On the nature of the chemical bond in heterobimetallic palladium(II) complexes with divalent 3d metals

Article abstract of DOI:10.1134/S0036023609060114

The complex Pd(μ-OOCMe)₄Cu(OH₂) ? 2Pd ₃(μ-OOCMe)₆ was synthesized and characterized by X-ray crystallography. In the heterometallic moiety of this complex, the Pd ^II and Cu^II atoms are at a

Full text of DOI:10.1134/S0036023609060114

ISSN 0036-0236, Russian Journal of Inorganic Chemistry, 2009, Vol. 54, No. 6, pp. 885–892. © Pleiades Publishing, Inc., 2009.
Original Russian Text © A.A. Markov, A.P. Klyagina, S.P. Dolin, N.S. Akhmadullina, N.Yu. Kozitsyna, N.V. Cherkashina, S.E. Nefedov, M.N. Vargaftik, I.I. Moiseev,
2
009, published in Zhurnal Neorganicheskoi Khimii, 2009, Vol. 54, No. 6, pp. 950–957.
COORDINATION
COMPOUNDS
On the Nature of the Chemical Bond in Heterobimetallic
Palladium(II) Complexes with Divalent 3d Metals
A. A. Markov, A. P. Klyagina, S. P. Dolin, N. S. Akhmadullina, N. Yu. Kozitsyna,
N. V. Cherkashina, S. E. Nefedov, M. N. Vargaftik, and I. I. Moiseev
Kurnakov Institute of General and Inorganic Chemistry, Russian Academy of Sciences,
Leninskii pr. 31, Moscow, 119991 Russia
E-mail: wahr36@gmail.com
Received December 17, 2008
Abstract—The complex Pd(µ-OOCMe) Cu(OH ) · 2Pd (µ-OOCMe) was synthesized and characterized by
4
2
3
6
II
II
X-ray crystallography. In the heterometallic moiety of this complex, the Pd and Cu atoms are at an extraor-
dinary short distance (2.521(3) Å). DFT quantum-chemical calculations of the geometric and electronic struc-
ture of a series of heterobinuclear paddlewheel complexes Pd M (µ-OOCMe) L (M = Zn , Ni , Cu , Co , Fe ;
L = OH and NCH) and their formate analogues Pd M (µ-OOCH) L (M = Zn , Ni , Fe ) showed that the
II II
II
II
II
II
II
4
II II
II
II
II
2
4
extraordinary short Pd···M distance in all these complexes is caused only by the tightening effect of carboxylate
bridges rather than by the metal–metal bond. The direct Pd–M interaction becomes possible only after removal
III
III 2+
of electrons from the antibonding orbitals and formation of oxidized complexes of the [Pd (µ-OOCMe) Ni ]
4
type.
DOI: 10.1134/S0036023609060114
Heterometallic transition metal complexes are of tance: the metal–metal bond (plus carboxylate bridges)
both theoretical and practical interest, in particular, as or only carboxylate bridges. In the case of homometal-
n+
initial compounds for synthesis of catalysts [1, 2] and
nanomaterials [3, 4]. Recently, we synthesized a series
lic complexes with the M2 moiety, the nature of the
metal–metal bond has been well studied by quantum-
chemical methods using the metal-based molecular
orbital scheme proposed by F.A. Cotton [9]. This
approach made it possible to adequately describe the
metal–metal bond for a large number of structurally
characterized homobinuclear complexes with a Chi-
nese lantern structure [10].
of heterobimetallic complexes Pd(µ-OOCR) ML with a
4
Chinese lantern structure
R
R
C
O
O
O
O
C
O
O
O
However, the Cotton method has not been applied to
heterobinuclear complexes. The existence of the metal–
metal bond in such compounds has been repeatedly
postulated (for example, for a Pd –Zn complex [11]).
The correspondence between the metal–metal distance
and bond order observed (according to X-ray crystallo-
graphic data) in homonuclear complexes is often used
as an argument for the existence of the metal–metal
Pd
M L
C
O
R
II
II
C
R
II
II
II
II
II
t
M = Co , Ni , Mn , Zn , Cu ; L = OH , NCMe; R = Me, Bu
2
in which the distance between the palladium(II) atom bond, although one-to-one correlations between these
and the divalent 3d metal atom was extraordinary short parameters are not fully justified [9, 10]. For heteronu-
(
Pd–M, 2.46–2.58 Å) and close to the sum of the cova- clear compounds, the validity of such a correlation is
lent radii and considerably shorter than the sum of the problematic.
van der Waals radii of the corresponding atoms [5–7].
