An acidity scale of tetrafluoroborate salts of phosphonium and iron hydride compounds in [D2]dichIoromethane

Article abstract of DOI:10.1002/chem.200601484

Equilibrium constants (K) for reactions between acids and the conjugate base forms of a number of phosphonium salts, [HPR₃][BF₄], and iron hydrides, [Fe(CO)₃H(PR₃)₂][BF ₄], in CD₂Cl₂ have been determined by means of ³¹P and ¹H NMR spectroscopy at 20°C. The anchor compound chosen for pK^CD2Cl2 determinations was [HPCy ₃][BF₄] with a pK^CD2Cl2 value of 9.7, as assigned by literature convention (Cy: cyclohexyl). A continuous scale of pK^CD2Cl2 values covering the range from 9.7 to -3 was created and correlated with the ΔH values reported by Angelici and co-workers and literature pK_a values. The pK_CD2Cl2 values for 15 other hydride or dihydrogen complexes of the iron group elements and of diethyl ether were also placed on this scale. The crystal structures of [Fe(CO) ₃H(PCy₂Ph)₂][BF₄] and [Fe(CO) ₃(PCy₂Ph)₂] revealed that the frans-oriented, bulky, unsymmetrical phosphane ligands distort the equatorial plane of the complexes. The acidity of iron carbonyl hydrides is an important feature of the reactions of iron hydrogenase enzymes.

Full text of DOI:10.1002/chem.200601484

DOI: 10.1002/chem.200601484
An Acidity Scale of Tetrafluoroborate Saltsof Phosphonium and Iron
Hydride Compoundsin [D ]Dichloromethane
2
Tianshu Li, Alan J. Lough, and Robert H. Morris*[a]
2
2
Abstract: Equilibrium constants (K)
for reactions between acids and the
conjugate base forms of a number of
scale of pKCD Cl values covering the
range from 9.7 to ꢀ3was created and
correlated with the DH values reported
by Angelici and co-workers and litera-
of diethyl ether were also placed on
this scale. The crystal structures of
[Fe(CO)3H
G
[BF4]
and
phosphonium salts, [HPR3]
A
[Fe(CO)3A(PCy2Ph)2] revealed that the
CHTREUNG
ture pKa values. The pKCD Cl values for
15 other hydride or dihydrogen com-
plexes of the iron group elements and
trans-oriented, bulky, unsymmetrical
phosphane ligands distort the equatori-
al plane of the complexes. The acidity
of iron carbonyl hydrides is an impor-
tant feature of the reactions of iron hy-
drogenase enzymes.
2
2
iron hydrides, [Fe(CO)3H(PR3)2]
A
ACHTREUNG
in CD2Cl2 have been determined by
1
means of 31P and H NMR spectrosco-
py at 208C. The anchor compound
chosen for pKCD Cl determinations was
2
2
Keywords: acidity
ligands · hydride ligands · iron ·
·
carbonyl
[HPCy3]ACHTREUNG
[BF4] with a pKCD Cl value of
2
2
9.7, as assigned by literature conven-
tion (Cy: cyclohexyl). A continuous
phosphanes
Introduction
phosphanes in which pKaaq values were reported to give a
1
pseudoaqueous value by using H and 31P NMR spectrosco-
We are studying the acid–base chemistry of transition metal
hydride ([M–H]) and dihydrogen complexes ([M–H2])[1] to
better understand the mechanism of catalytic hydrogenation
reactions[2] and the action of hydrogenases.[3–9] Tetrahydro-
furan (THF) is an excellent solvent to use when studying
weakly acidic hydrides, their strong conjugate base forms,
and weakly coordinated complexes, such as those containing
the dihydrogen ligand.[1,10–13] However, THF cannot be used
as a solvent for very acidic compounds because it becomes
protonated and undergoes ring-opening reactions. Dichloro-
methane is an excellent solvent for acidic hydride and dihy-
drogen complexes because it is weakly coordinating and it
dissolves many neutral and cationic metal complexes with-
out undergoing reaction. However, it is not suitable for dis-
solving hydrides that are too reducing or nucleophilic. Sev-
eral pKa values for cationic metal hydride and dihydrogen
complexes in CD2Cl2 have been previously reported by us
and the group of Jia.[14–19] The initial approach used was to
find equilibria between these complexes and protonated
py.[15,16,18] A second approach was to construct a continuous
set of overlapping equilibria to provide a more accurate or-
dering of acid strengths. For example, a ladder of acids with
pKCD Cl values ranging from 9.7 to 7.0 was built and refer-
2
2
enced arbitrarily to [HPCy3]CATHRE[UNG BPh4] (Cy: cyclohexyl, pKa =
9.7) (Table 1).[15] More recently, this ladder was extended by
using phosphonium tetraphenylborate salts, of which
[HPBu2Ph]
[BPh4] has a pKCD Cl value of 5.7, for example.[19]
2
2
The pKCD Cl values are generally consistent with those
2
values previously measured in THF.[1]
One limitation when using [BPh4]ꢀ salts is that acids with
pKCD Cl values of less than 6 react with this anion to pro-
duce benzene and BPh3.[19] The aim of the current work is to
make a continuous ladder based on cationic acids with the
[BF4]ꢀ anion in CD2Cl2. Iron hydride complexes [Fe(CO)3H-
2
2
AHCTREUNG
(PR3)2]+ were chosen, based upon previous work by Angeli-
ci and co-workers,[20] to extend the ladder in the direction of
stronger acids. We have strengthened the pKCD Cl scale by
adding new rungs to the ladder, which correspond to the
phosphonium salts and iron carbonyl hydride complexes re-
ported herein.
2
2
[a] Dr. T. Li, Dr. A. J. Lough, Prof. Dr. R. H. Morris
Department of Chemistry, University of Toronto
80 St. George Street, Toronto, ON M5S 3H6 (Canada)
Fax : (+1)416-978-6962
E-mail: rmorris@chem.utoronto.ca
3796
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3796 – 3803

FULL PAPER
Table 1. The pKCD Cl values for cationic acids [HB]+ arbitrarily refer-
Table 2. 31P NMR spectroscopic chemical shifts of phosphanes and phos-
phonium salts in CD2Cl2.
