Effect of the structure and concentration of diphosphine ligands on the rate of hydrocarbomethoxylation of cyclohexene catalyzed by palladium acetate/diphosphine/TsOH system

Article abstract of DOI:10.1016/j.molcata.2011.09.005

Cyclohexene hydrocarbomethoxylation catalyzed by the Pd(OAc)₂ - p-toluenesulfonic acid - diphosphine systems with broad variation of diphosphine structure and concentration was studied. It was shown that the hydrocarbon part of the structure and the mutual arrangement of the phosphine groups are the factors that control the activity of palladium-containing catalysts. By comparison of the promoting effects of mono and diphosphine ligands, it is demonstrated that bridging trans-diphosphines show higher efficiency with regard to both the kinetic (TOF) and concentration factors (low P/Pd ratios). In particular, their promoting activity is an order of magnitude higher than that for triphenylphosphine at lower P/Pd ratios (8-65 times). The results were interpreted from the standpoint of chelation effect and the geometric matching of the diphosphine structure to the arrangement of vacant s,d-orbitals of the Pd centre.

Full text of DOI:10.1016/j.molcata.2011.09.005

Journal of Molecular Catalysis A: Chemical 350 (2011) 64–68
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Effect of the structure and concentration of diphosphine ligands on the rate of
hydrocarbomethoxylation of cyclohexene catalyzed by palladium
acetate/diphosphine/TsOH system
I.E. Nifant’eva,d,∗, S.A. Batashevb, S.A. Tolorayac, A.N. Tavtorkind, N.T. Sevostyanovab, A.A. Vorobievb,
V.V. Bagrova, V.A. Averyanovb
a Department of Chemistry, M.V. Lomonosov Moscow State University, 119992 Moscow, Russian Federation
b L.N. Tolstoy Tula State Pedagogical University, Lenin Prospect 125, 300026 Tula, Russian Federation
c Department of Chemistry, Moscow State Pedagogical University, 1, M. Pirogovskaya, 119991 Moscow, Russian Federation
d A.V. Topchiev Institute of Petrochemical Synthesis, 29, Leninsky prospekt, 119991 Moscow, Russian Federation
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 18 July 2011
Received in revised form 5 September 2011
Accepted 9 September 2011
Available online 16 September 2011
Cyclohexene hydrocarbomethoxylation catalyzed by the Pd(OAc)2 – p-toluenesulfonic acid – diphosphine
systems with broad variation of diphosphine structure and concentration was studied. It was shown
that the hydrocarbon part of the structure and the mutual arrangement of the phosphine groups are
the factors that control the activity of palladium-containing catalysts. By comparison of the promoting
effects of mono and diphosphine ligands, it is demonstrated that bridging trans-diphosphines show
higher efficiency with regard to both the kinetic (TOF) and concentration factors (low P/Pd ratios). In
particular, their promoting activity is an order of magnitude higher than that for triphenylphosphine at
lower P/Pd ratios (8–65 times). The results were interpreted from the standpoint of chelation effect and
the geometric matching of the diphosphine structure to the arrangement of vacant s,d-orbitals of the Pd
centre.
Keywords:
Hydrocarboalkoxylation
Catalytic carbonylation
Pd-catalysis
Phosphine-based catalysts
Esters
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
Among the catalytic systems used in the hydrocarboalkoxylaton
of alkenes, of particular interest are palladium derivatives pro-
Carbonylation of olefins catalyzed by transition metal com-
plexes is a promising process for the manufacture of various
organic compounds, including pharmacological and agrochem-
ical agents [1]. An important carbonylation process is alkene
hydrocarboalkoxylation catalyzed by palladium complexes, which
represents a one-step route from accessible alkenes to diverse
esters [2]. This process has found use in industry. For example,
in 2008, Lucite commercialized hydrocarbomethoxylation of ethy-
lene [3] (Scheme 1).
