Synthesis and NMR spectra of derivatives of the polykis(diphenylphosphino)benzenes, (Ph2P)nC6H6-n [n=2 to 4]

Article abstract of DOI:10.1016/S0277-5387(99)00101-1

111 derivatives of the polykis(diphenylphosphino)benzenes, (Ph₂P)_nC₆H_6-n [n=2 to 4] obtained by oxidation or reaction with sulfur, selenium, iodomethane, M(CO)₆ [M=Cr, Mo, W], [codRh-μ-Cl]₂

Full text of DOI:10.1016/S0277-5387(99)00101-1

Polyhedron 18 (1999) 2117–2127
Synthesis and NMR spectra of derivatives of the
polykis(diphenylphosphino)benzenes, (Ph2P)nC6H62n [n52 to 4]
*
H. Christina E. McFarlane, William McFarlane
Department of Chemistry, University of Newcastle upon Tyne, Newcastle upon Tyne, NE1 7RU, UK
Received 1 March 1999; accepted 6 April 1999
Abstract
111 derivatives of the polykis(diphenylphosphino)benzenes, (Ph2P)nC6H62n [n52 to 4] obtained by oxidation or reaction with sulfur,
selenium, iodomethane, M(CO)6 [M5Cr, Mo, W], [codRh-m-Cl]2, or (PhCN)2MCl2 [M5Pd, Pt], are reported together with their 31P and
selected other NMR parameters. The reactions generally follow predictable courses, although stereochemical factors affect the range of
products obtained and can lead to significant structural distortion in extreme cases. The 31P chemical shifts and more particularly vicinal
coupling constants are also markedly influenced by such factors, to the extent that in species with three or more adjacent Ph2P moieties
they may be of limited diagnostic value.
1999 Elsevier Science Ltd. All rights reserved.
Keywords: Polyphosphine; 31P NMR; coupling constant; steric hindrance; metal complex
1. Introduction
We have previously reported [1,2] that the reaction
between sodium diphenylphosphide and the appropriate
polyfluorobenzene C6H62nFn in refluxing liquid ammonia
provides a convenient route to the polykis(diphenylphos-
phino)benzenes, (Ph2P)nC6H62n [n52–4], compounds 1
to 7. These polyphosphines have considerable potential as
multi- and ambi-dentate ligands with a rigid skeleton that
may facilitate or inhibit chelation to a metal centre or
bridging of several metal centres, and a few examples of
this diverse behaviour have been described [3,4]. We have
also shown [2] how the 31P chemical shifts and vicinal
coupling constants 3J(31P–31P) in the parent polyphos-
phines display trends that can be understood in terms of
preferred phosphorus electron lone-pair relative orienta-
tions determined by steric interactions between adjacent
diphenylphosphino groups.
In order to provide a basis for studying their many types
of coordination behaviour we have undertaken a survey of
the reactions of these polyphosphines with a range of metal
substrates and with the chalcogens. These reactions were
conveniently followed by 31P NMR spectroscopy, aug-
mented as necessary by 1H [including COSY and
ROESY], 13C [including HETCOR], selective 1H–h31Pj
decoupling, and 31P COSY experiments. In general they
proceeded in the expected sequential manner, and one of
our aims was to establish the extent to which the 31P
NMR. parameters can be used diagnostically.
In this paper we present 31P and selected other NMR.
data for 111 distinct derivatives of which 46 have been
isolated in analytically pure form. Some data were also
obtained on a considerable number of other species
detected in the reactions, but as their characterisation was
less certain they are not reported in detail. Two other
members of the series of parent phosphines, 1,3- and
*Corresponding author. Tel.: 144-191-222-7126; fax: 144-191-222-
6929.
E-mail address: william.mcfarlane@ncl.ac.uk (W. McFarlane)
0277-5387/99/$ – see front matter
PII: S0277-5387(99)00101-1
1999 Elsevier Science Ltd. All rights reserved.

2118
H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
1,4-bis(diphenylphosphino)benzene, have been made [5–7]
without recourse to the corresponding difluorobenzenes,
but are not included here since in absence of sterically
interacting diphenyl-phosphino groups their behaviour is
expected to resemble that of triphenylphosphine.
2.2. NMR spectroscopy
Pulsed Fourier transform 31P NMR spectra were ob-
tained
from
solutions
in
trichloromethane-d,
dichloromethane-d2, or dichloromethane contained in 10
mm o.d. NMR tubes at 24.2 MHz (JEOL FX 60) or 36.2
MHz (JEOL FX 90Q), or in 5 mm o.d. NMR tubes at
202.5 MHz (JEOL LAMBDA 500) and are referenced to
external 85% H3PO4. Proton (500.15 MHz), 13C (125.6
MHz), 77Se (95.4 MHz), and 195Pt (107.0 MHz) NMR
spectra were also obtained on the last instrument and are
referenced to internal Me4Si, internal Me4Si, external
Me2Se, and a notional J(195Pt)521.4 MHz respectively.
31P COSY spectra were acquired at 202.5 MHz from
unspun samples with broadband proton decoupling into a
1K31K data matrix (digital resolution approximately 10
Hz), a shifted sine-bell weighting function being applied in
each dimension prior to Fourier transformation. Other 2D
NMR experiments used standard pulse sequences and
methods of data treatment [8,9].
2. Experimental
2.1. Preparations
The starting polyphosphines 1 to 7 were made as
described previously [1,2], and most of their reactions
were conducted on a scale of 0.02 to 0.5 mmole. Those
with sulfur, selenium, iodomethane, [codRh-m-Cl]2, cis-
(PhCN)2PdCl2, or cis-(PhCN)2PdCl2 were performed at
room temperature in dichloromethane and final products
were recrystallised from dichloromethane/diethylether
prior to elemental and mass spectroscopic analysis. Oxida-
tions were conducted with ca. 3% aqueous hydrogen
peroxide in acetone/dichloromethane at room temperature
to give 1a and 5b, and by aerial oxidation in dichlorome-
thane for 5a or refluxing diglyme for 1d and 4a. Reactions
with M(CO)6 [M5Cr, Mo, W] were carried out under
nitrogen in refluxing diglyme for approximately 1 h and
the isolated products were recrystallised from dichlorome-
thane/methanol.
