Synthesis and structure of novel rhodium complexes of multi-functionalised amine-phosphine ligands

Article abstract of DOI:10.1039/b104523g

The coordination behaviour of a variety of amine containing phosphine ligands has been studied. Two new multifunctionalised ligands have been prepared and shown along with three known ligands to favour a variety of coordination modes when complexed with rhodium(III) centres. The hemilabile PN chelate ligands show three different types of behaviour: strong P,N chelation even in the presence of chloride ions, strong P,N chelation after a chloride ligand is removed (Ag salts), and weak coordination after chloride abstraction. Five of the complexes have been characterised by X-ray crystallography. Multinuclear NMR and IR spectroscopies were also highly diagnostic in assigning the coordination chemistry favoured by each ligand.

Full text of DOI:10.1039/b104523g

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis and structure of novel rhodium complexes of
multi-functionalised amine-phosphine ligands
Matthew L. Clarke, Alexandra M. Z. Slawin, Matthew V. Wheatley and J. Derek Woollins*
School of Chemistry, University of St. Andrews, St. Andrews, Fife, UK KY16 9ST.
E-mail: jdw3@st-andrews.ac.uk
Received 22nd May 2001, Accepted 3rd October 2001
First published as an Advance Article on the web 14th November 2001
The coordination behaviour of a variety of amine containing phosphine ligands has been studied. Two new
multifunctionalised ligands have been prepared and shown along with three known ligands to favour a variety
of coordination modes when complexed with rhodium() centres. The hemilabile P∧N chelate ligands show three
different types of behaviour: strong P,N chelation even in the presence of chloride ions, strong P,N chelation after a
chloride ligand is removed (Ag salts), and weak coordination after chloride abstraction. Five of the complexes have
been characterised by X-ray crystallography. Multinuclear NMR and IR spectroscopies were also highly diagnostic in
assigning the coordination chemistry favoured by each ligand.
prepared the tridentate P∧NN ligand 4 and shown it to form
Introduction
allyl complexes in which the amino-phosphine ligand adopts an
Bidentate ligands containing phosphorus and nitrogen donors
are extremely useful in both coordination chemistry and homo-
η3 coordination mode and the allyl ligand unusually adopts η1
coordination.11
genous catalysis.1 In fact there are a number of processes in
which complexes of P∧N bidentate ligands make them catalysts
of choice. Important examples are carbonylation of alkynes,2
asymmetric hydrogenation of highly substituted alkenes,3
asymmetric allylic alkylation,4 asymmetric hydroboration,5
Stille coupling,6 and nickel catalysed cycloaddition reactions.7
Pyridyl phosphines such as 2-diphenylphosphinopyridine 1 have
been applied successfully in several processes.8 Phosphorus–
nitrogen donors are particularly useful in catalytic processes as
one half of the ligand is capable of dissociating from an other-
wise stable chelate complex, allowing an organic substrate to
coordinate to the metal and undergo some transformation. In
other cases there is no stable chelate complex formed, but tran-
sient coordination of the nitrogen donor allows the formation
of stable coordinatively unsaturated species which would not
exist if a monodentate ligand were used. Thus, 2-PPh2Py and
Ph3P share similar electronic and steric properties, yet 16 eϪ
(2-PPh2Py)3Pt is the only zerovalent platinum species detected
even when a large excess of 2-PPh2Py is present. In contrast,
(Ph3P)4Pt is the platinum complex formed if an excess of Ph3P
is present. Another useful feature of ligands containing P,N
donor sets is the electronic asymmetry that is induced by the
ligands. This allows selective reactions to occur trans to the
stronger phosphorus ligand. This has been exploited with
considerable success in asymmetric catalysis.4 We wished to
develop some new P,N donor ligands which might improve
existing catalytic processes, and as the initial stage of this pro-
ject have sought to understand which coordination modes are
favoured by which structural motifs. A survey of the literature
and a recent study of our own shows that ligands with P,N
donor sets favour several different coordination modes depend-
ing on their exact structure. 2-Diphenylphosphinopyridine 1
seems to favour a bidentate bridging mode (although some
chelate complexes have been isolated).8,9 2-(Diphenylphos-
phino)aminopyridine 2 forms a strong P,N chelate with several
transition metals,10 while other pyridyl phosphine ligands like
3 prefer to bind in monodentate fashion.9 Some P,N ligands
which contain further functionality can bind to metals in tri-
dentate fashion. For example Van Leeuwen and coworkers have
The preferred mode of coordination has a very significant
impact on a ligand’s applications in catalysis and organometal-
lic chemistry. For example, a ligand that forms a very strong
chelate may not be sufficiently capable of partially dissociating
and allowing coordination of an organic substrate. In asym-
metric catalysis, on the other hand, a strong chelate (η2-P∧N)M
is necessary to prevent dilution of enantioselectivity by compet-
ing unselective monodentate complexes, (η1-P∧N)2M.12 In order
to clarify which co-ordination modes are favoured for different
ligands containing PNx (x = 1, 2 or 3) donors, we have studied
the coordination behaviour of five P,N ligands with rhodium
complexes.
Experimental
General
All manipulations were carried out under an atmosphere of
nitrogen, unless stated otherwise. All solvents were either freshly
distilled from an appropriate drying agent (THF, Et2O, DCM)
or obtained as anhydrous grade. 1H and 31P NMR spectra were
recorded using a “Varian 2000” 300 MHz spectrometer. IR
spectra were recorded as KBr discs (prepared in air) on a Perkin
Elmer PE1720 FTIR/RAMAN spectrometer. All significant
peaks > 800 cmϪ1 are quoted to serve as a fingerprint. Silver
salts, diphenylphosphine, 2-hydrazinopyridine, formaldehyde,
DOI: 10.1039/b104523g
J. Chem. Soc., Dalton Trans., 2001, 3421–3429
This journal is © The Royal Society of Chemistry 2001
3421

View Article Online
2-diphenylphosphinopyridine (Aldrich Chemical Co.), and 2-
piperidinopyridine (Lancaster Synthesis) were purchased and
used as received. Triethylamine and chlorodiphenylphosphine
were distilled prior to use. [Cp*RhCl2]2 and Ph2PNHPy 2
were prepared by literature methods.10,13 Diphenylphosphino-
methanol was obtained by heating neat diphenylphosphine (one
equiv.) and formaldehyde (one equiv.) in a Schlenk tube at 100 ЊC
for 3 h. This gives on cooling a quantitative yield of pure
material. (CAUTION: halide abstractions carried out within
this paper were carried out on a microscale, often using <0.020 g
of silver perchlorate. However, since organometallic perchlor-
ate salts can be explosive, the authors strongly suggest the use
of other silver salts that are less inherently dangerous, but often
equally effective.)
After stirring for 30 min, the solvent was evaporated to near
dryness and the resulting compound was washed twice with
either diethyl ether or light petroleum (bp 60–80 ЊC) to yield the
rhodium complexes in essentially quantitative yield. The com-
plexes obtained in this way were generally analytically pure, but
further purification was sometimes carried out by size exclusion
chromatography (in air, dichloromethane as eluent) or by
recrystallisation by slow diffusion of diethyl ether into dichloro-
methane solutions of the complexes.
