Cyanide Ligand Basicities in Cp′M(L)2CN Complexes (M = Ru, Fe). Correlation between Heats of Protonation and vCN

Article abstract of DOI:10.1021/ic971124o

Basicities of the cyanide ligands in a series of Cp′M(L)₂CN complexes were investigated by measuring their heats of protonation (-δH_CNH) by CF₃SO₃H in 1.2-dichloroethane solution at 25.0 °C to give Cp′M(L)₂(CNH)⁺CF₃SO₃ ^-, in which the N-H⁺ group is probably hydrogen-bonded to the CF₃SO₃^- anion. Basicities (-δH_CNH) of the CpRu(PR₃)₂CN complexes increase from 20.5 (PPh₃) to 22.4 (PMe₃) kcal/mol with increasing donor abilities of the phosphine ligands. Basicities of all the Cp′Ru(PR₃)₂CN complexes, where Cp′ = Cp or Cp*, are linearly correlated with their vCN values; the nonphosphine complexes. CpRu(l.10-phen)CN and CpRu(COD)CN, do not follow the same correlation. For a large number of Cp′M(L)₂CN complexes (M = Ru, Fe, L₂ = mono- and bidentate phosphines, CO, 1,10-phen, and COD), their vCN values parallel vCN values of their protonated Cp′M(L)₂(CNH)⁺ analogues. Also, ³¹P NMR chemical shifts of the unprotonated Cp′M(PR₃)₂-CN and protonated CpM(PR₃)₂(CNH)⁺ complexes are linearly related. Despite the high basicity of Ru in Cp*Ru-(PMe₃)₂Cl (30.2 kcal/mol), the CN^- in Cp*Ru(PMe₃)₂CN (25.0 kcal/mol) is the site of protonation: factors that determine whether protonation occurs at the Ru or the CN^- are discussed.

Full text of DOI:10.1021/ic971124o

1868
Inorg. Chem. 1998, 37, 1868-1875
Cyanide Ligand Basicities in Cp′M(L)2CN Complexes (M ) Ru, Fe). Correlation between
Heats of Protonation and νCN
Chip Nataro, Jiabi Chen,† and Robert J. Angelici*
Department of Chemistry and Ames Laboratory, Iowa State University, Ames, Iowa 50011
ReceiVed September 4, 1997
Basicities of the cyanide ligands in a series of Cp′M(L)2CN complexes were investigated by measuring their
heats of protonation (-∆HCNH) by CF3SO3H in 1,2-dichloroethane solution at 25.0 °C to give
Cp′M(L)2(CNH)+CF3SO3-, in which the N-H+ group is probably hydrogen-bonded to the CF3SO3 anion.
-
Basicities (-∆HCNH) of the CpRu(PR3)2CN complexes increase from 20.5 (PPh3) to 22.4 (PMe3) kcal/mol with
increasing donor abilities of the phosphine ligands. Basicities of all the Cp′Ru(PR3)2CN complexes, where Cp′
) Cp or Cp*, are linearly correlated with their νCN values; the nonphosphine complexes, CpRu(1,10-phen)CN
and CpRu(COD)CN, do not follow the same correlation. For a large number of Cp′M(L)2CN complexes (M )
Ru, Fe, L2 ) mono- and bidentate phosphines, CO, 1,10-phen, and COD), their νCN values parallel νCN values
of their protonated Cp′M(L)2(CNH)+ analogues. Also, 31P NMR chemical shifts of the unprotonated Cp′M(PR3)2-
CN and protonated CpM(PR3)2(CNH)+ complexes are linearly related. Despite the high basicity of Ru in Cp*Ru-
(PMe3)2Cl (30.2 kcal/mol), the CN- in Cp*Ru(PMe3)2CN (25.0 kcal/mol) is the site of protonation; factors that
determine whether protonation occurs at the Ru or the CN- are discussed.
Introduction
for the reaction (eq 1) of the cyanide complex with CF3SO3H
The cyanide ligand in transition metal complexes undergoes
a variety of reactions, particularly those with electrophiles.1 The
nitrogen is alkylated by R3O+ and RX reagents to give alkyl
isocyanide complexes. Reactions with other transition metal
complexes have yielded numerous cyanide-bridged di- and
polynuclear compounds.1,2 The cyanide ligand is also readily
protonated to give hydrogen isocyanide (CNH) complexes.
Despite the existence of a large number of protonated cyanide-
ligand complexes,1 only two quantitative measurements of
cyanide ligand donor ability have been reported.3 In one study,
DCE
Cp′M(L)2CN + CF3SO3H
8
25 °C
Cp′M(L)2(CNH)+CF3SO3 ; ∆HCNH (1)
-
in 1,2-dichloroethane (DCE) solvent at 25.0 °C. Previously,
heats of protonation of the metal (∆HHM) in the related non-
cyanide complexes Cp′M(PR3)2X, where M ) Ru or Os, Cp′
) Cp or Cp*, and X ) Cl, Br, I, or H, were reported.4 It was
observed
3a
the complexes M(bipy)2(CN)2, where M ) Fe, Ru, or Os,
+
were protonated by o-ClC6H4NH3 in acetic acid solvent.
Equilibrium constants for protonation of the first CN- ligand
increased slightly with the metal in the following order: Fe
(0.9) < Os (1.3) < Ru (2.8). Protonation of the second CN-
occurred with substantially smaller equilibrium constants and
the metal-dependence changed to the following: Fe (0.005) <
Ru (0.008) < Os (0.026). In the other study,3b the basicities of
the M(CN)64- complexes in water were found to increase with
the metals, as indicated by the pKa of their protonated forms
(in parentheses), in the following order: Ru (3.2) < Os (3.3)
< Fe (3.4). Thus, the basicity trend is different in the two
systems, but the effect of the metal is not large.
In the present paper, we describe studies of the basicity of
the cyanide ligand in the family of Cp′M(L)2CN complexes,
where M ) Ru or Fe, Cp′ ) Cp or Cp*(η5-C5Me5), (L)2 )
phosphines, 1,5-cyclooctadiene (COD), or 1,10-phenanthroline
(1,10-phen). The basicity of the cyanide ligand in these
that the basicity (-∆HHM) of the Ru increased by 9.0 kcal/mol
when the Cp in CpRu(PMe3)2Cl (21.2(4) kcal/mol) was
substituted by Cp* in Cp*Ru(PMe3)2Cl (30.2(2) kcal/mol). A
large increase in basicity was also noted when the PPh3 ligands
in CpOs(PPh3)2Br (16.3(1) kcal/mol) were replaced by PMe3
in CpOs(PMe3)2Br (29.4(4) kcal/mol). In addition, the replace-
ment of Br- in CpOs(PPh3)2Br (16.3(1) kcal/mol) by H- in
CpOs(PPh3)2H (37.3(1) kcal/mol) resulted in an enormous
increase (21.0 kcal/mol) in the basicity of the Os. Also,
replacement of the Ru in CpRu(PMe3)2Br (20.9(3) kcal/mol)
by Os in CpOs(PMe3)2Br (29.4(4) kcal/mol) increased the
basicity of the metal center by 8.5 kcal/mol. Thus, variations
complexes is defined as the enthalpy of protonation (∆HCNH
)
† Permanent address: Shanghai Institute of Organic Chemistry, Chinese
Academy of Sciences, Shanghai, PRC.
(1) Fehlhammer, W. P.; Fritz, M. Chem. ReV. 1993, 93, 1243.
(2) Darensbourg, D. J.; Yoder, J. C.; Holtcamp, M. W.; Klausmeyer, K.
K.; Reibenspies, J. H. Inorg. Chem. 1996, 35, 4764.
(3) (a) Schilt, A. A. J. Am. Chem. Soc. 1963, 85, 904. (b) Poulopoulo, V.
G.; Taube, H. Inorg. Chem. 1997, 36, 4782.
(4) Rottink, M. K.; Angelici, R. J. J. Am. Chem. Soc. 1993, 115, 7267.
S0020-1669(97)01124-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 03/31/1998

CN Ligand Basicities in Cp′M(L)2CN Complexes
Inorganic Chemistry, Vol. 37, No. 8, 1998 1869
Table 1. IR Data (CH2Cl2 Solvent) for the Cp′M(L)2CN and
Cp′M(L)2(CNH)+ Complexes
in the Cp′, M, PR3, and X units of the Cp′M(PR3)2X complexes
lead to large changes in the basicity (-∆HHM) of the metal
center.
complex
Cp′ML2
ν(CN)
complex
ν(CN)
In the cyanide complexes, Cp′ML2CN, there is also the
possibility that protonation would occur at the metal rather than
at the cyanide ligand. It is known,5 for example, that the trigonal
bipyramidal complexes (L)RhCN, where L ) N(CH2CH2PPh2)3
or P(CH2CH2PPh2)3, are protonated by CF3SO3H at the Rh
instead of the CN to give (L)Rh(CN)(H)+. Although CpRu-
(PPh3)2CN is known6,7 to protonate at the cyanide, it seemed
possible that protonation might occur at the metal if its basicity
were sufficiently enhanced by strongly donating ligands, as in
Cp*Ru(PMe3)2CN.