3. Results
31P chemical shifts and coupling constants for identified
species are in Tables 1–7, and microanalytical data,
molecular weights determined by mass spectrometry and
melting points for isolated products are in Table 8.
Table 1
31P NMR parameters of derivatives of 1,2-bis-(diphenylphosphino)benzene
No.
Compound
d(31P)a
3J(31P1 –31P2)b
1
1
213.3
31.8
43.1, 217.5
47.2
27.5, 41.0
34.6, 218.2
38.2
22.2, 214.9
83.3
59.8
46.5
62.3
63.8
40.1
38.0
145.2
–
31.7
–
4.9
36.6
7.3
26.9
–
–
–
–
–
1a
1b
1c
1d
1e
1f
1g
1h
1i
1j
1k
1l
1m
1n
1h1,2-O2j
1h1-Sj
1h1,2-S2j
1h1-O,2-Sejc
1h1-Sejd
1h1,2-Se2je
1h1-MeIj
1h1,2-Cr(CO)4j
1h1,2-Mo(CO)4j
1h1,2-W(CO)4jf
[1h1,2-Rh-m-Clj]2
1h1,2-PdCl2j
1h1,2-PtCl2jh
g
–
–
[12h1,19,2,29-Ptj]21[PtCl4]2-i
a In ppm 60.2 ppm. When two values are given the first relates to P1 and the second to P2.
b In Hz 60.2 Hz. The absence of an entry indicates that the coupling constant is unavailable owing to symmetry or accidental isochronicity.
c
1J(77Se–31P)5743 Hz.
d
1J(77Se–31P)5739 Hz.
e
1J(77Se–31P)5754 Hz.
g
1J(103Rh–31P)5134.3 Hz.
f
1J(183W–31P)5232.2 Hz.
h
1J(195Pt–31P)53606 Hz.
i In the 195Pt spectrum the resonance at 2558 ppm was a quintet with 1J(195Pt–31P)52502 Hz.

H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
2119
Table 2
31P NMR parameters of derivatives of 1,2,3-tris-(diphenylphosphino)benzene
No
Compound
d(31P1 )a
d(31P2)a
d(31P3)a
3 J(31P1 –31P2)b
4 J(31P1 –31P3 )b
3 J(31P2 –31P3 )b
2
2
212.7
45.8
46.3
50.1
37.3
37.4
21.6
86.5
63.9
48.3
28.9
29.9
25.8
48.3
210.6
27.0
21.9
89.7
212.7
215.5
46.3
50.1
215.2
37.4
214.4
217.1
216.0
217.0
97.5
55.5
36.6
11.0
63.0
42.1
46.4
22.0
6.1
–
4.9
–
–
5.3
–
4.9
3.6
3.7
3.7
97.5
|0
36.6
11.0
|1
42.1
|0
13.5
14.7
15.8
2a
2b
2c
2d
2e
2f
2g
2h
2i
2h1-Sj
2h1,3-S2j
2h1,2,3-S3j
2h1-Sejc
2h1,3-Se2jd
2h1-MeIj
2h1,2-Cr(CO)4j
2h1,2-Mo(CO)4j
2h1,2-W(CO)4je
66.3
50.5
3.7
a In ppm 60.2 ppm.
b In Hz 60.2 Hz. The absence of an entry indicates that the coupling constant is unavailable owing to symmetry or accidental isochronicity.
c
1J(77Se–31P)5735 Hz.
d
1J(77Se–31P1)5740 Hz.
e
1J(183W–31P1 )5230 Hz, 1J(183W–31P2 )5228 Hz.
3.1. Notation
derivatives were isolated in analytically pure form [see
Table 8], and satisfactory microanalytical data were also
obtained for the monosulfide 5c. In other cases the
intermediate products were identified from the patterns of
their 31P chemical shifts and 31P–31P coupling constants
together with consideration of the sequence of their
appearance as the reaction progressed. It was noted that the
spectra continued to change over a period of hours after
the initial reaction was complete, indicating that that the
first products were kinetically determined and that there
were subsequent (probably intramolecular) transfers of
sulfur to give the thermodynamically preferred products.
This approach is exemplified in Fig. 1 which shows part of
the PV region of proton-decoupled 31P spectra resulting
from successive additions of sulfur to 1,2,4-(Ph2P)3C6H3.
Assignments were aided by a proton-decoupled 31P COSY
spectrum of which the PV part is shown in Fig. 2. This
spectrum also showed appropriate correlations with the PIII
region (0 to 220 ppm). These spectra do not unequivo-
cally distinguish 3a and 3b or 3d and 3e, and the latter pair
Each reaction product is specified by the number of its
parent polyphosphine (as in the introductory section of this
paper) printed in bold followed by the position(s) and
nature of the added substituent(s) of phosphorus in curly
brackets.
Thus
1,2,4-tris(diphenylthiophosphino)-5-
1,2,4-[Ph2P(S)]3-5-Ph2P-
diphenylphosphinobenzene,
C6H2, is designated 7 h1,2,4-S3j, and other examples of
the use of this system are as follows:
5
are assigned on the basis that J(31PV–31PV) is likely to be
greater than 4J(31PV–31PV) as we have established in the
case of 6e [10]. The same reasoning was applied in other
cases of pairs showing four- or five-bond 31PV–31PV
couplings, but is less well established for the corre-
sponding 31PIII–31PV couplings.