[Cp*RhCl(2-dppap-P,N)]Cl 8. Crude material was pure as
judged by spectroscopy, and obtained in 100% yield. Recrystal-
lisation by slow diffusion (CH2Cl2–Et2O) gave crystals of the
dichloromethane solvate complex suitable for X-ray diffraction
study. C27H30N2PCl2RhؒCH2Cl2 requires: C, 50.90; H, 4.71; N,
4. 09. Found: C, 51.17; H, 4.77; N, 4.17%. IR (νmax/cmϪ1): 3416,
3051, 2985, 1612, 1572, 1475, 1435, 1377, 1330, 1158, 1102,
1019. 31P NMR (121.4 MHz, CDCl3), δ 89.1(JP–Rh = 145 Hz). 1H
NMR (300 MHz, CDCl3), δ 1.5 (15H, d, J = 3.6 Hz), 6.8 (1H, t,
J = 7.0 Hz), 7.2–7.6 (10H, m), 7.9–8.1 (3H, m), 11.9 (1H, s, br).
FAB MS: m/z 551/553 (M Ϫ Cl)ϩ, 515/517 (M Ϫ 2Cl)ϩ. HRMS
(ESϩ): found: m/z 551.0886, (M Ϫ Cl)ϩ requires 551.0890.
2-Diphenylphosphinohydrazinopyridine 5
Chlorodiphenylphosphine (2.33 ml, 13 mmol) was added to a
solution of 2-hydrazinopyridine (1.42 g, 13 mmol) and tri-
ethylamine (2.01 ml, 14.3 mmol) in THF (40 ml) and stirred
overnight under nitrogen. The reaction mixture was filtered
to remove the triethylamine hydrochloride and solvent was
removed in vacuo leaving a yellow solid. Crude yield 2.79 g,
73%. 2.16 g of the solid was recrystallised by cooling a concen-
trated chloroform solution in the fridge overnight (yield: 1.90 g,
6.5 mmol, 50%). C17H16N2P requires: C, 69.60; H, 5.50;
N, 14.33. Found: C, 69.09; H, 5.29; N, 14.01%. νmax/cmϪ1: 3313,
Cp*RhCl2(2-PPh2Py) 9. The compound was obtained in
analytically pure form by removing solvent and washing with
Et2O, and thus in quantitative yield. C27H29NPRhCl2 requires:
C, 56.66; H, 5.11; N, 2.45. Found: C, 56.12; H, 4.91; N, 2.34%.
IR (νmax/cmϪ1): 3150, 1638, 1618, 1571, 1482, 1435, 1421, 1372,
1
3200, 1602, 989. 31P NMR (121.4 MHz, CDCl3), δ 49.6. H
NMR (300 MHz, CDCl3), δ 7.9 (1H, d, J = 8 Hz, pyC[6]H), 7.1–
7.6 (12H, m, aromatic), 6.7 (1H, d, J = 16 Hz, aromatic) 6.5
(1H, t, J = 8 Hz), 6.1 (1H, s, NH[py]), 4.5 (1H, d, J = 12 Hz,
NH[P]).
1159, 1094, 1026. 31P NMR (121.4 MHz, CDCl3), δ 28.9 (JP–Rh
=
1
141 Hz). H NMR (300 MHz, CDCl3), δ 1.4 (15H, d, J =
3.6 Hz), 7.1–7.6 (10H, m), 7.9–8.1 (3H, m) 8.7 (1H, s, br). FAB
MS: m/z 536/538 (M Ϫ Cl)ϩ, 500/502 (M Ϫ 2Cl)ϩ. HRMS
(ESϩ): found: m/z 536.0780, (M Ϫ Cl)ϩ requires 536.0781.
2-(N-Diphenylphosphino)piperazinopyridine 6
A solution of diphenylphosphinomethanol (2.386 g, 11.035
mmol) in acetonitrile (40 mL) was added to 2-piperizino-
pyridine (1.68 mL, 1.80 g, 11.035 mmol). This solution was then
heated to reflux overnight. On cooling to room temperature,
analytically pure microcrystals of pippyphos 6 formed in the
reaction flask. These were filtered off very briefly in air and
dried in vacuo (yield: 2.04 g, 5.63 mmol, 51%). This compound
has also been prepared directly from diphenylphosphine in
toluene by other workers.14 C22H24N3P requires: C, 73.11; H,
6.69; N, 11.63. Found: C, 73.33; H, 6.32; N, 11.80%.
Cp*RhCl2(pippyphos-P) 11. The compound was obtained in
analytically pure form by removing solvent and washing with
Et2O, and thus in quantitative yield. Crystals suitable for X-ray
diffraction were grown by slow diffusion (Et2O–CH2Cl2).
C32H39N3PRhCl2 requires: C, 57.33; H, 5.86; N, 6.27. Found:
C, 56.90; H, 5.77; N, 6.15%. IR (νmax/cmϪ1): 3051, 2813, 1592,
1565, 1480, 1436, 1377, 1312, 1242, 1099, 1008. 31P NMR
(121.4 MHz, CDCl3), δ 32.5 (JP–Rh = 139 Hz). 1H NMR
(300 MHz, CDCl3), δ 1.35 (15H, d, J = 3.3 Hz), 2.26 (4H, t,
J = 4.94 Hz), 3.10 (4H, s, br), 4.0 (2H, app s?), 6.48 (1H, d, J =
8.8 Hz), 6.54 (1H, dd, J = 8.0, 4.9 Hz), 7.4–7.6 (8H, m), 8.0–8.3
(4H, m). 13C NMR (75.5 MHz, CDCl3), δ 5.3, 42.0, 51.5, 53.9
(d, J = 40.0 Hz), 95.3, 104.1, 110.1, 124.7 (d, J = 9.8 Hz), 125.3
(d, J = 42.20 Hz), 127.8, 131.8 (d, J = 8.7 Hz), 134.1, 144.7. FAB
MS: m/z 670 (MH)ϩ, 633/5 (M Ϫ Cl)ϩ, 597/9 (M Ϫ 2Cl)ϩ.
HRMS (FAB): found: m/z 670.1407, C32H40N3PCl2Rh requires
670.1392.
2-Diphenylphosphinobenzylideneaminopiperidine 7
2-Diphenylphosphinobenzaldehyde (0.3 g, 1.033 mmol) and N-
aminopiperidine (0.113 mL, 0.105 g, 1.044 mmol) were heated
in THF at reflux for 5 h. The reaction mixture was then cooled
and solvent removed to give a quantitative yield of 7 as a
colourless wax. Compound 7 was further purified by cooling a
solution (Et2O, 1 ml: hexane, 3 ml) in a freezer overnight to give
colourless crystals of 2-dppbap 7 (yield: 0.170 g, 0.456 mmol,
44%). C24H25N2Pؒ0.25Et2O requires: C, 76.80; H, 7.09; N, 7.17.