Although the cyanide ligand protonation reaction (eq 1)
suggests that the ∆HCNH value simply represents the Lewis
basicity of the ligand, there is excellent X-ray evidence7 for
hydrogen bonding between the CN-H+ proton and an oxygen
of the CF3SO3- anion (Ru-CN-H- - -OSO2CF3) in the solid-
state structure of CpRu(PPh3)2(CNH)+-O3SCF3, which is one
of the complexes studied in the present investigation. In
addition, there are other examples of hydrogen bonding between
protonated cyanide ligands and the oxygen atom in ethers,
as for example in the tetrahydrofuran adduct (CO)5Cr-CN-
H- - -THF.1,8 If such hydrogen-bonding were to occur between
the Cp′M(L)2(CNH)+ and -O3SCF3 ions in solution under the
conditions of the calorimetric studies, the measured ∆HCNH
values would include the enthalpies associated with hydrogen
bonding. It was therefore important to determine whether
hydrogen bonding was a consideration in the interpretation of
the ∆HCNH results. Finally, these investigations required the
synthesis of a broad range of Cp′M(L)2CN complexes which
2CN
5CN
6CN
7CN
8CN
9CN
CpRu(PEt3)2
2065 (m) 2CNH+
2065 (m) 5CNH+
2069 (m) 6CNH+
2071 (m) 7CNH+
2072 (m) 8CNH+
2105 (m) 9CNH+
2129 (m) 10CNH+
2015 (vw)
2017 (vw)
2017 (vw)
2021 (vw)
2023 (vw)
2092 (vw)
obscured
2016 (vw)
2019 (vw)
2021 (vw)
2032 (vw)
2021 (vw)
2033 (vw)
2023 (vw)
2022 (vw)
2055 (vw)
2024 (vw)
2015 (vw)
2001 (vw)
CpRu(PMe3)2
CpRu(PMe2Ph)2
CpRu(PMePh2)2
CpRu(PPh3)2
CpRu(CO)(PPh3)
CpRu(CO)2
10CN
11CN
12CN
13CN
14CN
15CN
17CN
18CN
19CN
20CN
21CN
23CN
5*CN
18*CN
19*CN
20*CN
CpRu(P(p-MeOC6H4)3)2 2067 (m) 11CNH+
CpRu(P(p-MeC6H4)3)2
2069 (m) 12CNH+
CpRu(P(m-MeOC6H4)3)2 2069 (m) 13CNH+
CpRu(P(p-CF3C6H4)3)2
CpRu(P(p-FC6H4)3)2
CpRu(dppm)
CpRu(dppe)
CpRu(dppp)
CpRu(COD)
CpRu(1,10-phen)
CpFe(dppe)
Cp*Ru(PMe3)2
Cp*Ru(dppe)
Cp*Ru(dppp)
Cp*Ru(COD)
2080 (m) 14CNH+
2075 (m) 15CNH+
2076 (m) 17CNH+
2075 (m) 18CNH+
2067 (m) 19CNH+
2100 (m) 20CNH+
2073 (m) 21CNH+
2063 (m) 23CNH+
2057 (m) 5*CNH+
2065 (m) 18*CNH+ 2019 (vw)
2057 (m) 19*CNH+ 2017 (vw)
2092 (m) 20*CNH+ 2046 (vw)
either Strem Chemical Co. or Aldrich Chemical Co. Dicyclopentadiene
was purchased from Aldrich. NaCN and 1,10-phenanthroline were
purchased from Fisher Chemical Co. The complexes CpRu(P(OMe)3)2Cl
(1Cl),10 CpRu(PMe3)2Cl (5Cl),11 CpRu(PPh3)2Cl (8Cl),12 CpRu(PPh3)2-
CN (8CN),13 CpRu(CO)(PPh3)Cl (9Cl),14 CpRu(CO)2CN (10CN),15
CpRu(dppm)Cl (17Cl),10 CpRu(dppe)Cl (18Cl),10 CpRu(dppe)CN
(18CN),16 CpRu(COD)Cl (20Cl),11 CpFe(CO)2Br (22Br),17 CpFe(dppe)-
CN (23CN),18 Cp*Ru(PMe3)2Cl (5*Cl),19 Cp*Ru(dppe)Cl (18*Cl),20
Cp*Ru(dppp)Cl (19*Cl),20 and Cp*Ru(COD)Cl (20*Cl)20,21 were
prepared by literature methods.
provided an opportunity to examine correlations among ∆HCNH
,
Elemental analyses were performed on a Perkin-Elmer 2400 Series
II CHNS/O analyzer at Iowa State University or at the Shanghai Institute
of Organic Chemistry. IR spectra were measured on a Nicolet 710
FTIR or Magna-IR560 spectrophotometer; values for the v(CN) band
νCN, 31P NMR chemical shifts, and other ligand parameters.
Experimental Section
-
of the Cp′M(L)2(CN) and Cp′M(L)2(CNH)+ CF3SO3 complexes are
General Procedures. All preparative reactions, chromatography
and manipulations were carried out under an atmosphere of nitrogen
or argon using standard Schlenk techniques. Solvents were purified
under nitrogen using standard methods.9 Diethyl ether (Et2O) and THF
were distilled over sodium benzophenone, while benzene, hexanes, and
CH2Cl2 were distilled from CaH2. The C2H5OH and CH3OH were
deoxygenated using four or five freeze-pump-thaw cycles and stored
under N2. Acetone was deoxygenated and dried with 4 Å molecular
sieves; CD2Cl2 was stored over molecular sieves under nitrogen. 1,2-
Dichloroethane (DCE) was purified by washing with concentrated
sulfuric acid, distilled deionized water, 5% NaOH, and again with water.
The solvent was then predried over anhydrous MgSO4 and stored in
amber bottles over molecular sieves (4 Å). The DCE was distilled
from P4O10 under argon immediately before use. Triflic acid (CF3-
SO3H) was purchased from 3M Co. and purified by fractional distillation
under argon prior to use. The neutral Al2O3 (Brockman, Activity I,
80-100 mesh) used for chromatography was deoxygenated under high
vacuum at room temperature for 16 h, deactivated with 5% (w/w) N2-
saturated water, and stored under N2. Chromatography columns were
1.5 × (5-15) cm. The phosphines, PPh3, P(p-MeOC6H4)3, P(p-
MeC6H4)3, P(m-MeC6H4)3, P(p-FC6H4)3, P(p-CF3C6H4)3, Ph2PCH2PPh2
(dppm), Ph2PCH2CH2PPh2 (dppe), Ph2PCH2CH2CH2PPh2 (dppp), PMe3,
PEt3, PMePh2, PMe2Ph, PCy3, PEt3, and P(i-Pr)3, were purchased from
1
given in Table 1. All H NMR spectra were recorded on compounds
dissolved in CD2Cl2 solution with TMS (δ ) 0.00) as the internal
reference using a Bruker AC 200 MHz spectrometer; the abbreviation
pst refers to a pseudo-triplet. 31P{1H} NMR spectra were obtained in
CD2Cl2 on a Bruker AC 200 MHz spectrometer with H3PO4 (δ ) 0.00)
as the external reference. Electron ionization mass spectra (EIMS) were
run on a Finnigan 4500 spectrometer at 70 eV, and fast atom
bombardment (FAB) spectra were run on a Kratos MS-50 mass
spectrometer with samples in a 3-nitrobenzyl alcohol/CH3NO2 matrix.
Melting points were recorded on compounds in sealed N2-filled
capillaries and are uncorrected. Conductivity measurements were made
with a YSI 3100 Conductivity Instrument and are corrected to 25.0
°C.
Preparation of CpRu[P(OMe)3]2CN (1CN). A mixture of 1Cl
(0.51 g, 1.1 mmol) and NaCN (0.22 g, 4.5 mmol) in 50 mL of CH3OH
(10) Ashby, G. S.; Bruce, M. I., Tomkins, I. B.; Wallis, R. C. Aust. J.
Chem. 1979, 32, 1003.
(11) Albers, M. O.; Robinson, D. J.; Shaver, A.; Singleton, E. Organo-
metallics 1986, 5, 2199.
(12) Bruce, M. I.; Hameister, C.; Swincer, A. G.; Wallis, R. C. Inorg. Synth.
1990, 28, 270.
(13) Laidlaw, W. M.; Denning, C. G. J. Organomet. Chem. 1993, 463,
199.
(14) Davies, S. G.; Simpson, S. J. J. Chem. Soc., Dalton Trans. 1984, 993.
(15) Haines, R. J.; du Preez, A. L. J. Organomet. Chem. 1975, 84, 357.
(16) Baird, G. J.; Davies, S. G. J. Organomet. Chem. 1984, 262, 215.
(17) Hallam, B. F.; Pauson, P. L. J. Chem. Soc. 1956, 3030.
(18) Treichel, P. M.; Molzahn, D. C. Synth. React. Inorg. Met.-Org. Chem.
1979, 9, 21.
(19) Tilley, T. D.; Grubbs, R. H.; Bercaw, J. E. Organometallics 1984, 3,
274.
(20) Oshima, N.; Suzuki, H.; Moro-oka, Y. Chem. Lett. 1984, 1161.
(21) Koelle, U.; Kossakowski, J. J. Organomet. Chem. 1989, 362, 383.
(5) Bianchini, C.; Laschi, F.; Ottaviani, M. F.; Peruzzini, M.; Zanello,
P.; Zanobini, F. Organometallics 1989, 8, 893.
(6) Bruce, M. I.; Wallis, R. C. Aust. J. Chem. 1981, 34, 209.
(7) Sapunov, V. N.; Mereiter, K.; Schmid, R.; Kirchner, K. J. Organomet.
Chem. 1997, 530, 105.
(8) Ba¨r, E.; Fuchs, J.; Rieger, D.; Aguilar-Parrilla, F.; Limbach, H.-H.;
Fehlhammer, W. P. Angew. Chem., Int. Ed. Engl. 1991, 30, 88.