2
2
enced to [HPCy3]+ salts at 9.7 in CD2Cl2.[15,16,19]
Acid[a]
pKCD Cl of [BPh4]ꢀ
salt
pKCD Cl of [BF4]ꢀ
salt
Base
d [ppm]
Acid
[HPCy3]
[HPBu3]
[HPEt3][BF4]
[HPCy2Ph][BF4]
[HPMe2Ph][BF4]
[HPBu2Ph][BF4]
[HPCyPh2][BF4]
[HPMePh2][BF4]
[HPEtPh2][BF4]
[HP(tol)3][BF4]
[HPPh3][BF4]
d [ppm]
2
2
2
2
PCy3
PBu3
PEt3
PCy2Ph
PMe2Ph
PBu2Ph
PCyPh2
PMePh2
PEtPh2
9.8
ꢀ31.1
ꢀ18.8
ꢀ2,9
A
G
29.5
13.3
22.6
27.1
0.6
14.9
15.6
2.0
A
9.7
9.7
–
–
8.5
–
G
ACHTREUNG
G
8.6(1)
U
ACHTREUNG
A
N
(PPh3)2]+ 8.6(1)
E
ACHTREUNG
trans-[Ru(Cp)
trans-[Ru(Cp)
U
–
ꢀ45.0
ꢀ24.6
ꢀ4.1
E
ACHTREUNG
(dppp)(H)2]+ 8.4
N
G
trans-[Ru(Cp)(H)2-
(PPh3)2]+
[HPtBu2Ph]+
trans-[Ru(Cp)
[Ru(Cp)
(dppm)(H2)]+
[Ru(Cp)
(dppe)(H2)]+
[RuCl(H2)
[HPCy2Ph]+
[HPBu2Ph]+
8.3–
G
ACHTREUNG
H
ꢀ27.1
ꢀ11.8
ꢀ8.1
E
ACHTREUNG
G
8.0(1)
–
G
E
12.5
3.85
12.0
(dppe)(H)2]+
–
–
–
7.3
7.1
7.0
–
–
–
P
A
U
ACHTREUNG
G
ꢀ5.5
C
ACHTREUNG
ACHTREUNG
E
{tmeP2(NH)2}]+ 6.9(3)
E
6.6(3)
5.7(3)
product was recrystallized by using CH2Cl2/hexanes (40–70%). 31P NMR
spectroscopy chemical shifts are reported in Table 3.
E
[a] Abbreviations: dach: (1R,2R)-(ꢀ)-diaminocyclohexane; Cp: h5-C5H5;
[Fe(CO)3H
added to [Fe(CO)
less or light yellow solution. The mixture was stirred for 1 h before the
G
G
ACHTRU(NEG Et2O)]ACHRTE[UGN BF4] (1 equiv) in CH2Cl2 (2 mL) was
dape: (tol)2PCH2CH2PACHTREUNG(tol)2, (tol: p-tolyl); dppp: Ph2PACHRTECUNG H2AHCERTCUNG H2ACHTERCNGU H2ACHTREPUNG Ph2;
CTHREUNG
dppm:
{tmeP2(NH)2}: PPh
PPh2CH2PPh2;
2AC6H4ACH
dppe:
NHCMe
1,2-bis(diphenyl
ACHTREUNG
G
A
G
C
E
Table 3. 31P and 1H NMR spectroscopic chemical shifts of [Fe(CO)
CD2Cl2.
3A
ACHRTUNEG(PR3)2]ACHTREUNG[BF4] in
Experimental Section
Base
d
A
Acid
d
A
d(1H) [ppm]
General: All preparations and manip-
ulations were carried out under a ni-
trogen or an argon atmosphere by
using standard Schlenk, vacuum line,
and glove-box techniques in dry,
oxygen-free solvents. THF, diethyl
ether (Et2O), and hexanes were dried
and distilled from sodium benzophe-
none ketyl. Dichloromethane was
dried and distilled from calcium hy-
dride. Deuterated solvents were de-
gassed and dried over activated molec-
[Fe(CO)
[Fe(CO)
[Fe(CO)
[Fe(CO)
3A
3A(PCy2Ph)2]
3A(PCyPh2)2]
3A{PPh2A(tol)}2]
(tol)3}2]
(PPh3)2]
[Fe(CO)3{P(p-FC6H4)3}2]
84.7
84.1
82.2
80.1
78.3[Fe(CO)
81.1
79.2
[Fe(CO)3H
[Fe(CO)3H
[Fe(CO)3H
[Fe(CO)3H
3H{P
[Fe(CO)3H
N
U
67.5
65.1
56.4
54.0
52.5
54.8
53.2
ꢀ9.1
ꢀ8.4
ꢀ8.5
ꢀ7.8
ꢀ7.8
ꢀ7.8
ꢀ8.0
G
ACHTREUNG
U
ACHTREUNG
R
G
ACHTREUNG
[Fe(CO)3{P
[Fe(CO)3A
U
ACHTREUNG
T
ACHTREUNG
G
ACHTRE[UNG BF4]
ular sieves. NMR spectra were recorded by using a Varian Unity-500
(500 MHz for 1H), a Varian Unity-400 (400 MHz for 1H), or a Varian
Gemini 300 MHz spectrometer (300 MHz for 1H and 121.5 MHz for 31P).
All 31P NMR spectroscopic chemical shifts were measured relative to
solvent was removed under vacuum. Hexanes (10 mL) were added to
give a white or light yellow precipitate. The product was obtained by
means of filtration, washed with hexanes, and dried under vacuum. The
product was recrystallized by using CH2Cl2/hexanes (ꢁ80%). The 1H
(hydride) and 31P NMR spectroscopic hydride chemical shifts are report-
ed in Table 3.
10% PACHTREUNG(OMe)3 in C6D6. This reference solution was placed in a 2 mm ca-
pillary tube inside the 5 mm NMR tube containing the sample. 1H NMR
spectroscopic chemical shifts were measured relative to partially deuter-
ated solvent peaks, but are reported relative to tetramethylsilane. The
phosphane compounds were received from commercial suppliers (Al-
drich and Strem) and were used without further purification. [Fe-
Determination of equilibrium constants in CD2Cl2: Samples were mixed
under N2 as described in Tables 4 and 5. In general, equilibrium constants
were determined by means of 1H and 31P NMR spectroscopy. Usually, sig-
nals for all of the species in equilibrium could be located and integrated
in the 31P NMR spectra (with inverse gated decoupling for quantitative
integration), and also in the 1H NMR spectra for the hydride complexes.