3.2. Reactions with elemental sulfur
These reactions were readily followed by 31P NMR.
because the phosphorus chemical shifts of the resulting
tertiary phosphine sulfides are some 50 ppm to high
frequency of those of the parent tertiary phosphine. They
were generally conducted in toluene or dichloromethane or
occasionally chloroform-d at room temperature, initially
with a single atomic equivalent of sulfur in an attempt to
achieve mono-addition, and then with stepwise addition of
more sulfur. With 1, 3, 4, 6, and 7 the fully sulfurised
3.3. Reactions with elemental selenium
The behaviour was similar to that with sulfur, signals in
the 31P(Se) spectral region (32 to 40 ppm) having satellites
due to 1J(77Se–31P) of 731 to 772 Hz. The reactions
generally proceeded with greater difficulty, and the com-
position of the reaction mixtures continued to change over
a period of several days. It was not possible to form
species with three adjacent Ph2P(Se) groups even though

2120
H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
Table 3
31P NMR parameters of derivatives of 1,2,4-tris-(diphenylphosphino)benzene
No
Compound
d(31P1)a
d(31P2)a
d(31P4)a
3J(31P1 –31P2 )b
5J(31P1 –31P4 )b
4 J(31P2 –31P4)b
3
3
213.4
44.1
217.6
211.9d
44.0
218.3
47.3f
211.8d
34.9g
218.9
39.5e
24.1
213.6
212.7
23.4
214.9
217.9
44.4
26.7
24.2
25.6
42.9
42.6
43.0
153.3
29.7
30.7
147.0
31.0
29.8
7.0
147.0
35.3
35.4
7.3
24.8
24.8
140.2
24.8
21.7
–
,3
1.0
0
,3
0
0
0
0
1.8
0.5
,5
1.4
3.2
,0.5
0
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
3q
3r
3s
3h1-Sjc
3h2-Sjc
3h4-Sj
213.9d
217.9
44.4
2.0
4.7
1.8
4.6
,5
1.4
5.1
5.0
0
3h1,4-S2je
3h2,4-S2je
3h1,2,4-S3je
3h4-Sej
47.3f
213.9d
218.5
34.7g
39.1e
214.2
23.4
210.7
212.0
24.3
82.7
83.1
42.2
34.9g
35.1h
34.6i
35.1
3h1,4-Se2jc
3h2,4-Se2jc
3h1,2,4-Se3jj
3h1-Meljc
22.3
23.6
22.4
22.7
23.3
26.1
56.6
26.3
38.6
26.4
21.6l
25.9
26.2
24.6
25.3
24.1
3h2-MeIjc
0
0
0
3h4-MeIj
1.5
|2
5.4
,3
,3
,5
3.7
,3
,4
,5
,5
,5
,5
,5
,5
,5
3h1,4-(MeI)2jc
3h2,4-(MeI)2jc
3h1,2-Cr(CO)4j
3h1,2-Cr(CO)4,4-Cr(CO)5j
3h1,2-Mo(CO)4j
3h1,2-Mo(CO)4,4-Mo(CO)5j
3h1,2-W(CO)4j
3h1,2-W(CO)4,4-W(CO)5j
3.1
|2
,3
,3
,5
,3
,3
,4
,5
,5
,5
,5
,5
,5
,5
211.7
82.7
83.1
–
–
59.9
59.9
59.6
60.2
4.9
–
,4
–
–
–
–
–
–
–
3t
45.0k
45.6l
61.0
45.0k
45.2l
61.0
3u
3v
3w
3x
3y
3z
3a
3b
m
[3h1,2-Rh-m-Clj]2
[3h1,2-Rh-m-Cl,4-Oj]2
3h1,2-PdCl2j
m
62.4
63.3
53.5
63.2
213.4
40.8
62.4
63.3
53.5
63.2
213.4
40.8
n
[3h1,2-PdCl2j]2
[3h1,2-PdCl2j]2h4,49-PdCl2jo
p
[3h4,49-PtCl2j]2
13.8
15.4
[3h1,2-PtCl2j]2h4,49-PtCl2jo,q
a In ppm 60.2 ppm.
b In Hz 60.2 Hz. The absence of an entry indicates that the coupling constant is unavailable owing to symmetry or accidental isochronicity.
c It may be necessary to interchange the relative assignments of 3a and 3b, of 3h and 3i, of 3k and 3l, and of 3n and 3o.
d It may be necessary to interchange d(P1) and d(P2).
e Assigned on the basis that 5J(P1V –PV4 ).4J(P2V –PV4 ), see text.
f Complex multiplet simulated to yield d(31P2)2d(31P1)50.06 ppm.
g
1J(77Se–31P)5745 Hz.
h
1J(77Se–31P1, P2)5749 Hz.
i
1J(77Se–31P4)5750 Hz.
j
1J(77Se–31P1)5765 Hz; d(77Se)52191 ppm; 1J(77Se–31P2 )5762 Hz; d(77Se)52195 ppm. 1J(77Se–31P4)5754 Hz; d(77Se)52296 ppm.
k
1J(183W–31P)523062 Hz.
l
1J(183W–31P1,2 )523462 Hz; 1J(183W–31P4 )524362 Hz.
m
1J(103Rh–31P1,2 )5133.561 Hz.
n Probable structure, see text.
o See text for structures.
p
1J(195Pt–31P)53638 Hz.
q
1J(195Pt–31P1,2 )5360761 Hz, d(195Pt)54 ppm; 1J(195Pt–31P4)53636 Hz, d(195Pt)584 ppm.
the corresponding sulfides (2c, 6e and 6f) were identified.
6f was made by sulfurisation of 6g.
it was easy to add only one atomic equivalent of chal-
cogen, in each case in the 1-position, so that the products
were readily identifiable in spite of the unusual sizes of
their vicinal 31P–31P coupling constants (see Fig. 3).
3.4. Oxidation reactions
Very marked variations in susceptibility to oxidation
were noted for these species. Thus one phosphorus atom of
the hindered tetraphosphine 5 was easily oxidised in
dichloromethane solution by air at room temperature but
hydrogen peroxide was required to yield the tetroxide 5b
as the final product. In general with the highly hindered 5
3.5. The addition of iodomethane
This proceeded sequentially and led to the quaternisation
of one or more phosphorus atoms as evidenced by the
appearance of 31P resonances in the range 20 to 27 ppm;

H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
2121
Table 4
recorded. This showed clear NOE correlations between the
methyl resonances and the ortho-resonances of the phenyl
groups on the same phosphorus atom, but unfortunately
there were no correlations with the proton resonances of
the central C6 ring, and so the spectrum did not provide the
required information. In general the products of the
quaternisation reactions were of limited stability, but solid
materials with satisfactory elemental analyses were iso-
lated in the cases of 4i, 4j, and 6l.