(9) Perrin, D. D.; Armarego, W. L. F.; Perrin, D. R. Purification of
Laboratory Chemicals, 2nd ed.; Pergamon: New York, 1980.

1870 Inorganic Chemistry, Vol. 37, No. 8, 1998
Nataro et al.
was refluxed with stirring for 1.5 h, during which time the orange
solution gradually turned yellow. The solvent was removed under
vacuum, and the residue was extracted with CH2Cl2 (2 × 10 mL). The
extract was reduced to about 5 mL, and hexanes (ca. 50 mL) were
added to precipitate the product. The resulting mixture was cooled at
-20 °C overnight to give 0.32 g (64%, based on 1Cl) of yellow crystals
of 1CN. Mp 133-134 °C (dec). 1H NMR (CD2Cl2): δ 4.81 (s, 5H,
Cp), 3.64 (pst, 18H, CH3). 31P NMR (CD2Cl2): δ 38.45 (s). MS
(EI): m/e 440 (M+). Anal. Calcd for C12H23NO6P2Ru: C, 32.73; H,
5.27; N, 3.18. Found: C, 32.88; H, 5.10; N, 3.22. Complex CpRu-
[P(i-Pr)3]2CN(3CN) was prepared by a similar procedure (see Sup-
porting Information).
Preparation of CpRu(PEt3)2Cl (2Cl). To a suspension of 20Cl
(0.50 g, 1.6 mmol) in 50 mL of acetone was added 0.50 g (4.2 mmol)
of PEt3. The mixture was stirred at room temperature for 1.5 h, during
which time the solid material dissolved and the solution became orange-
red in color. After removal of the solvent, the residue was recrystallized
from hexanes at -20 °C to give orange-red crystals of 2Cl. Yield:
0.60 g (85%, based on 20Cl). Mp 66-68 °C (dec). 1H NMR (CD2-
Cl2): δ 4.45 (s, 5H, Cp), 1.98 (m, 6H, -(CH2)-), 1.57 (m, 6H,
-(CH2)-), 1.09 (m, 18H, CH3). 31P NMR (CD2Cl2): δ 34.79 (s). MS
(FAB): m/e 438 (M+). Anal. Calcd for C17H35ClP2Ru: C, 46.62; H,
8.06. Found: C, 46.60; H, 8.30. The complexes CpRu[P(i-Pr)3]2Cl
(3Cl), CpRu[P(Cy)3]2Cl (4Cl), CpRu(PMe2Ph)2Cl (6Cl), and CpRu-
(PMePh2)2Cl (7Cl) were prepared by similar procedures.
Preparation of CpRu(PEt3)2CN (2CN). To a solution of 2Cl (0.60
g, 1.4 mmol) in 50 mL of CH3OH was added 0.17 g (3.5 mmol) of
NaCN. The red solution quickly turned yellow. The mixture was
stirred under reflux for 0.5 h. After cooling, the solvent was removed
in vacuo and the residue was extracted with CH2Cl2 (2 × 10 mL). The
combined CH2Cl2 extracts were evaporated to a volume of ca. 5 mL,
to which hexanes (ca. 50 mL) were added to precipitate the light yellow
product. The solid product was removed by filtration, washed with
hexanes and then dried under vacuum. Yield: 0.42 g (71%, based on
2Cl). Mp 84-86 °C (dec). 1H NMR (CD2Cl2): δ 4.74 (s, 5H, Cp),
1.95 (m, 6H, -(CH2)-), 1.45 (m, 6H, -(CH2)-), 1.05 (m, 18H, CH3).
31P NMR (CD2Cl2): δ 42.25 (s). MS (FAB): 428, 429 (M+). Anal.
Calcd for C18H35NP2Ru: C, 50.45; H, 8.23; N, 3.27. Found: C, 50.46;
H, 8.51; N, 3.17. The complexes CpRu[P(Cy)3]2CN (4CN), CpRu-
(PMe3)2CN (5CN), CpRu(PMe2Ph)2CN (6CN), CpRu(PMePh2)2CN
(7CN), and CpRu(CO)(PPh3)(CN) (9CN) were prepared by similar
methods.
Preparation of CpRu[P(p-MeOC6H4)3]2Cl (11Cl). The reaction
was carried out under an N2 atmosphere in a 500 mL, three-necked,
round-bottom flask equipped with a 50 mL dropping funnel and a reflux
condenser topped with a nitrogen bypass. The P(p-MeOC6H4)3 (2.40
g, 6.81 mmol) was dissolved in 150 mL of absolute ethanol by heating.
Hydrated ruthenium trichloride, RuCl3‚3H2O, (0.45 g, 1.7 mmol) was
dissolved in ethanol (15 mL) by bringing the mixture to a boil and
then allowing the solution to cool to room temperature. Freshly distilled
cyclopentadiene (C5H6) (0.86 mL, 0.69 g, 10 mmol) was added to the
RuCl3 solution, and the mixture was transferred to the dropping funnel.
The dark-brown solution was then added to the P(p-CH3OC6H4)3
solution over a period of 10 min while the temperature was maintained
at reflux. An additional 1.5 h of refluxing caused the solution to lighten
to a dark red-orange. The solution was filtered quickly while hot. The
filtrate was evaporated under vacuum to about two-thirds of its volume
and then cooled overnight at -20 °C. The resulting fine orange crystals
were filtered, washed with ethanol (ca. 10 mL) and then with hexanes
(10 mL), and dried in vacuo. Yield: 1.10 g (86%, based on
RuCl3‚3H2O). Mp 140-141 °C (dec). 1H NMR (CD2Cl2): δ 7.32-
7.23 (m, 12H, C6H4), 6.69-6.64 (m, 12H, C6H4), 4.08 (s, 5H, Cp),
3.77 (s, 18H, CH3O). 31P NMR (CD2Cl2): δ 38.90 (s). MS (FAB):
m/e 906 (M+). Anal. Calcd for C47H47ClO6P2Ru: C, 62.28; H, 5.23.
Found: C, 62.72; H, 5.43. The complexes CpRu[P(p-MeC6H4)3]2Cl
(12Cl), CpRu[P(m-MeC6H4)3]2Cl (13Cl), CpRu[P(p-CF3C6H4)3]2Cl
(14Cl), and CpRu[P(p-FC6H4)3]2Cl (15Cl), were prepared by similar
procedures.
became a bright yellow solution after 10 min, refluxing was continued
for a total of 2 h to ensure complete conversion. After the solvent
was removed under vacuum, the residue was extracted with CH2Cl2 (2
× 15 mL). The combined CH2Cl2 extracts were evaporated to a small
volume (ca. 5 mL). To this solution was added an excess of hexanes
(ca. 50 mL) which caused the precipitation of a yellow powder. After
filtration, washing with hexanes and drying under vacuum, 11CN was
isolated in 80% yield (0.55 g) as a yellow powder. Mp 148-150 °C
(dec). 1H NMR (CD2Cl2): δ 7.34-7.24 (m, 12H, C6H4), 6.70-6.66
(m, 12H, C6H4), 4.37 (s, 5H, Cp), 3.77 (s, 18H, CH3O). 31P NMR
(CD2Cl2): δ 49.71 (s). MS (FAB): m/e 897 (M+). Anal. Calcd for
C48H47NO6P2Ru: C, 64.28; H, 5.28; N, 1.56. Found: C, 64.11; H,
5.36; N, 1.54. The complexes CpRu[P(p-MeC6H4)3]2CN (12CN),
CpRu[P(m-MeC6H4)3]2CN (13CN), CpRu[P(p-CF3C6H4)3]2CN (14CN),
and CpRu[P(p-FC6H4)3]2CN (15CN) were prepared by similar proce-
dures.
Preparation of CpRu[P(OPh)3]2Cl (16Cl). A mixture of 8Cl (1.50
g, 2.07 mmol) and P(OPh)3 (1.92 g, 6.19 mmol) in decalin (60 mL)
was heated under reflux for 20 min, during which time the solution
turned from brick-red to yellow. The cooled solution was directly
chromatographed on Al2O3 (neutral). After the decalin and excess
P(OPh)3 were washed out of the column with hexanes, the yellow band
was eluted with hexanes/CH2Cl2/Et2O (20:1:1) and collected. The
solvent was removed, and the residue was recrystallized from hexanes/
CH2Cl2 at -20 °C to afford 0.40 g (24%, based on 8Cl) of yellow
crystals of 16Cl. Mp 142-144 °C (dec). 1H NMR (CD2Cl2): δ 7.34-
6.86 (m, 30H, OPh), 4.02 (s, 5H, Cp). 31P NMR (CD2Cl2): δ 55.65
(s). MS (FAB): m/e 822 (M+). Anal. Calcd for C41H35ClO6P2Ru:
C, 59.89; H, 4.29. Found: C, 60.08; H, 4.11.
Preparation of CpRu[P(OPh)3]2CN (16CN). Similar to the
synthesis of 1CN, 0.32 g (0.39 mmol) of 16Cl and 0.067 g (1.4 mmol)
of NaCN in 50 mL of CH3OH were refluxed for 2 h, during which
time the orange solution turned light yellow. Further treatment of the
resulting solution as described in the preparation of 1CN gave 0.27 g
(84%, based on 16Cl) of 16CN as a yellow powder. Mp 135-137 °C
(dec). 1H NMR (CD2Cl2): δ 7.34-7.10 (m, 30H, OPh), 4.02 (s, 5H,
Cp). 31P NMR (CD2Cl2): δ 55.63 (s). MS (FAB): 813, 814 (M+).
