Table 4 13C-{1H} NMR data a
Compound
δC
I
2a
29.2 [1 C, d, 1J(PC) 21.3, CH2P], 161.1 [1 C, d, 2J(PC) 8.2, C᎐N]
᎐
30.7 (6 C, s, 2CMe3), 31.2 (6 C, s, 2CMe3), 35.8 [2 C, d, J(PC) 4.1, 2CMe3], 37.6 [2 C, d, J(PC) 13.7, 2CMe3], 54.3 (2 C, dt,b
3
3
2CHPd), 77.2 [2 C, d, J(PC) 53.0, 2CHPF , 153.4 (2 C, m, 2CN), 182.3 [2 C, d, J(PC) 17.9, 2CN]
3
3
4a c
4c
28.6 (6 C, s, 2CMe3), 30.1 (6 C, s, 2CMe3), 38.4 [2 C, d, J(PC) 3.6, 2CMe3], 40.2 [2 C, d, J(PC) 1.6, 2CMe3], 43.8 [2 C, dd,
J(PC) 23.8 and ≈2, 2CH2P], 57.3 (2 C, see text, 2CHPd), 185 (4 C, m, 4CN)
3
3
28.6 (6 C, s, 2CMe3), 30.1 (6 C, s, 2CMe3), 38.5 [2 C, d, J(PC) 3.9, 2CMe3], 40.2 [2 C, d, J(PC) 1.8, 2CMe3], 44.6 [2 C, d,
1J(PC) 23.3, 2CH2P], 57.4 (2 C, m, 2CHPd), 185 (4 C, m, 4CN)
4d
28.7 (6 C, s, 2CMe3), 30.1 (6 C, s, 2CMe3), 38.3 [2 C, d, 3J(PC) 3.2, 2CMe3], 40.4 (2 C, m, 2CMe3), 44.1 [2 C, d, 1J(PC) 23.8, 2CH2P],
57.3 [2 C, dd, 1J(PC) 10.0, 2J(PC) 72.7, 2CHPd], 184.4 (2 C, m, 2CN), 185.5 (2 C, m, 2CN)
6a
30.0 (6 C, s, 2CMe3), 30.7 (6 C, s, 2CMe3), 35.3 [2 C, d, 3J(PC) 3.0, 2CMe3], 37.1 [2 C, d, 3J(PC) 14.4, 2CMe3], 54.1 (2 C, m, 2CHBr),
76.2 (2 C, m, 2CHPd), 181.8 [2 C, d, 2J(PC) 16.7, 2CN], 185.5 (2 C, m, 2CN)
6bc
21.5 [2 C, d, 2J(PC) 6.3, 2CHMe], 28.6 (6 C, s, 2CMe3), 30.8 (6 C, s, 2CMe3), 39.5 (2 C, s, 2CMe3), 40.0 (2 C, s, 2CMe3),
51.3 [2 C, d, 1J(PC) 23.4, 2CHMe], 60.2 (2 C, m, 2CHPd), 186.5 (2 C, m, 2CN), 190.0 (2 C, m, 2CN)
a
Recorded at 62.9 MHz, unless shown otherwise, chemical shifts (δ) in ppm relative to SiMe4, J values are in Hz, solvent CDCl3 unless otherwise
stated. An attached proton test was applied to the carbon resonances to establish whether the carbons were CH3, CH2, CH or CN and for complexes
2a and 4a 13C᎐1H COSY experiments were conducted to establish connectivities. b The doublet spacing was 80.8 Hz and the separation of the peaks
of the 1:2:1 triplets was 14.4 Hz; this pattern was probably a doublet of deceptively simple 1:2:1 triplets, i.e. with N = |1J(PSC) ϩ 2J(PSЈC)| = 28.8
Hz. c Recorded at 100.6 MHz in CD2Cl2.
data are in Table 3 but the complex was rather insoluble and a
carbon-13 NMR spectrum was not recorded.
with N = |JAX ϩ JAXЈ| = 54 Hz, as shown in Fig. 4. The proton-
ation was reversible, and on addition of a solution of sodium
methoxide the deep red neutral species 2a was reformed (NMR
evidence). The diprotonated complex was isolated as the tri-
fluoroacetate 4a as yellow microcrystals, by evaporation and
addition of diethyl ether to the residue but the product con-
tained solvent of crystallisation (chloroform). Similarly the
phenyl analogue 2b reacted with trifluoroacetic acid to give the
diprotonated derivative 4b. Treatment of 2a with hydrochloric
acid gave the cationic dichloride 4c and treatment of 2a or 2b
with picric acid gave the corresponding dipicrate salts 4d and
4e, respectively. Further details are in the Experimental section
with elemental analytical and mass spectral data, and NMR
data in Tables 2–4.
We suggest that the novel and stable dipalladium() com-
plexes of type 2 are formed by deprotonation of a complex of
type 1 with 1 equivalent of Li[N(SiMe3)2] to give a carbanion
which loses a chloride ion and rearranges to give a species con-
taining an unco-ordinated PPh2 group. This PPh2 group then
rapidly attacks the palladium of another molecule of the same
species in an intermolecular reaction to give the five fused-
chelate-ring product of type 2; see Scheme 2.