31P NMR parameters of derivatives of 1,3,5-tris-(diphenylphosphino)
benzene
No
Compound
d(31P1)a
d(31P3)a
d(31P5 )a
4
4
25.4
23.1
42.1
42.0
42.4
34.3
34.2
34.7
21.5
22.4
22.8
39c
25.4
23.1
25.3
42.0
42.4
25.7
34.2
34.7
24.3
22.4
22.8
25.6
39c
25.4
25.2
25.3
25.2
42.4
25.7
25.5
34.7
24.3
23.5
22.8
25.6
25.0
40.1
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
4m
4n
4h1,3-O2j
4h1-Sj
4h1,3-S2j
4h1,3,5-S3j
4h1-Sej
4h1,3-Se2jb
4h1,3,5-Se3j
4h1-MeIj
3.6. Reactions with group six metal carbonyls
4h1,3-(MeI)2j
4h1,3,5-(MeI)3j
4h1-Mo(CO)5j
4h1,3-Mo(CO)5j2
4h1,3,5-Mo(CO)5j3
In refluxing diglyme (1608) either one or two molecules
of carbon monoxide were displaced from the hexacar-
bonyls of chromium, molybdenum, and tungsten to yield
complexes with monodentate and/or bidentate chelating
polyphosphine. Characterisation is based mainly on the
known variations with ring size and metal of 31P chemical
shifts in analogous complexes of other phosphines and
diphosphines [12], together with satisfactory elemental
analyses and mass spectra in the cases of 1l, 1m, 1n, 2g,
2i, 3q, 3s, 3w, 6m, 6n, 6p, 7n, 7p, and 7r.
39c
40.1
22.8
40.1
22.8
d
4h1,3,5-W(CO)5j3
22.8
a In ppm 60.2 ppm.
b
1J(77Se–31P)5746 Hz.
c Overlapped signal, 61 ppm.
d
1J(183W–31P)5244 Hz.
only in the case of 1 was a persistent species found with a
pair of adjacent Ph2MeP1 groups. In the cases of 3n and
3o the identities of the products were confirmed by their
methyl proton resonances which showed characteristic
3.7. Complexes with platinum group metals
2J(31PV–H) and J(31PIII–H) couplings of ca. 14 and 2 Hz
5
The 31P NMR spectra of dichloromethane solutions
containing a polyphosphine and [codRh-m-Cl]2 showed the
presence of many transient species with Rh–P bonds
[1J(103Rh–31P)|130 to 190 Hz], but only 1g, 3v, and 3w
(from adventitious aerial oxidation during the preparation
of 3v) were characterised with sufficient confidence to be
included in the tables. The palladium and platinum com-
plexes were made by reacting cis-(PhCN)2PdCl2 and cis-
(PhCN)2PtCl2 respectively with the polyphosphine in
dichloromethane at room temperature. The 31P NMR
spectra of the reaction mixtures changed considerably over
periods of several days and indicated transient formation of
respectively, and their methyl 13C resonances which had
1J(31PV–13C) and 4J(31PIII–13C) couplings of ca. 57 and
24 Hz respectively. A H/ 13C 2D chemical shift correla-
1
tion experiment showed the signs of 1J(31PV–13C) and
2J(31PV–H) to be opposite, presumably positive and
negative respectively [11]. The relative assignments of 3n
and 3o are based on 5J(31P–31P) being greater than
4J(31P–31P) as in the case of 3d and 3e above, and in an
attempt to put this on a more secure basis a proton 2D
ROESY spectrum with a spin-locking (mixing) time of
250ms of the reaction mixture containing 3n and 3o was
Table 5
31P NMR parameters of derivatives of 1,2,3,4-tetrakis-(diphenylphosphino)benzene
No Compound
d(31P1)a
d(31P2)a
d(31P3)a
d(31P4)a
3J(P1 –P2)b
4J(P1 –P3)b
5J(P1 –P4)b
3J(P2 –P3)b
4J(P2 –P4)b
3J(P3 –P4)b
5
5
213.5
31.3
38.9
45.9
37.1
20.5
48.0
56.8
34.7
22.5
3.4
22.5
24.8
32.4c
25.2
25.3
23.7
26.5
26.8
27.1
213.5
213.6
38.9
215.2
215.1
214.9
215.3
216.2
217.8
81.0
37.7
,3
62.5
70.6
52.5
,10
,10
,8
4.7
5.0
,3
6.6
6.7
20.3
,3
,3
1.2
35.3
,3
,3
0
4.7
,3
,3
1.7
2.1
,3
,5
,10
,8
81.0
,3
,3
,0.5
0
,3
,5
,10
,8
5a 5 h1-Oj
5b 5 h1,2,3,4-O4j
5c 5 h1-Sj
32.4c
21.6
22.6
7.6
5d 5 h1-Sejd
0
0
5e 5 h1-MeIj
6.1
,3
,5
,10
,8
,3
26g
24g
12g
5f 5 h1,2-W(CO)4je
5g 5 h1,2-PdCl2j
5h 5 h1,2-PtCl2jf
57.6
77.4
41.4
,10
,10
,8
a In ppm 60.2 ppm.
b In Hz 60.2 Hz. The absence of an entry indicates that the coupling constant is unavailable owing to symmetry or accidental isochronicity.
31
c The
P
peaks are broader.
2,3
1J (77Se–31P)5731 Hz.
d
1J (183W–31P)5230 Hz.
e
f
1J (195Pt–31P1)53603 Hz; 1J(195Pt–31P2 )53456 Hz.
g 61 Hz.