Found: C, 76.63; H, 6.46; N, 7.20%. IR (νmax/cmϪ1): 3051, 1584,
1550, 1570, 1477, 1464, 1448, 1375, 1348, 1259, 1242, 1164,
[RhCp*Cl2(Ph2PNHNHPy-P)] 10. Analytically pure
material was obtained by removing most of the solvent and
precipitating out the red solid using hexane (yield 93 mg, 57%)
C27H31N3PCl2Rh requires: C, 53.84; H, 5.19; N, 6.98; Found: C,
53.57; H, 5.12; N, 7.05%. 31P NMR (121.4 MHz, CDCl3), δ 77.1
(JP–Rh = 148 Hz). 1H NMR (300 MHz, CDCl3), δ 1.4 (15H, d),
5.75 (1H, d, J = 34.6 Hz, NH), 6.05 (1H, s, br, NH), 6.4 (1H, t,
J = 4.1 Hz), 6.75 (1H, d, J = 8.5 Hz), 7.15 (1H, t, J = 5.5 Hz), 7.4
(6H, m), 7.8 (1H, d, J = 3.9 Hz), 7.95 (4H, m). IR (KBr) ν/cmϪ1:
3318 (N–H stretch), 3056 (N–H stretch), 1598 (C–N [py]
stretch), 987 (P–N stretch). FAB MS: m/z 602 (Mϩ), 566
(M Ϫ Cl)ϩ, 531 (M Ϫ 2Cl)ϩ.
1
1119. 31P NMR (121.4 MHz, CDCl3), δ Ϫ12.9. H NMR (300
MHz, CDCl3), δ 1.4 (2H, m), 1.58 (4H, m), 2.94 (4H, appt, J =
5.6 Hz), 6.88 (1H, m), 7.0 (1H, t, J = 7.6 Hz), 7.3 (1H, m), 7.86,
(1H, m), 8.0 (1H, d, J = 4.7 Hz). 13C NMR (75.4 MHz, CDCl3),
δ 20.8, 21.6, 48.6, 122.3 (d, J = 4.3 Hz), 124.6, 125.2, 125.4
(d, J = 7.6 Hz), 125.8, 130.2 (d, J = 24 Hz), 130.3, 133.7 (J =
9.8 Hz), 133.9 (d, J = 8.0 Hz), 137.1 (d, J = 9.8 Hz). HRMS
(ESϩ): found: m/z 373.1834, (MH)ϩ requires 373.1834.
General procedure for synthesis of compounds of type
Cp*Rh(L)Cl2
[RhCp*Cl (Ph P(o-C H CH᎐N-c-NC H )-P)] 12. C H N -
᎐
2
2
6
4
5
10
34 40
2
PRhCl2 requires: C, 59.92; H, 5.92; N, 3.66. Found: C, 59.03;
H, 5.74; N, 3.66%. IR (νmax/cmϪ1): 3053, 1956, 1637, 1567,
1481, 1435, 1373, 1353, 1258, 1163, 1120, 1085. 31P NMR
A dichloromethane solution of ligand was added dropwise over
ca. 1 min to a solution of [Cp*RhCl2]2 in the same solvent.
3422
J. Chem. Soc., Dalton Trans., 2001, 3421–3429

View Article Online
(121.4 MHz, CDCl3), δ 30.5 (JP–Rh = 145 Hz). 1H NMR
(300 MHz, CDCl3), δ 1.37 (15H, d, J = 3.0 Hz), 1.6 (s, br), 1.7
(app. s), 2.7–2.9 (m, br), 7.2–7.8 (m, br), 8.1 (d, br, J = 8.0 Hz).
HRMS (ESϩ): found: m/z 645.1666, (M Ϫ Clϩ requires
645.1679 (spectrum was broadened and uninformative).
Results and discussion
(a) Synthesis of ligands
Ligand 2-dpphp 5 contains a P–N bond, and is prepared by the
addition of chlorodiphenylphosphine to 2-hydrazinopyridine in
the presence of triethylamine (Scheme 1). This ligand can be
General procedure for reaction of Cp*RhCl2L compounds with
silver salts
To a dichloromethane solution of the monodentate rhodium
complexes, 9–12 was added one (or in some cases two) equiv-
alents of silver salts. After stirring in the dark overnight, the
solutions were either filtered through Celite or centrifuged to
remove silver chloride. After removal of solvent and drying
in vacuo, the compounds were characterised spectroscopically.
In some cases, the complexes were isolated in analytically pure
form by addition of diethyl ether to concentrated dichloro-
methane solutions of the relevant compound.
Scheme 1
obtained in good purity after removal of triethylamine hydro-
chloride by-product, providing the reactions are carried out
under anhydrous, anaerobic conditions. Slow addition of
chlorodiphenylphosphine is required, to prevent disubstitution
of the amine. This ligand, which is fairly tolerant to moisture
and air, can be purified by recrystallisation from cold chloro-
form to give a moderate yield of analytically pure material. This
ligand was of considerable interest to us as it can provide a
comparison to 2-Ph2PNHPy, which we have studied in some
detail.10 Furthermore, the presence of the extra NH functional-
ity opens up the possibility for tridentate coordination and
supramolecular structures resulting from hydrogen bonding
between pyridine and NH moieties.
[Cp*RhCl(2-Ph2PPy-P,N)]ClO4 13. The compound was
obtained from 9 in analytically pure form by removing solvent
and washing with Et2O, and thus in quantitative yield. A good
quality crystal (suitable for X-ray crystallography) was grown
by slow diffusion (CH2Cl2–Et2O). C27H29NPRhCl2O4 requires:
C, 50.96; H, 4.59; N, 2.20. Found: C, 50.60; H, 4.79; N,
2.00%. IR (νmax/cmϪ1) 3058, 1638, 1615, 1571, 1482, 1436, 1377,
1262, 1095, 1020. 31P NMR (121.4 MHz, CDCl3), δ Ϫ11.6
(JP–Rh = 116 Hz). 1H NMR (300 MHz, CDCl3), δ 1.66 (15H, d,
J = 4.67 Hz), 7.0–8.2 (13H, m), 8.6 (1H, d, J = 5.2 Hz). FAB MS:
m/z 536/538 (M Ϫ ClO4)ϩ, 500/502 (M Ϫ ClO4 Ϫ Cl)ϩ. HRMS
(ESϩ): found m/z 536.0778, (M Ϫ Cl)ϩ requires 536.0781.
The synthesis of pippyphos 6 is even more straightforward
(Scheme 2). After heating a solution of 2-piperidinopyridine
[RhCp*Cl(Ph2PNHNHPy-P,N)]ClO4 14. This compound
obtained from 10 was purified by removing most of the solvent,
adding hexane and cooling in the fridge overnight. Crystals
could be obtained by layering a DCM solution with diethyl
ether. (Yield 43 mg, 26%.) There can be no doubt as to the
identification and high purity (spectroscopy, XRD evidence) of
14. However, it repeatedly gave chemical analysis with too much
carbon. C27H37Cl2N3O4PRh requires: C, 47.14; H, 4.63; N, 6.20.
Found: C, 48.67; N, 4.69; N, 6.31%. IR (νmax/cmϪ1) 3319, 3058,
Scheme 2
1599, 987. 31P NMR (121.4 MHz, CDCl3), δ 88.7 (d, JP–Rh
=
1
143 Hz). H NMR (300 MHz, CDCl3), δ 1.2 (15H, d, 1.3 Hz),
4.6 (1H, d, J = 4.6 Hz), 6.25 (1H, d, J = 3.4 Hz ), 7–8 (13H, m,
ArH), 8.7 (1H, d, J = 6.0 Hz). FAB MS: m/z 566 (M Ϫ ClO4)ϩ,
530 (M Ϫ Cl Ϫ ClO4)ϩ.
with diphenylphosphinomethanol overnight, analytically pure
pippyphos crystallises out of the reaction mixture on cooling.