We attempted to synthesize the analogous diplatinum()
complex by treating the neutral platinum complex 1c3 with 2
equivalents of Li[N(SiMe3)2] in tetrahydrofuran. On adding the
base there was an immediate change from orange to dark brown
but, after 3 d, no substantial amount of a single product could
be observed by 31P-{1H} NMR spectroscopy. However, treat-
ment of 1c with an excess of potassium tert-butoxide in dry
dimethyl sulfoxide (dmso) gave a dark solution whose 31P-{1H}
NMR spectrum showed two phosphorus singlets, each with
We studied complex 4a in some detail by NMR spec-
troscopy. The 31P-{1H} data are in Table 1. The 1H-{31P} NMR
spectrum of 4a showed two tert-butyl resonances at δ 0.50 and
1.16, an AB pattern due to the non-equivalent methylene pro-
3
tons at δ 3.48 and 4.46, J(HH) = 15.8 Hz, and a singlet at δ
1
1
satellites, one at δ 23.3, J(PtP) = 2118 Hz, and the other at δ
7.03 due to the PCH᎐ protons. In the H NMR spectrum the
᎐
3
Ϫ11.4, J(PtP) = 112 Hz. The resonance at δ 23.3, indicates a
methylene hydrogens were further coupled to one phosphorus
giving an ABX pattern; for δH 3.48, J(PH) = 14.0 Hz and for
δH 4.46, J(PH) = 5.5 Hz. A selective decoupling experiment
established that the methylene hydrogens were coupled to the
phosphorus directly bonded to platinum and trans to carbon,
whilst the small value of J(PtP) = 112 Hz for the resonance at
δ Ϫ11.4 and the negative chemical shift is indicative of phos-
phorus not directly bonded to platinum. We tentatively suggest
structure 3 for this complex. There was no observable coupling
between the two phosphorus atoms. After 2 d there was no
noticeable change in the 31P-{1H} NMR spectrum. The product
did not react with either carbon monoxide or triphenylphos-
phine as evidenced by a 31P-{1H} NMR study. Attempts to iso-
late 3 gave an orange product whose 31P-{1H} NMR spectrum
showed an AB pattern at δ 49.9 and 26.0 [2J(PP) = 394,
1J(PtPA) = 2886 at 1J(PtPB) = 2773 Hz]. However, this de-
composed on attempted purification.
phosphorus, δ 51.4, i.e. P . The ᎐CHPd protons were coupled
᎐
P
F
to both PF and PS and gave an apparent triplet (a deceptively
simple 1:2:1 triplet). Carbon-13, APT and heteronuclear
COSY experiments established that the protons of the tert-
butyls absorbing at δ 0.50 were attached to the carbons absorb-
ing at δ 28.6 and those absorbing at δ 1.16 were attached to the
carbons absorbing at δ 30.1. The methylene carbons absorbed
at δ 43.8 as a doublet of doublets with J(PC) = 23.8 Hz and ≈2
Hz; such a chemical shift (δ ca. 44) is typical of a methylene
carbon in a five-membered chelate ring.1–4 The PCH᎐ carbons
᎐
The dinuclear palladium() complexes 2a and 2b contain
absorbed at δ 57.3 as a second-order quartet with a doublet
separation of N 73.5 Hz and two equally intense outer peaks
95.3 Hz apart.
The protonations of complexes 2a and 2b were reversible. In
the presence of sodium methoxide 4a–4e were deprotonated
readily at the methylene carbon to give the neutral dipal-
ladium() complexes 2a or 2b, as indicated by the discharge of
the yellow colour to red (or orange) and subsequently con-
firmed by NMR studies. Although 4a, 4c and 4d contain the
same dipalladium dications the chemical shifts of the CH2P and
enamine type (C᎐C᎐N), i.e. ene–hydrazone, moieties. Enamines
᎐
react with electrophiles in what is a very useful and selective
synthetic method in organic chemistry,15–19 and we have
reported that platinum complexes of type 1 react with a proton
or with methyl iodide or bromine with electrophilic attack at an
ene–hydrazone carbon (C᎐C᎐N᎐N) on the ligand backbone.3
᎐
We have since studied electrophilic substitution reactions on the
backbone of palladium complexes of type 1.20 We have now
investigated the action of electrophiles on complexes 2a and 2b.
Treatment of a deuteriochloroform solution of 2a with 2 mol of
trifluoroacetic acid (one per Pd atom) at ca. 20 ЊC gave an
immediate reaction, with a change from deep red to yellow. The
31P-{1H} NMR spectrum of the resultant solution was of the
AAЈXXЈ type with the two chemical shifts at δ 51.4 and 14.8
the ᎐CHPd protons differ a lot between the three species, e.g.
᎐
for the CH2P protons δ 3.48 and 4.46 (4a), 3.29 and 3.61 (4c)
and 3.24 and 3.49 (4d), and for the PdCH proton δ 7.03 (4a),
6.24 (4c) and 5.81 (4d), i.e. ion-pair interaction has a consider-
able effect on chemical shifts.
2616
J. Chem. Soc., Dalton Trans., 1997, Pages 2613–2620