2122
H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
Table 6
31P NMR parameters of derivatives of 1,2,3,5-tetrakis-(diphenylphosphino)benzene
No Compound
d(31P1)a
d(31P2)a
d(31P3)a
d(31P5)a
3J(P1–P2)b
4J(P1–P3)b
5J(P1–P5)b
3J(P2–P3)b
4J(P2–P5)b
3J(P3–P5)b
6
6
212.7
29.3
212.7
26.0
41.4
41.4
41.2
41.4
42.3
33.3
33.7
33.8
33.5
33.5
20.6
22.8
56.7
37.5
33.2
21.0
24.8
34.0
93.0
–
–
,3
93.0
,3
,3
,3
,5
,5
4.9
4.9
,4
,1
,2
,2
3.5
3.7
3.7
3.4
4.6
3.7
,4
,4
,3
,3
,3
,5
,5
,1
,2
,4
,1
,2
,2
2.2
1.8
1.8
1.5
1.5
1.5
,4
,4
c
c
c
c
c
c
6a 6h5-Sj
6b 6h1,5-S2j
6c 6h2,5-S2j
6d 6h1,3,5-S3j
6e 6h1,2,3,5-S4j
6f 6h1,2,3-S3,5-Sej
6g 6h5-Sejd,e
6h 6h1,5-Se2jf
46.3
213.1
46.4
51.6
50.4
212.3
37.4
46.4
38.0
21.7
26.8
87.5
63.9
64.1
48.9
59.7
38.2
210.1
46.1
215.8
213.1
46.4
53.7
36.6
35.4
10.9
9.8
4.9
–
,3
,5
,5
,1
,2
,4
1.8
,2
,2
3.3
1.8
–
2.4
36.6
35..4
10.9
9.8
25.3
53.8
–
51.6
–
52.4
50.4
–
28.5
211.1
25.4
26.0
22.5
12.4
90.8
212.3
216.2
46.4
94
–
94
61.0
34.1
39.5
45.5
27.3
23.8
8.5
4.9
–
2.4
6i
6h1,3-S2,5-Sej
34.1
39.5
2.2
6j 6h1,3,5-Se3j
6k 6h1,5-(MeI)2j
6l
38.0
–
213.8
26.8
4.5
–
6h1,3,5-(MeI)3j
27.3
12.8
14.4
13.7
15.3
,4
6m 6h1,2-Cr(CO)4,5-Cr(CO)5j
6n 6h1,2-Mo(CO)4,5-Mo(CO)5j
6o 6h1,2-Mo(CO)4,5-Sejg
216.2
217.0
217.0
216.6
217.1
217.3
3.7
3.4
3.4
3.4
,4
,4
66.3
–
66.8
8.2
–
6p 6h1,2-W(CO)4,5-W(CO)5jh
51.0
1.2
–
i
6q 6h1,2-PdCl2j2h5,5-PdCl2j2
70.3
,4
,4
,4
6r 6h1,2-PtCl2,5-Sejj
47.2
,4
12.2
a In ppm 60.2 ppm.
b In Hz 60.2 Hz. The absence of an entry indicates that the coupling constant is unavailable owing to symmetry or accidental isochronicity.
c Transient species; peaks in P(III) region overlapped by other signals.
d From analysis as AB2X system.
e
1J(77Se–31P)5748 Hz.
f
1J(77Se–31P1)5741 Hz; 1J(77Se –31P5)5754 Hz.
g
1J(77Se–31P5)5750 Hz.
h
1J(183W–31P1 )5235 Hz; 1J(183W–31P2 )5228 Hz; 1J(77Se–31P5)5243 Hz.
i Broad signals.
j
1J(195Pt–31P1 )|3590 Hz; 1J(195Pt–31P2 )|3535 Hz; 1J(77Se–31P5)5772 Hz.
species other than those listed. These probably included
polymeric and cyclophane-like species such as [1,3,5-
(Ph2P)3C6H3]2(PtCl2)3 whose successful characterisation
was dependent upon the isolation of suitable crystals [3,4].
Assignments of a cis or trans geometry at the metal are
transition metal substrates to yield products that are
generally of the type expected by analogy with the
behaviour of such ligands as dppe, but which may vary
according to reaction conditions (i.e. for kinetic reasons) as
well as their inherent thermodynamic stability. In par-
ticular, reactions involving attack on the central Ph2P
moiety of three adjacent ones are significantly hindered.
For example, species such as 2h2-MeIj, and 2h2-Sej were
not observed, and 2c and 6e formed only with difficulty.
Exceptions to this occur with species such as 2g–i, 5f–h,
and 6m–r where additional stability is provided by chelate
ring formation, but in these cases it was not then possible
to attack the Ph2P group in the 3-position. It was also
found that the highly hindered 5 was considerably less
robust than the other polyphosphines and underwent
spontaneous decomposition over a period of months at 08
C.
1
based upon values of J(195Pt–31P) of approximately 3500
or 2500 Hz respectively for the platinum complexes
[13,14], and by analogy with these in the case of pal-
ladium. The structure of 1k was confirmed by the presence
of a quintet from coupling to 31P with a 1:4:6:4:1 intensity
pattern in its 195Pt spectrum. The reaction of 3 with
cis-(PhCN)2PtCl2 initially gave mainly 3a which is pre-
sumably a chlorine-bridged dimer, but after several days
3b was the major product and was identified from its
characteristic 195Pt NMR spectrum consisting of a pair of
phosphorus-coupled triplets in a 2:1 intensity ratio together
with 31P resonances in the appropriate regions and allied
195Pt satellites indicative of cis geometry at each type of
platinum.
It is of interest to consider the ways in which the
different degrees of strain encountered in these species are
reflected in their NMR (primarily 31P) parameters, par-
ticularly with reference to predictive and diagnostic value.