The chemical shifts for the pure phosphanes and phosphonium salts
ACHTREUNG
(bda)(CO)3][21] (bda: benzylideneacetone) was prepared according to lit-
erature procedures. Phosphonium salts that have not already been report-
ed in the literature were found to have the correct elemental analyses by
the Analyst Laboratory of the University of Toronto. CCDC-624440 and
-624441 contain the supplementary crystallographic data for this paper.
These data can be obtained free of charge from the Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_/cif.
(Table 2) were determined and referenced to PCATHRE(UGN OMe)3 at d=141.5 ppm.
In some cases, in which the chemical shifts of two species are very simi-
lar, mass-balance arguments could be used to estimate the equilibrium
concentration of the species from their starting concentrations. The re-
sulting equilibrium constants in CD2Cl2 are shown in Tables 4 and 5.
A
ACHTREUNG[BF4]: All phosphonium salts [HPR3]ACHTREUNG
AHCTREUNG
X-ray diffraction structure determination of [Fe(CO)
3ACHTREU(GN PCy2Ph)2] and
were isolated according to the reported preparation,[15] with typical yields
of 80–90%. The 31P NMR spectroscopic chemical shifts are reported in
Table 2.
[Fe(CO)3H(PCy2Ph)2][BF4]: Crystals suitable for X-ray diffraction were
G
ACHTREUNG
obtained by means of vapor diffusion between a solution of the crystals
in CH2Cl2 and hexanes. Data were collected by using a Nonius Kappa-
CCD diffractometer with MoKa radiation (l=0.71073Š). CCDC data
were integrated and scaled by using the DENZO-SMN software package,
[Fe(CO)
(2.2 equiv; PR3: PCy3, PCy2Ph, PCyPh2, PPh
3ACHTREUNG ACHTRENUG
(PR3)2]:[22] [Fe
N
G
ACHTREUNG
[40]
and the structures were solved and refined by using SHELXTL
FC6H4)3) in THF (10 mL) were stirred for 24 h under Ar at room temper-
ature to give a light yellow solution or yellow precipitate. Half of the sol-
vent was removed under vacuum. Hexanes (20 mL) were added to the
mixture to give a yellow precipitate. The product was isolated by means
of filtration, washed with hexanes, and then dried under vacuum. The
V6.0.[41] The crystallographic data are listed in Table 6. The hydride was
located and refined with isotropic thermal parameters.
Chem. Eur. J. 2007, 13, 3796 – 3803
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3797

R. H. Morris et al.
Keq
þ
ꢀG
H
B1þ½B2HŠ ½BF4Š
Table 4. Equilibrium constants for acid–base reaction of phosphonium salts in CD2Cl2.
ð3Þ
Acid
[HPCy3]
[HPBu3]
[HPBu3]
[HPCy2Ph]
[HPMe2Ph]
[HPMe2Ph]
[HPEtPh2][BF4]
[HPEtPh2][BF4]
[HPCyPh2][BF4]
[HPCyPh2][BF4]
[HPCyPh2][BF4]
[BF4]
[BF4]
c [mmolLꢀ1
]
Base
c [mmolLꢀ1
]
t[a] [h]
Keq
Error [%][b]
pKeq
½B1HŠ ½BF4ŠꢀþB2
þ
G
U
49
34
83PEt
44
44
62
50
50
56
54
3
PBu3
PCy2Ph
50
38
85
45
45
60
50
40
50
54
33
47
50
<2
<1
<1
<2
<1
<1
<1
<2
<1
<4
<1
<1
<1
0.0310
0.0310
0.81
0.14
1.35
0.045
0.50
4.7
0.135
27.5
1.5
1.5
G
A
The equilibrium constants
(Keq) for reactions in CD2Cl2
were measured by means of
quantitative 31P and 1H NMR
spectroscopy at room tempera-
ture. Keq values in the range 103
to 10ꢀ3 can be measured with
10 to 30% error. The difference
R
G
5
5
0.1
0.9
3
N
U
PBu2Ph
A
A
PBu2Ph
PCyPh2
PMePh2
PCyPh2
PMePh2
PBu2Ph
ꢀ0.1
U
G
10
5
5
1.4
U
G
0.3
ꢀ0.7
U
G
R
G
0.9
E
U
10
10
5
ꢀ1.4
A
N
4
A
P
0.044
7.4
3.8
1.4
2
2
[HP
[HP
E
N
51
50
PEtPh2
PMePh2
ꢀ0.9 in pKCD Cl between the two
acids [B1H]
CHTREUNG
+A[BF4]ꢀ and [B2H]+
G
5
ꢀ0.6
[BF4]ꢀ was calculated from the
[a] Time taken to reach equilibrium. [b] The error in the equilibrium constant determination is estimated from
the magnitude of the constant and the signal/noise ratio of each species in the spectrum of the equilibrium
mixture.
Keq value by using Equa-
tion (4):
1
þ
ꢀ
DpKCD Cl ¼ pKeq ¼pKCD Cl ð½B HŠ ½BF4Š Þ
2
2
2
2
Results and Discussion
ð4Þ
ꢀpKCD Cl ð½B HŠ ½BF4Š Þ
2
þ
ꢀ
2
2
Preparation of [HPR3]ACHTERU[NG BF4]: [HPR3]ACHTREU[GN BF4] can be readily
prepared through protonation of PR3 (PR3: PCy3, PBu3,
PCy2Ph, PBu2Ph, PMe2Ph, PCyPh2, PEtPh2, PPh3) with [H-
A
ACHTREUNG[BF4] in diethyl ether as shown in Equation (1):
Table 6. Crystallographic data for [Fe(CO)
(PCy2Ph)2][BF4].
