in the reaction mixtures
containing 1 makes this organozinc reagent unsuited
for reactions with this category of electrophiles as they
react with DMF directly. In general, a 6-16-fold excess
of freshly distilled ECl3 (E ) P, As) was mixed with THF
and cooled to -10 to -60 °C. Then 2 or 3 was added
dropwise, after which the reaction mixture was allowed
to warm to room temperature.
Hz, J (PW) ) 273.8 Hz):12:9:7, was 1.3:0.2:0.2:1. This
implies 85% relative abundance of 3 before the reaction
and 48% of 11 after the reaction. As 12 was already
present in the metallation mixture of 3, it is difficult to
ascertain whether it was also formed during the reaction
with PCl3. Nevertheless, the low relative abundance of
11 (48%) made the reaction unattractive for preparative
purposes. Altogether, it is difficult to mechanistically
explain the rather disappointing behavior of 2 and 3 in
reactions with PCl3.
Tr a n sm eta la tion s w ith Tin Der iva tives. These
reactions have been carried out predominantly with 1
and only incidentally with 2 and 3. The transmetala-
tions with 1 were focused on reactions with ClSnPh3
and Cl2Sn(CH3)2 (Scheme 3).
The reactions of 2 and 3 with AsCl3 are presented in
Scheme 1 and showed the expected monosubstitution
with the selective formation of 4 (δ(31P) 198.6 ppm) and
1
5 (δ(31P) 165.4 ppm, J (PW) ) 268.4 Hz), respectively.
The purification of 4 turned out to be difficult as the
yellow crystals, initially formed by cooling a pentane
extract to -70 °C, slowly decomposed. Therefore, the
attention was focused on the purification of 5. The
formation of 5 via 3 was selective and the ratio of 5:7
(δ(31P) 154.8 ppm, 1J (PW) ) 268.6 Hz) was 8.6:1, which
is satisfactory considering that 7 was already present
in the solution of 3 (3:7 ) 8.7:1). Crystallization of 5
by cooling a pentane extract to -70 °C did not afford 5
The reaction of 1 with ClSnPh3 at room temperature
2
afforded 13 (δ(31P) 222.0 ppm, J (P119Sn) ) 317.3 Hz);
it was the only detectable phosphorus containing prod-
uct, but it was isolated in 37% yield only. In solution,
13 decomposes slowly with concomitant formation of a
yellow precipitate. This slow decomposition explains
the moderate yield and the slight deviations in the
elemental analysis (see Experimental Section). Mathey
et al. reported similar problems with functionalized
phosphinines.3c
1
in pure form as 6 (δ(31P) 159.5 ppm, J (PW) ) 268.6
Hz) was formed by exchange reactions of 5 with ZnClI.
After several unsuccessful purification procedures, the
ratio of 5:6:7 was 188:71:21. Apart from NMR spectra,
MS evidence (see Experimental Section) supported the
identity of 5 and of the impurities.
The reaction of 1 with Cl2Sn(CH3)2 was performed to
obtain 14, which was expected to possess bidentate
coordinating properties. The isolation of 14 (δ(31P) 218.5
2
The reactions of 2 and 3 with phosphorus electro-
philes are presented in Scheme 2. The course of the
reaction of 2 with PCl3 was rather disappointing as
formation of 9 and (surprisingly!) 10 accompanied the
nucleophilic substitution on phosphorus. Before the
reaction with PCl3, 2 and 9 were present in a ratio of
41:1. After the reaction, 8 (δ(31P(Ar)) 204.8 ppm,
δ(31P(Cl2)) 165.3 ppm, 2J (PP) ) 242.4 Hz), 9 (δ(31P) 192
ppm), and 10 (δ(31P) 216.4 ppm) were present in a ratio
ppm, J (P119Sn) ) 330.8 Hz) in pure form was difficult
as decomposition with concomitant formation of a yellow
precipitate was substantially faster than in the case of
13. Therefore, 14 was transformed into its W(CO)5
complex by reaction with (acetonitrile)pentacarbonyl-
tungsten (15). In this way, 16 (δ(31P) 176.8 ppm, 1J (PW)
) 262.8 Hz, 2J (P119Sn) ) 213.5 Hz) was isolated in pure
form. It was obtained in a yield of 7% relative to Cl2Sn-
(CH3)2 and fully characterized. The low yield of 16 does

804 Organometallics, Vol. 15, No. 2, 1996
Teunissen and Bickelhaupt
Sch em e 6a
Sch em e 4
a
In structures 21, L, L′, or L′′ indicates coordination of
solvent or Lewis base, which in an unidentified fashion
depends on the system used.
reaction of 12 with in situ prepared, highly reactive
Sch em e 5
metallic copper, in analogy with a procedure developed
by Rieke.8
A
31P NMR spectrum of the crude reaction
mixture showed the presence of 20 (δ(31P) 171 ppm,
1J (PW) ) 244 Hz), 7, and a component with δ(31P) 227
ppm, in a ratio of 10.6:1:2.6. The decomposition of 20,
obtained from 3 or 12, proceeded in a similar fashion.
Tr a n sm eta la tion s w ith Silver Sa lts. Transmeta-
lations of 1, 2, and 3 with silver salts were carried out
with limited success, as the organosilver derivatives
were unstable at room temperature. Whereas the
transmetalation of 3 with the THF-soluble salt
AgBr‚2LiBr resulted in decomposition of 3, the trans-
metalations involving 1 and 2 were more successful
(Scheme 6).