The replacement of the axial L ligand by bidentate
N-donor ligands (Ó-phenanthroline, bipyridine) is
accompanied by the cleavage of one of the four carbox-
ylate bridges, but the short Pd–M distance remains vir-
tually unaltered [8].
In this work, we synthesized and structurally char-
acterized a new complex Pd(µ-OOCMe) Cu(OH ) ·
2
4
2
Pd (µ-OOCMe) (1). The crystals of this complex con-
3 6
tain both hetero- and homometallic moieties. To decide
what factors are responsible for the short Pd···M dis-
tance and whether the metal–metal bond can form in
In this context, the question arises as to what factor the heterometallic moiety of complex 1, as well as in
is responsible for the extraordinary short Pd–M dis- the previously synthesized heterobinuclear complexes
8
85

8
86
MARKOV et al.
4
Table 1. Atomic coordinates (×10 ) and equivalent isotro- [5–7], we performed density functional theory (DFT)
2
3
pic factors (Å × 10 ) for complex 1
quantum-chemical calculations of the geometric and
II
electronic structure of the paddlewheel complexes Pd
Atom
Pd(1)
x
y
z
Ueq
II
M (µ-OOCMe) L (M= Zn (2), Ni (3), Cu (4), Co (5), Fe
4
(
6)) with ligands L = H O and HCN (the latter mimics
MeCN) and their formate analogues Pd M (µ-
OOCH) L (M = Zn (2‡), Ni (3‡), Fe (6‡)).
5000
5186(1)
2500
23(1)
24(1)
25(1)
28(1)
20(1)
25(3)
25(3)
24(3)
26(3)
29(4)
29(3)
29(3)
34(3)
34(3)
34(3)
42(3)
30(3)
33(3)
33(3)
40(3)
30(3)
38(3)
24(4)
29(4)
24(4)
27(4)
32(3)
32(3)
21(4)
24(4)
19(4)
58(6)
30(4)
34(4)
14(3)
30(4)
24(4)
47(5)
2
II II
Pd(2)
Pd(3)
Pd(4)
Cu(1)
O(1)
1713(1)
2359(1)
894(1)
5589(1) –3357(1)
5393(1) –1408(1)
5410(1) –2096(1)
4
EXPERIMENTAL
5000
3468(2)
5156(7)
3634(7)
5132(7)
3613(7)
2017(9)
2500
Palladium(II) acetate Pd (OOCMe) was obtained
3
6
4190(4)
4167(4)
5426(4)
5374(4)
5000
1587(6)
1642(6)
1595(6)
1553(6)
2500
by oxidation of palladium black (prepared by reducing
PdCl (chemically pure, Reakhim. Russia) with sodium
2
O(2)
borohydride) by concentrated nitric acid in glacial ace-
tic acid as described in [12], purified from impurities of
nitrito complexes by repeated boiling in glacial acetic
acid in the presence of fresh portions of palladium
O(3)
O(4)
O(5)
black until the termination of NO evolution, and
2
recrystallized from acetic acid. Copper(II) acetate
O(6)
2371(4)
2770(4)
2310(5)
2858(5)
1042(5)
669(5)
4688(6) –3294(6)
4429(7) –1847(6)
6587(7) –3126(6)
6319(7) –1757(6)
6519(7) –3727(6)
6617(7) –2622(7)
4580(7) –3849(6)
4772(7) –3184(6)
Cu(OOCMe) · H O (chemically pure, Acros, Belgium)
2
2
O(7)
was used as purchased.
O(8)
Elemental analysis was carried out on a Carlo Erba
CHN analyzer. IR spectra were recorded as KBr pellets
on a Specord M-80 spectrophotometer.