(PCy2Ph)2] and [Fe(CO)3H-
[Fe(CO)3H(PCy2Ph)2][BF4]
G
ACHTREUNG
[Fe(CO)
3A
E
ACHTREUNG
ð1Þ
formula
Mr
C39H54FeO3P2
688.61
C39H55BF4FeO3P2
776.43
pKCD Cl determination: The acid dissociation equilibrium
constant (Ka) of an acid ([BH]+) in CD2Cl2 is defined by
using Equation (2):
2
2
space group
a [Š]
b [Š]
c [Š]
T [K]
P2(1)/c
16.7920(5)
10.3300(3)
21.8130(6)
150(1)
Pmn2(1)
14.5561(4)
10.1860(3)
16.6463(4)
150(1)
90
90
90
1.045
½BHŠ Ð ½HðCD2Cl2ފþþB
ð2Þ
þ
a [8]
90
b [8]
g [8]
106.205(2)
90
1.259
3633.38(18)
4
However, the concentration of the acid form in the solvent,
1calcd [mgmꢀ3
V [Š3]
Z
]
([HACHTREUNG
(CD2Cl2)]+), is not readily determined. Therefore, the
2468.1(1)
2
0.0819
relative acidity (pKCD Cl ) of phosphonium salts and metal
hydrides was studied. The acid ([BH]+) and base (B) reac-
tion equilibria are based on Equation (3):
2
2
R1 (all data)
wR2
0.1027
0.1809
0.2121
Table 5. Equilibrium constants for the reactions of iron hydride complexes in CD2Cl2.
Acid
c
A
Base
c
t[a] [h]
Keq
Error
[%]
pKeq
[mmolLꢀ1
]
[mmolLꢀ1
ACHTREUNG ]
[Fe(CO)3H
[Fe(CO)3H
C
N
13PPh
15
20
18
22
15
15
15
18
13
17
12
15
14
7
<1
<1
<1
<4
<1
<1
<1
<1
<4
<4
<1
<4
<1
<1
<1
0.0017
0.14
5.55
0.069
0.032
0.25
0.50
609
15.8
0.18
0.06310
0.37
20
5
5
10
10
5
2.8Æ0.2
3
PEtPh2
[Fe(CO)
0.85
A
N
N
ꢀ0.74
[Fe(CO)3H
[Fe(CO)3H
A
13[Fe(CO)
G
1.2Æ0.1
G
18
22
14
17
10
18
16
8
PPh3
1.5Æ0.1
A
N
[Fe(CO)
[Fe(CO)
[Fe(CO)
[Fe(CO)
[Fe(CO)
[Fe(CO)
[Fe(CO)
[Fe(CO)3{P
Et2O
[Fe(CO)3{P
3A
3A(PCyPh2)2]
3A(PCy3)2]
3A(PCyPh2)2]
3A{PPh2A(tol)2}]
3A(PPh3)2]
3A{PPh2A(tol)2}]
(p-FC6H5)3}2]
0.6[b]
A
N
5
0.30
A
N
20
10
5
ꢀ2.8Æ0.2
ꢀ1.2Æ0.1
0.74
[Fe(CO)3H{P
[Fe(CO)3H{P
[Fe(CO)3H{P
[Fe(CO)3H
[Fe(CO)3H
U
[BF4]
G
U
1.2
Æ0.1
G
5
20
20
0.43
N
15
15
30
11
50
15
0.0054
0.024
93
2.3Æ0.2
1.6Æ0.2
ꢀ1.8Æ0.2
[Fe(CO)3H{P
[HOEt2]
G
ACHTREUNG[BF4]
A
N
20
[a] Time taken to reach equilibrium. [b] A side reaction occurs. The PCy2Ph ligand is substituted by PEtPh2.
3798
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3796 – 3803

Acidity Scale in [D2]Dichloromethane
FULL PAPER
CD2 Cl2
Ideally, we wanted to obtain the pK
values for the free
errors relative to the reference value of 9.7 and the estimat-
ed errors for K.
fi
ions [BH]+, independent of the counter ion ([BF4]ꢀ), and
not values for the ion-paired species [BH]
+A
We have not attempted to estimate corrections for ion
pairing. We assume that the phosphonium salts have similar
interion distances and similar ion-pair energetics so that
these factors cancel out in the equilibrium shown in Equa-
tion (6).
thermodynamics of ion pairing, solvation, and hydrogen
bonding is the same for [B1H]
C
+A
CD2 Cl2
then we expect that the DpKCD Cl and DpKfi
values will
2
also be the same, as shown in Equation (5):
The presence of a more electron-donating substituent on
the phosphonium salt results in a decrease in acidity as ex-
CD2 Cl2
CD2 Cl2
DpKCD Cl ¼ DpKfi
¼ pK
ð½B1HŠþÞꢀpKfCiD2 Cl2 ð½B2HŠþÞ
2
2
fi
pected (i.e., the pKCD Cl values increase as Me<Et<Bu<
2
2
ð5Þ
Cy). By using [BF4]ꢀ salts, the acidity ladder reaches a
pKCD Cl value of 2.7 for [HP
ACTHERNGU(tol)3]ACHTRE[UGN BF4].
2
2
2 Cl2
[BF4] with a free ion (pKCD
+
ACHTREUNG[HPCy3]ACHTREUNG ACHTRE(UNG HPCy3 )=9.7) was
fi
chosen as the arbitrary anchor for the scale in CD2Cl2 in a
similar manner to the THF scale.[1] Overlapping equilibria
were examined with more acidic compounds to create a con-
tinuous ladder of values. For examples, see Equations (6)
and (7):
Synthesis and structures of iron complexes: [Fe(CO)
and [Fe(CO)3H(PR3)2][BF4] complexes were chosen to
extend the acidity ladder measured in CD2Cl2 in the more
3ACHTREUNG(PR3)2]
G
ACHTREUNG
acidic direction. Yellow complexes [Fe(CO)3ACHTRE(UGN PR3)2] were re-
ported by Angelici and Wang,[20] and Nolan and co-workers.
pKeq¼1:5
[22–25]
These complexes are conveniently synthesized from
½HPCy Š½BF ŠþPBu
PCy þ½HPBu Š½BF Š
ð6Þ
ð7Þ
3G
H
3
4
3
3
4
[Fe
ACHTRE(UGN bda)(CO)3] and PR3 in which PR3 is PPh3, PMe2Ph,
þ
pKCD Cl ½HPBu3Š ¼ 9:7ꢀpKeq ¼ 8:2
2
2
PMePh2, and PMe3 (Scheme 1). We adapted the same
An acidity scale ladder of phosphonium tetrafluoroborate
salts in CD2Cl2 (Figure 1) was built by means of the overlap-
ping equilibria method. The estimated errors for the equilib-
Scheme 1. Preparation of [Fe(CO)3ACHTRE(UNG PR3)2] complexes.
method and prepared [Fe(CO)
PR3 is PCy3, PCy2Ph, PCyPh2, PPh
(p-FC6H4)3).