moted by strong protic acids and free phosphines [4–7]. In the series
of phosphines, diphosphines are used more and more often. While
functioning as chelating agents with respect to the catalyst palla-
dium site, they form more stable complexes than usual phosphines,
which markedly affects the kinetics and regioselectivity of the
hydroalkoxycarbonylation of alkenes [3,8–10]. Despite the obvi-
ous topicality of studies dealing with the effect of the diphosphine
ligand structure on the efficiency of palladium complexes in the
hydrocarboalkoxylation of alkenes, no systematic comparative data
on this topic can be found in the literature. Therefore, we undertook
a special study to find out how the structure of diverse diphosphines
affects the kinetic parameters of the model hydrocarbomethoxyla-
tion reaction. As this reaction, we chose hydrocarbomethoxylation
of cyclohexene, as all of its reaction sites are chemically equivalent,
and methyl cyclohexanecarboxylate is formed as the only hydro-
carbomethoxylation product. In addition, for cyclohexene we did
not expect a noticeable copolymerization with CO, and, hence, the
course of hydrocarbomethoxylation was expected to characterize
the efficiency of the chosen catalytic system “as such”.
R1
R3
R2
R4
R1
R3
R2
R4
[Pd]/phosphine
H
H
COOR
ROH
CO
Scheme 1.
As diphosphines that promote the hydrocarbomethoxylation of
cyclohexene, we studied bis-diphenylphosphinoalkanes 1–3 con-
taining di-, tri-, and tetramethylene bridges, respectively, and an
∗
Corresponding author at: Department of Chemistry, M.V. Lomonosov Moscow
State University, 119992 Moscow, Russian Federation.
E-mail address: inif@org.chem.msu.ru (I.E. Nifant’ev).
1381-1169/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2011.09.005

I.E. Nifant’ev et al. / Journal of Molecular Catalysis A: Chemical 350 (2011) 64–68
65
1H NMR (CDCl3): 7.54 (m, 10H), 7.38 (m, 10H), 6.14 (s, 2H),
3.06 (s, 2H), 2.41 (s, 4H), 2.28 (m,2H), 2.20 (m,2H), 1.43 (d, 1H),
1.17(d, 1H).
extensive series of bis-diphenylphosphines with a four-membered
bridge 4–12 able to form bite-angle palladium complexes appro-
priate for effective catalysis of hydrocarbomethoxylation [11].
Systematic research into the effect of the nature of the substituent
at phosphorus is the subject of a separate project and will be
reported elsewhere.
13C NMR (CDCl3): 139.6 (d), 138.4 (d), 135.5 (s), 133.0 (d),
132.6 (d), 128.5 (d), 128.38 (s), 128.34 (d), 48.6 (s), 47.3 (d), 39.54
(d), 39.4(d), 29.1 (d).
The structure of the synthesized diphosphines 1–12 is shown
below.
b) Adduct
of
trans-2,3-Bis-(diphenylphosphinomethyl)-
norbornane and Pd(OAc)2, 5·Pd(OAc)2
trans-2,3-Bis(diphenylphosphinomethyl)norbornane
PPh2
Ph2P
PPh2
Ph2P
(100 mg, 0.2 mol) prepared by
a reported procedure [6]
Ph2P
PPh2
was added to a solution of palladium acetate (50 mg, 0.2 mol)
in 2 mL of CH2Cl2. The reaction mixture was stirred for 2 h.
Then the solvent was removed in vacuo and the residue was
recrystallized from ether. Yield 100 mg (70%).
3
2
1
PPh2
PPh2
PPh2
31P NMR (CDCl3): 25.7, 20.6.
1H NMR (CDCl3): 7.96 (m, 4H), 7.36 (m, 16H), 2.58 (m, 1H),
2.39 (m, 2H), 2.23 (m, 1H), 2.05 (m, 1H), 1.93 (br.s., 1H), 1.86
(br.s., 1H), 1.38 (m, 10H), 1.22 (d, 2H), 1.08 (m, 1H).
PPh2
PPh2
PPh2
6
4
5
2.2. Procedure of catalytic experiments
PPh2
PPh2
The hydrocarbomethoxylation of cyclohexene was studied in
the batch reactor described previously [19]. The experiments
were carried out in toluene at constant temperature (105 ◦C)
and CO pressure (2.1 × 106 Pa). The invariability of the temper-
ature was ensured by high-temperature organic heat medium
circulating through the reactor shell. A mixture of cyclohexene
(0.51 mL), methanol (0.91 mL) and o-xylene, the internal standard
for chromatography (0.30 mL) in toluene (48.28 mL) was placed in
autoclave. A catalytic mixture of (CH3COO)2Pd (0.05 mmol), TsOH
(0.60 mmol) and one of the ligands (the concentration of one of
the ligands being varied within each series from 0 to 7.50 mmol)
was placed in special glace “basket” fastened to the autoclave lid.