The transmetalation of 1 was achieved with AgNO3
at room temperature. The solution became black, and
a grey precipitate was formed. The 31P NMR spectrum
showed, apart from small amounts of 9 and 10 in a ratio
of 1:0.29, two large, broad signals at 208 ppm (line width
100 Hz) and 194 ppm (line width 120 Hz) in a ratio of
3.4:1.0, which are assigned to two forms of 21. The
organosilver derivatives were thermally stable at -20
°C, whereas at room temperature, decomposition oc-
curred rather quickly with concomitant formation of a
grey precipitate. The formation of 21 was supported by
a quench reaction with iodine, which furnished 10; after
this quench reaction, the ratio 9:10 was 1:15.2. This
does not unambiguously prove that Ag is bonded to the
2-position of the phosphinine ring, as 1 would show the
same reaction with iodine while the new 31P NMR
signals could be due to the formation of η1-P-Ag
complexes. The arguments in favor of the formation of
21 are that silver nitrate is frequently used in trans-
metalation reactions of organozinc compounds to give
the corresponding organosilver species.9 Furthermore,
the low thermal stability of 21 suggests the presence of
an organosilver species, as 1 is thermally very stable.
A conceivable alternative for the transmetalation is the
formation of an η1-Ag complex of 1. However, η1-
coordination of a phosphinine to Ag has been reported
to give stable complexes.10
not reflect the real yield of 14 in view of the numerous
manipulations.
Tr a n sm eta la tion s w ith Cop p er Sa lts. Trans-
metalations of 2 and 3 with copper salts showed that
the organocopper derivatives were rather unstable. As
shown in Scheme 4, the organocopper phosphinine 17
was formed in situ at -80 °C. A solution of 2 was
reacted with a THF solution of CuCN‚2LiBr, which also
contained 2-cyclohexen-1-one and ClSi(CH3)3. By Michael
addition of 17, which in a separate experiment was
shown not to occur in reaction with 2 alone, the enolate
18 was formed and converted directly to the silyl ether
19 (δ(31P) 190.2 ppm) by ClSi(CH3)3. Together with 19,
an appreciable amount of 9 (Scheme 2) was formed. The
purification of 19 was difficult, not only because of the
presence of 9, but mainly due to the presence of zinc
and copper salts and their TMEDA complexes (see
Experimental Section). After purification, nearly pure
19 was obtained as a brown oil in a yield of 20% relative
to 10, the precursor of 2. The purity of 19 was not 100%;
1
the H NMR spectrum showed the presence of traces of
9 and several signals between 0 and 2 ppm not belong-
ing to 19. Therefore, the δ(1H) of the aliphatic protons
of 19 was determined from its CH COSY spectrum. The
lack of purity was also evident from the elemental
analysis, in which the experimental value for carbon
was 2.08% higher than the calculated one.
The reaction of 3 with a THF soluble copper salt at
-60 °C furnished 20 (δ(31P) 171.2 (1J (PW) ) 249.9 Hz))
which, together with 7, was present in a ratio of 13.3:1
(Scheme 5). Before the reaction, the ratio of 3 to 7 was
18.8:1. Compound 20 decomposed completely at room
temperature within 6 h; in the 31P NMR spectrum, two
new signals appeared around 55 ppm and one at 37
ppm. In an alternative approach, 20 was obtained by
The transmetalation of 2 was carried out with AgNO3
and with the THF-soluble salt AgBr‚2LiBr; it showed
remarkable differences from that of 1 (Scheme 6). Only
(8) Stack, D. E.; Dawson, B. T.; Rieke, R. D. J . Am. Chem. Soc. 1991,
113, 4672.
(9) van Koten, G.; Noltes, J . G. In Comprehensive Organometallic
Chemistry; Wilkinson, G., Ed.; Pergamon Press: New York, 1982; Vol.
2, p 709 vv.
(10) Kanter, H.; Dimroth, K. Tetrahedron Lett. 1975, 541.

Organozinc Derivatives of Phosphinines
Organometallics, Vol. 15, No. 2, 1996 805
Sch em e 7
The stability of 22, prepared from 2, was completely
different from that of 22, prepared from 1. A solution
of 2 was cooled to -60 °C; then TMEDA and finally solid
HgCl2 (0.5 equiv) were added. Next, the reaction
mixture was warmed slowly to room temperature,
during which a brown precipitate was formed. Subse-
quent 31P NMR analysis showed the presence of 22
(δ(31P) 212.8 ppm, 2J (PHg) ) 786.1 Hz) and 9 in a ratio
of 10:1. After 2 h at room temperature, this ratio had
decreased to 5:1. After the addition of pentane (THF:
pentane ) 1:1), more precipitate was formed. The
supernatant was separated and cooled to -70 °C. After
cooling, the extract showed the formation of light brown
crystals, contaminated with some grey material. The
1
crystals of 22 were analyzed by H, 13C, and 31P NMR
spectroscopies. Compound 22 decomposed slowly in the
CDCl3 solution: approximately 15% in 4 days. The 13
C
NMR spectrum of 22 strongly supported its structural
composition. The signal of C2 (194.7 ppm) was a double
doublet with 1J (PC) ) 79.4 Hz and 3J (PC) ) 7.2 Hz.
This indicates the connection of two phosphinine moi-
eties via C2 and is only compatible with the structure
assignment of 22; the conceivable alternative, i.e., the
formation of a monoarylated mercury chloride (RHgCl),
is excluded. The proton spectra of 22 showed the
presence of TMEDA signals which were deshielded
relative to a reference sample of TMEDA in CDCl3: 0.40
ppm for the methyl groups and 0.36 ppm for the
methylene groups. In view of the known differences
between zinc15 and mercury16 in their complexing ability
toward TMEDA, these signals may confidently be as-
signed to zinc salt complexes. The molar ratio of
TMEDA:22 was 1.5:1, thus indicating the presence of
an appreciable amount of zinc salts, which were also
observed in the mass spectra. Presumably these salts
play a role in the thermal decomposition of 22.
one broad signal (line width 50-70 Hz) of the organo-
silver derivatives 21′ (δ(31P) 199 ppm) and 21′′ (δ(31P)
200 ppm), respectively, was observed. Westmijze et al.
reported that the use of AgBr‚2LiBr suppresses the
decomposition of organosilver compounds.11 They sug-
gested that LiBr is built into the alkylsilver clusters by
analogy of the stabilizing influence of CuBr on some
arylcopper cluster compounds. In our case, a stabilizing
influence of LiBr on 21′′ relative to 21′ was not observed.