The X-ray crystal structure of [Fe(CO)
shown in Figure 2 and the bond lengths and angles are listed
in Table 7. [Fe(CO)3A(PCy2Ph)2] has a distorted trigonal bi-
(PR3)2] complexes in which
2A(tol), P(tol)3, PPh3, and
PACHTREUNG
3ACHTREN(UG PCy2Ph)2] is
CTHREUNG
pyramidal geometry. The two phosphane ligands are located
in axial positions with a P-Fe-P angle of 158.66(4)8. They
adopt a conformation in which the cyclohexyl rings, one
from each ligand, are face to face in the equatorial plane of
the complex. The C(2)-Fe(1)-C(3) angle is widened to
Figure 1. An acidity scale ladder for phosphonium tetrafluoroborate salts
in CD2Cl2 showing pKCD Cl values.
2
2
rium constants are listed in Tables 4 and 5. The cumulative
error in the pKCD Cl value is calculated by using Equa-
2
2
tion (8):[1]
Cumulative error in pKCD Cl ¼ Æ0:08jpKCD Cl ꢀ9:7j
2
2
2
2
ð8Þ
The errors in the last digit are shown in parentheses in
Figure 1. They represent a combination of the cumulative
Figure 2. X-ray crystal structure of [Fe(CO)3ACHTERU(NG PCy2Ph)2].
Chem. Eur. J. 2007, 13, 3796 – 3803
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3799

R. H. Morris et al.
Table 7. Selected bond lengths [Š] and angles [8] for [Fe(CO)3-
(PCy2Ph)2].
ACHTREUNG
ꢀ
ꢀ
Fe(1) C(1)
1.774(4)
1.764(4)
2.2337(9)
Fe(1) C(2)
1.770(3)
2.2378(9)
ꢀ
ꢀ
Fe(1) C(3)
Fe(1) P(1)
ꢀ
Fe(1) P(2)
C(3)-Fe(1)-C(2)
C(2)-Fe(1)-C(1)
C(2)-Fe(1)-P(2)
C(3)-Fe(1)-P(1)
C(1)-Fe(1)-P(1)
146.1(2)
C(3)-Fe(1)-C(1)
C(3)-Fe(1)-P(2)
C(1)-Fe(1)-P(2)
C(2)-Fe(1)-P(1)
P(2)-Fe(1)-P(1)
107.2(2)
106.6(2)
87.1(1)
86.0(1)
101.0(1)
87.2(1)
100.3(1)
87.4(1)
158.66(4)
146.1(2)8 by this intrusion of the cyclohexyl groups. The
other C-Fe-C angles diminish, but the FeC3 grouping re-
mains planar so that the sum of the three C-Fe-C angles
adds up to 3608. There is approximate Cs symmetry in which
a mirror plane cuts though the carbonyl ligands and iron.
Figure 3. Structure of the cation of [Fe(CO)3H
mined by means of X-ray diffraction.
ACHRTUNEG(PCy2Ph)2]CAHTRE[UGN BF4] as deter-
Table 8. Selected bond lengths [Š] and angles [8] for [Fe(CO)3H-
(PCy2Ph)2][BF4].
The complex exhibits
a singlet at d=84 ppm in the
A
ACHTREUNG
31P NMR spectrum.
ꢀ
ꢀ
Fe(1) C(1)
1.815(8)
1.791(9)
2.2920(9)
Fe(1) C(2)
1.790(6)
2.2920(9)
1.48(7)
There is a report of such distortions in trans-(tricarbonyl)-
bis(trans-2,2,3,4,4-pentamethyl-1-phenylphosphetane)iron,
which contains a P-Fe-P angle of 162.48 and one C-Fe-C
ꢀ
ꢀ
Fe(1) C(3)
Fe(1) P(1)
ꢀ
ꢀ
Fe(1) P(1a)
Fe(1) H(1)
C(3)-Fe(1)-C(2)
C(2)-Fe(1)-C(1)
C(2)-Fe(1)-P(1a)
C(3)-Fe(1)-P(1)
C(1)-Fe(1)-P(1)
H(1)-Fe(1)-C(2)
H(1)-Fe(1)-C(1)
H(1)-Fe(1)-P(1)
157.8(3)
C(3)-Fe(1)-C(1)
C(3)-Fe(1)-P(1a)
C(1)-Fe(1)-P(1a)
C(2)-Fe(1)-P(1)
P(1a)-Fe(1)-P(1)
H(1)-Fe(1)-C(3)
H(1)-Fe(1)-P(1a)
102.1(4)
angle of 133.38.[26] Surprisingly, [Fe(CO)
3ACHTRE(UGN PPh2Cy)2] does not
100.1(3)
89.12(5)
88.17(6)
97.05(3)
91(4)
88.17(6)
97.05(3)
89.12(5)
165.88(6)
66(4)
have steric hindrance in the equatorial plane, and therefore,
has a P-Fe-P angle of 178.45(6)8 and C-Fe-C angles of 118–
1228.[22]
The pale yellow or white hydride [Fe(CO)3H
complexes were prepared through protonation of [Fe(CO)3-
ACHTREUNG(PR3)2] in CH2Cl2 (Scheme 2), which is formally an oxidative
ACHRTU(NEG PR3)2]ACHTRE[UGN BF4]
169(4)
83.02(8)
83.02(8)
An acidity scale for iron hydride complexes in CD2Cl2:
Overlapping equilibria between [Fe(CO)3A(PR3)2] and
[Fe(CO)3H(PR3)2][BF4] were established to create a contin-
CTHREUNG
A
ACHTREUNG
uous acidity scale for iron hydride complexes in CD2Cl2
(Figure 4).
Scheme 2. Preparation of the hydride complexes.
To anchor our [Fe(CO)3H
were established between [HPEtPh2]
and [Fe(CO)3A(PCy3)2] (Figure 4). We have not corrected the
ACHTRUNEG(PR3)2]ACHRTE[UGN BF4] scale, equilibria
AHCTREUNG
[BF4] (pKCD Cl =3.6)
2
2
addition reaction (Fe0/FeII). THF cannot be used as a sol-
CHTREUNG
vent when PR3 is PPh3 and P
N
equilibrium constant for the differences between the ion-
pair energetics of the phosphonium and iron hydride salts.