4. Discussion
4.1. Phosphorus chemical shifts
As already pointed out, the entries in the tables represent
only a proportion of species detected in this work, and it is
clear that the polyphosphines 1 to 7 can undergo a wide
range of reactions with chalcogens, iodomethane, and
As indicated in the introduction, steric interactions
between adjacent Ph2P groups are likely to have a major
effect upon d(31P), with electronic effects involving more

H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
2123
Table 7
31P NMR parameters of derivatives of 1,2,4,5-tetrakis-(diphenylphosphino)benzene
No
Compound
d(31P1)a
d(31P2)a
d(31P4)a
d(31P5)a
3J(31P1 –31P2)b
3J(31P4 –31P5)b
7
7
214.2
43.2
42.8
42.8
45.0
46.0
33.7
34.6
33.3
35.6
36.2
22.1
23.3
82.7
83.4
60.0
59.7
45.0
44.0
61.0
38.4
214.2
218.5
219.7
219.7
45.0
214.2
213.5
42.8
219.7
42.2
46.0
213.4
34.6
214.2
213.5
219.7
42.8
219.3
46.0
213.6
220.1
33.3
219.6
36.2
214.0
23.3
213.8
83.4
213.9
59.7
214.4
44.0
–
29.3
29.3
29.3
–
–
–
29.3
29.3
25.4
–
7a
7b
7c
7d
7e
7f
7g
7h
7i
7 h1-Sj
7 h1,4-S2jc
7 h1,5-S2jc
7 h1,2,4-S3j
7 h1,2,4,5-S4j
46.0
–
7 h1-Sejd
219.3
220.1
219.3
35.6
35.4
34.2
33.3
–
–
7 h1,4-Se2jc
34.2
32.2
31.7
–
22.0
18.0
–
–
–
–
–
7 h1,5-Se2jc
219.3
33.6
36.2
7 h1,2,4-Se3j
7j
7k
7l
7 h1,2,4,5-Se4je
7 h1,4-(MeI)2jc
7 h1,5-(MeI)2jc
7 h1,2-Cr(CO)4j
7 h1,2-Cr(CO)4, 4,5-Cr(CO)4j
7 h1,2-Mo(CO)4j
7 h1,2-Mo(CO)4, 4,5-Mo(CO)4j
7 h1,2-W(CO)4jf
7 h1,2-W(CO)4, 4,5-W(CO)4jg
7 h1,2-PdCl2, 4,5-PdCl2jh
7 h1,2-PtCl2, 4,5-PtCl2ji
36.2
–
214.0
213.3
82.7
83.4
60.0
59.7
45.0
44.0
61.0
22.1
22.0
18.0
–
–
–
–
–
–
–
213.3
213.8
83.4
213.9
59.7
214.4
44.0
61.0
7m
7n
7o
7p
7q
7r
7s
7t
–
–
–
61.0
38.4
38.4
38.4
–
a In ppm 60.2 ppm.
b In Hz 60.2 Hz. The absence of an entry indicates that the coupling constant is unavailable owing to symmetry or accidental isochronicity.
c It may be necessary to interchange the relative assignments of 7b and 7c, 7g and 7h, and 7k and 7l.
d
1J(77Se –31P1)5742 Hz.
e
1J(77Se–31P)5766 Hz.
f
1J(183W–31P)5234 Hz.
g
1J(183W–31P)5238 Hz.
h
1J(195Pt–31P)53620 Hz.
i
1J(195Pt–31P)53618 Hz.
widely separated groups being much less important. This is
supported by the observation that the data for derivatives
of 4 (Table 4) and the isolated Ph2P groups of 3 (Table 3)
and 6 (Table 6) are all very similar to those of the
corresponding derivatives of triphenylphosphine itself.
Similarly, the isolated Ph2PCCPPh2 fragments in 1, 3, and
7, show closely parallel behaviour to each other. Thus
mono-sulfurisation increases d(31P) by approximately 57
ppm (cf. 48 ppm for Ph3P [15]) and disulfurisation of
adjacent Ph2P groups by approximately 61 ppm, these
larger figures being indicative of a significant steric effect.
The corresponding figures for mono- and di-selenisation
and quaternization with MeI are also greater at 48, 52 and
39 ppm respectively compared with 40, 40, and 27 ppm for
isolated Ph2P groups. The M(CO)4 [M5Mo, W]
derivatives of chelating 1 have been referred to previously
[16–18], but apparently not fully characterised and the
coordination chemical shifts (i.e. the changes in d(31P)
accompanying their formation) of 96.6, 73.1, and 59.8 ppm
for M5Cr, Mo, and W respectively are all some 6 ppm
greater than for the corresponding complexes of dppe
which also gives a five-membered chelate ring. This
perhaps implies a greater degree of ‘‘ring strain’’ in the
former, and it is noteworthy that its greater conformational
mobility permits dppe to achieve a smaller angle of bite
than 1. However, measurements of heats of formation from
norbornadieneMo(CO)4 [16] suggest that 1m and
dppeMo(CO)4 are of comparable stability. Similar remarks
apply to the five-membered chelate rings in the M(CO)4,
PdCl2 and PtCl2 complexes of 3 and 7. The foregoing
additions also have minor effects upon the chemical
shift(s) of the non-reacting Ph2P group(s). Thus sulfurisa-
tion in an ortho position reduces d(31P) of the unreacted
Ph2P group by approximately 4 ppm and selenisation has a
somewhat larger effect, while quaternisation leads to small
increases. The additional strain that might be expected to
be present in derivatives of 2 and 6 with three adjacent
Ph2P groups has little further effect upon d(31P), and
indeed this generally also applies to the severely hindered
5.
4.2. Phosphorus–phosphorus coupling
In the parent polyphosphines 3J(31PIII–31PIII) is large
and varies considerably with steric factors but nonetheless
has considerable diagnostic value. As expected, 4J(31PIII–
31PIII) and 5J(31PIII–31PIII) are much smaller and any
steric dependence is not readily apparent. In the derivatives
this also applies to the four and five bond coupling
constants but not always to the three-bond couplings. In
3
general terms it can be stated that J(31PV–31PV) is always
less than 11 Hz, the only apparent exceptions to this being

2124
H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
Table 8
Microanalytical data for selected compounds
No
%C
%H
%P
MW
MP
8C
Calc.