The ratio 21′:9 decreased within 50 min at room
temperature from 2.8:1 to 1.7:1. The analogous ratio
of 21′′:9 decreased from 3:1 to 1.9:1 under the same
conditions. The decomposition of 21′′ was followed at 4
°C by 31P NMR spectroscopy, which indicated that the
decomposition with concomitant formation of 9 was
complete within 5 days. Organosilver derivative 21′′
was also characterized by 13C NMR spectroscopy at -25
°C. The signals originating from 21′′ were distinguished
from those of 9 as the signals of 21′′ disappeared upon
standing.
As the goal of an isolable organomercury derivative
of a phosphinine derivative was not yet achieved, the
attention was focused on the synthesis of 23, the
pentacarbonyltungsten complex of 22. As shown in
Scheme 7, 23 was prepared by the reaction of HgCl2
with 3 at 0 °C. Compound 23 (δ(31P) 175.1 ppm, 1J (PW)
Tr an sm etalation s with Mer cu r y Salts. The trans-
metalations reactions of 1, 2, and 3 with mercury salts
were the only category of transmetalations for which
all organozinc derivatives provided unambiguous posi-
tive results (Scheme 7).
2
) 261.6 Hz, J (PHg) ) 627.7 Hz) was isolated in pure
form. The purification was simple as 23 precipitated
from the reaction mixture and washing with THF
removed all zinc salts. This compound was also char-
acterized with (HR)MS spectroscopy; it showed a mo-
lecular ion with the appropriate isotope pattern. In the
mass spectrum, zinc salts were absent, indicating that
the removal of zinc salts in the purification of 23 is no
problem. The stability of 23 is higher than that of 22,
probably due to the absence of zinc salts. Nevertheless,
the light yellow green powder became inhomogeneous
after a few weeks storage at 4 °C as some grey material
was formed. In spite of the slow decomposition, the
results of the elemental analysis of 23 were very good
and in agreement with the proposed composition.
In the reaction of 1 with HgCl2 at room temperature,
a colorless precipitate was formed quickly and the 31P
NMR spectrum showed the presence of 22 (δ(31P) 213.1
ppm, 2J (PHg) ) 797 Hz), 1, and small amounts of 9 and
10 in the dark red solution. The ratio of 22:9 was 11.7:
1. After evaporation of DMF under vacuum, the residue
was extracted with toluene. After removal of toluene
under vacuum, the residue was extracted with C6D6
1
after which 22 was characterized by H and 31P NMR
spectroscopy. It turned out to be unstable at room
temperature; decomposition was complete within sev-
eral hours. This decomposition is possibly catalyzed by
zinc salts such as ZnICl‚(DMF)2. Their presence was
indicated by 1H NMR signals of DMF which were shifted
relative to those of free DMF (see Experimental Section);
note that zinc,12,13 contrary to mercury,14 is known to
coordinate to DMF.
Con clu sion s
The reactions of organozinc derivatives 2 and 3 with
PCl3 or AsCl3 furnished the expected monosubstitution
(11) Westmijze, H.; Kleijn, H.; Vermeer, P. J . Organomet. Chem.
1979, 172, 377.
(12) Ishiguro, S.; Ozutsumi, K.; Ohtaki, H. Bull. Chem. Soc. J pn.
1987, 60, 531.
(13) Goggin, P. L. In Comprehensive Coordination Chemistry;
Wilkinson, G., Ed.; Pergamon Press: Oxford, U.K., 1987; Vol. 2, p 491.
(14) Dean, P. A. W. Prog. Inorg. Chem. 1978, 24, 109.
(15) Armstrong, R. S.; Aroney, M. J .; Duffin, R. K.; Stootman, H.
J .; Le Fe`vre, R. J . W. J . Chem. Soc., Perkin Trans. 2 1973, 1272.
(16) Bell, N. A.; Nowell, I. W.; Reynolds, P. A.; Lynch, R. J . J .
Organomet. Chem. 1980, 193, 147.

806 Organometallics, Vol. 15, No. 2, 1996
Ta ble 1. 1H NMR Da ta of 2-R-4,5-Dim eth ylp h osp h in in esa
4-CH3 H3
5J (PH) (Hz)
Teunissen and Bickelhaupt
H6
no.
R
δ (ppm)
5-CH3 δ (ppm)
δ (ppm)
3J (PH) Hz)
δ (ppm)
2J (PH) (Hz)
b
19
8
4
R1
2.03
1.72
1.74
1.64
1.86
1.95
1.82
2.04
2.35
2.36
3.1
3.5
3.5
5.9
3.4
3.3
6.2
3.5
3.5
6.2
2.04
1.77
1.81
1.53
2.00
1.99
1.74
2.09
2.43
2.46
7.55
8.12
8.11
8.31
8.13
8.10
7.75
7.50
7.89
7.95
6.2
5.7
6.0
19.9
12.3
12.8
29.8
15.3
14.8
32.5
8.34
8.01
7.99
7.54
8.48
8.45
8.23
8.62
8.67
8.52
38.3
40.4
40.8
27.7
36.0
35.7
24.0
34.4
34.5
22.8
PCl2
AsCl2
AsCl2
c
5
13
14
16
22
22
23
SnPh3
SnMe2
SnMe2
Hgd
c
Hge
Hgc,f
a
b
C6D6 was used as solvent unless otherwise stated. R1 refers to the 3-(trimethylsiloxy)-2-cyclohexen-1-yl substituent. The numbering
of the phosphinine ring is adapted to that of other phosphinines and differs from the numbering in the experimental part. c P-W(CO)5
complex. Prepared from 1. e Prepared from 2, NMR in CDCl3. f NMR in THF-d8.
d
Ta ble 2. 13C NMR Da ta of 2-R-4,5-d im eth ylp h osp h in in esa
C2
1J (PC)
C6
1J (PC)
C3
2J (PC)
C5
2J (PC)
C4
3J (PC)
4-CH3
5-CH3
δ
δ
δ
δ
δ
δ
4J (PC)
δ
3J (PC)
no.