Previously, this correction was done for a pKTaHF scale of
metal hydrides and phosphonium salts[1] and protonated ni-
trogen bases[10,12] by using the Fuoss equation, but this ap-
proach is inaccurate.[27] The continuous acidity ladder of all
of the [BF4]ꢀ salts from this work is shown in Table 9.
hydride complexes are deprotonated by THF. Attempts to
exchange [BF4]ꢀ for [BPh4]ꢀ in the hydride complex through
treatment with NaBPh4 resulted in the formation of
[Fe(CO)
The structure of the cation of [Fe(CO)3H
is shown in Figure 3and selected bond lengths and angles
are listed in Table 8. [Fe(CO)3H(PCy2Ph)2][BF4] has octahe-
3ACHTREUNG(PR3)2], BPh3, and benzene.
Angelici and co-workers reported the heats of protona-
tion of phosphane compounds and iron carbonyl complexes
in dichloroethane (DCE) by using trifluoromethylsulfonic
acid (HOTf).[29,30] Some of the ꢀDHMH and ꢀDHPH values
available are also included in Table 9. A plot of ꢀDHMH and
dral geometry, as expected for low-spin FeII, but there are
distortions owing to the face-to-face cyclohexyl groups form-
ing a sandwich with the hydride ligand. In essence, the con-
formations of the ligands have changed very little from the
unprotonated structure shown in Figure 2.
ꢀDHPH values versus pKCD Cl is shown in Figure 5, and
Equation (9) shows the conversion equation (R2 =0.98). By
2
2
The hydride complex exhibits a triplet at d=ꢀ8.4 ppm in
1
using this equation, we can predict the pKCD Cl values of
compounds with known ꢀDHMH or ꢀDHPH values.
2
2
the H NMR spectrum for the hydride ligand and a singlet
at d=65 ppm in the 31P NMR spectrum for the two equiva-
lent phosphane ligands.
3800
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3796 – 3803

Acidity Scale in [D2]Dichloromethane
FULL PAPER
Figure 5. A plot of ꢀDHFeH and ꢀDHPH values in kcalmolꢀ1 (the anion is
[OTf]ꢀ) versus pKCD Cl (the anion is [BF4]ꢀ).
2
2
values were determined by using [BF4]ꢀ. The slope of the
plot of Figure 5 is 1.8. pK values are related to free energies,
whereas DH values are enthalpies, which means that differ-
ences in entropy are not accounted for in such a plot. A plot
of free energy versus ꢀlogK should have a slope of
2.303RT=1.35. The difference between these two slopes
must represent continuous changes in entropy and differen-
ces in solvation, ion pairing, and hydrogen bonding between
the more basic and more acidic compounds.
Figure 4. An acidity scale ladder of the iron hydride complexes and phos-
phonium tetraphenylborate salts in CD2Cl2 showing DpKCD Cl values.
2
2
A set of pKa values are used by convention in the litera-
ture to describe the general acidic properties of protonated
phosphanes. Most of these values have been obtained by
means of extrapolation from nonaqueous or mixed solvents
to aqueous conditions. These pKa values are shown in
Table 9. The complete acidity ladder of [Fe(CO)3H
A
ACHTREU[NG BF4] and
[HPR3]
values.
ACHTREUNG
Table 9 and a plot of the pKa values versus the pKCD Cl
2
2
2
2
Acid
pKCD Cl
pKa
ꢀDHHP or ꢀDHHM
values is shown in Figure 6 as diamond-shaped points. A
best-fit line through ten points gives Equation (10) (R2 =
0.99):
[a]
[kcalmolꢀ1
]
A
9.7
8.2
8.1
6.7
5.8
5.7
4.4
9.7
8.4
8.7
7.4
–
34.2
–
–
–
–
ACHTREUNG
G
U
þ
ꢀ
pKa ¼ 0:85 pKCD Cl ð½HPR3Š ½BF4Š Þþ1:60
2
2
ð10Þ
E
C
6.8
–
–
–
G
The pKa values used for PCy3, PBu3, PEt3, PPh3, and
PMe2Ph were measured in nitromethane by Streuli and ex-
A
4.35.1
3.6
3.4
–
G
4.9
–
–
–
A
A
3.3
2.7
1.6
0.1
4.6
3.8[b]
2.7[b]
–
–
–
24.7
–
21.2
–
–
14.1
–
G
G
[Fe(CO)3H{P
[Fe(CO)3H{P
[Fe(CO)3H
E
G
ꢀ0.6
G
ꢀ1.1
[Fe(CO)3H(P
G
ꢀ3.4
–
[a] Determined in dichloroethane. [b] These values are now known to be
at least 1.6 units too high.[28]
DCE
FeH or PH
ꢀDH
ðHBþOTfꢀÞ ¼
ð9Þ
1:8p KCD Cl ð½HBŠ ½BF4Š or ½HFeŠ ½BF4Š Þþ17:48
þ
ꢀ
þ
ꢀ
2
2
Figure 6. Correlation between pKa and pKCD Cl scale. The points marked
with diamonds are from Table 9, whereas those marked with squares
refer to pKa values determined in CH3CN/H2O (v/v=85:15) and are
from ref. [28].
2
2
The plot suggests that the solvation effects are similar in
CD2Cl2 and DCE. The ꢀDHFeH and ꢀDHHP values were
measured for cations by using [OTf]ꢀ, whereas the pKCD Cl
2
2
Chem. Eur. J. 2007, 13, 3796 – 3803
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3801