Fd.
Calc.
Fd.
Calc.
Fd.
Calc.
Fd.
479
510
542
526
526i
–
1aa
1ca
1da
1eb
1fc
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
13.1
11.7
10.0
12.8
–
–
–
–
–
12.8
–
7.0
5.8
–
14.6
13.9
–
–
–
–
–
–
–
–
–
–
–
–
–
12.7
12.2
9.8
12.9
–
–
–
–
–
12.2
–
7.2
5.8
–
14.0
13.0
–
478
510
542
526
606
–
–
–
–
585
711
–
710
795
926
727
–
146–147
266–268
178–180
209–211
205–208
d.235
d.235
d.230
d.275
d .140
d.240
d.360
215–216
d.215
240–245
280–283
d.160
d.160
d.180
d.160
d.250
220–221
261–263
140–142
158–160
160–164
d.190
239–242
d.265
d.260
259–260
227–230
255–258
285–290
d.160
d.180
d.205
d.275
d.280
d.270
d.320
d.270
d.250
d.260
d.260
d.360
70.6
66.6
68.6
59.6
63.3
66.9
62.4
55.0
61.6
50.6
50.6
71.1
69.5
59.6
69.4
62.1
57.5
49.0
64.2
49.1
69.4
58.2
51.1
42.7
73.8
76.6
72.6
65.4
60.0
68.8
72.6
66.7
55.2
60.1
64.6
52.7
60.0
68.8
66.7
57.4
65.2
60.1
52.9
55.4
48.2
70.0
66.6
68.8
59.0
62.4
66.8
62.4
54.7
61.8
50.5
49.6
69.2
71.3
58.2
69.1
61.8
57.2
49.0
62.9
48.6
67.2
58.1
51.3
42.8
71.7
75.2
72.1
65.0
58.2
68.6
71.8
66.3
55.5
59.6
63.3
52.8
59.5
67.7
66.1
57.3
65.2
60.1
52.7
55.5
46.7
4.74
4.47
4.60
4.00
4.63
3.96
3.70
3.26
4.14
3.40
3.40
4.69
4.19
3.59
4.58
3.37
3.10
2.66
4.24
2.95
4.58
3.83
2.49
2.08
4.82
5.00
4.74
4.27
3.92
4.49
4.74
4.35
4.14
3.36
3.62
2.95
3.92
4.49
4.35
3.74
3.70
3.41
3.01
3.62
3.14
4.47
4.25
4.47
4.01
4.41
3.98
3.53
3.07
4.16
3.29
3.23
4.76
4.40
3.51
4.68
3.45
3.30
2.67
4.10
3.12
4.44
3.90
2.72
2.16
5.05
5.12
5.09
4.00
3.72
4.47
4.82
4.27
4.35
3.41
3.66
2.88
3.92
4.60
4.14
3.88
3.71
3.47
2.99
3.50
2.89
1ga
1hd
1ia
–
–
–
1jb
1kd
1me
1ne
2db
2gf,h
2if
549g
712
–
710
794
926
727
–
3fe
3qd
3sb
3ub
3wd
3ab
4da
4ea
4ie
–
–
–
–
785
–
727
867
–
786
–
727
867
–
4je
–
–
5ba
5cb
5di
5gb
5hb
6eb
6gd
6hd
6lf
879
847
893
992
1081
943
894
–
–
–
–
–
877
847
814i
991
1081
942
894
–
–
–
–
–
–
–
13.1
13.9
12.7
10.0
9.8
10.6
8.6
–
13.1
12.7
11.0
10.8
10.0
8.8
–
12.9
13.2
12.4
9.6
9.9
10.7
8.5
–
12.6
12.6
11.2
10.3
10.2
8.6
–
6nb
6md
6pd
6qd
7ea
7g,hd
7jf
–
–
942
973
–
–
–
–
–
–
942
973
–
–
–
–
–
–
7nd
7pd
7rd
7sc
7te
–
–
a White.
b Light yellow.
c Light tan.
d Yellow.
e Off-white.
f Orange.
g MW(Calc.)2MW(Fd.);1 Cl.
h Unstable compound.
i MW(Calc.)2MW(Fd.);1 Se.
3
the Cr(CO)4 derivatives 2g and 6m in which there is an
additional coupling pathway via the metal atom. In the
absence of serious steric constraints (i.e. in the derivatives
of 1, 3 and 7) J(31PV–31PIII) lies in the range 18.0 to 36.5
Hz, the largest values being associated with Ph2PSe,
intermediate ones with Ph2PS, and the smallest with

H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
2125
Fig. 1. P5S regions of proton-decoupled 31P NMR spectra at 202.5 MHz of reaction mixtures resulting from the addition of elemental sulfur to
1,2,4-(Ph2P)3C6H3 in CDCl3. (a) 1 atomic equivalent of sulfur. (b) 1.5 atomic equivalents of sulfur. (c) 2 atomic equivalents of sulfur. (d) 2.5 atomic
equivalents of sulfur. The resonances from 3f at 47.3 ppm are off-scale.
Ph2MeP1, in a clear parallel with effective electronegativi-
ty. In addition 3J(31PIII–31PIII) remains large and of
comparable size to its value in the parent polyphosphine.