Unbeknown to us, this ligand had already been prepared
directly from diphenylphosphine.14 However, its coordination
chemistry has not been studied to any extent.
X-Ray crystallography
Given the huge number of amines that are readily available
and the mild reaction conditions employed, the condensation
reaction between Ph2PCH2OH and amines is undoubtedly a
useful and under used method for phosphine synthesis. It is
complimentary to P–N bond formation, and allows the effect
of increasing ring size to be evaluated. It has recently been
applied by Smith and coworkers in the preparation of pyridine
functionalised diphosphines.16 Another interesting method to
prepare phosphines with P,N donor sets is the condensation of
phosphine-substituted aldehydes with amines/hydrazines. The
phosphino-hydrazone ligand 7 was synthesised by reaction of
commercially available 2-diphenylphosphinobenzaldehyde with
N-aminopiperidine (Scheme 3). The reaction proceeds very
Crystal structures were obtained using a Bruker SMART
diffractometer with graphite-monochromated Mo-Kα radi-
ation (λ = 0.71073 Å). Intensity data were collected using 0.3
or 0.15Њ width ω steps accumulating area detector frames
spanning a hemisphere of reciprocal space for all structures.
All data were corrected for Lorentz, polarisation and long
term intensity fluctuations. Absorption effects were corrected on
the basis of multiple equivalent reflections. Structures were
solved by direct methods and refined by full-matrix least
squares against F 2 (SHELXTL).15 In 11 there were serious
disorder problems associated with an acetone solvate; the
acetone was refined isotropically in two 50% occupancy orient-
ations with a common C᎐O group, the solvate protons were
᎐
not included in the refinement. In 13 the hydrogen atoms of
the CH2Cl2 solvate were not included in the refinement. In 14
the perchlorate anion was disordered and refined in two 50%
orientations.
Crystal data for all the compounds are given in Table 1.
CCDC reference numbers 165059–165063.
See http://www.rsc.org/suppdata/dt/b1/b104523g/ for crystal-
lographic data in CIF or other electronic format.
Scheme 3
J. Chem. Soc., Dalton Trans., 2001, 3421–3429
3423

View Article Online
Table 1 Crystal data for compounds 8, 10, 11, 13 and 14
8
10
11
13
14
Empirical formula
M
C28H32Cl4N2PRh
672.4
C27H31Cl2N3PRh
602.33
C35H46Cl2N3O1.5PRh
737.53
C27.50H30Cl3NO4PRh
678.76
C31H41Cl2N3O5PRh
740.45
Crystal system
Space group
Triclinic
P1
Monoclinic
P21/n
Monoclinic
P21/n
Monoclinic
C2/c
Monoclinic
P21/n
¯
a/Å
b/Å
c/Å
α/Њ
β/Њ
γ/Њ
U/Å
8.5863(4)
10.3041(5)
18.7795(9)
97.6060(10)
94.2100(10)
112.4040(10)
1508.60(12)
2
20.1794(3)
14.2417(4)
20.7001(2)
90
110.993(2)
90
5554.1(2)
8
0.885
13.0569(3)
19.4344(5)
13.9166(3)
90
91.1350
90
3530.69(14)
4
0.713
32.668(2)
8.1102(5)
24.401(2)
90
102.928(2)
90
6300.9(7)
8
0.878
16.6792(11)
9.7256(6)
21.2011(13)
90
97.3360(10)
90
3411(4)
4
0.746
Z
µ/mmϪ1
0.993
Reflections measured
Independent reflections
Final R 1,
9484
4353
0.0447
0.1102
23956
7976
0.0639
0.1019
14889
5036
0.0589
0.1414
13306
4515
0.0574
0.1460
14412
4891
0.0411
0.0820
wR2[I > 2σ(I)]
Table 2 31P NMR data and coordination modes of the six ligands studied
a
a
Ligand
δP
Coordination ϩ [Cp*RhCl2]2
Bidentate (8)
δP(1JP–Rh
)
Coordination ϩ AgClO4
δP(1JP–Rh
)
26.2
89.1 (145)
N/A
—
Ϫ4.6
Monodentate (9)
Monodentate (10)
28.9 (141)
77.1 (148)
Bidentate (13)
Bidentate (14)
Ϫ11.6 (116)
49.6
88.7 (143)
Ϫ27.1
Ϫ12.9
Monodentate (11)
Monodentate (12)
32.5 (139)
30.5 (145)
Mixture (2 compounds)
Mixture (2 compounds)
27.1 (148)[40%]
Ϫ13.3 (128) [60%]
35.5 (138) [70%]
33.1 (140) [30%]
a Chemical shifts in ppm relative to external H3PO4, coupling constants in Hz.
cleanly to give a quantitative yield of essentially pure material.
The ligand can be purified further by recrystallisation. This
ligand was of additional interest as a related phosphino-imine
ligand, in which the c-NC5H10 group of 7 was replaced with a
c-C6H12 group, has very recently been found to be the best
ligand for palladium catalysed carbostannylation of alkynes.6
Other nitrogen substituents were found to be significantly less
selective or efficient. Phosphino-hydrazone ligands have also
been used in Pd catalysed allylic alkylation.17 The remaining
two ligands, 2-Ph2PPy and 2-dppap have been prepared pre-
viously, with 2-diphenylphosphinopyridine, being commercially
available. The 31P NMR chemical shift of each of the ligands is
shown in Table 2.
phosphorus NMR spectra on coordination to the rhodium
centres (Table 2). In the case of 2-dpphp and 2-PPh2Py, the com-
plexes display a broadened doublet in their 31P NMR spectra,
which may suggest either transient coordination to form a
cationic chelate complex or perhaps the presence of two inter-
converting structures which differ only in the conformation of
the ligand. The other complexes all show sharp doublets with
coupling constants of 139–148 Hz. The 1H NMR spectra of the
monodentate complexes 9–12 broadly resemble those of the
corresponding ligands. In particular, there is no large downfield
shift associated with protons adjacent or near to the nitrogen
substituents, which could be indicative of N coordination.
There is also no great change in the position of C᎐N stretching
᎐
vibrations in the IR spectrum in going from free ligand to
monodentate complex (although in the case of 12 an extra band
is observed in the IR spectrum at 1637 cmϪ1). All the com-
plexes were additionally characterised by elemental analysis
which gave good agreement with the proposed formulae in each
case. The FAB mass spectra of the complexes generally gave
(b) Reaction with [Cp*RhCl2]2
Four of the ligands react with [Cp*RhCl2]2 to give monodentate
complexes 9–12 of type Cp*RhCl2(L) (Scheme 4). All ligands
show a downfield shift (between 35 and 50 ppm) in their
3424
J. Chem. Soc., Dalton Trans., 2001, 3421–3429

View Article Online
Table 3 Selected bond lengths (Å) and angles (Њ) for 8
Rh(1)–N(14)
P(1)–N(1)
2.104(4)
1.679(4)
2.286(1)
2.217(6)
2.169(5)
2.186(6)
2.160(6)
2.389(1)
1.339(7)
N(14)–Rh(1)–P(1)
P(1)–Rh(1)–Cl(1)
C(13)–N(1)–P(1)
N(14)–C(13)–N(1)
N(1)–P(1)–Rh(1)
80.65(1)
93.37(5)
120.1(4)
117.3(4)
100.8(2)
Rh(1)–P(1)
Rh(1)–C(21)
Rh(1)–C(22)
Rh(1)–C(23)
Rh(1)–C(24)
Rh(1)–Cl(1)
C(13)–N(14)
Scheme 4
ions due to (M Ϫ Cl)ϩ and (M Ϫ 2Cl)ϩ [and in several cases
(M ϩ Na)ϩand (MH)ϩ] and therefore also fully supports the
structures 9–12.