The autoclave was purged three times with 0.6 × 106 Pa of CO and
then pressurized to 1.6 × 106 Pa at room temperature. The system
was heated to 105 ◦C and then pressurized to 2.1 × 106 Pa. Then
the glace “basket” with catalytic system was immersed in reac-
tion solution. During kinetic experiments, the reaction mixture
was sampled at particular time intervals, and the samples were
analyzed by gas liquid chromatography on a Tsvet 162 chromato-
graph with a flame ionization detector. Analysis was carried out
using 3000 mm × 3 mm glass columns. The separation was per-
formed with Chromosorb W (80/100 mesh) on the stationary phase
OV-275 – 3% at a carrier gas (argon) flow rate of 30 mL/min and
evaporator temperature of 225 ◦C in the temperature programmed
mode from 75 to 205 ◦C at a heating rate of 8 ◦C/min. Chromato-
graphic calculations were performed using MultiChrom software.
The chromatographic peaks of the reactants and product were iden-
tified based on the retention times, and the component contents
were determined using chromatography of artificial mixtures with
known contents of the components.
PPh2
PPh2
PPh2
7
8
PPh2
PPh2
PPh2
10
9
PtBu2
PtBu2
PPh2
PPh2
12
11
2. Experimental
2.1. Synthesis of diphosphines
The diphosphines used in this work were prepared by the
reported procedures: 1, 2, 3 [12], 4 [13], 5 and 6 [14], 7 [15], 8
[16], 9 [17], 11–12 [18].
The previously unknown compounds 10 and 5·Pd(OAc)2 were
obtained as follow:
3. Results and discussion
a) cis-2,3-Bis-(diphenylphosphinomethyl)-norbornene, 10
Finely divided Li (1 g, 143 mmol) was added to a solution of
triphenylphosphine (12.45 g, 47.5 mmol) in 70 mL of anhydrous
THF. The reaction mixture was stirred at room tempera-
ture overnight. Then a solution of tosylate (12.2 g, 26.4 mmol)
obtained as reported [12] in 30 mL of anhydrous THF was added.
The reaction mixture was stirred at room temperature for 20 min
and then treated with an aqueous solution of NH4Cl, 10% Н2SO4,
and saturated brine. The organic layer was separated and dried
with Na2SO4, and the solvent was removed in vacuo. The residue
was recrystallized from ethanol. Yield 5 g (42%).
The influence of the diphosphine ligand structure and con-
centration on the rate of cyclohexene hydrocarbomethoxylation
was studied in 12 series of experiments using (CH3COO)2Pd,
p-TsOH, and diphosphines 1–12 as catalyst components. An addi-
tional series of experiments with the triphenylphosphine ligand
was carried out to compare the behaviors of diphosphine and
monophosphine ligands.
Typical results of kinetic experiments are presented in Fig. 1 as
dependences of the methyl cyclohexanecarboxylate concentration
on the reaction time for catalytic systems. It can be seen that each
kinetic curve has an autocatalytic period, which is indicative of the
31P NMR (CDCl3): −17.14.

66
I.E. Nifant’ev et al. / Journal of Molecular Catalysis A: Chemical 350 (2011) 64–68
Fig. 4. Effect of the concentration of ligand
5 on the cyclohexene hydrocar-
bomethoxylation rate. (a) Reaction rate vs. concentration of free ligand 5 in the
presence of (CН3COO)2Pd (1.0 × 10−3 mol/L). (b) Reaction rate vs. the total con-
centration of ligand 5 in the presence of the complex [(CН3COO)2Pd·ligand 5]
(1.0 × 10−3 mol/L).
Fig. 1. Kinetic curves for the accumulation of methyl cyclohexanecarboxylate with
time at the diphosphine concentration of 3.0 × 10−3 mol/L. (a) ligand 4, (b) ligand 5,
(c) ligand 6, and (d) ligand 8.
be seen that for ligands 3–9 all of the presented dependences pass
through extrema. Meanwhile, the catalytic systems with ligands
1, 2, and 10–12 were virtually inert over the whole concentration
range of the latter. It is of interest that the position of the max-
imum rate for triphenylphosphine corresponds to much higher
P/Pd ratio (∼65) than that for diphosphine ligands (1–8). More-
over, diphosphines 3 and 8 demonstrate the P/Pd maximum close
to 1, and compounds 4–7 and 9 exhibit the maximum activity
at much higher P/Pd ratio (∼8). The values of the maximum rate
(r0 × 103, mol/L) were higher for ligands 3–9 than for PPh3 (0.150).