Anal. Calcd for C42H35NO6P2Ru: C, 62.07; H, 4.34; N, 1.72. Found:
C, 62.45; H, 4.08; N, 1.31.
Preparation of CpRu(dppm)CN (17CN). NaCN (0.31 g, 6.3
mmol) was added to a suspension of 17Cl (1.12 g, 1.91 mmol) in 60
mL of CH3OH. The mixture was heated under reflux for 0.5 h, during
which time the bright red solution faded to yellow. The solvent was
removed under vacuum and the resulting solid was extracted with CH2-
Cl2 (2 × 10 mL). After the combined CH2Cl2 extracts were evaporated
to a small volume (ca. 5 mL), 50 mL of hexanes were added to
precipitate a yellow solid. The crude product was recrystallized from
CH2Cl2/hexanes at -20 °C to give 0.45 g (41%, based on 17Cl) of
17CN as orange-red crystals. Mp 244-246 °C (dec). 1H NMR (CD2-
Cl2): δ 7.40-7.19 (m, 20H, Ph), 4.99 (s, 5H, Cp), 4.35 (m, 2H, -CH2-
). 31P NMR (CD2Cl2): δ 19.82 (s). MS (FAB): m/e 576, 578 (M+).
Anal. Calcd for C31H27NP2Ru: C, 64.58; H, 4.72; N, 2.43. Found:
C, 64.26; H, 4.71; N, 2.28. The complex CpRu(dppp)CN (19CN) was
prepared by a similar method.
Preparation of CpRu(dppp)Cl (19Cl). A mixture of 8Cl (1.50 g,
2.07 mmol) and dppp (0.85 g, 2.1 mmol) in 250 mL of benzene was
refluxed for 8 h, during which time the brown-yellow solution turned
orange. The solution was reduced to ca. 40 mL in vacuo and a 4:1
Et2O/hexanes mixture was added until a light yellow precipitate formed
(ca. 40-50 mL of Et2O/hexanes). After filtration, further addition of
hexanes to the filtrate gave 0.65 g (52%, based on 8Cl) of 19Cl as a
yellow powder. Mp 140-146 °C (dec). 1H NMR (CD2Cl2): δ 7.65
(m, 4H, Ph), 7.34 (m, 8H, Ph), 7.16 (m, 8H, Ph), 4.43 (s, 5H, Cp),
2.88-2.69 (m, 4H, -(CH2)3-), 2.41-2.23 (m, 2H, -(CH2)3-). 31P
NMR (CD2Cl2): δ 41.99 (s). MS (FAB): m/e 614 (M+). Anal. Calcd
for C32H31ClP2Ru: C, 62.59; H, 5.09. Found: C, 62.47; H, 5.01.
Preparation of CpRu(COD)CN (20CN). A mixture of 20Cl (0.40
g, 1.3 mmol) and NaCN (0.19 g, 3.9 mmol) in 50 mL of CH3OH was
heated under reflux for 1 h. After removal of the solvent, the residue
was chromatographed on Al2O3 (neutral) with CH2Cl2/Et2O (10:1) as
eluent. The yellow band was collected. After removal of the solvent,
Preparation of CpRu[P(p-MeOC6H4)3]2CN (11CN). A mixture
of 11Cl (0.70 g, 0.77 mmol) and NaCN (0.19 g, 3.9 mmol) in 60 mL
of CH3OH was refluxed with stirring. Although the orange suspension

CN Ligand Basicities in Cp′M(L)2CN Complexes
Inorganic Chemistry, Vol. 37, No. 8, 1998 1871
0.18 g (46%, based on 20Cl) of 20CN was obtained as a yellow powder.
Mp 173-175 °C (dec). 1H NMR (CD2Cl2): δ 5.04 (s, 5H, Cp), 4.64
(m, 2H, C8H12), 3.91 (m, 2H, C8H12), 2.60 (m, 2H, C8H12), 2.23 (m,
2H, C8H12), 2.15 (m, 2H, C8H12), 1.91 (m, 2H, C8H12). MS (FAB):
m/e 300 (M+). Anal. Calcd for C14H17NRu: C, 55.98; H, 5.71; N,
4.66. Found: C, 55.55; H, 5.84; N, 4.53.
Preparation of CpRu(1,10-phen)Cl (21Cl). A solution of 20Cl
(0.50 g, 1.61 mmol) and 1,10-phenanthroline (0.32 g, 1.61 mmol) in
50 mL of acetone was stirred at room temperature for 7.5 h. The dark
purple precipitate that formed was filtered off, washed with Et2O (2 ×
25 mL), and dried in vacuo. This gave 0.55 g (89%, based on 20Cl)
of the dark purple product, 21Cl. Mp > 250 °C (dec). 1H NMR (CD2-
Cl2): δ 9.90 (dd, J ) 5 and 1 Hz, 2H), 8.28 (dd, J ) 8 and 1 Hz, 2H),
7.90 (s, 2H), 7.72 (dd, J ) 8 and 5 Hz, 2H), 4.32 (s, 5H, Cp). MS
(FAB): m/e 382 (M+). Anal. Calcd for C17H13ClN2Ru: C, 53.48; H,
3.43; N, 7.34. Found: C, 53.49; H, 3.37; N, 7.30.
Preparation of CpRu(1,10-phen)CN (21CN). A mixture of 21Cl
(0.52 g, 1.4 mmol) and NaCN (0.13 g, 3.7 mmol) in 50 mL of CH3OH
was refluxed for 2 h, during which time the dark purple solution turned
red-purple. After evaporation of the solvent, the dark purple residue
was washed with EtOH/H2O (3:1) (2 × 10 mL) and then with EtOH
(10 mL), and finally dried in vacuo to give 0.45 g (88%, based on
21Cl) of brick-red crystals of 21CN. Mp > 280 °C (dec). 1H NMR
(CD2Cl2): δ 9.63 (dd, J ) 5 and 1 Hz, 2H), 8.30 (dd, J ) 8 and 1 Hz,
2H), 7.93 (s, 2H), 7.63 (dd, J ) 8 and 5 Hz, 2H), 4.61 (s, 5H, Cp). MS
(FAB): m/e 372, 373 (M+). Anal. Calcd for C18H13N3Ru: C, 58.06;
H, 3.52; N, 11.28. Found: C, 57.71; H, 3.50; N, 11.31.
Preparation of Cp*Ru(PMe3)2CN (5*CN). To a solution of 5*Cl
(0.60 g, 1.4 mmol) in 40 mL of CH3OH was added 0.13 g (2.7 mmol)
of NaCN. The solution turned from orange to yellow after stirring for
10 min at room temperature. The mixture was refluxed for 5 min to
ensure complete conversion. The resulting light yellow solution was
evaporated to dryness, and the residue was extracted with CH2Cl2 (2
× 20 mL). After the combined extracts were reduced to a volume of
about 3 mL, hexanes (50 mL) were added to precipitate 5*CN as a
light yellow powder. Yield: 0.45 g (76%, based on 5*Cl). Mp 166-
168 °C (dec). 1H NMR (CD2Cl2): δ 1.80 (s, 15H, Cp*), 1.40 (pst,
18H, CH3). 31P NMR (CD2Cl2): δ 9.41 (s). MS (EI): m/e 414, 415
(M+). Anal. Calcd for C17H33NP2Ru: C, 49.26; H, 8.03; N, 3.38.
Found: C, 48.74; H, 8.28; N, 3.34. The complexes Cp*Ru(dppe)CN
(18*CN), Cp*Ru(dppp)CN (19*CN), and Cp*Ru(COD)CN (20*CN),
were prepared by similar methods.
the acid. 1H NMR (CD2Cl2): δ 5.00 (s, 5H, Cp), 1.95 (m, 6H,
-(CH2)-), 1.56 (m, 6H, -(CH2)-), 1.06 (m, 18H, CH3). 31P NMR
(CD2Cl2): δ 39.36 (s).
-
[CpRu(P(i-Pr)3)2(CNH)+]CF3SO3 (3CNH+CF3SO3-). 1H NMR
(CD2Cl2): δ 4.86 (s, 5H, Cp), 2.29 (m, 6H, CH(CH3)2), 1.25 (m, 36H,
CH3). 31P NMR (CD2Cl2): δ 76.50 (s).
[CpRu(PMe3)2(CNH)+]CF3SO3 (5CNH+ CF3SO3-). 1H NMR
-
(CD2Cl2): δ 4.69 (s, 5H, Cp), 1.49 (pst, 18H, CH3). 31P NMR (CD2-
Cl2): δ 11.77 (s).
[CpRu(PMe2Ph)2(CNH)+]CF3SO3- (6CNH+ CF3SO3-). 1H NMR
(CD2Cl2): δ 7.35 (m, 10H, Ph), 4.83 (s, 5H, Cp), 1.60 (pst, 12H, CH3).
31P NMR (CD2Cl2): δ 20.21 (s). IR (DCE): ν(CN) (cm-1) 2013 (vw).
[CpRu(PMePh2)2(CNH)+]CF3SO3- (7CNH+ CF3SO3-). 1H NMR
(CD2Cl2): δ 7.23 (m, 20H, Ph), 4.87 (s, 5H, Cp), 1.46 (pst, 6H, CH3).
31P NMR (CD2Cl2): δ 35.79 (s).