O(9)
O(10)
O(11)
O(12)
O(13)
O(14)
O(15)
O(16)
O(17)
C(1)
The X-ray diffraction experiment was carried out on
a Bruker SMARTAPEX II diffractometer (λMo, graph-
ite monochromator, ω scan, 2θmax = 50°). For 1:
C H CuO Pd , space group ë2/c, a = 22.594(6) Å, b =
1132(5)
410(5)
32
50
33
7
1
4.667(4) Å, c = 16.329(4) Å, α = 90°, β = 108.113(6)°,
3
2047(4)
1046(5)
2106(5)
1102(5)
3928(6)
3294(6)
5510(7)
5761(6)
2741(7)
3171(6)
2764(6)
3208(6)
713(7)
4423(7)
–801(6)
γ = 90° (293 K), V = 5143(2) Å , Z = 4; the number of
measured reflections is 4528, the number of unique
reflections with F > 2σ(I) is 1781; ρ = 2.287 g cm ,
µ = 2.891 cm , R1 = 0.0545, wR2 = 0.0951. The struc-
4196(7) –1511(7)
2
–3
calc
6318(7)
6061(7)
–764(6)
–970(6)
–1
ture was solved using the SHELXTL PLUS program
package (PC version) [13] and refined with the
SHELXTL97 program [14]. The atomic coordinates
and equivalent isotropic factors and selected bond
lengths and bond angles for 1 are summarized in Tables 1
and 2.
4387(12) 1332(9)
4399(10) 667(8)
4362(11) 1288(9)
4382(11) 558(8)
C(2)
C(3)
C(4)
Synthesis of Pd(m-OOCMe) Cu(OH ) · 2Pd (m-
4
2
3
C(5)
4242(10) –2632(12)
3568(10) –2807(10)
6739(10) –2471(11)
7500(9) –2478(9)
6871(10) –3330(10)
7668(11) –3741(12)
4463(11) –3758(10)
3835(10) –4420(9)
OOCMe) (1). A solution of Cu(OOCMe) · H O
6
2
2
(
33 mg, 0.15 mmol) in 2 mL of glacial acetic acid was
C(6)
added to a solution of Pd (µ-OOCMe) (100 mg,
3
6
C(7)
0
.45 mmol as calculated per palladium atom) in 3 mL
of acetic acid. The solution was kept at room tempera-
ture in an open flask for 4 days. The resulting yellow-
green crystals were separated from the supernatant by
decantation, washed with cold benzene and hexane, and
dried in an argon flow. The yield of 1 was 64 mg (56%
in terms of Pd).
C(8)
C(9)
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
315(7)
599(8)
167(7)
For Pd CuC H O anal. calcd. (%): C, 21.70; H,
7
32 50 33
2
.85.
1558(7)
1543(7)
1593(8)
1580(7)
3988(9)
3159(9)
–941(9)
–426(8)
Found (%): C, 21.87; H, 3.09.
–1
IR (KBr), ν, cm : 3464 m, 2940 m (ν(CH)), 1712 s,
616 vs br (ν (CO)), 1428 s br (ν (CO)), 1336 w, 1024
6474(10) –644(9)
7180(10) 57(11)
1
as
s
w, 932 m, 888 m, 676 s, 624 s.