It is significant that trans-diphosphine 5 proved to be an order of
magnitude more effective promoter than the cis-isomer 6. In this
respect, it is noteworthy that all other diphosphine ligands with
cis- or “cis-like”-arranged phosphine groups (10–12) had almost
no accelerating effect on the hydrocarbomethoxylation of cyclo-
hexene. On the other hand, attention is attracted by the fact that
a double bond influences the efficiency of diphosphines. Indeed,
unsaturated diphosphine 7 was a twice less efficient promoter than
the hydrogenated analog 5.
Fig. 2. Effect of the triphenylphosphine concentration on the cyclohexene hydro-
carbomethoxylation rate.
In order to study the lability of palladium phosphine catalysts,
we compared hydrocarbomethoxylation of cyclohexene catalyzed
by the Pd(OAc)2 – 5 – TsOH system and by the pre-synthesized
adduct 5·Pd(OAc)2. The results of these experiments are presented
in Fig. 4. It can be seen that the Pd(OAc)2 – 5 catalytic system
has a higher catalytic activity than the complex 5·Pd(OAc)2 over
the whole range of variation of the ligand 5 concentration. This
fact prompts the idea that the active intermediates of the reaction
using Pd(OAc)2 and 5 do not involve the adduct 5·Pd(OAc)2, but the
reaction follows an alternative route.
formation of active complexes responsible for the catalytic action.
Ligand 5 functions almost without an induction period.
The initial reaction rates were determined by differentiating
the initial sections of the kinetic curves after completion of the
autocatalytic period. The results of determinations are presented
in Figs. 2 and 3 as the dependence of the initial hydrocarboalkoxy-
lation rate on the concentration of the phosphine ligands. It can
The obtained data can be summarized as the turnover frequency
(TOF) for Pd(OAc)2 – p-TsOH–diphosphine (monophosphine) sys-
tems at the maximum rate (Table 1).
The curve of the catalytic activity of the Pd(OAc)2–PPh3 sys-
tem presented in Fig. 2 differs markedly from the plots for the
Pd(OAc)2–diphosphine ligand systems, which pass through a max-
imum (Fig. 3). Two sections can be distinguished in this curve.
One characterizes a steep increase in the activity from 0 to
1.1 × 10−4 mol/(L min) in the range of PPh3 concentrations from
0 to 8.0 × 10−4 mol/L and the other one shows a slightly sloping
dependence of the activity on the PPh3 concentration that passes
through a maximum in the concentration range of 8.0 × 10−4 to
0.15 mol/L. It is easy to see that the hydrocarbomethoxylation rate
of 1.1 × 10−4 mol/(L min) attained in the first section differs little
from the rate at the extremum (1.5 × 10−4 mol/(L min)). This indi-
cates, in our opinion, that the formation of the palladium phosphine
complexes responsible for the catalysis is a reversible reaction. At
Fig. 3. Effect of the concentration of diphosphine ligands on the cyclohexene
hydrocarbomethoxylation rate. (a) Ph2P–(CH2)4–PPh2 (ligand 3), (b) trans-
Ph2P–CH2–C4H6–CH2–PPh2 (ligand 4), (c) trans-C7Н10(CН2PPh2)2 (ligand 5),
(d) cis-C7Н10(CН2PPh2)2 (ligand 6), (e) trans-C7Н8(CН2PPh2)2 (ligand 7), (f)
C8Н12(CН2PPh2)2 (ligand 8), and (g) (C6H4CHCHCH2PPh2)2 (ligand 9).

I.E. Nifant’ev et al. / Journal of Molecular Catalysis A: Chemical 350 (2011) 64–68
67
Table 1
Turnover frequency values for Pd(OAc)2–p-TsOH–diphosphine (monophosphine) systems at the maximum rate of hydrocarbomethoxylation.
Ligand No.