In the case of the ligand 2-dppap 2 a bidentate complex is
formed in which one of the chloride ions is found outside the
coordination sphere. The particularly large downfield shift from
the free ligand (δ 26.2 to 89.1) is consistent and diagnostic of
the formation of a five-membered chelate ring.10,18,19 The IR
spectrum of this complex also shows ν(C᎐N) moving from 1587
᎐
(free ligand) to 1612 cmϪ1. In a study which focused on the
coordination behaviour of this ligand with a wide selection of
other metals, a large downfield shift in the phosphorus NMR
Fig. 1 Crystal structure of 8.
spectra and an increase in the value of ν(C᎐N) were always
᎐
ligand [length of hydrogen bond: N(1)H1–Cl(2) 2.126(8),
N(1)–Cl(2) 3.101(4) Å]. This type of hydrogen bond has been
observed previously in complexes of 2-dppap.10 The nitrogen
that is bound to phosphorus shows planar geometry and the
P–N bond length is somewhat shorter in 8 than that found in
the free ligand [1.679(4) vs. 1.705(3) Å]. Rh–Cl and Rh–P bond
lengths are very similar to those found in Cp*Rh(P(OEt)3)Cl2.21
The Rh–C bond distances vary from 2.160(6) to 2.238(6) Å,
which reflects whether the carbon atom is located directly
opposite the phosphorus donor (strong trans influence) or the
chloride ligand. Selected bond lengths and angles are found in
Table 3.
Complex 10 was also characterised by X-ray diffraction. The
molecular structure of 10 is particularly interesting, as the
complex forms hydrogen bonded dimers via the acidic NH
functionalities on the ligand (a good H-bond donor) and both
the pyridine and chloride ligands (H-bond acceptors). This
arrangement is shown in Fig. 2. The structure shows how the 2-
aminopyridine units are complementary to each other. It is the
presence of these hydrogen bonds that enables the two com-
plexes to associate together. The aromatic NH groups [N(2)H
and N(32)H in Fig. 2(b)] form an intermolecular bond to the
pyridine nitrogen belonging to the other complex in the dimer
[N(2)H ؒ ؒ ؒ N(44) 2.37(6), N(32)H ؒ ؒ ؒ N(14) 2.13(2) Å; N(32)–
N(32)H ؒ ؒ ؒ N(14) 172(6), N(2)–N(2)H ؒ ؒ ؒ N(44) 130(6)Њ].
There is a slight difference between the two members of the
dimer regarding the H-bonding between the P bound NH and
the chloride ligands. In molecule 1 [contains labels Rh(1), P(1)
etc.], the hydrogen bond is a simple bond, between one Cl and
one H. In molecule 2 [denoted with labels Rh(2), P(2) etc.], the
hydrogen bonding is less strong, and bifurcated between two Cl
and one H. The relevant parameters for the intramolecular
bonds are N(1)H ؒ ؒ ؒ Cl(1) 2.53(7), N(31)H ؒ ؒ ؒ Cl(3) 2.88(9),
N(31)H–Cl(4) 2.92(1) Å; N(1)–N(1)HؒؒؒCl(1) 119(6), N(31)–
N(31)H ؒ ؒ ؒ Cl(3) 106(6), N(31)–N(31)H ؒ ؒ ؒ Cl(4) 92(6)Њ. The
structures of each of the two rhodium complexes shown in Fig.
2 are both approximately octahedral in geometry and do not
show any particularly unusual bond lengths or angles (Table 4).
indicative of P,N chelation. Another useful diagnostic tool to
assign N coordination is a considerable downfield shift for the
proton ortho to the aromatic nitrogen. This is observed in com-
plex 8 (ortho H shifted from δ 7.97 in 2 to 11.87 in 8). The
structure of this complex demonstrates the strong preference
for chelation shown by this ligand. It has been found previously
that even non-symmetrical diphosphine ligands can exhibit
monodentate coordination to Cp*RhCl2 fragments, so 2 can be
considered to form quite a strong chelate. It is likely that the
formation of a five membered ring enhances the strength of a
phosphine/pyridine chelate ring.10,18,19 In actual fact, it has
recently been found that the related ligand Ph2POPy shows iden-
tical chemistry with this metal fragment. An X-ray crystal struc-
ture determination of [Cp*Rh(Cl)(η2-Ph2POPy)]Cl revealed
similar molecular parameters and the presence of the strong
five-membered P,N chelate.20
(c) X-Ray crystal structures of Cp*RhCl2(L) complexes
Recrystallisation of complex 8 from CH2Cl2–Et2O (slow dif-
fusion) yielded crystals suitable for
a crystal structure
determination. The crystal structure of compound 8 (Fig. 1)
shows the rhodium centre to be approximately octahedral with
the Cp* ligand occupying three facial positions of this octa-
hedron. The structure confirms the η2 chelation mode for ligand
2 and the expected η5-coordination of the Cp* ligand. The
crystallographic bite angle of this angle, P(1)–Rh(1)–N(14) is
80.65(12)Њ, which is significantly smaller than the idealised 90Њ
for octahedral geometry. It is therefore clear that the formation
of a somewhat constrained metallocyclic ring is actually bene-
ficial to chelate strength, as ligand 2 is clearly the strongest
chelating ligand of those we have studied (none of the other
ligands can displace chloride from the coordination sphere).
The angle N(14)–Rh(1)–Cl(1), which is 84.43(12)Њ is also sig-
nificantly smaller than in idealised octahedral geometry. The
structure clearly shows the location of the chloride counter-ion,
which is found outside the coordination sphere, and hydrogen
bonded to the NH functionality present within the phosphine
J. Chem. Soc., Dalton Trans., 2001, 3421–3429
3425

View Article Online
Table 4 Selected bond lengths (Å) and angles (Њ) for 10
Table 5 Selected bond lengths (Å) and angles (Њ) for 11
Rh(1)–P(1)
Rh(1)–Cl(2)
P(1)–C(7)
N(1)–N(2)
Rh(2)–P(2)
Rh(2)–Cl(4)
P(2)–C(31)
N(31)–N(32)
Rh(1)–Cl(1)
P(1)–N(1)
2.314(3)
2.415(3)
1.800(10)
1.396(10)
2.324(3)
2.395(3)
1.807(9)
1.393(10)
2.403(3)
1.679(8)
1.839(9)
1.39(1)
P(1)–Rh(1)–Cl(2)
P(1)–Rh(1)–Cl(1)
Cl(1)–Rh(1)–Cl(2)
N(1)–P(1)–Rh(1)
N(2)–N(1)–P(1)
C(13)–N(2)–N(1)
N(14)–C(13)–N(2)
P(2)–Rh(2)–Cl(4)
Cl(3)–Rh(2)–Cl(4)
P(2)–Rh(2)–Cl(3)
N(31)–P(2)–Rh(2)
N(32)–N(31)–P(2)
C(43)–N(32)–N(31)
N(44)–C(43)–N(32)
89.14(9)
86.69(9)
91.23(10)
108.2(3)
119.5(6)
119.1(8)
115.2(9)
86.27(9)
91.61(10)
88.52(9)
106.8(3)
123.1(7)
119.5(8)
114.8(9)
Rh(1)–P(1)
2.339(1)
2.428(1)
2.411(1)
2.166(5)
2.156(5)
2.229(5)
2.193(5)
2.174(5)
1.519(6)
1.339(7)
P(1)–Rh(1)–Cl(1)
C(13)–N(14)–C(15)
Cl(1)–Rh(1)–Cl(2)
P(1)–Rh(1)–Cl(2)
89.86(5)
108.2(4)
88.53(5)
87.53(5)
Rh(1)–Cl(1)
Rh(1)–Cl(2)
Rh(1)–C(26)
Rh(1)–C(27)
Rh(1)–C(28)
Rh(1)–C(29)
Rh(1)–C(30)
C(13)–N(13)
C(20)–N(21)
P(1)–C(1)
N(2)–C(13)
Rh(2)–Cl(3)
P(2)–N(31)
P(2)–C(37)
C(43)–N(44)
[P(1)–Rh(1)–Cl(1) 89.86(5), P(1)–Rh(1)–Cl(2) 87.53(5), Cl(2)–
Rh(1)–Cl(1) 88.53(5)Њ]. Selected bond lengths and angles are
shown in Table 5.