R
(ppm)
(Hz)
(ppm)
(Hz)
(ppm)
(Hz)
(ppm)
(Hz)
(ppm)
(Hz)
(ppm)
(Hz)
(ppm)
(Hz)
b
19
8
4
R1
177.7
167.0
c
151.7
163.9
167.7
161.3
177.7
194.7
c
47.9
63.0
c
15.5
79.2
80.1
24.5
74.5
79.4
c
154.8
154.3
154.1
149.0
156.7
156.5
154.0
152.4
155.4
151.0
49.9
57.2
58.9
18.1
61.0
61.3
11.0
30.6
60.7
8.2
136.3
139.3
140.1
142.2
144.9
143.8
147.5
149.3
144.3
148.1
12.8
12.6
18.7
15.7
15.0
14.2e
21.0
22.0
15.2
20.8
141.2
147.5
c
145.0
144.1
143.5
148.5
142.8
143.7
147.8
15.8
13.4
c
15.4
14.5
14.5f
15.8
16.6
14.8
15.9
138.7
140.1
c
138.7
138.6
138.0
136.6
136.1
138.7
136.9
17.2
17.2
c
28.9
24.8
25.6
35.8
36.8
26.4
37.4
22.4
21.7
21.7
21.4
22.0
22.1
21.7
23.8
22.5
22.0
2.2
1.9
1.9
3.7
1.7
g
4.2
g
2.0
g
22.9
23.1
23.0
22.9
23.3
23.4
22.9
22.5
23.4
22.2
3.4
3.6
3.4
9.8
2.7
g
8.4
12.8
2.7
8.2
PCl2
AsCl2
AsCl2
d
5
13
14
16
SnPh3
SnMe2
SnMe2
d
21′′ Agh
22
23
Hgi
Hgd,j
a
b
NMR in C6D6 unless otherwise mentioned. R1 refers to the 3-(trimethylsiloxy)-2-cyclohexen-1-yl substituent. The numbering of the
phosphinine ring is adapted to that of the other phosphinines and differs from the numbering in the experimental part. c Quaternary
aromatic carbon atom(s) not detectable. P-W(CO)5 complex. e AXX′ system: 2J (PC) + 4J (PC) ) 14.2 Hz. f AXX′ system: 2J (PC) + 6J (PC)
d
g
h
i
j
) 14.5 Hz. Too small to be detected. NMR in THF. Prepared via 2, NMR in CDCl3. NMR in THF-d8.
a Bruker MSL 400, operating at 400 (1H), 100.64 (13C), or 161.9
products. Due to limited stability of the products and/
or the formation of byproducts none of the functionalized
phosphinines could be isolated in pure form. The
transmetalations of 1, 2, and 3 with tin, copper, silver,
and mercury derivatives were successful, although the
thermal lability of the products in several cases ham-
pered their isolation and characterization. The organo-
copper and organosilver derivatives were less well
characterized; the organotin and organomercury deriva-
tives were isolable as their pentacarbonyltungsten
complexes and showed sufficient, but not absolute,
thermal stability under ordinary conditions.
MHz (31P). The coupling constants (J ) are given in hertz. The
diagnostic data are collected in Table 1 (1H) and Table 2 (13C).
The assignments are based on routine 2D NMR techniques.
Ratios of different phosphinines present in reaction mixtures
were determined from the relative intensities of their 31P NMR
signals. This is justified as these phosphinine derivatives are
structurally so closely related that they may be expected to
have a similar relaxation behavior.
The direct inlet mass spectra were measured with a Finni-
gan MAT 90 (Bremen, FRG), Source temperature 150 °C (70
eV IP). The GC/MS spectra were measured with a Hewlett-
Packard 5970 MSD equipped with a 5890 GC with 50 m CP
Sil 5CB column. In the mass spectra, the appropriate isotope
pattern was observed unless otherwise stated.
Elemental analyses were carried out by Micro Analytisches
Labor Pascher (Remagen, Germany). Melting points (uncor-
rected) were determined with melting point equipment of
Pleuger after Dr. Tottoli.
A general conclusion from both our results and those
of Mathey et al.3c may be that functionalized phosphin-
ines appear to be less stable than initially expected. The
reasons for this are not clear at the moment.
Exp er im en ta l Section
2-(Dich lor oa r sin o)-4,5-d im eth ylp h osp h in in e (4). A so-
lution of AsCl3 (10 g, 55.2 mmol) in THF (10 mL) was cooled
to -10 °C. Then a THF solution (8.5 mL) containing 2 (4.0
mmol) was added dropwise. The resulting dark brown reaction
mixture was warmed to room temperature. After evaporation
of the reaction mixture under reduced pressure, the residue
was extracted three times with pentane (50 mL). Several
crystallizations did not furnish pure 4. Upon cooling, brown
and yellow crystals were formed; the yellow crystals became
brown upon standing, even at -20 °C. 4: NMR (see also
Tables 1 and 2) (C6D6) δ(13C{1H}) 21.7 (ddq, {3J (CH) ) 5.3},
{1J (CH) ) 127.0}, 4-CH3), 23.0 (ddq, {3J (CH) ) 6.3}, {1J (CH)
) 127.3}, 5-CH3), 154.1 (ddq, {1J (CH) ) 154.9}, {3J (CH) ) 5.4},
C6), quaternary carbon atoms were not detected; δ(31P{1H})
198.6. HRMS calcd for C7H8As35Cl2P 267.8957, found 267.896;
MS (EI) m/e 268 ([M]+, 62), 233 ([M - Cl]+, 100), 198 ([M -
2Cl]+, 10), 123 ([M - AsCl2]+, 65).
Gen er a l P r oced u r es. All oxygen- and/or water-sensitive
reactions were carried out under dry nitrogen with oven-dried
glassware and oxygen-free, dry solvents. THF was distilled
first from NaH and finally from Na/benzophenone. Pentane
was distilled from LiAlH4. PCl3, chlorotrimethylsilane, and
AsCl3 were distilled before use. LiCl and LiBr were dried by
heating under vacuum. The starting compound (acetonitrile)-
pentacarbonyltungsten17 was prepared according to a reported
procedure.