R. H. Morris et al.
Table 10. Converted selected pKCD Cl values of metal hydride and dihydrogen complexes measured in CD2Cl2
trapolated to water.[31] The pKa
value used for P(tol)3 was ob-
tained experimentally by
2
2
on the basis of the aqueous reference to pKCD Cl values (sorted by the acidity in CD2Cl2 from strong to
2
2
ACHTREUNG
weak).[a]
pKCD Cl
pKCD Cl
2
2
2
2
Allman and Goel.[32] Giering
and co-workers estimated pKa
values for PCyPh2 and PCy2Ph
by means of interpolation be-
tween the values for PPh3 and
PCy3.[33] Kochi and coworker
estimated pKa values for
PMePh2 and PEtPh2 by using
Hammett correlation analy-
ses.[34]
Dihydrogen or dihydride
Anion
Base
literature converted
trans-[Os
[Ru(CO)
(HCn)(H2)
[Ru(CO)(H2)
(PPh3)(Tp)]+
(dppe)(H2)]2+
(PPh3)2(H2)]2+
(pmp)
(PPh3)]+
trans-[RuCl
(dppe)2(H2)]+
[Ru(Cp)
(dppe)(H2)]+
[Ru(Cp)
(dppm)(H2)]+
C
G
2
2
U
Et2O
ꢀ2Æ1[36]
ꢀ1.3[17]
ꢀ0.6[17]
3.8[17]
ꢀ4Æ1
ꢀ3Æ1
ꢀ2Æ1
2.5
3.2
4.0
4.7
7.0
7.1
A
Et2O
G
[Ru(CO)ACTHEURNG ACHTREUNG
(HCn)H
[Ru
[Ru
A
G
2
2
U
trans-[RuCl
trans-[RuCl
ACHTREUNG
A
4.5[17]
G
P
A
5.1[14]
ꢀ
PF6
[BF4]ꢀ
[BF4]ꢀ
PEtPh2
6.0[15,16]
7.0[15,16]
7.1[15,16]
E
U
[Ru(Cp)H
[Ru(Cp)H
[Ru(Cp)H
[Ru(Cp)H
[Ru(Cp)
[Ru(Cp)
C
U
E
A
G
A
E
7.2[14]
7.2[14]
However, there is recent evi-
dence that, at least for the triar-
ylphosphanes, these literature
pKa values are too high. For ex-
ample, [HPPh3]+ and [HP-
R
N
A
7.6[16]
7.6
7.9
8.1
8.3
8.4
8.9
E
G
G
7.9[16]
A
R
[Ru(Cp)H
(PPh3)2]
8.1[15]
U
H
8.3[15,16]
8.4[15,16]
8.9[17]
A
E
[Ru(Cp)H
[Ru(Cp)HCAHTREUNG
ACHTREUNG
(tol)3]+ have pKa values in
A
N
C
[a] Abbreviations: HCn: 1,4,7-triazacyclononane, MeCn: 1,4,7-trimethyl-1,4,7-triazacyclononane, Tp: hydrotris(-
pyrazolyl), pmp: 2,6-(Ph2PCH2)2C5H3N.
CH3CN/H2O (v/v=85:15) of
1.05 and 2.18, respectively.[28]
The
hypothetical
aqueous
values would be even lower.
more equilibria with [Fe(CO)
3ACHTRE(NUG PPh3)2] or [Fe(CO)3{PACHTREUNG(p-
Acidity of diethyl ether: The protonation equilibrium of di-
ethyl ether was measured in aqueous sulfuric acid by Per-
doncin and Scorrano and the pKa of [HOEt2]+ was reported
FC6H4)3}2] need to be established.
Iron hydride acidity and hydrogenase: The mechanism of
action of iron hydrogenases involves reversible protonation/
deprotonation of iron centers coordinated by carbonyl, thio-
late, and in all known cases except one,[37] cyanide li-
gands.[38,39] These acid–base reactions occur at around pH 7
in the interior of the protein, which has a low dielectric con-
stant. The present work emphasizes the acidifying effect
that carbonyl ligands have on iron hydrides.
as ꢀ2.4.[35] This value has been used to estimate the pKCD Cl
2
2
values of some very acidic complexes in CD2Cl2, such as
trans-[Os
(HCn)(H2)
no)ethane, HCn: 1,4,7-triazacyclononane).[17] Therefore, the
A
[(BF4)2][36] and [Ru(CO)-
ACHTREUNG(dppe)2(H2)]ACHTREUGN
A
N
ACHTREUNG
pKCD Cl value of [HOEt2]
ACHTRE[UNG BF4] is important to allow a
2
2
better estimate of the pKCD Cl values of these complexes.
[Fe(CO)3H{P(p-FC6H4)3}2][BF4], the most acidic complex on
2
2
A
ACHTREUNG
the acidity ladder (Table 9), reacts with diethyl ether and at
equilibrium, a Keq value of 0.024 was obtained, as shown in
Equation (11):
Conclusion
A continuous scale of very acidic compounds, [HPR3]ACHTREUNG[BF4]
and [Fe(CO)3HACTHREU(GN PR3)2]ACHTREU[NG BF4], in CD2Cl2 was constructed
Keq
½FeðCOÞ HfPðp-FC H Þ g Š½BF ŠþEt O
G
¼0:018 H
3
6
4
3
2
4
2
ð11Þ
with a pKCD Cl range of 9.7 to ꢀ3. The pKCD Cl value of di-
2
2
2
2
½FeðCOÞ3fPðp-FC6H4Þ3g2Šþ½HOEt2Š½BF4Š
ethyl ether was found to be ꢀ5. Existing pKCD Cl values of
selected metal hydride and dihydrogen complexes in CD2Cl2
were converted to the new scale. When an absolute acid
standard in CD2Cl2 becomes available, these values will all
have to be shifted to more positive values. Thus for this
reason, and owing to the difficulties in determining the ion-
2
2
The reverse equilibrium reaction between [HOEt2][BF4] and
AHCTREUNG
[Fe(CO)3{P(p-FC6H4)3}2] was also studied and a Keq value of
ACHTREUNG
0.011 was obtained. Therefore, the pKCD Cl value of
2
2
[BF4] was found to be ꢀ4.7Æ1 when the pKCD Cl
2
2
[HOEt2]
value of [Fe(CO)3H{P
(p-FC6H4)3}2]
[BF4] is ꢀ3Æ1.
pairing contribution, the pKCD Cl values reported here are
not thermodynamically accurate pKa values, but are useful
nevertheless for estimating the relative acidity of com-
pounds in CD2Cl2. The crystal structures of [Fe(CO)3-
2
2
Correcting the existing pKCD Cl values: As mentioned in the
2
2
Introduction, pKCD Cl values of metal hydride and dihydro-
gen complexes in CD2Cl2 have been reported and anchored
to the pKa value of phosphonium salts, Et2O, or metal hy-
dride complexes. In this study, we established a new acidity
2
2
AHCTREGNU(PCy2Ph)2] and [Fe(CO)3HACHETR(UNG PCy2Ph)2]ACHTREU[NG BF4] are reported and
found to be distorted by the conformation adopted by the
scale in [D2]dichloromethane. The existing pKCD Cl values of
the selected metal hydride and dihydrogen complexes in
CD2Cl2 were converted to the new dichloromethane scale
(shown in Table 10). To locate the very acidic complexes,
bulky, unsymmetrical phosphane ligands. [BPh4]ꢀ and [BF4]ꢀ
2
2
anions do not have a large effect on pKCD Cl values, but the
[BF4]ꢀ salts allow a much wider range of acid strengths to
be measured.