-
[CpRu(PPh3)2(CNH)+]CF3SO3 (8CNH+CF3SO3-). 1H NMR
(CD2Cl2): δ 7.24 (m, 30H, Ph), 4.63 (s, 5H, Cp). 31P NMR (CD2Cl2):
δ 48.91 (s).
[CpRu(PPh3)(CO)(CNH)+]CF3SO3- (9CNH+CF3SO3-). 1H NMR
(CD2Cl2): δ 7.41 (m, 15H, Ph), 5.19 (s, 5H, Cp). 31P NMR (CD2Cl2):
δ 51.23 (s). IR (CH2Cl2): ν(CO) (cm-1) 2004 (s).
[CpRu(CO)2(CNH)+]CF3SO3- (10CNH+CF3SO3-). The color of
the solution changed from light yellow to orange-yellow upon addition
of the acid. 1H NMR (CD2Cl2): δ 5.70 (s, 5H, Cp). IR (CH2Cl2):
ν(CO) (cm-1) 2084 (s), 2024 (s).
[CpRu(P(p-MeOC6H4)3)2(CNH)+]CF3SO3 (11CNH+CF3SO3-).
-
1H NMR (CD2Cl2): δ 7.57 (m, 12H, Ph), 7.33 (m, 12H, Ph), 4.77 (s,
5H, Cp), 2.31 (s, 18H, CH3). 31P NMR (CD2Cl2): δ 45.25 (s).
[CpRu(P(p-MeC6H4)3)2(CNH)+]CF3SO3- (12CNH+CF3SO3-). 1H
NMR (CD2Cl2): δ 7.05 (m, 24H, C6H4), 4.60 (s, 5H, Cp), 2.34 (s,
18H, CH3). 31P NMR (CD2Cl2): δ 46.80 (s).
[CpRu(P(m-MeC6H4)3)2(CNH)+]CF3SO3- (13CNH+CF3SO3-). 1H
NMR (CD2Cl2): δ 7.01 (m, 24H, C6H4), 4.63 (s, 5H, Cp), 2.18 (s,
18H, CH3). 31P NMR (CD2Cl2): δ 48.57 (s).
[CpRu(P(p-CF3C6H4)3)2(CNH)+]CF3SO3- (14CNH+CF3SO3-). 1H
NMR (CD2Cl2): δ 7.57 (m, 12H, C6H4), 7.33 (m, 12H, C6H4), 4.77 (s,
5H, Cp). 31P NMR (CD2Cl2): δ 51.68 (s).
-
[CpRu(P(p-FC6H4)3)2(CNH)+]CF3SO3 (15CNH+CF3SO3-). 1H
NMR (CD2Cl2): δ 7.19 (m, 12H, C6H4), 7.01 (m, 12H, C6H4), 4.68 (s,
5H, Cp). 31P NMR (CD2Cl2): δ 48.11 (s).
-
[CpRu(dppm)(CNH)+]CF3SO3 (17CNH+CF3SO3-). The color
of the solution changed from orange-yellow to yellow upon addition
of the acid. 1H NMR (CD2Cl2): δ 7.45 (m, 20H, Ph), 5.25 (s, 5H,
Cp), 4.36 (m, 2H, -(CH2)-). 31P NMR (CD2Cl2): δ 13.31 (s).
Preparation of CpFe(CO)2CN (22CN). To a suspension of
CpFe(CO)2Br (0.50 g, 2.0 mmol) in 40 mL of CH3OH was added 0.28
g (5.7 mmol) of NaCN. The mixture was refluxed for 15 min during
which time the mixture changed to an orange-yellow solution. After
removal of the solvent, the residue was extracted with CH2Cl2 (2 × 10
mL). The combined extracts were evaporated to a volume of about 5
mL, and hexanes (50 mL) were added to precipitate a brown-yellow
solid. The product was filtered off and dried in vacuo to give 0.30 g
(75%, based on CpFe(CO)2Br) of 22CN as a brown-yellow powder.
Mp 104-107 °C (dec). 1H NMR (CD2Cl2): δ 5.16 (s, Cp). MS (EI):
203 (M+). Anal. Calcd for C8H5O2NFe: C, 47.34; H, 2.48; N, 6.90.
Found: C, 47.30; H, 2.51; N, 6.59.
[CpRu(dppe)(CNH)+]CF3SO3 (18CNH+CF3SO3-). 1H NMR
-
(CD2Cl2): δ 7.48 (m, 20H, Ph), 4.96 (s, 5H, Cp), 2.59 (d, J ) 24.0
Hz, 4H, -(CH2)-). 31P NMR (CD2Cl2): δ 83.14 (s).
[CpRu(dppp)(CNH)+]CF3SO3- (19CNH+CF3SO3-). The color of
the solution changed from yellow to faint green upon addition of the
acid. 1H NMR (CD2Cl2): δ 7.33 (m, 20H, Ph), 4.91 (s, 5H, Cp), 2.68
(m, 4H, -(CH2)-), 2.53 (m, 2H, -(CH2)-). 31P NMR (CD2Cl2): δ
42.69 (s).
[CpRu(COD)(CNH)+]CF3SO3- (20CNH+CF3SO3-). The color of
the solution changed from yellow to faint green upon addition of the
acid. 1H NMR (CD2Cl2): δ 5.20 (s, 5H, Cp), 4.98 (m, 2H, C8H12),
4.18 (m, 2H, C8H12), 2.56 (m, 2H, C8H12), 2.41 (m, 2H, C8H12), 2.35
(m, 2H, C8H12), 2.17 (m, 2H, C8H12).
Protonation Reactions. Compounds 2CN, 3CN, 5CN-15CN,
17CN-21CN, 23CN, 5*CN, and 18*CN-20*CN were protonated for
characterization of the [CpML2(CNH)]+CF3SO3- products by dissolving
approximately 10 mg of the complex in 0.50 mL of either CD2Cl2 (for
NMR) or CH2Cl2 (for IR) in an NMR tube under nitrogen. To the
solution was added 1 equiv of CF3SO3H through the rubber septum
using a gas-tight microliter syringe. The solutions did not change color
significantly, unless otherwise noted. Yields of the protonated products
were quantitative as indicated by the disappearance of the reactant
signals and appearance of the product signals in IR and 1H NMR spectra
[CpRu(1,10-phen)(CNH)+]CF3SO3- (21CNH+CF3SO3-). The color
of the solution changed from purple to orange upon addition of the
acid. 1H NMR (CD2Cl2): δ 9.41 (dd, J ) 5 and 1 Hz, 2H), 8.46 (dd,
J ) 8 and 1 Hz, 2H), 8.01 (s, 2H), 7.75 (dd, J ) 8 and 5 Hz, 2H), 4.86
(s, 5H, Cp).
-
[CpFe(dppe)(CNH)+]CF3SO3 (23CNH+CF3SO3-). 1H NMR
(CD2Cl2): δ 7.49 (m, 20H, Ph), 4.52 (s, 5H, Cp), 2.54 (m, 4H, -(C-
1
of the solutions. No precipitates formed, and no H NMR signals for
H2)-). 31P NMR (CD2Cl2): δ 98.30 (s).
[Cp*Ru(PMe3)2(CNH)+]CF3SO3- (5*CNH+CF3SO3-). The color
of the solution changed from yellow to faint green upon addition of
the acid. 1H NMR (CD2Cl2): δ 1.83 (s, 15H, Cp*), 1.46 (pst, 18H,
CH3). 31P NMR (CD2Cl2): δ 6.43 (s).
other products were detected. The absence of a signal at fields higher
than δ 0.0 ppm showed that protonation did not occur at the metal. No
resonance6,7 for the proton on the CNH ligand was observed, as was
previously reported for CpRu(PPh3)2(CNH)+.
[CpRu(PEt3)2(CNH)+]CF3SO3- (2CNH+CF3SO3-). The color of
the solution changed from light yellow to faint green upon addition of
[Cp*Ru(dppe)(CNH)+]CF3SO3- (18*CNH+CF3SO3-). The color
of the solution changed from yellow to faint green upon addition of

1872 Inorganic Chemistry, Vol. 37, No. 8, 1998
Nataro et al.
the acid. 1H NMR (CD2Cl2): δ 7.56 (m, 20H, Ph), 2.46 (m, 4H,
-(CH2)-), 1.55 (s, 15H, Cp*). 31P NMR (CD2Cl2): δ 78.98 (s).
[Cp*Ru(dppp)(CNH)+]CF3SO3- (19*CNH+CF3SO3-). The color
of the solution changed from yellow to faint green upon addition of
the acid. 1H NMR (CD2Cl2): δ 7.33 (m, 20H, Ph), 2.59 (m, 6H,
-(CH2)-), 1.43 (s, 15H, Cp*). 31P NMR (CD2Cl2): δ 39.85 (s).
[Cp*Ru(COD)(CNH)+]CF3SO3- (20*CNH+CF3SO3-). The color
of the solution changed from yellow to faint green upon addition of
the acid. 1H NMR (CD2Cl2): δ 3.72 (m, 4H, C8H12), 2.51 (m, 4H,
C8H12), 2.30 (m, 2H, C8H12), 2.06 (m, 2H, C8H12), 1.75 (s, 15H, Cp*).