NMR spectra were recorded at a Bruker AC 200 spectrom-
eter at 200 MHz (1H) or 50.32 MHz (13C). For 31P NMR
spectroscopy, a Bruker WM 250 was used operating at 100.26
MHz. 1H, 13C, or 31P NMR experiments were carried out with
(17) Marinetti, A.; Bauer, S.; Ricard, L.; Mathey, F. Organometallics
1990, 9, 793.

Organozinc Derivatives of Phosphinines
Organometallics, Vol. 15, No. 2, 1996 807
[η1-2-(Dich lor oa r sin o)-4,5-d im eth ylp h osp h in in e]p en -
ta ca r bon yltu n gsten (5). A solution of AsCl3 (1.3 g, 7.2
mmol) in THF (5 mL) was cooled to -40 °C. Then a THF
solution (3 mL) containing 3 (1.2 mmol) was added dropwise.
The resulting black reaction mixture was allowed to warm to
room temperature. Concentration to 3 mL was followed by
mixing with pentane (100 mL). A large amount of dark brown
precipitate was formed. Separation of the supernatant and
cooling to -70 °C led to the formation of orange crystals, which
consisted of three components with δ(31P{1H}) of 165.6, 159.5
and 154.8 ppm in the following ratio 146.2:53.0:17.5 (31P{1H}).
It is assumed that the most abundant component is 5. 5:
(dd, 4J (PC) ) 1.2, 2J (SnC) ) 37.1, {1J (CH) ) 159.0}, SnPh3
o-C), 138.6 (d, 3J (PC) ) 3.4, SnPh3 ipso-C), 144.9 (ddq, 2J (SnC)
) 14.7, {1J (CH) ) 155.3}, {3J (CH) ) 5.2}, C3), 156.7 (ddq,
3J (119SnC) ) 56.3, 3J (117SnC) ) 55.0, {1J (CH) ) 152.0}, {3J (CH)
) 5.2}, C6); δ(31P{1H}) 222.0 (2J (P117Sn) ) 303.0, 2J (P119Sn)
) 317.3); δ(119Sn) -125.4. HRMS calcd for C25H23P120Sn
474.0562, found 474.0588; MS (EI) m/e 474 ([M]+, 29), 397 ([M
- Ph]+, 31), 351 ([M - C7H8P]+, 100). Anal. Calcd for C25H23
-
PSn: C, 63.47; H, 4.90; P, 6.55; Sn, 25.09. Found: C, 62.57;
H, 4.93; P, 5.99; Sn, 24.1.
Bis(4,5-dim eth ylph osph in in yl-KC2)dim eth yltin (14) an d
bis((4,5-d im eth ylp h osp h in in yl-1KP , 2KC2)p en ta ca r bon yl-
tu n gsten )d im eth yltin (16). Syn th esis of 14. To a solution
of freshly sublimed Cl2Sn(CH3)2 (0.53 g, 2.42 mmol) in THF
(3 mL), 6.5 mL of a DMF solution containing 4.8 mmol of 1,
prepared as described above, was added dropwise while
stirring. After 3.5 h of stirring at room temperature, the
solution was evaporated under vacuum. The residue was
extracted three times with toluene (25 mL); after filtration,
the filtrate was evaporated under vacuum. Extraction of the
residue with pentane was followed by crystallization by cooling
the extract to -70 °C. After several futile attempts to obtain
pure crystals, it became evident that 14 was too unstable to
be isolated in pure form. The initially colorless crystals turned
yellowish rather rapidly upon standing. 14: NMR (see also
Tables 1 and 2) (C6D6) δ(1H) 0.74 (s, 2J (119SnH) ) 55.9,
2J (117SnH) ) 53.6, 6H, SnMe2), 8.10 (d, 3J (SnH) ) 66.2, 1H,
H3), 8.45 (d, 4J (SnH) ) 24.0, 1H, H6); δ(13C{1H}) -8.1 (tq,
2
NMR (see also Tables 1 and 2) (C6D6) δ(13C) 194.2 (d, J (PC)
1
2
) 8.8, J (WC) ) 124.5, CO (cis)), 197.5 (d, J (PC) ) 29.6, CO
(trans)); δ(31P{1H}) 165.4 (1J (PW) ) 268.4). HRMS calcd for
C
12H8As35Cl2O5P182W 589.8185, found 589.819, MS (EI) m/e
594 ([M]+, 22), 566 ([M - CO]+, 14), 485 ([M - 4CO]+, 10),
454 ([M - 5CO]+, 100), impurities (31P NMR, vide supra); 7
448 ([M]+), 392 ([M - 2CO]+), 364 ([M - 3CO]+), 336 ([M -
4CO]+) ((η1-3,4-dimethylphosphinine)pentacarbonyltungsten);
6 686 ([M]+), 658 ([M]+ - CO), 600 ([M]+ - 3CO) ([η1-2-
(chloroiodoarsino)-4,5-dimethylphosphinine]pentacarbonyl-
tungsten).
2-(Dich lor op h osp h in o)-4,5-d im et h ylp h osp h in in e (8).
A solution of PCl3 (14 mL, 0.16 mol) in THF (10 mL) was cooled
to -60 °C. Then a THF solution (20 mL) containing 2 (10
mmol) was added dropwise while stirring. A red precipitate
was formed, and after completion of the addition, the reaction
mixture was allowed to warm to room temperature. Removal
of volatile components under reduced pressure was followed
by two extractions of the residue with pentane (100 mL).