TOF (h−1
1, 2, 10–12
PPh3
9.0
3
4
5
6
7
8
9
)
0
21.0
42.0
150.0
15.0
54.0
63.0
57.0
point a, this reaction is not far from equilibrium, which results in
sharp retardation of the formation of the complexes after this point
has been passed. In the second section, as the triphenylphosphine
concentration increases, the fraction of ligand exchange reactions
with its participation, resulting in the formation of less active com-
plexes, starts to grow. Finally, gradual retardation of the reversible
formation of active complexes and acceleration of ligand exchange
reactions with their participation leads to a kink in the dependence
of the hydrocarbomethoxylation rate on the PPh3 concentration
and to the observed maximum.
The pattern of the extreme curves of the reaction rate vs. the
diphosphine concentration attests rather to purely kinetic control
of the concentration of active complexes. Apparently, the reaction
of their formation proceeds under conditions where it is nearly irre-
versible. This accounts for the high slope of the ascending part of
the curves shown in Fig. 3 in the region of low concentrations of
diphosphines until the maximum is reached. In addition, the fairly
steep decline after the maximum in these curves attests to a high
rate of ligand exchange involving diphosphines, resulting in the
formation of less active complexes.
High promoting activity of diphosphine ligands with trans-
phosphine groups at bridging structures prompts the idea of a
favorable spatial arrangement of these groups with respect to
vacant s,d-orbitals of palladium. This is in line with the reported
fact that linear diphosphine ligands are bound in the 1,3-position in
the square plane configuration of palladium complexes [20]. Cor-
respondingly, the bridging bidentate ligands with cis-phosphine
groups have either low promoting activity (ligand 6) or no such
activity at all (ligand 10). Similarly, the lack of activity of ligands
11, 12 attests to an unfavorable positions of their phosphine groups
relative to the Pd vacant s,d-orbitals. The relatively low activity of
ligand 7 containing a double bond compared with the hydrogenated
analog 5 implies that 7 may form oligomers, which are coordinated
to the palladium site, thus making it spatially less accessible for
reagents.
The virtually inert behavior of ligands 1 and 2 in the car-
bomethoxylation of cyclohexene and the noticeable promoting
effect of ligand 3 reflect, in our opinion, the role of the chelation
effect where the formation of sterically non-strained metal rings
upon the coordination of a linear bidentate ligand to the com-
plexing metal is energetically favorable [21,22]. In this case, the
competing formation of inactive polynuclear complexes in which
the metal atoms are linked by diphosphine bridges is more probable
for diphosphines 1 and 2.
This stability was found to be caused by the concerted action of
the entropy and energy factors upon the chelate formation [21]. As
a consequence, the formation of chelate type active intermediates
requires much lower concentrations of phosphine groups than the
formation of active complexes with monophosphine ligands [22].
Thus, the use of palladium chelates based on trans-diphosphine
ligands as catalysts in the carbonylation opens up prospects for the
development of highly efficient processes economical as regards
the cost of the catalyst for the manufacture of esters and other
valuable oxygenate products. Owing to the stability of chelate com-
plexes, the possibility of their multiple use in carbonylation is
predicted.
4. Conclusions
It was shown that variation of the structure of the bridge and
mutual arrangement of the phosphine groups in diphosphine lig-
ands is a factor of control over the activity of palladium phosphine
catalysts in the hydrocarbomethoxylation of cyclohexene. trans-
Diphosphines are much superior over cis-diphosphines as regards
the promoting activity. Ligand 5 is most efficient among the studied
diphosphines.
A comparison of the kinetic behavior of mono- and di-phosphine
ligands in the hydrocarbomethoxylation of cyclohexene showed
that bridging trans-diphosphines are more efficient as regards both
the kinetic (high reaction rates) and concentration factors (low P/Pd
ratios). In particular, the promoting activity of trans-diphosphines
is an order of magnitude higher than this value for triphenylphos-
phine at P/Pd ratios 8–65 times lower than required when the latter
ligand is used.
The obtained results were interpreted from the standpoint of
chelation effect and the geometric matching of the ligand structure
to the arrangement of vacant s,d-orbitals of the Pd centre.
It is concluded that palladium chelates with trans-diphosphines
would be useful as catalysts of carbonylation.
Acknowledgments
This work was supported by the Federal Target Program
“Research and Pedagogical Cadres of Innovative Russia” for
2009–2013 (state contract No. 02.740.11.0266) and the Russian
Foundation for Basic Research (Project No. 09-08-00890).