Calorimetric Studies. Heats of protonation (∆HCNH) of the CpML2-
CN complexes were determined with 0.1 M CF3SO3H in DCE solvent
at 25.0 °C. Titrations were performed using a Tronac model 458
Preparation of Compounds 1CN-7CN, 9CN, 11CN-
17CN, 19CN-22CN, 5*CN, 19*CN and 20*CN. The reaction
of CpRu(PPh3)2Cl and NaCN in refluxing methanol gives CpRu-
(PPh3)2CN (8CN) in high yield.13 Related cyanide complexes
have been prepared by a similar method except KCN was used
instead of the sodium salt.9 We have found that the synthesis
developed by Laidlaw and Denning13 is a useful general
procedure for the preparation of the CpRuL2CN compounds (eq
6). Successful synthesis occurs when the L2 ligands are a variety
CpRuL2Cl + xs NaCN methanol8 CpRuL2CN + NaCl (6)
reflux
isoperibol calorimeter as originally described22 and then modified.23
A
of monodentate and bidentate phosphines or other four-electron
donors such as COD or 1,10-phenanthroline. Yields are
typically greater than 70% and the pure products are obtained
by column chromatography.
typical calorimetric run consisted of three sections:24 initial heat capacity
calibration, titration, and final heat capacity calibration. Each section
was preceded by a baseline acquisition period. During the titration,
1.2 mL of a 0.1 M CF3SO3H solution (standardized to a precision of
(0.0002 M) in DCE was added at a rate of 0.3962 mL/min to 50 mL
of a 2.6 mM solution of the complex (5-10% excess) in DCE at 25.0
°C. Infrared spectra of the titrated solutions exhibited ν(CN) bands
characteristic of the CpM(L)2(CNH)+ products.
Protonation of the Cp′ML2CN Complexes. The reactions
of 2CN, 3CN, 5CN-15CN, 17CN-21CN, 23CN, 5*CN, and
18*CN-20*CN with 1 equiv of triflic acid in CD2Cl2 gave
quantitatively the N-protonated cationic isocyanide complexes
2CNH+, 3CNH+, 5CNH+-15CNH+, 17CNH+-21CNH+,
23CNH+, 5*CNH+, and 18*CNH+-20*CNH+. This proto-
nation causes the Cp resonances in the 1H NMR spectra to shift
downfield by approximately 0.3 ppm, while protonation of the
Cp* complexes causes the Cp* resonances to shift downfield
by approximately 0.03 ppm. The 31P signals of the compounds
containing phosphorus ligands shift upfield approximately 5.0
ppm upon protonation of the CN- ligand. At the same time,
the ν(CN) bands move approximately 50 cm-1 to lower
wavenumbers. IR data in the ν(CN) region for the protonated
compounds are shown in Table 1. Previously, [CpRu(PPh3)2-
Two separately standardized acid solutions were used for determining
∆HCNH values of each complex. The reported values are the average
of at least four titrations and as many as five. The reaction enthalpies
were corrected for the heat of dilution (∆Hdil) of the acid in DCE (-0.2
kcal/mol).23 The reported error in ∆HCNH is the average deviation from
the mean of all of the determinations. Titrations of 1,3-diphenylguani-
dine (GFS Chemicals) with CF3SO3H in DCE (-36.9 ( 0.3 kcal/mol;
lit.22 -37.2 ( 0.4 kcal/mol) were used to monitor the performance of
the calorimeter before each set of determinations.
Results
Preparations of 2Cl-4Cl, 6Cl, 7Cl, 11Cl-16Cl, 19Cl and
21Cl. Three different routes were used to prepare these
CpRuL2Cl complexes. Compounds 2Cl-4Cl, 6Cl, and 7Cl
were prepared by using 2 equiv of the appropriate phosphine
to displace cyclooctadiene from CpRu(COD)Cl (20Cl) (eq 3).
-
(CNH)]+ PF6 was isolated and characterized by elemental
analysis and IR and NMR spectroscopy;6 recently [CpRu-
(PPh3)2(CNH)]+ CF3SO3- was isolated and its structure estab-
lished by X-ray diffraction studies.7 Also, the isolation and
-
characterization of [CpFe(dppe)(CNH)]+ BF4 have been
reported.26
CpRu(COD)Cl + 2PR3 acetone8 CpRu(PR3)2Cl + COD (3)
Complexes 2CNH+, 3CNH+, 5CNH+-15CNH+, 17CNH+-
21CNH+, 23CNH+, 5*CNH+, and 18*CNH+-20*CNH+ are
deprotonated quickly and quantitatively with 1 equiv of diphe-
nylguanidine (DPG) in CH2Cl2 or DCE solution to give the
original complexes. The compounds were separated from the
DPGH+CF3SO3- by passing the solution through a short (5 cm)
alumina column using CH2Cl2 as the eluent. Evaporation of
the solution to dryness gave the pure compounds. Compound
3CN is sparing soluble in CH2Cl2 and DCE and therefore, the
basicity of this compound could not be determined by calorim-
etry. The reactions of CF3SO3H with 1CN, 4CN, 16CN, and
22CN in CD2Cl2 gave products other than CpML2(CNH)+ as
or THF
20Cl
This route was developed by Albers et al.11 to prepare CpRuL2-
1
Cl (L ) CO, PPh3, PMe3,
/ dppe, P(OMe)3, or PPh2H).
2
Compounds 11Cl-15Cl were prepared in a two-step synthesis
developed by Bruce et al. for the synthesis of CpRu(PPh3)2-
Cl.12 In the first step, RuCl3‚3H2O and C5H6 were heated to
reflux in ethanol to give [CpRu(Cl)2]n,12 which was then reacted
with approximately 3 equiv of the desired triarylphosphine to
give CpRu(PR3)2Cl (eq 4). Compound 12Cl was prepared by
RuCl3‚3H2O + xs C5H6 + xs PR3 ethanol8 CpRu(PR3)2Cl
1
indicated by their H NMR spectra.
(4)
Calorimetry Studies. Heats of protonation (∆HCNH) deter-
mined by calorimetric titration of complexes 5CN, 8CN, 12CN,
18CN, 20CN, 21CN, 5*CN, 18*CN, and 23CN with CF3SO3H
in DCE solvent at 25.0 °C according to eq 1 are presented in
Table 2. Plots of temperature vs amount of acid added during
the titrations were linear, indicating that the protonations
occurred rapidly and stoichiometrically.24 Normal pre- and post-
titration traces were evidence that no decomposition of the
neutral or protonated species occurred.
a similar route reported previously;25 however, our synthesis
resulted in a higher yield. Compounds 16Cl, 19Cl, and 21Cl
were prepared using a modification of the reported10 preparations
of CpRuL2Cl (L ) P(OMe)3, 1/2 dppm or 1/2 dppe). Both PPh3
ligands in 8Cl were displaced by P(OPh)3 (16Cl), dppp (19Cl),
or 1,10-phen (21Cl) to give the desired product (eq 5).
acetone
benzene8 CpRuL2Cl + 2PPh3 (5)
CpRu(PPh3)2Cl + L2
(24) Eatough, D. J.; Christensen, J. J.; Izatt, R. M. Experiments in
Thermometric and Titration Calorimetry; Brigham Young Univer-
sity: Provo, UT, 1974.
8Cl
(25) Bruce, M. I.; Windsor, N. J. Aust. J. Chem. 1977, 30, 1601.
(26) Weigand, W.; Nagel, U.; Beck, W. J. Organomet. Chem. 1988, 352,
191.
(22) Bush, R. C.; Angelici, R. J. Inorg. Chem. 1988, 27, 681.
(23) Sowa, J. R., Jr.; Angelici, R. J. J. Am. Chem. Soc. 1991, 113, 2537.

CN Ligand Basicities in Cp′M(L)2CN Complexes
Inorganic Chemistry, Vol. 37, No. 8, 1998 1873
Table 2. Heats of Protonation (∆HCNH) of Cp′M(L)2CN Complexes
Discussion
Effects of Phosphine Ligands on Cyanide Basicity (-∆HC
metal complex
-∆HCNH,a,b (kcal/mol)
-
Cp*Ru(PMe3)2CN, 5*CN
Cp*Ru(dppe)CN, 18*CN
CpRu(PMe3)2CN, 5CN
CpRu(P(p-CH3C6H4)3)2CN, 12CN
CpFe(dppe)CN, 23CN
CpRu(PPh3)2CN, 8CN
CpRu(COD)CN, 20CN
CpRu(dppe)CN, 18CN
25.0(2)
22.6(4)
22.4(2)
21.0(3)
20.9(3)
20.5(2)
20.2(3)
19.5(5)
13.3(3)
NH) in CpRuL2CN. For reactions of the phosphine complexes
CpRuL2CN with CF3SO3H (eq 1), the -∆HCNH values (Table
2) increase in the following order: 18CN (dppe, 19.5 kcal/mol)
< 8CN (2 PPh3, 20.5 kcal/mol) < 12CN (2 P(p-C6H4Me)3, 21.0
kcal/mol) < 5CN (2 PMe3, 22.4 kcal/mol). The phosphine
ligand effect in the Cp*RuL2CN complexes is much the same
(18*CN (dppe, 22.6 kcal/mol) < 5*CN (2PMe3, 25.0 kcal/mol)).
Particularly notable is the fact that the -∆HCNH values are
relatively insensitive to the nature of the phosphine ligands. For
example, the difference in -∆HCNH values for CpRu(PPh3)2-
CN and CpRu(PMe3)2CN is only 1.9 kcal/mol, whereas for
protonation at the metal in CpOs(PR3)2Br, the -∆HHM value
increases by 13.1 kcal/mol when PPh3 is substituted by PMe3.4
The relative insensitivity of -∆HCNH to the donor ability of
the phosphine may be due to the substantial distance between
the phosphine and the site of protonation at the CN- nitrogen
atom. However, it may also be due to the effect of hydrogen
bonding. The overall -∆HCNH may be considered (eq 7) as
CpRu(1,10-phen)CN, 21CN
a For protonation with 0.1 M CF3SO3H in DCE solvent at 25.0 °C.
b Numbers in parentheses are average deviations from the mean of at
least four titrations.