Several crystallizations did not furnish pure 8. The formation
of 10 during the reaction was a complicating factor as well as
the fact that 8 decomposed slowly at room temperature. 8:
1
1
3J (PC) ) 4.9, J (119SnC) ) 364.9, J (117SnC) ) 355.0, {1J (CH)
) 130.2}, SnMe2), 22.1 (dq, {3J (CH) ) 4.8}, {1J (CH) ) 126.5},
4-CH3), 23.4 (ddq, {3J (CH) ) 6.3}, {4J (CH) ) 2.5}, {1J (CH) )
126.6}, 5-CH3), 138.0 (d, 3J (PC) ) 25.6, C4), 143.8 (AXX′ {ddq},
3
{1J (CH) ) 153.9}, {3J (CH) ) 5.2}, C3), 156.5 (ddq, J (SnC) )
54.3, {1J (CH) ) 151.5}, {3J (CH) ) 4.9}, C6), 167.7 (dd, 3J (PC)
4
2
NMR (see also Tables 1 and 2) (C6D6) δ(1H) 8.01 (dd, J (PH)
) 4.4, C2); δ(31P{1H}) 218.5 (2J (P119Sn) ) 330.8, J (P117Sn) )
) 6.2, 1H, H6), 8.12 (dd, J (PH) ) 12.7, 1H, H3); δ(13C{1H})
316.1). HRMS calcd for C16H22P2120Sn 396.022, found 396.023;
MS (EI) m/e 396 ([M]+, 5), 381 ([M - CH3]+, 24), 366 ([M -
2CH3]+, 7), 273 ([M - C7H8P]+, 13).
3
21.7 (dddq, {3J (CH) ) 5.4}, {4J (CH) ) 0.8}, {1J (CH) ) 127.0},
4-CH3), 23.1 (ddq, 5J (PC) ) 1.8, {1J (CH) ) 127.2}, 5-CH3),
139.3 (dddq, 2J (PC) ) 24.4, {1J (CH) ) 162.8}, {3J (CH) ) 5.5},
Syn th esis of 16. All crude 14 (vide supra) was combined
and dissolved in THF (10 mL) after which 15 (1.0 g, 2.7 mmol)
was added. This mixture was stirred at room temperature
for 30 h and evaporated; the residue was extracted with
pentane (80 mL). Crystallization of the pentane extract
furnished pure 16 as yellow crystals (mp 157 °C) in a yield of
7% (0.17 g, 0.16 mmol) relative to 10. 16: NMR (see also
Tables 1 and 2) (C6D6) δ(1H) 0.88 (s, 2J (119SnH) ) 56.9,
2J (117SnH) ) 54.5, 6H, SnMe2), 7.75 (d, 2J (SnH) ) 66.4, 1H,
H3), 8.23 (d, 4J (SnH) ) 23.9, 1H, H6); δ(13C{1H}) -4.6 (tq,
3J (PC) ) 1.8, {1J (CH) ) 130.8}, SnMe2), 21.7 (ddq, {3J (CH) )
4.6}, {1J (CH) ) 126.9}, 4-CH3), 22.9 (dq, {1J (CH) ) 126.3},
3
3
C3), 140.1 (dd, J (PC) ) 6.0, C4), 154.3 (dddq, J (PC) ) 20.6,
{1J (CH) ) 154.9}, {3J (CH) ) 5.0}, C6), 167.0 (td, 1J (PC) )
63.0, {3J (CH) ) 9.1}, C2); δ(31P{1H}) 165.3 (P(Cl2)), 204.8
(P(ar)) (2J (PP) ) 242.4). HRMS calcd for C7H835Cl2P2 223.9478,
found 223.947; MS (EI) m/e 224 ([M]+, 100), 189 ([M - Cl]+,
43).
[η1-2-(Dich lor op h osp h in o)-4,5-d im eth ylp h osp h in in e]-
p en ta ca r bon yltu n gsten (11). A solution of PCl3 (excess,
freshly distilled) in THF (3 mL) was cooled to -60 °C. Then
a THF solution (2 mL) containing 3 (0.70 mmol) was added
dropwise. After the addition of 3, a brown precipitate was
formed and the reaction mixture was warmed to room tem-
perature. 11: NMR (THF) δ(31P{1H}) 174.2 (P(ar)), 158.4
3
2
5-CH3), 136.6 (d, J (SnC) ) 43.9, C4), 147.5 (ddq, J (SnC) )
11.1, {1J (CH) ) 156.1}, {3J (CH) ) 5.9}, C3), 148.5 (d, 4J (SnC)
) 10.7, C5), 154.0 (dd, 3J (SnC) ) 38.6, {1J (CH) ) 157.9}, C6),
2
(P(Cl2)), J (PP) ) 278.1 ,1J (PW) ) 273.8.
3
2
1
161.3 (dd, J (PC) ) 4.4, C2), 195.9 (d, J (PC) ) 9.3, J (WC) )
124.8, CO (cis)), 198.2 (d, 2J (PC) ) 28.1, CO (trans)); δ(31P{1H})
176.8 (1J (PW) ) 262.8, 2J (P119Sn) ) 213.5, 2J (P117Sn) ) 205.2).
2-(Tr ip h en ylst a n n yl)-4,5-d im et h ylp h osp h in in e (13).
The organozinc reagent 1 was prepared from zinc powder (0.68
g, 10.4 mmol) and 10 (1.23 g, 4.92 mmol) in DMF (7.5 mL) by
stirring for 18 h at room temperature. This solution was added
dropwise over 5 min to a solution of ClSnPh3 (1.89 g, 4.9 mmol)
in DMF (2 mL). After stirring for 15 min, a white precipitate
was formed. The solvent was removed under vacuum, and the
residue was extracted with toluene (40 mL) and filtered. After
evaporation of the filtrate under reduced pressure, the residue
was recrystallized several times from pentane to furnish pure
13 as colorless crystals (mp 122-3 °C) in a yield of 37% (0.86
g, 1.82 mmol) relative to 10. 13: NMR (see also Tables 1 and
2) (C6D6) δ(1H) 7.05-7.35 (m, 9H, m-, p-SnPh3), 7.75-7.85 (m,
6H, o-SnPh3), 8.13 (d, 3J (SnH) ) 69.6, 1H, H3), 8.48 (d, 4J (SnH)
) 26.8, 1H, H6); δ(13C{1H}) 22.0 (dddq, {3J (CH) ) 5.4}, {4J (CH)