RUSSIAN JOURNAL OF INORGANIC CHEMISTRY Vol. 54 No. 6 2009

ON THE NATURE OF THE CHEMICAL BOND
887
Table 2. Interatomic distances and bond angles for complex 1
Bond d, Å
Pd(1)–O(1)#1
Bond
d, Å
1.970(9)
1.999(9)
2.521(3)
1.968(9)
Pd(1)–O(1)
1.970(9)
1.999(9)
Pd(1)–O(3)
Pd(1)–Cu(1)
Pd(2)–O(6)
Pd(2)–O(10)
Pd(2)–Pd(4)
Pd(3)–O(9)
Pd(3)–O(14)
Pd(4)–O(11)
Pd(4)–O(15)
Cu(1)–O(2)
Cu(1)–O(4)
Cu(1)–O(5)
Pd(1)–O(3)#1
Pd(2)–O(8)
Pd(2)–O(12)
Pd(2)–Pd(3)
Pd(3)–O(7)
Pd(3)–O(16)
Pd(3)–Pd(4)
Pd(4)–O(17)
Pd(4)–O(13)
Cu(1)–O(2)#1
Cu(1)–O(4)#1
1.947(10)
1.975(10)
3.0751(17)
1.947(10)
1.975(10)
3.1472(18)
1.994(10)
2.006(9)
1.989(10)
3.1790(18)
1.961(10)
1.984(10)
1.964(11)
2.000(10)
1.981(9)
1.982(9)
1.992(10)
2.127(13)
1.992(10)
Angle
ω, deg
Angle
ω, deg
O(1)#1Pd(1)O(1)
O(1)Pd(1)O(3)
O(1)Pd(1)O(3)#1
O(1)#1Pd(1)Cu(1)
O(3)Pd(1)Cu(1)
O(8)Pd(2)O(6)
177.4(6)
89.3(4)
90.6(4)
88.7(3)
87.7(3)
91.6(4)
85.1(4)
165.6(4)
78.0(3)
113.1(3)
116.4(3)
75.8(3)
60.40(4)
168.1(4)
87.0(4)
93.1(4)
76.4(3)
117.1(3)
121.8(3)
71.7(3)
85.8(5)
91.7(4)
161.4(4)
105.2(3)
79.4(3)
76.4(3)
109.5(3)
58.16(4)
88.7(4)
89.8(4)
167.7(6)
97.1(3)
96.1(3)
82.9(3)
83.9(3)
O(1)#1Pd(1)O(3)
O(1)#1Pd(1)O(3)#1
O(3)Pd(1)O(3)#1
O(1)Pd(1)Cu(1)
O(3)#1Pd(1)Cu(1)
O(8)Pd(2)O(12)
O(8)Pd(2)O(10)
O(12)Pd(2)O(10)
O(6)Pd(2)Pd(3)
O(10)Pd(2)Pd(3)
O(6)Pd(2)Pd(4)
O(10)Pd(2)Pd(4)
O(7)Pd(3)Pd(9)
O(9)Pd(3)O(16)
O(9)Pd(3)O(14)
O(7)Pd(3)Pd(2)
O(16)Pd(3)Pd(2)
O(7)Pd(3)Pd(4)
O(16)Pd(3)Pd(4)
Pd(2)Pd(3)Pd(4)
O(11)Pd(4)O(15)
O(11)Pd(4)O(13)
O(15)Pd(4)O(13)
O(17)Pd(4)Pd(3)
O(13)Pd(4)Pd(3)
O(17)Pd(4)Pd(2)
O(13)Pd(4)Pd(2)
O(2)Cu(1)O(2)#1
O(2)#1Cu(1)O(4)
O(2)#1Cu(1)O(4)#1
O(2)Cu(1)O(5)
90.6(4)
89.3(4)
175.4(6)
88.7(3)
87.7(3)
167.2(4)
87.7(4)
92.5(4)
77.0(3)
116.7(3)
119.6(3)
73.2(3)
90.7(4)
86.2(4)
165.4(4)
79.3(3)
111.1(3)
116.3(3)
74.9(3)
61.44(4)
173.9(4)
92.1(4)
88.5(4)
75.8(3)
122.4(3)
122.7(3)
74.4(3)
165.9(6)
89.8(4)
88.7(4)
97.1(3)
96.1(3)
82.9(3)
83.9(3)
180.000(1)
O(6)Pd(2)O(12)
O(6)Pd(2)O(10)
O(8)Pd(2)Pd(3)
O(12)Pd(2)Pd(3)
O(8)Pd(2)Pd(4)
O(12)Pd(2)Pd(4)
Pd(3)Pd(2)Pd(4)
O(7)Pd(3)O(16)
O(7)Pd(3)O(14)
O(16)Pd(3)O(14)
O(9)Pd(3)Pd(2)
O(14)Pd(3)Pd(2)
O(9)Pd(3)Pd(4)
O(14)Pd(3)Pd(4)
O(11)Pd(4)O(17)
O(17)Pd(4)O(15)
O(17)Pd(4)O(13)
O(11)Pd(4)Pd(3)
O(15)Pd(4)Pd(3)
O(11)Pd(4)Pd(2)
O(15)Pd(4)Pd(2)
Pd(3)Pd(4)Pd(2)
O(2)Cu(1)O(4)
O(2)Cu(1)O(4)#1
O(4)Cu(1)O(4)#1
O(2)#1Cu(1)O(5)
O(4)#1Cu(1)O(5)
O(2)#1Cu(1)Pd(1)
O(4)#1Cu(1)Pd(1)
O(4)Cu(1)O(5)
O(2)Cu(1)Pd(1)
O(4)Cu(1)Pd(1)
O(5)Cu(1)Pd(1)
RUSSIAN JOURNAL OF INORGANIC CHEMISTRY Vol. 54 No. 6 2009

8
88
MARKOV et al.