The average Rh–P bond length from the monodentate struc-
tures [2.326(3) Å] is clearly considerably longer than that found
in bidentate 8 [2.286(1)]. This is presumably a reflection of the
increased bond strength that is found in chelate complexes. The
2.417(3)
1.663(9)
1.811(9)
1.32(1)
cationic nature of 8 may also enhance the P
Rh donor inter-
action. This trend is also present, but somewhat less pro-
nounced, in the Rh–Cl bond lengths [average for 10 and 11 is
2.411(3) vs. 2.389(1) Å]. Our results clearly show that the major-
ity of P,N ligands will form a monodentate complex on reaction
with [Cp*RhCl2]2, but that a bidentate cationic complex can
be formed if the ligand is quite strongly chelating. It was now
of interest to determine the affinity of the nitrogen donors for
the Cp*RhCl2 fragment by removal of the competing chloride
ligands.
(d) Addition of silver salts to monodentate Cp*RhCl2(L)
complexes
L ؍ 2-PPh2Py. When silver perchlorate was added to a
dichloromethane solution of monodentate 9 complete conver-
sion to a new species was observed by 31P NMR spectroscopy
(Scheme 5). The formation of a chelate complex was deduced
Fig. 2 (a) Crystal structure of 10 and (b) of hydrogen-bonded dimers
formed in 10.
Scheme 5
from the following spectroscopic data. The 31P NMR spectrum
was especially informative as it shows a doublet that is shifted
significantly upfield from the starting complex 9 (δ Ϫ11.6 vs.
ϩ28.9). An upfield shift in the 31P NMR spectrum (relative to a
monodentate complex) is known to be diagnostic for a four-
membered chelate complex.19,22 In addition, the coupling
constant JP–Rh was reduced in magnitude from that found in 9
(116 vs. 141 Hz). Although there are no comparable rhodium
systems known, it seems likely that a reduction in coupling
constant may also be a diagnostic feature of a four-membered
rhodium–phosphorus containing ring (see remaining discus-
sion). Unusually small coupling constants have been found in
the four-membered ring system (dppm)Pt(CH3)2 [dppm = bis-
(diphenylphosphino)methane] when compared to other diphos-
Fig. 3 Crystal structure of 11.
All the nitrogen atoms found in the structure are planar, as
a result of the sp2 character of P–N, C᎐N and N᎐C–N–N–P
᎐
᎐
bonding environments.
The crystal structure of 11 is shown in Fig. 3. This confirms
the mononuclear, monodentate structure of this compound in
which the piperazine ring adopts a chair conformation. The
rhodium centre is octahedral and the angles between phos-
phorus and chlorine ligand are fairly close to the ideal 90Њ
1
phine dimethylplatinum complexes.22 The H NMR spectra of
both monodentate and chelate complex are fairly similar in
appearance (although the aromatic protons show more signals
in the chelate complex). This is somewhat surprising as it has
3426
J. Chem. Soc., Dalton Trans., 2001, 3421–3429

View Article Online
Table 6 Selected bond lengths (Å) and angles (Њ) for one of the
independent molecules of 13
Rh(1)–P(1)
Rh(1)–Cl(1)
Rh(1)–N(14)
Rh(1)–C(21)
Rh(1)–C(22)
Rh(1)–C(23)
Rh(1)–C(24)
Rh(1)–C(25)
C(13)–N(14)
C(13)–P(1)
2.338(2)
2.394(2)
2.132(6)
2.171(8)
2.195(8)
2.195(8)
2.142(8)
2.171(8)
1.33(10)
1.819(7)
P(1)–Rh(1)–N(14)
N(14)–C(13)–P(1)
P(1)–Rh(1)–Cl(1)
N(14)–Rh(1)–Cl(1)
C(13)–P(1)–Rh(1)
C(13)–N(14)–Rh(1)
67.3(2)
102.0(5)
90.94(8)
84.4(2)
84.4(3)
105.8(5)
Fig. 5 Crystal structure of 14.
to the rhodium centre. The IR spectrum of this new species is
also informative and confirms the chelate structure 14. In par-
ticular, the position of ν(C᎐N) which is found at 1599 cmϪ1 in
᎐
10 and the free ligand, has shifted in frequency to 1611 cmϪ1.
This type of shift is diagnostic of aromatic N coordination for
2-dppap, and it likely that this is also the case for 2-dpphp. The
identity of this complex was further confirmed by the FAB
mass spectrum which shows peaks due to (M Ϫ ClO4)ϩ and
(M Ϫ Cl Ϫ ClO4)ϩ.
Fig. 4 Crystal structure of 13.
been noted that N-coordinated pyridylphosphines sometimes
show a shift of the ortho-N proton to higher frequency. The IR
spectra of pyridyl phosphines is not especially informative in
X-Ray crystal structure of [Cp*RhCl(2-dpphp-P, N)]ClO4 14.
Since the crystal structure of the monodentate complex 10 dis-
played an interesting hydrogen bonded dimer structure, it was
of considerable interest to determine the crystal structure of
complex 14. This structure also allowed us to make a com-
parison with structural data obtained in the other chelate com-
plexes studied and to fully corroborate the proposed structure.