) 0.7}, {1J (CH) ) 127.4}, 4-CH3), 23.3 (dddq, {3J (CH) ) 6.5},
HRMS calcd for
C
26H22O10P2118Sn184W2 1041.8726, found
1041.874; MS (EI) m/e 1042 ([M]+, 1), 690 ([M - W(CO)6]+,
3), 634 ([M - W(CO)6 - 2CO]+, 7), 578 ([M - W(CO)6 - 4CO]+,
13). Anal. Calcd for C26H22O10P2SnW2: C, 29.95; H, 2.13; P,
5.94. Found: C, 30.37; H, 2.29; P, 5.91.
4,5-Dim eth yl-3′-(tr im eth ylsiloxy)-1′,4′,5′,6′-tetr a h yd r o-
2-p h osp h a bip h en yl (19). A solution of 2 (6.08 mmol) and
2-cyclohexen-1-one (1.76 g, 18.3 mmol) in THF (13 mL) was
cooled to -90 °C. Then a light green THF solution (8 mL)
containing CuCN‚2LiBr (prepared from CuCN (0.66 g, 7.37
mmol) and LiBr (1.52 g, 17.5 mmol)) and ClSi(CH3)3 (2.4 g, 22
mmol) was added over 30 min. During 3 h the temperature
was maintained between -80 and -100 °C. Then the black
reaction mixture was allowed to warm to room temperature.
Concentration to 15 mL was followed by extraction with
pentane (75 mL). The upper layer was removed, and THF (8
3
{4J (CH6) ) 0.7}, {1J (CH) ) 126.7}, 5-CH3), 129.1 (d, J (SnC)
) 52.3, {1J (CH) ) 154.9}, SnPh3 m-C), 129.5 (dt, 4J (SnC) )
11.6, {1J (CH) ) 158.9}, {3J (CH) ) 14.1}, SnPh3 p-C), 137.8

808 Organometallics, Vol. 15, No. 2, 1996
Teunissen and Bickelhaupt
mL) was added to the lower layer. Extraction of the THF layer
was carried out with pentane (75 mL), and the extraction
procedure was repeated once more. The combined extracts
were concentrated to 50 mL and cooled to -70 °C. Some dark
brown oil was formed which did not contain phosphorus
compounds as evidenced by a 31P NMR spectrum. Addition
of pentane to the upper layer lead to separation of salts as a
dark oil. This cycle of operations was continued until a salt-
free pentane extract containing 9 and 19 was obtained. The
resulting extract (5 mL) was evaporated under vacuum to give
nearly pure 19 as a brown oil in a yield of 20% (0.35 g, 1.20
mmol) relative to 10. 19: NMR (see also Tables 1 and 2) (C6D6)
δ(1H) 0.27 (s, 9H, 3′-OSi(CH3)3), 1.35 (m, 1H, H4′), 1.50 (m,
1H, H6′), 1.65 (m, 1H, H5′), 1.90 (m, 1H, H6′), 2.10 (m, 1H,
H4′), 2.10 (m, 1H, H5′), 3.85 (m, 1H, H1′), 5.24 (m, 1H, H2′);
δ(13C{1H}) 0.7 (q, {1J (CH) ) 118.6}, 3′-OSi(CH3)3), 21.9 (t,
{1J (CH) ) 125.3}, C5′), 22.4 (ddq, {3J (CH) ) 5.7}, {1J (CH) )
126.3}, 5-CH3), 22.9 (ddq, {3J (CH) ) 6.5}, {1J (CH) ) 126.3},
191) and a grey precipitate. 21′′: NMR (see also Table 2) (THF
+ C6
D6) δ(13C{1H}) 46.8 (s, NCH3), 57.6 (s, NCH2); δ(31P{1H})
194.1; TMEDA (THF + C6D6): δ(13C) 45.5 (s, NCH3), 58.1 (s,
NCH2).
Bis(4,5-d im eth ylp h osp h in in yl-KC2)m er cu r y (22). Syn -
th esis via 1. To 1 mL of a DMF solution containing 0.9 mmol
of 1 was added HgCl2 (0.10 g, 0.37 mmol); a pale white
precipitate was formed immediately. After evaporation of the
solution under vacuum, the residue was extracted with toluene
(10 mL). Evaporation of the extract under vacuum was
followed by extraction of the residue with C6D6 (1 mL). As
decomposition occurred within several hours, 22 could not be
1
isolated in pure form. In the H NMR spectrum of 22, signals
originating from DMF were present. Relative to a reference
sample of DMF in C6D6, the amide proton was shifted 0.26
ppm downfield and one methyl group was shifted 0.16 ppm
upfield; the other methyl group did not show any shift. 22:
NMR (see also Table 1) (C6D6) δ(1H) 1.95 (s, 3H, NCH3), 2.25
(s, 3H, NCH3), 7.95 (s, 1H, CHO); δ(31P{1H}) 213.1 (2J (PHg)
) 799.7). Reference spectrum of DMF (C6D6) δ(1H) 1.95 (s,
3H, NCH3), 2.41 (s, 3H, NCH3), 7.69 (s, 1H, CHO).
3
4-CH3), 30.3 (t, {1J (CH) ) 125.5}, C4′), 35.4 (dt, J (PC) ) 6.3,
{1J (CH) ) 132.6}, C6′), 45.0 (dd, 2J (PC) ) 28.0, {1J (CH) )
128.1}, C1′), 136.3 (ddq, {1J (CH) ) 150.7}, {3J (CH) ) 5.3},
C6), 154.8 (ddq, {1J (CH) ) 151.0}, {3J (CH) ) 4.9}, C3);
δ(31P{1H}) 190.2. HRMS calcd for C16H25OP28Si 292.1412,
found 292.141; MS (EI) m/e 292 ([M]+, 27), 277 ([M - CH3]+,
20), 202 ([M - HOSi(CH3)3]+, 2), 169 ([M - C7H8P]+, 100).