The foregoing does not apply to the derivatives of the
more severely hindered 2, 4 and 5, and some striking
variations in the three-bond couplings are apparent, al-
though it is noteworthy that those in derivatives of 2 are
always similar to those of the corresponding derivatives of
5 as expected on the basis that an isolated Ph2P group will
have little electronic and no steric influence. Thus
3J(31PIII–31PIII) is close to zero in 2a and 6b, while
3J(31PV–31PIII) is 55.5 and 53.7 Hz respectively, results
which might cast doubts upon the authenticity of these
species had not 6e, the product of complete sulfurisation of
6 been fully characterised by elemental analysis and
single-crystal X-ray diffraction [10]. It is known that in 6e
the bulk of the three adjacent Ph2P(S) groups leads to large
structural distortions with dihedral angles of as much as

2126
H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
Fig. 2. Part of the P5S region of the 2D proton-decoupled 31P COSY NMR spectrum at 202.5 MHz of the reaction mixture resulting from the addition of
2.5 atomic equivalents of sulfur to 1,2,4-(Ph2P)3C6H3 in CDCL3. The resonance at 47.3 ppm from P1 in 3f is off-scale but showed a correlation with the P4
resonance at 42.2 ppm, and the resonances from 3d and 3e at 44.1 and 44.4 ppm were also correlated with corresponding signals in the PIII region of the
spectrum.
708 for the nominally planar P–C–C–P fragments and that
these probably persist in solution. Comparable deviations
from planarity can be expected to occur in other deriva-
tives of 2 and 6 and to affect the vicinal 31P–31P couplings
both by direct modification of the through-bond pathway
and by modifying electron lone-pair interactions. It is
difficult to quantify these effects at present, but it may be
noted that the values of 3J(31PIII–31PIII) in 1, cis-
Ph2PCH5CHPPh2, trans- Ph2PCH5CHPPh2, and dppe
are 145.2, 105.5, 13.4, and 33.6 Hz respectively [19],
which illustrates the large potential for variation of this
coupling constant; indeed it is possible that the contribu-
tions from through-bond and direct lone-pair interactions
are of opposite sign.
The highly hindered 5 has smaller (81.0 and 35.3 Hz)
vicinal 31P–31P coupling constants than any of the other
polyphosphines studied here, and its derivatives display
some remarkably small or even zero values of 3J(31PIII–
31PIII) in conjunction with abnormally large 3J(31PV–
31PIII), e.g. 70.6 Hz in 5d. It is particularly striking that in
5a–e which are substituted in the 1-position not only is the
adjacent vicinal PIII–PIII coupling constant almost zero,
but so is that between P3 and P4. This is consistent with
there being quite severe distortions of the P4C6 skeleton in

H.C.E. McFarlane, W. McFarlane / Polyhedron 18 (1999) 2117–2127
2127
Fig. 3. Proton-decoupled 31P NMR spectrum at 24.2 MHz of 1-Ph2P(Se)-2,3,4-(Ph2P)3C6H2 in CH2CL2 showing the pattern of unusual coupling
constants.
addition to specific electron lone-pair orientations which
are normally regarded as being the main determining
factors in vicinal 31PIII–31PIII couplings.
Acknowledgements
We thank the Leverhulme Trust and EPSRC for support.
4.3. Other couplings to phosphorus
References
The coupling constants 1J(M–31P) [M577Se, 103Rh,
183W, 195Pt] all lie in the ranges expected for derivatives of
less sterically demanding ligands, and indicate that the
hybridisation and effective nuclear charge on the phos-
phorus atoms are essentially normal thus confirming that
the unusual features shown by the vicinal 31P–31P cou-
plings arise primarily from special dihedral angles and/or
[1] H.C.E. McFarlane, W. McFarlane, Polyhedron 2 (1983) 303.
[2] H.C.E. McFarlane, W. McFarlane, Polyhedron 1988 (1875) 7.
[3] P.M. Van Calcar, M.M. Olmstead, A.L. Balch, Chem. Commun.
(1996) 2597.
[4] D. Imhof, U. Burckhardt, K.–H. Dahmen, F. Joho, R. Nesper, Inorg.
Chem. 36 (1997) 1813.
[5] A. Michaelis, L. Bleichmann, Chem. Ber. 15 (1882) 801.
[6] H. Zorn, H. Schindelbauer, H. Hogan, Chem. Ber. 98 (1965) 2431.
[7] R.A. Baldwin, M.T. Cheng, J. Org. Chem. 32 (1967) 1572.
[8] A. Bax, D.G. Lewis, J. Magn. Reson. 63 (1985) 207.
[9] A. Bax, R.H. Griffey, B.L. Hawkins, J. Magn. Reson. 55 (1983)
301.
1
coupling paths. These considerations also apply to J(31P–
13C) of the methyl groups in the quaternary phosphonium
iodides.
[10] W. Clegg, A.J. Edwards, H.C.E. McFarlane, W. McFarlane, Poly-
hedron 17 (1998) 3515.
[11] W. McFarlane, Proc. Roy. Soc., A 306 (1968) 185.
[12] P.A. Garrou, Chem. Rev. 81 (1981) 229.
5. Conclusions
The polykis(diphosphino)benzenes without more than
two adjacent Ph2P moieties react with chalcogens and
iodomethane and with transition metal substrates in a
predictable manner and the patterns of 31P chemical shifts
and coupling constants are a valuable guide to the nature
of the products. However, the presence of more than two
Ph2P proximate moieties leads to considerable steric strain
in the products and this is relieved by significant distor-
tions from planarity of the PnC6 skeleton which are
reflected in abnormal values of 3J(31PIII–31PIII) and
3J(31PV–31PIII) to the extent that these are of less immedi-
ate and direct value as guides to identity.
[13] A. Pidcock, R.W. Richards, L.M. Venanzi, Proc. Chem. Soc. (1962)
184.
[14] S.O. Grim, R.L. Keiter, W. McFarlane, Inorg. Chem. 6 (1967) 1133.
[15] L. Maier, Inorg. Nucl. Chem. 24 (1962) 1073.
[16] S.L. Mukerjee, S.P. Nolan, C.D. Hoff, R. LopezdelaVega, Inorg.
Chem. 27 (1988) 81.
¨
[17] W.A. Schenk, D. Kummerle, J. Organometal. Chem. 303 (1986)
C25.
¨
[18] W.A. Schenk, D. Kummerle, T. Schwietske, J. Organometal. Chem.
349 (1988) 163.
[19] I.J. Colquhoun, W. McFarlane, J. Chem. Soc., Dalton Trans. (1982)
1915.