(CH2)n
PPh2
(CH2)n
PPh2
References
[1] K. Nozaki, I. Ojima, in: I. Ojima (Ed.), Catalytic Asymmetric Synthesis, 2nd ed.,
Wiley, New York, 2000, p. 429.
[2] E. Drent, P.H.M. Buzelaar, Chem. Rev. 96 (1996) 663–681.
[3] T. Fanjul, G. Eeastham, N. Fey, A. Hamilton, A.G. Orpen, P.G. Pringle, M. Waugh,
Organometallics 29 (2010) 2292–2305.
[4] T.E. Kron, M.I. Terekhova, Yu.G. Noskov, E.S. Petrov, Kinet. Katal. 42 (2001)
204–211.
[5] V.A. Averyanov, S.A. Batashev, N.T. Sevostianova, N.M. Nosova, Kinet. Katal. 47
(2006) 381–390.
(CH2)n PPh2 Pd PPh2(CH2)nPPh2 Pd PPh2 (CH )
~
2 n
~
PPh2
PPh2
(CH2)n
(CH2)n
[6] G. Kiss, Chem. Rev. 101 (2001) 3435–3456.
[7] V.A. Averyanov, N.M. Nosova, S.A. Batashev, Neftekhimiya 46 (2006) 116–119.
[8] I. del Rio, N. Ruiz, C. Claver, L.A. van der Veen, P.W.N.M. Leeuwen, J. Mol. Catal.
A: Chem. 161 (2000) 39–48.
[9] W. Clegg, G.R. Eastham, M.R. Elsegood, R.P. Tooze, X.L. Wang, K. Whiston, Chem.
Commun. 187 (1999) 1877–1878.
The considerable difference of the P/Pd ratios at the maximum
observed for diphosphines and triphenylphosphine deserves spe-
cial attention. In our opinion, this is due to high stability of chelates
compared with the complexes containing monodentate ligands.

68
I.E. Nifant’ev et al. / Journal of Molecular Catalysis A: Chemical 350 (2011) 64–68
[10] E. Guiu, M. Caporali, D. Munoz, C. Muller, M. Lutz, A.L. Spek, C. Claver, P.W.N.M.
van Leeuwen, Organometallics 25 (2006) 3102–3104.
[11] P.W.N.M. van Leeuwen, P.C.J. Kamer, J.N.H. Reek, P. Dierkes, Chem. Rev. 100
(2000) 2741–2769.
[12] N. Bricklebank, S.M. Godfrey, C.A. McAuliffe, J. Chem. Soc., Dalton Trans. 237
(1998) 2379–2382.
[13] T. Hayashi, M. Tanaka, Y. Ikeda, I. Ogata, Bull. Chem. Soc. Jpn. 52 (1979)
2605–2608.
[14] T. Hayashi, Y. Kawabata, T. Isoyama, I. Ogata, Bull. Chem. Soc. Jpn. 54 (1981)
3438–3446.
[17] Chr. Doebler, H.-J. Kreuzfeld, J. fuer Prakt. Chem. 325 (1983) 1021.
[18] T. Hayashi, Y. Kawabata, T. Isoyama, I. Ogata, Bull. Chem. Soc. Jpn. 54 (1981)
2408.
[19] A.R. El’man, V.A. Matveev, E.V. Slivinskii, S.M. Loktev, Khim. -Farm. Zh. 3 (1990)
47–49.
[20] O.V. Gusev, A.M. Kalsin, M.G. Peterleither, P.P. Petrovskii, K.A. Lissenko,
Organometallics 21 (2002) 3637.
[21] M. Yu, N.A. Kiselev, Dobrynina Khimiya Koordinatsionnykh Soedinenii (Chem-
istry of Coordination Compounds), Tsentr Akademiya, Moscow, 2007, p. 147.
[22] V.V. Skopenko, A.Y. Tsivadze, L.I. Savranckii, A.D. Garnovskii, Koordinatsion-
naya Khimiya (Coordination Chemistry), IKTs Akademkniga, Moscow, 2007,
p. 398.
[15] Chr. Doebler, H.-J. Kreuzfeld, J. Fuer Prakt. Chem. 323 (1981) 667.
[16] T.P. Dang, J.-C. et Kagan, H.B. Poulin, J. Organomet. Chem. 91 (1975) 105–115.