C(8B)
C(6B)
C(5B)
O(9B)
C(7B)
O(8B)
Pd(2B)
O(16B)
O(10B)
C(9B)
O(6B)
O(7B)
Pd(3B)
O(14B)
O(12B)
C(11B)
C(2AA)
C(1AA)
C(4A)
C(13B)
C(14B)
O(2AA)
O(3A)
C(3A) O(15B)
O(4A)
C(16B)
O(1AA)
O(17B)
C(10B)
C(12B)
O(11B)
Pd(4B)
H(1AA)
O(13B)
Pd(1A)
O(5A)
Cu(1A)
H(1A)
O(1A)
O(13A)
O(17A)
O(11A)C(10A)
Pd(4A)
O(3AA)
C(3AA)
O(2A)
C(14A)
C(1A)
O(4AA)
C(12A)
C(11A)
C(9A)
C(16A)
O(15A)
C(13A)O(10A)
C(15A)
O(16A)
C(2A)
O(14A)
C(4AA)
O(12A)
Pd(2A)
Pd(3A)
O(8A)
C(7A)
O(6A)
O(7A) O(9A)
C(5A)
C(6A)
C(8A)
Fig. 1. Structure of the complex Pd(µ-OOCMe) Cu(OH ) · 2Pd (µ-OOCMe) (1) (X-ray diffraction data).
4
2
3
6
Computational procedure. Quantum-chemical copper(II) atom (Cu–O, 2.127(13) Å) form hydrogen
calculations were performed by the DFT method with the bonds with the bridging oxygen atoms of the acetate
B3LYP three-parameter exchange correlation potential anions of two Pd (µ-OOCMe) molecules (O(5A)–
3
6
[
15, 16] using the CEP-121G and 6-31G*/SDD basis sets, O(15A)(O(15B)), 2.998(15) Å, H–O, 2.21 Å). The
common for systems of this type, with the use of the
Gaussian 03 program package [17]. Calculations were
performed with full optimization of all geometric
parameters without symmetry constraints.
geometry of the trinuclear Pd (µ-OOCMe) moieties is
3
6
nearly the same as in palladium(II) acetate without sol-
vating molecules (Pd···Pd 3.0751(17), 3.1472(18), and
3.1790(18) Å, Pd–O 1.947(10)–2.006(9) Å). The Pd···Cu
distance in the heterobimetallic moiety differs only
slightly from a distance of 2.5318(12) Å found in the
RESULTS AND DISCUSSION
linear 1D polymer [Pd(µ-OOCMe) Cu(OH )
·
4
2
2
HOOCMe] , synthesized by boiling an equimolar
Slow evaporation of acetic acid at room temperature
n
for a solution containing Pd (µ-OOCMe) and mixture of Pd (µ-OOCMe) and Cu(OOCMe) · H O in
3
6
3 6 2 2
Cu(OOCMe) · H O in a molar ratio of 3 : 1 leads to the acetic acid and formally having a similar system of
2
2
intermolecular hydrogen bonds with acetic acid mole-
cules of solvation [18].
formation of single crystals of complex 1, which con-
sists of the heterobinuclear paddlewheel complex Pd(µ-
OOCMe) Cu(OH ) connected with two homonuclear
4
2
Quantum-chemical calculations of the geometric
Pd (µ-OOCMe) moieties. X-ray crystallography
3
6
II II
and electronic structure of Pd M (µ-OOCMe) L com-
plexes (M = Zn (2), Ni (3), Cu (4), Co (5); L = H O
and HCN modeling MeCN), as well as of the hypothet-
ical complex with M = Fe (6) and their formate ana-
4
(
Fig. 1, Tables 1, 2) shows that the heterometallic moi-
II
II
II
II
ety in 1 has a four-bridge structure typical of a paddle-
2
wheel dimer (Pd–O, 1.970(9)–1.999(9) Å; Cu–O,
1
.981(9)–1.992(10) Å) with the metal atoms being at a
II II
short distance (Pd···Cu 2.521(3) Å) and that the hydro- logues Pd M (µ-OOCH) L (M = Zn (2‡), Ni (3‡), Fe
gen atoms of the water molecule coordinated to the (6‡), were performed by the DFT method.