2
2
3802
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3796 – 3803

Acidity Scale in [D2]Dichloromethane
FULL PAPER
[19] T. Li, A. J. Lough, C. Zuccaccia, A. Macchioni, R. H. Morris, Can. J.
Chem. 2006, 84, 164.
Acknowledgement
[20] D. Wang, R. J. Angelici, J. Am. Chem. Soc. 1996, 118, 935.
[21] M. Brookhart, G. O. Nelson, J. Organomet. Chem. 1979, 164, 193.
[22] C. Li, E. D. Stevens, S. P. Nolan, Organometallics 1995, 14, 3791.
[23] L. Luo, S. P. Nolan, Inorg. Chem. 1993, 32, 2410.
[24] C. M. Haar, S. P. Nolan, Inorg. Chim. Acta 1999, 291, 3 2.
[25] S. A. Serron, S. P. Nolan, Inorg. Chim. Acta 1996, 252, 107.
[26] D. W. Bennett, D. S. Grubisha, S. E. Cremer, A. C. Peterson, J. Crys-
tallogr. Spectrosc. Res. 1992, 22, 83 .
[27] A. Streitwieser, A. E. McKeown, F. Hasanayn, N. R. Davis, Org.
Lett. 2005, 7, 1259.
[28] O. Pestovsky, A. Shuff, A. Bakac, Organometallics 2006, 25, 2894.
[29] R. C. Bush, R. J. Angelici, Inorg. Chem. 1988, 27, 681.
[30] J. R. Sowa, Jr., V. Zanotti, G. Facchin, R. J. Angelici, J. Am. Chem.
Soc. 1991, 113, 9185.
This work was supported in part by a discovery grant to R.H.M. from
NSERC Canada.
[1] K. Abdur-Rashid, T. P. Fong, B. Greaves, D. G. Gusev, J. G. Hinman,
S. E. Landau, A. J. Lough, R. H. Morris, J. Am. Chem. Soc. 2000,
122, 9155.
[2] S. E. Clapham, A. Hadzovic, R. H. Morris, Coord. Chem. Rev. 2004,
248, 2201.
[3] X. M. Liu, S. K. Ibrahim, C. Tard, C. J. Pickett, Coord. Chem. Rev.
2005, 249, 1641.
[4] A. Kayal, T. B. Rauchfuss, Inorg. Chem. 2003, 42, 5046.
[5] R. M. Henry, R. K. Shoemaker, D. L. DuBois, M. R. DuBois, J. Am.
Chem. Soc. 2006, 128, 3002.
[6] D. J. Evans, C. J. Pickett, Chem. Soc. Rev. 2003, 32, 268.
[7] M. Y. Darensbourg, E. J. Lyon, J. J. Smee, Coord. Chem. Rev. 2000,
206, 533.
[8] F. A. Armstrong, Curr. Opin. Chem. Biol. 2004, 8, 133.
[9] T. P. Fong, C. E. Forde, A. J. Lough, R. H. Morris, P. Rigo, E. Roc-
chini, T. Stephan, J. Chem. Soc., Dalton Trans. 1999, 4475.
[10] I. Kaljurand, T. Rodima, A. Pihl, V. Mäemets, I. Leito, I. A. Koppel,
M. Mishima, J. Org. Chem. 2003, 68, 9988.
[11] I. Leito, T. Rodima, I. A. Koppel, R. Schwesinger, V. M. Vlasov, J.
Org. Chem. 1997, 62, 8479.
[12] T. Rodima, I. Kaljurand, A. Pihl, V. Mäemets, I. Leito, I. A. Koppel,
J. Org. Chem. 2002, 67, 1873.
[31] C. A. Streuli, Anal. Chem. 1959, 31, 1652.
[32] T. Allman, R. G. Goel, Can. J. Chem. 1982, 60, 716.
[33] M. M. Rahman, H. Liu, A. Prock, W. P. Giering, Organometallics
1989, 8, 1.
[34] P. M. Zizelman, C. Amatore, J. K. Kochi, J. Am. Chem. Soc. 1984,
106, 3771.
[35] G. Perdoncin, G. Scorrano, J. Am. Chem. Soc. 1977, 99, 6983.
[36] M. Schlaf, A. J. Lough, R. H. Morris, Organometallics 1993, 12,
3808.
[37] S. Shima, E. J. Lyon, R. K. Thauer, B. Mienert, E. Bill, J. Am. Chem.
Soc. 2005, 127, 10430.
[38] J. W. Peters, W. N. Lanzilotta, B. J. Lemon, L. C. Seefeldt, Science
1998, 282, 1853.
[39] Y. Nicolet, C. Cavazza, J. C. Fontecilla-Camps, J. Inorg. Biochem.
2002, 91, 1.
[40] Z. Otwinowski, W. Minor, Methods in Enzymolog. 1997, 276, 307.
[41] SHELXTL (V6.0), Bruker AXS, Madison, WI, USA, 2001.
[13] E.-I. Room, I. Kaljurand, I. Leito, T. Rodima, I. A. Koppel, V. M.
Vlasov, J. Org. Chem. 2003, 68, 7795.
[14] G. Jia, H. M. Lee, I. D. Williams, C.-P. Lau, Y. Chen, Organometal-
lics 1997, 16, 3941.
[15] G. Jia, R. H. Morris, J. Am. Chem. Soc. 1991, 113, 875.
[16] G. Jia, R. H. Morris, Inorg. Chem. 1990, 29, 581.
[17] S. M. Ng, Y. Q. Fang, C. P. Lau, W. T. Wong, G. Jia, Organometallics
1998, 17, 2052.
Received: October 18, 2006
[18] G. C. Jia, C. P. Lau, Coord. Chem. Rev. 1999, 192, 83 .
Published online: January 24, 2007
Chem. Eur. J. 2007, 13, 3796 – 3803
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3803