The crystal structure of 14 clearly confirms the bidentate P,N
coordination of 2-dpphp and is shown in Fig. 5. The are two
independent molecules within the structure, but the molecular
parameters for each are very similar. The structure also shows
that this ligand is still capable of forming a mono-dimensional
hydrogen bonding network even when the pyridine nitrogen,
which is a key hydrogen bond acceptor in the solid state struc-
ture of [Cp*RhCl2(2-dpphp-P)]), is bound to the metal. The
complexes form infinite chains in which an intermolecular
hydrogen bond [N(1)–Cl(1) 3.415(4), N(1)H–Cl(1) 2.47(2) Å]
between the P bound N(1)H and the chloride ligands on a
separate rhodium complex link the chains together. The N-
functionality which is bound to the pyridine ring, N(2)H,
additionally forms a hydrogen bond to the perchlorate counter-
ion [N(2)–O(2) 2.944(7), N(2)H–O(2) 1.99(1) Å]. These com-
plexes represent an interesting example of a supramolecular
architecture that can be modified from strong H-bonded dimers
to infinite chains by altering the coordination mode of the
ligand. The coordination chemistry of 2-dpphp and its deriv-
atives will therefore be studied further in due course. Complex
14 has a shorter Rh–P bond length than in the monodentate
structure 10 [Rh(1)–P(1) 2.305(1) vs. 2.314(3), 2.324(3) Å]. The
Rh–Cl bond distance on the other hand is slightly elongated
[Rh(1)–Cl(1) 2.417(1) Å in 14 vs. average bond length of
2.408(3) Å in 10]. This is in contrast with the other two chelate
structures and is presumably a reflection of the hydrogen bond-
ing that is taking place between the chloride ligands and NH
from an adjacent molecule. Other relevant bond lengths and
angles are shown in Table 7.
assigning coordination behaviour as ν(C᎐N) does not normally
᎐
shift to higher frequency as is found with 2-dppap. The IR
spectrum of 13 shows bands at 1638, 1615 and 1571 cmϪ1,
which are presumably due to C᎐N and C᎐C vibrations. This is
᎐
᎐
fairly similar to the monodentate complex, (1636, 1615 and
1587 cmϪ1) although there is clearly a shift for the band found
at the lowest wavenumber. There is also a very intense band
Ϫ
(1095 cmϪ1) attributed to the ClO4 counter-ion. The formu-
lation of 13 was also supported by microanalytical and mass
spectral data.
X-Ray crystal structure of [Cp*RhCl(2-PPh2Py-P,N)]ClO4.
As the spectroscopic data was not quite as conclusive as we may
have hoped, crystals of 13 were grown by slow diffusion
(CH2Cl2–Et2O). The crystal structure of 13, which is shown in
Fig. 4, clearly confirms the bidentate, mononuclear structure
that was suggested from the spectroscopic data. The P,N
chelate shows, as might be expected for a four-membered ring, a
severe distortion from idealised octahedral geometry [N(14)–
Rh(1)–P(1) 67.3(2)Њ]. The Rh–Cl bond length is significantly
shorter than that found in the neutral monodentate complex 10.
However, the Rh–P bond length, which was much shorter in 8,
is of similar length to that found in monodentate 10 and 11
[Rh(1)–P(1) 2.338(2) vs. average bond length of 2.326(3) Å].
These values support the idea that the shortened bond length
found in 8 is predominantly due to its particularly strong
chelate ring. The strained chelate ring obtained here also has a
longer Rh–N bond length than that of 8 [Rh(1)–N(14) 2.104(4)
Å in 8 vs. 2.132(6) Å in 13]. Selected bond lengths and angles
are shown in Table 6.
L ؍ 2-dpphp. On treatment of the monodentate complex 10
with one equivalent of silver perchlorate, a single new species is
formed as judged by 31P NMR spectroscopy. This spectrum
shows a doublet (δ 88.7, J = 143 Hz) that is shifted slightly down
field from the monodentate complex. The position of the aro-
matic proton ortho to the pyridine nitrogen has moved down-
field from δ 7.8 (complex 10) to δ 8.7 in the cationic complex
which we assign structure 14. It is reasonable to assume that
this downfield shift could only be due to nitrogen coordination
L ؍ pippyphos. On addition of either one or two equivalents
of AgBF4, the starting complex 11 is converted into two new
species as monitored by 31P NMR spectroscopy. One of these
species appears as a broadened doublet that can be found sig-
nificantly upfield from the 31P NMR resonance of the starting
1
material (δ Ϫ13.3 vs. 32.5). This doublet has a smaller JP–Rh
J. Chem. Soc., Dalton Trans., 2001, 3421–3429
3427

View Article Online
Table 7 Selected bond lengths (Å) and angles (Њ) for 14
(broadened NMR spectra with poor signal to noise ratio). For
a given transition metal, a hemilabile ligand should perhaps be
further defined according to whether it shows: (a) unassisted
stable chelate coordination, (b) assisted stable chelate coordin-
ation (assisted by removal of competing ligands), or (c) assisted
fluxional weak chelate coordination. These three classes are
not, by any means, peculiar to Cp*RhCl2 centres. For example,
various researchers, including ourselves, have observed that
some P,N ligands sometimes react with platinum precursors
such as (cod)PtCl2 (cod = cycloacta-1,5-diene) or (PhCN)2PtCl2
to give chelate complexes (η2-L)PtCl2 and [(η2-L)2Pt]2Cl (class
a).10,12b,18 Other ligands form monodentate complexes (η1-L)2-
PtCl2, which on treatment with silver salts give chelates, [(η2-
L)2Pt]2ϩ and [(η2-L)(η1-L)PtCl]ϩ (class b). Some P,N ligands do
not yield isolable chelate complexes that are stable in the pres-
ence of competing ligands (e.g. water) even when halide ligands
are removed (class c). These distinctions are useful when con-
sidering the design of phosphorus ligands in catalysis: a strong
chelate is required in asymmetric catalysis or as replacement as
a diphosphine. A conventional hemilabile ligand can often
increase the reactivity of catalyst in comparison to a kinetically
inert diphosphine system. In some instances a very weak chelate
is required to generate a very reactive transition metal complex.
We now hope that we will be able to apply a specific type of
hemilabile ligand to its ideal task in organometallic chemistry
and catalysis. Indeed, since the completion of this work, we
have prepared weakly chelating (class c) ligands that contain a
bulky and electron rich phosphine moiety and shown palladium
complexes derived from them to catalyse cross-coupling
reactions of aryl chlorides: a reaction that does not occur if
normal monodentate or bidentate phosphine ligands are used.23
Further studies on the application of functionalised phos-
phino-amines in catalysis and coordination chemistry are now
underway.
Rh(1)–P(1)
Rh(1)–Cl(1)
Rh(1)–N(14)
Rh(1)–C(21)
Rh(1)–C(22)
Rh(1)–C(23)
Rh(1)–C(24)
Rh(1)–C(25)
C(13)–N(14)
N(2)–N(1)
2.305(1)
2.417(1)
2.15(4)
P(1)–Rh(1)–N(14)
P(1)–N(1)–N(2)
P(1)–Rh(1)–Cl(1)
N(14)–Rh(1)–Cl(1)
N(1)–N(2)–C(13)
N(1)–P(1)–Rh(1)
88.0(1)
112.4(3)
95.90(5)
87.5(1)
117.9(4)
107.7(2)
2.183(5)
2.167(5)
2.235(5)
2.221(5)
2.161(5)
1.348(6)
1.424(6)
1.697(4)
P(1)–N(1)
coupling constant than 11 (128 vs. 139 Hz). We have already
discussed how this NMR data is highly diagnostic of a four-
membered chelate complex in a variety of other systems. We
therefore can suggest, on this evidence alone, that a P,N chelate
complex is likely to account for this signal in the 31P NMR
spectrum. Unfortunately, we have not been able to separate the
1
two products formed in this reaction, and both the H NMR
spectrum, which is broad and messy, and IR spectrum are not
informative in fully corroborating the proposed structure. The
other compound detected from this reaction displays a sharp
doublet in its 31P NMR spectrum. Its chemical shift (δ ϩ27.1)
1
and JP–Rh (148 Hz) are more reminiscent of the starting com-
plex 11. There are several possible structures which could be
formed in this case: a simple P-monodentate complex in which
BF4 binds weakly to the Rh centre, a cationic aqua complex
(formed from adventitious water), or P,N-coordinated Rh
complexes in which either the pyridine or the other piperazine
nitrogen coordinate. In addition, compounds in which the
nitrogen donors bridge two rhodium centres are also eminently
feasible. Despite several attempts, we have been unable to make
this halide abstraction go cleanly or purify the two products
formed. Thus we can only speculate as to the structures formed.