Since the CNH+ ligand is known to form hydrogen bonds
with ethers,1 we sought to determine if H-bonded species were
formed during titrations of the CpRuL2CN complexes with CF3-
SO3H. In a recent study,7 the addition of 1 equiv of CpRu-
(PPh3)2CN to 1 equiv of CpRu(PPh3)2(CNH)+ CF3SO3 gave
-
-
[(CpRu(PPh3)2(CN))2(µ-H)]+ CF3SO3 in which there is a
C-N-H-N-C hydrogen bonding bridge between the two CN-
ligands. This dimer, which was structurally characterized by
X-ray diffraction studies,7 is easily detected by IR spectroscopy
(poly(chlorotrifluoroethylene) mull) as the ν(CN) band is located
at higher wavenumbers (2084 cm-1) than those in either
-∆Hp
-∆HHB
-
M-CN + CF3SO3H
8 M-CN-H+CF3SO3
8
M-CN-H- - -OSO2CF3 (7)
CpRu(PPh3)2CN (2070 cm-1) or CpRu(PPh3)2(CNH)+CF3SO3
-
the sum of the simple enthalpy of protonation (∆Hp) and the
enthalpy of hydrogen bonding ∆HHB. More strongly donating
phosphines (e.g., PMe3) will make -∆Hp more positive but the
proton in M-CN-H+ will be less acidic which will make
-∆HHB less positive. Thus, the increased basicity provided by
PMe3 in the first step (-∆Hp) is at least partially canceled by
weaker hydrogen bonding (-∆HHB) in the second step. That
hydrogen bonding between an acid and base becomes less
favorable as the acidity of the acid decreases is well-documented
in a variety of other systems.28
Despite the opposing effects of -∆Hp and -∆HHB, the
-∆HCNH values for the CpRuL2CN complexes increase as the
donor ability (-∆HHP) of the phosphine increases. Thus, the
phosphine basicity, as defined by the enthalpy of protonation
(-∆HHP) of the free monophosphine (eq 8) increases,29 PPh3
(2016 cm-1). To determine whether such a species exists in
solution during the titration, 2.6 mM solutions (the same
concentrations as in the titrations) of CpRu(PMe2Ph)2CN (6CN)
and CpRu(PMe2Ph)2(CNH)+ CF3SO3 (6CNH+CF3SO3-) in
-
DCE were prepared. After the solution of 6CNH+ was added
to the 6CN solution, the resulting mixture was allowed to sit
for 10 min. An IR spectrum of the solution showed peaks only
for 6CN (2071 cm-1) and 6CNH+ (2013 cm-1), but no peak
was observed at higher frequency which would indicate the
presence of a hydrogen-bridging dimer. This evidence, in
addition to the observation that the calorimetric titration plots
were linear over the entire titration, strongly suggests that
[(CpRuL2(CN))2(µ-H)]+CF3SO3 is not formed during the
-
titrations.
As noted in the Introduction, the recently determined7
structure of CpRu(PPh3)2(CNH)+CF3SO3 shows hydrogen
bonding CpRu(PPh3)2CN-H‚‚‚OSO2CF3 between the proton on
the cyanide and an oxygen atom of the triflate, as indicated by
the short (2.747 Å) N‚‚‚O distance. We sought to determine
using conductivity measurements if there was hydrogen bonding
DCE
-
+
-
PR3 + CF3SO3H 25.0 °C8 HPR3 CF3SO3 ; ∆HHP (8)
(21.2 kcal/mol) < P(p-CH3C6H4)3 (23.2 kcal/mol) < PMe3 (31.6
kcal/mol), in the same order as the basicity of their CpRu(PR3)2-
CN complexes. Thus, the protonation step (-∆Hp) in eq 7
appears to dominate the trend in -∆HCNH values.
-
between the CNH ligand and CF3SO3 in the products of the
titrations (eq 1) in DCE solution in this study. A 0.020M
solution of [Bu4N]I in DCE was used as the standard; its Λ
value was measured as 18 cm2/Ω mol which is in good
agreement with 19 cm2/Ω mol obtained previously for [Bu4N]-
Br in DCE.27 In general, 1:1 electrolytes in DCE give values
of Λ in the range 14-30 cm2/Ω mol. The Λ value determined
Since an increase in phosphine basicity is expected7 to lower
the νCN values for the CpRu(PR3)2CN complexes by reducing
CN- σ donation to the metal and increasing π back-bonding
from the metal to the CN-, a correlation between the phosphine
basicity (-∆HHP) and νCN is expected and observed (Figure
1). This correlation (eq 9) (r ) 0.9611) allows one to estimate
-
for a 0.020 M solution of 15CNH+CF3SO3 was 2.4 cm2/Ω
mol. This low value suggests that 15CNH+CF3SO3- does not
exist as ions in solution but is present primarily in a hydrogen-
bonded form similar to that found in the solid-state structure7
of [CpRu(PPh3)2(CNH)+]CF3SO3- (8CNH+CF3SO3-). On the
basis of these results, it is likely that the other Cp′M(L)2-
ν(CN) (cm-1) ) 2088.7 - 0.731 55(-∆HHP)
(9)
basicities (-∆HHP) for the range of phosphines whose
CpRu(PR3)2CN complexes have reported ν(CN) values (Table
-
CNH+CF3SO3 products formed in the calorimetric titrations
(28) (a) March, J. AdVanced Organic Chemistry; 3rd ed.; John Wiley &
Sons: New York, 1985; p 72. (b) Epshtein, L. M. Russ. Chem. ReV.
1979, 48, 854.
also exist at least in part in hydrogen-bonded forms.
(27) Geary, W. J. Coord. Chem. ReV. 1971, 7, 81 and references therein.
(29) Angelici, R. J. Acc. Chem. Res. 1995, 28, 51.

1874 Inorganic Chemistry, Vol. 37, No. 8, 1998
Nataro et al.
Figure 3. Correlation of ν(CN) for Cp′M(PR3)2CN and for Cp′M-
(PR3)2(CNH)+ for 2CN, 5CN-8CN, 11CN-21CN, 23CN, 5*CN,
18*CN, and 19*CN and their protonated analogues.
Figure 1. (a) Correlation (filled circles) of ν(CN) and phosphine
basicity (-∆HHP) (left axis) for the CpRu(PR3)2CN complexes 2CN,
5CN-8CN, 12CN, 14CN, and 15CN. (b) Correlation (open squares)
of ν(CN) and cyanide basicity (-∆HCNH) (right axis) for the mono-
and diphosphine Cp′Ru(L)2CN complexes 5CN, 8CN, 12CN, 18CN,
5*CN, and 18*CN.
Figure 4. Correlation of the 31P chemical shifts for Cp′M(PR3)2CN
and for Cp′M(PR3)2(CNH)+ for 2CN, 5CN-9CN, 11CN-19CN,
23CN, 5*CN, 18*CN, and 19*CN and their protonated analogues.
Figure 2. Correlation of the Hammett parameter (σ) and ν(CN) for
the CpRu[P(aryl)3]2CN complexes, 8CN and 11CN-15CN.
Cp′ML2CN and protonated Cp′ML2(CNH)+ complexes. This
correlation includes complexes of both Ru and Fe, Cp and Cp*,
and L2 ligands as varied as phosphines, COD and 1,10-phen.
Thus, electronic changes in the metal, as well as the Cp′ and
L2 ligands, affect ν(CN) for the unprotonated and protonated
complexes in a very similar manner.
1). Similarly, there is a correlation (eq 10) (r ) 0.971 88)
ν(CN) (cm-1) ) 2072.2 + 16.82σ
(10)
between the Hammett σ parameters30 of the substituent in the
aryl phosphines and the ν(CN) values of their CpRu[P(aryl)3]2CN
complexes (Figure 2). As their donor abilities increase, their σ
values decrease in the order p-CF3 (0.54) > p-F (0.06) > H
(0.00) > m-CH3 (-0.07) > p-CH3 (-0.17) > p-MeO (-0.27).
Of particular interest is the correlation (eq 11) (r ) 0.996 75)
Finally, there is an excellent (r ) 0.999 39) correlation (eq
13) between the 31P chemical shifts of the phosphine-containing
31P[MCNH+] ) -4.022 + 0.995 34(31P[MCN]) (13)
-∆HCNH ) 653.13 - 0.305 39(ν(CN) (cm-1)) (11)
Cp′ML2CN complexes and their protonated analogues Cp′ML2-
(CNH)+; this correlation (Figure 4) includes complexes of both
Ru and Fe as well as Cp and Cp*.
between -∆HCNH and ν(CN) (cm-1) (Figure 1), which shows
that an increase in the phosphine basicity increases the basicity
(-∆HCNH) of the CN- ligand and decreases its ν(CN) value in
a linear fashion. This correlation, which includes both the Cp
and Cp* derivatives as well as mono- and diphosphine Cp′Ru-
(PR3)2CN complexes, allows one to estimate -∆HCNH values
for other Cp′Ru(PR3)2CN complexes whose ν(CN) values are
known (Table 1). The -∆HCNH values are very sensitive to
changes in ν(CN) since a decrease of 1.0 cm-1 in ν(CN)
increases -∆HCNH by approximately 0.3 kcal/mol.