Anal. Calcd for C16H25OPSi: C, 65.72; H, 8.62. Found: C,
67.80; H, 9.08.
Syn th esis via 2. A solution of 2 (1.24 mmol) and TMEDA
(0.20 g, 1.72 mmol) in THF (10 mL) was cooled to -60 °C. Then
HgCl2 (0.18 g, 0.66 mmol) was added, after which the dark
red solution was warmed slowly to room temperature, during
which a brown precipitate was formed. Addition of pentane
(10 mL) led to the formation of more voluminous precipitate.
The supernatant was separated and cooled, and the resulting
light brown crystals, contaminated with some grey powder,
were separated and dried. As MS indicated the presence of
zinc salts and NMR spectroscopy indicated slow decomposition
at room temperature, 22 could not be isolated in pure form.
22: NMR (see also Tables 1 and 2) (CDCl3) δ(1H) 7.89 (d,
(η1 -2-C u p r i -4,5-d i m e t h y lp h o s p h i n i n e )p e n t a c a r -
bon yltu n gsten (20). Syn th esis of 20 via 3. A solution of
CuCN‚2LiCl in THF (3 mL), prepared from LiCl (0.20 g, 4.65
mmol) and CuCN (0.06 g, 0.67 mmol) was cooled to -60 °C.
Then a THF solution (2 mL) containing 3 (0.70 mmol) was
added dropwise. After addition of 3, the resulting light brown
solution was warmed to room temperature. Decomposition of
20 at room temperature was complete within 6 h; in the 31P
NMR spectrum, two new signals appeared around 55 ppm and
one at 37 ppm. 20: NMR (THF) δ(31P{1H}) 171.2 (1J (PW) )
249.9).
4
3J (HgH) ) 143.1, 1H, H3), 8.67 (d, J (HgH) ) 43.6, 1H, H6),
2.63 (s, 12H, NCH3), 2.74 (s, 4H, NCH2), TMEDA:22 ) 1.46:1;
3
δ(13C) 48.3 (s, NCH3), 56.8 (s, NCH2), 194.7 (dd, J (PC) ) 7.2,
C2); δ(31P{1H}) 212.8 (2J (PHg) ) 786.1). Reference spectrum
of TMEDA (CDCl3) δ(1H) 2.23 (s, 12H, NCH3), 2.38 (s, 4H,
NCH2); δ(13C) 45.6 (s, NCH3), 57.4 (s, NCH2).
Syn th esis of 20 via 12.8 The highly reactive Cu(0) was
prepared by dropwise addition of a solution of CuCN‚2LiBr
(1.0 mmol) in THF (1 mL) to lithium naphthalenide (1.1 mmol)
in THF (2 mL) at -100 °C. After 5 min stirring at -100 °C,
a solution of 12 (0.25 mmol) in THF (0.6 mL) was added
dropwise. Then the reaction mixture was warmed slowly to
room temperature. Subsequent measurement of a 31P NMR
spectrum showed the presence of 20, 7, and a component with
δ(31P) 227 ppm, in a ratio of 10.6:1:2.6. 20: (δ(31P) 171 ppm,
1J (PW) ) 244). The decomposition of 20, obtained from 3 or
12, proceeded in a similar fashion: two signals appeared
around 55 ppm; but the signal at 37 ppm did not appear.
2-Ar gen tio-4,5-d im eth ylp h osp h in in e (21′′). A solution
of 2 was prepared from 10 (1.35 g, 5.4 mmol), TMEDA (1.37
g, 11.8 mmol), and zinc powder (0.7 g, 10.7 mmol) in THF (5
mL). Of this solution, 0.9 mL was added dropwise to a solution
of AgBr‚2LiBr (prepared from AgBr (0.19 g, 1.0 mmol) and
LiBr (0.19 g, 2.2 mmol) in THF (3 mL)), which was cooled -40
°C. The reaction mixture became red and was allowed to
slowly warm to room temperature. At room temperature, a
grey precipitate formed slowly. NMR measurements showed
the presence of 21′′. After that, the NMR sample was stored
at 4 °C. After 5 days, 21′′ was completely decomposed with
concomitant formation of 3,4-dimethylphosphinine (δ(31P{1H})
B is -[(4,5-d im e t h y lp h o s p h in in y l-1KP ,2KC 2 )p e n t a -
ca r bon yltu n gsten ]m er cu r y (23). A solution of 3 (3.52
mmol) in THF (10 mL) was cooled to 0 °C. Then a solution of
HgCl2 (0.48 g, 1.77 mmol) in THF (2 mL) was added quickly
while stirring. A large amount of yellow precipitate was
formed in the dark red solution. After addition of pentane,
the reaction mixture was filtered and washed twice with THF
(30 mL) and finally with pentane (30 mL) and dried. This
procedure furnished pure 23 (mp 192 °C dec) as a very light
yellow green powder in a yield of 50% (0.97 g, 0.89 mmol)
relative to 10. 23: NMR (see also Tables 1 and 2) (THF-d8)
3
δ(1H) 7.95 (d, J (HgH) ) 147.5, 1H, H3), 8.52 (d, 4J (HgH) )
48.0, 1H, H6); δ(13C{1H}) 195.2 (d, 2J (PC) ) 9.4, CO (cis));
1
δ(31P{1H}) 175.1 (2J (PHg) ) 627.7, J (PW) ) 261.6). HRMS
calcd for C24H16202HgO10P2182W184W 1093.8918, found 1093.890;
MS (EI) m/e 1094 ([M]+, 18), 726 ([M - W(CO)6 - CH3 - H
-CO]+, 13), 698 ([M - W(CO)6 - CH3 - H - 2CO]+, 17), 670
([M - W(CO)6 -CH3 -H -3CO]+, 29). Anal. Calcd for
C24H16HgO10P2W2: C, 26.33; H, 1.47; Hg, 18.32; P, 5.66. Found
C, 25.70; H, 1.54; Hg, 19.4; P, 5.42.
OM9506936