We suggest that pippyphos is a weaker chelating ligand than
those discussed thus far, and that this prevents us from isolating
a chelate complex.
Acknowledgements
We wish to thank the EPRSC for financial support (M.L.C),
Johnson Matthey for loan of rhodium() chloride, the EPRSC
mass spectrometry service (Swansea), the JREI for equipment
grants, and Dr Steve Aucott, who donated a large sample of
2-dppap, and [Cp*RhCl2]2, thus making this work possible.
L ؍ 2-dppbap. On addition of one equivalent of silver tetra-
fluoroborate to complex 12, there are two new doublet peaks
observed at fairly similar shift in the 31P NMR spectra of the
resulting solution. Removal of the solvent gives an orange
solid, which displays several bands in the region typical for
References
ν(C᎐N) stretching vibrations. This is tentative evidence that
᎐
the two complexes formed may differ in the coordination of
the nitrogen atoms. The mass spectrum of this solid supports
a cationic complex of formulation [Cp*Rh(Cl)(2-dppbap)]-
BF4. Unfortunately, attempts to separate these two complexes
by size exclusion chromatography or recrystallisation failed.
Although it is tempting to assign this mixture as the two
1 For reviews, see: K. N. Gavrilov and A. I. Polosukhin, Russ. Chem.
Rev., 2000, 69, 661; P. J. Guiry, M. McCarthy, P. M. Lacey, C. P.
Saunders, S. Kelly and D. J. Connolly, Curr. Org. Chem., 2000, 4,
821; P. Espinet and J. K. Soulantica, Coord. Chem. Rev., 1999, 193,
499.
2 E. Drent, P. Arnoldy and P. H. M. Budzelaar, J. Organomet.
Chem., 1993, 455, 247 and references therein; A. Dervisi, P. G.
Edwards, P. D. Newman and R. P. Tooze, J. Chem. Soc., Dalton.
Trans., 2000, 523.
complexes that would form if both sp2 C᎐N and sp3 C–N
᎐
coordination could occur, the evidence we have obtained is not
sufficient to characterise these products.
3 A. Lightfoot, P. Schnider and A. Pfaltz, Angew. Chem., Int. Ed.,
1998, 37, 2897.
4 J. M. J. Williams, Synlett, 1996, 705 and references therein; G. J.
Dawson, C. G. Frost and J. M. J. Williams, Tetrahedron Lett., 1993,
34, 3149; J. Sprintz and G. Helmchen, Tetrahedron Lett., 1993, 34,
1769; P. Von Matt and A. Pfaltz, Angew. Chem., Int. Ed. Engl., 1993,
32, 566; M. Sawamura and Y. Ito, Chem. Rev., 1992, 92, 857.
5 J. M. Brown, D. I. Hulmes and T. P. Layzell, J. Chem. Soc., Chem.
Commun., 1993, 1673; I. Beletskaya and A. Pelter, Tetrahedron,
1997, 35, 4957.
6 H. Yoshida, E. Shirakawa, T. Kurahashi, Y. Nakao and T. Hiyama,
Organometallics, 2000, 19, 5671 and references therein.
7 T. Tsuda, S. Morikawa and T. Saegusa, J. Chem. Soc., Chem.
Commun., 1989, 9.
8 For a review, see: H. Cheng and Z. Zhang, Coord. Chem. Rev., 1996,
147, 1; see also: J. P. Farr, M. M. Olmstead and A. L. Balch, Inorg.
Chem., 1983, 22, 1229.
9 A. S. C. Chan, C. Chen, R. Cao, M.-R. Lee, S. M. Peng and G. H.
Lee, Organometallics, 1997, 16, 3469.
(e) Conclusions and summary
The complexes prepared in this study have allowed a greater
understanding of the varying affinity of P,N ligands to chelate
rhodium centres.
We have synthesised two new multifunctional ligands which
consist of PNx (x = 1,2 or 3) donor atoms. These ligands, along
with the three known ligands, smoothly react with [Cp*RhCl2]2
to give complexes of type Cp*RhCl2(L). In one instance, a
bidentate complex is formed indicating that the ligand 2-dppap,
strongly favours a chelate coordination mode. The other ligands
all form monodentate complexes. On addition of one equiv-
alent of AgClO4 (or AgBF4), a variety of coordination
behaviours were observed: a relatively strong chelate (well
defined NMR spectra), and weak fluxional chelates (probably)
3428
J. Chem. Soc., Dalton Trans., 2001, 3421–3429

View Article Online
16 S. E. Durran, M. B. Smith, A. M. Z. Slawin and J. W. Steed, J. Chem.
10 S. M. Aucott, A. M. Z. Slawin and J. D. Woollins, J. Chem. Soc.,
Dalton Trans., 2000, 2559 and references therein.
Soc., Dalton Trans., 2000, 2771.
11 P. Wehman, R. E. Rulke, V. E. Kaasjager, P. C. J. Kamer,
H. Kooijmann, A. Spek, C. J. Elsevier, K. Vrieze and P. W. N. M.
Van Leeuwen, J. Chem. Soc., Chem. Commun., 1995, 331.
12 (a) A. M. Porte, J. Reibenspies and K. Burgess, J. Am. Chem. Soc.,
1998, 120, 9180; see also: (b) A. J. Blacker, M. L. Clarke, M. S.
Humphries, M. S. Loft, M. F. Mahon and J. M. J. Williams, Chem.
Eur. J., 2000, 6, 353.
13 C. White, A. Yates and P. M. Maitlis, Inorg. Synth., 1992, 29, 228.
14 S.-M. Kuang, Z. Z. Zhang, Q. G. Wang and T. C. Mak, Inorg.
Chem., 1998, 37, 6090.
17 T. Mino, W. Imiya and M. Yamashita, Synlett, 1997, 5, 583.
18 G. K. Anderson and R. Kumar, Inorg. Chem., 1984, 23, 4064.
19 P. E. Garrou, Chem. Rev., 1981, 81, 229.
20 P. Cheshire, A. M. Z. Slawin and J. D. Woollins, unpublished
results.
21 S. Ogo, T. Suzuki, Y. Ozawa and K. Isobe, Inorg. Chem., 1996, 35,
6093.
22 S. Hietkamp, D. J. Stufkens and K. Vrieze, J. Organomet. Chem.,
1979, 169, 107.
23 M. Rodriguez
i Zubiri, M. L. Clarke, D. F. Foster, D. J.
15 SHELXTL, structure solution and determination package, Bruker
AXS, Madison, WI, 1999.
Cole-Hamilton, A. M. Z. Slawin and J. D. Woollins, J. Chem. Soc.,
Dalton Trans., 2001, 969.
J. Chem. Soc., Dalton Trans., 2001, 3421–3429
3429