Effects of 1,10-phen and COD Ligands on Cyanide
Basicity (-∆HCNH) in CpRuL2CN. The basicities (-∆HCNH
)
of the CpRuL2CN complexes containing the 1,10-phenanthroline
and COD ligands are not predicted from their ν(CN) values
using the ν(CN) vs -∆HCNH correlation (eq 11, Figure 1) that
fit the CpRu(PR3)2CN phosphine complexes so well. By using
eq 11, we predict CpRu(1,10-phen)CN (21CN) to have a
-∆HCNH value (20.1 kcal/mol) very similar to that (20.5 kcal/
mol) of CpRu(PPh3)2CN (8CN); however, its actual value (Table
2) is only 13.3 kcal/mol. There is no obvious reason the basicity
of 21CN is so unusually low unless there is substantially
different hydrogen bonding or other interactions between the
cation and anion (eq 7) and solvent than occurs in the phosphine
complexes.
It is also interesting that there is a correlation (eq 12) (r )
νCN[MCNH+] ) -177.54 + 1.0623(νCN[MCN]) (12)
0.948 21) (Figure 3) between ν(CN) values for the unprotonated
The predicted (eq 11) -∆HCNH value for CpRu(COD)CN
(20CN) is only 11.8 kcal/mol, but the actual value is 20.2 kcal/
(30) Lowry, T. H.; Richardson, K. S. Mechanism and Theory in Organic
Chemsitry, 3rd ed.; Harper Collins Publishers: New York, 1987.

CN Ligand Basicities in Cp′M(L)2CN Complexes
Inorganic Chemistry, Vol. 37, No. 8, 1998 1875
mol. In view of the much lower donor ability of COD as
compared with phosphines in CpIrL2 complexes,29 it is surpris-
ing that the CN- ligand in 20CN is as basic as that in
CpRu(PPh3)2CN (20.5 kcal/mol). Clearly, we do not understand
the factors that determine CN- ligand basicities in the non-
phosphine CpRuL2CN complexes.
CpRuL2CN complexes with mono- and diphosphine ligands.
The basicity of the CN- ligand is increased by 2.6-3.1 kcal/
mol when Cp in CpRuL2CN is replaced by Cp*, and the CpM-
(dppe)CN complex is slightly (1.4 kcal/mol) more basic when
M is Fe rather than Ru.
One of the questions raised in the Introduction was whether
it was possible to increase the basicity of the metal center
sufficiently to make it, rather than the CN- ligand, the site of
protonation. It is evident that the addition of strongly donating
ligands to the complexes increases the basicity of the metal more
than it increases the basicity of the CN-. Thus, it seemed
possible to make a Cp′RuL2CN complex that would be
protonated at the Ru. In these studies, the most electron-rich
unit is Cp*Ru(PMe3)2. Its chloride complex Cp*Ru(PMe3)2Cl
is protonated (eq 14) at the Ru with -∆HHM ) 30.2 kcal/mol.
Effect of Cp and Cp* Ligands on Cyanide Basicity
(-∆HCNH) in CpRuL2CN. Evidence of the more strongly
donating nature of Cp* as compared with Cp is seen (Table 2)
in the 2.6 kcal/mol higher basicity (-∆HCNH) of Cp*Ru(PMe3)2-
CN (25.0 kcal/mol) as compared with CpRu(PMe3)2CN (22.4
kcal/mol). Similarly, Cp*Ru(dppe)CN (22.6 kcal/mol) is 3.1
kcal/mol more basic than CpRu(dppe)CN (19.5 kcal/mol). The
2.6-3.1 kcal/mol difference in -∆HCNH between the Cp* and
Cp complexes is significantly smaller than the 9.0 kcal/mol
difference in -∆HHM values between Cp*Ru(PMe3)2Cl (30.2
kcal/mol) and CpRu(PMe3)2Cl (21.2 kcal/mol) and the 5.5 kcal/
mol difference between Cp*Ru(PPh3)2H (35.2 kcal/mol) and
CpRu(PPh3)2H (29.7 kcal/mol), where protonation occurs at the
Ru (eq 2). Protonation at the CN- ligand is expected to be
less sensitive to a change in the Cp′ ligand than protonation at
the metal because of the greater distance of the Cp′ from the
site of protonation. Also, hydrogen bonding (∆HHB, eq 7) in
the Cp′Ru(PR3)2CN protonations is expected, as discussed
above, to reduce the difference between -∆HCNH values for
the Cp and Cp* complexes.
Cp*Ru(PMe3)2Cl + CF3SO3H f
-
Cp*Ru(PMe3)2(Cl)(H)+CF3SO3 ;
∆HHM ) -30.2 kcal/mol (14)
Cp*Ru(PMe ) (CNH)+-O3SCF3;
Cp*Ru(PMe ) CN
f
3 2
3 2
18*CNH+
18*CN
∆HCNH ) -25.0 kcal/mol (15)
The analogous cyanide complex is protonated (eq 15) at the
nitrogen with -∆HCNH ) 25.0 kcal/mol. Given the favorable
protonation at the metal in eq 14, why does protonation occur
at the nitrogen, rather than the metal, in eq 15? Complex
18*CNH+ appears to be the thermodynamic product since it is
stable in DCE solution for at least 2 days at room temperature
without rearranging to a metal-protonated isomer. If the metal-
protonated form were more stable, it seems unlikely that there
would be a significant kinetic barrier to this rearrangement.
Therefore, it is likely that 18*CNH+ is the thermodynamic
product. That protonation of 18*CN occurs at the CN- must
mean that the equilibrium constant for protonation at the Ru is
significantly smaller than that for protonation at the CN-. This
is possible if CN- were a less strongly donating ligand than
Cl-. That this is likely is suggested by Lever’s ligand
electrochemical parameter, EL(L),33 which is significantly more
negative for Cl- (-0.24) than CN- (0.02). Thus, it is reasonable
to conclude that the Ru is less electron-rich in Cp*Ru(PMe3)2-
CN than in Cp*Ru(PMe3)2Cl, which possibly explains why
protonation occurs at the CN- ligand. However, it is also
possible, although not obvious, that the entropy associated with
protonation at the CN- is significantly more positive than
protonation at the Ru, which might also contribute to the
observed protonation at the CN- ligand.
Effect of the Metal (Fe vs Ru) on Cyanide Basicity
(-∆HCNH) in CpML2CN. The somewhat higher basicity
(-∆HCNH) of the Fe complex CpFe(dppe)CN (20.9 kcal/mol)
than the Ru analogue CpRu(dppe)CN (19.5 kcal/mol) is
unexpected considering the well-known31 strong π-donating
ability of RuII in its complexes with π-accepting ligands.
Molecular orbital calculations1,7 indicate that CN- is a weakly
π-accepting ligand, but one might expect the hydrogen isocya-
nide ligand (CNH) in the protonated CpRu(dppe)(CNH)+
product to be stabilized by π donation from the metal as in
alkyl and aryl isocyanide complexes.32 If this occurs, it is not
evident in -∆HCNH values for the Fe and Ru complexes, 23CN
and 18CN. On the other hand, the νCN value for CpFe(dppe)-
CN (2063 cm-1) is lower than that for CpRu(dppe)CN (2075
cm-1) which suggests (eq 11) that the Fe complex (23CN)
should be more basic than the Ru (18CN), as observed. As
noted in the Introduction, the Ru analogue of M(bipy)2(CN)2 is
4-
more basic than that of Fe, while the Fe analogue of M(CN)6
is slightly more basic than that of Ru.3 In both of these cases,
as well as for CpM(dppe)CN, the differences in basicities
between the Fe and Ru derivatives are not large.
Concluding Comments
In this first systematic investigation of the basicity of the
cyanide ligand in a series of related complexes (eq 1), we
observe that the basicities (-∆HCNH) of the CpRu(PR3)2CN
complexes increase as the donor abilities of the phosphines
(-∆HHP) increase. However, the range of basicities is relatively
small, e.g., 1.9 kcal/mol between CpRu(PMe3)2CN and CpRu-
(PPh3)2CN, in part because of hydrogen bonding between the
CNH ligand in the protonated product and the CF3SO3- anion.
Since the ν(CN) values for the complexes decrease with
increasing donor ability of the phosphine, there is a very good
correlation (Figure 1) between -∆HCNH and ν(CN) for the
Acknowledgment. We thank the National Science Founda-
tion (Grant CHE-9414242) and the Office of Basic Energy
Sciences, Chemical Sciences Division, of the U.S. Department
of Energy under Contract W-7405-Eng-82 to Iowa State
University for support of this work. We also thank Johnson-
Matthey for the generous loan of RuCl3‚nH2O.
Supporting Information Available: Experimental procedures for
the syntheses of compounds 3CN, 3Cl, 4Cl, 6Cl, 7Cl, 4CN, 5CN,
6CN, 7CN, 9CN, 12Cl, 13Cl, 14Cl, 15Cl, 12CN, 13CN, 14CN, 15CN,
19CN, 18*CN, 19*CN, and 20*CN (7 pages). Ordering information
is given on any current masthead page.
(31) Krogh-Jespersen, K.; Zhang, X.; Ding, Y.; Westbrook, J. D.; Potenza,
J. A.; Schugar, H. J. J. Am. Chem. Soc. 1992, 114, 4345.
IC971124O
(32) (a) Yamamoto, Y. Coord. Chem. ReV. 1980, 32, 193. (b) Singleton,
E.; Oosthuizen, H. E. AdV. Organomet. Chem. 1983, 22, 130.
(33) Lever, A. B. P. Inorg. Chem. 1990, 29, 1271.