4
RUSSIAN JOURNAL OF INORGANIC CHEMISTRY Vol. 54 No. 6 2009

ON THE NATURE OF THE CHEMICAL BOND
889
II II
Table 3. Comparison of the interatomic distances (Å) in the Pd M (µ-OOCMe) L complexes calculated by the DFT method
4
(
B3LYP with the CEP-121G and 6-31G*/SDD basis sets) with experimental X-ray crystallographic data
II II
Pd M (µ-OOOCMe) L
R(Pd–M)
R(M–L)
R(M–O)
R(Pd–O)
Method
4
(s) M = Zn, L(Zn) = OH2 2.651
2.025
2.075–2.100
2.45–2.055
CEP-121G
2
.5811(6)
1.963(3)
2.000(9)
2.089
2.049(3)–2.135(3) 1.998(3)–2.003(3) XRD [6]
2.040(4), 2.074(4) 1.984(4), 2.015(4) XRD [6]
L(Zn) = HOOCMe
2.576(1)
(
(
(
(
s) M = Zn, L(Zn) = NCH 2.667
s) M = Ni, L(Ni) = OH2 2.609
s) M = Ni, L(Ni) = NCH 2.608
t) M = Ni, L(Ni) = OH2 2.555
.472(3)
t) M = Ni, L(Ni) = NCH 2.574
2.084
2.049
CEP-121G
CEP-121G
CEP-121G
CEP-121G
2.289
1.922–1.932
1.921–1.932
2.043–2.066
2.058–2.066
2.058–2.066
2.042–2.052
2.467
2.023
2
2.000(13) 2.044(9), 2.056(8) 2.028(8), 2.053(8) XRD[6]
(
2.023
2.054
2.160(6)
2.110
2.194
2.072
2.132
2.025(7)
1.872
2.102
2.079
2.054–2.055
2.052–2.053
2.045–2.046
2.056–2.057
CEP-121G
2
2
.518
6-31G*/SDD
.483(2)
2.028(6)–2.076(6) 2.02.(5)–2.036(4) XRD [6]
(
d) M = Co, L(Co) = OH2 2.561
.506
q) M = Co, L(Co) = OH2 2.604
1.963–1.975
1.960–1.973
2.062–2.072
2.066–2.075
2.050–2.058
2.060–2.071
2.047–2.052
2.056–2.066
CEP-121G
2
6-31G*/SDD
CEP-121G
(
2
2
.556
6-31G*/SDD
.607(2)
2.084(5), 2.131(5) 2.012(6), 2.065(5) XRD [6]
(
d) M = Co, L(Co) = NCH 2.476
2.084–2.085
1.971–1.972
2.069–2.071
2.037(5)–2.085
2.012–2.016
2.020–2.027
2.036–2.037
2.061–2.061
2.058–2.060
CEP-121G
2
2
2
.531
.562
6-31G*/SDD
6-31G*/SDD
.459(1), 2.515(3) 2.057(5)
1.994(4)–2.017(4) XRD [6]
(
d) M = Cu, L(Cu) = OH2 2.636
2.133
2.053–2.055
2.064–2.068
CEP-121G
2
2
.578
2.174
6-31G*/SDD
.521(3)
2.100(8)
2.211
1.981(4), 2.012(4) 1.978(5), 2.036(4) XRD [this work]
2.010–2.012 2.053–2.054 CEP-121G
(
d) M = Cu, L(Cu) = NCH 2.645
Note: s is singlet, d is doublet, t is triplet, and q is quartet.
II II
Geometry optimization for the Pd M (µ- tively). This is consistent with the experimental Pd···M
OOCMe) L complexes gives satisfactory agreement distances (Table 2) and magnetic behavior of these
4
complexes.
with the experimental geometric parameters (Table 3).
The data in Table 4 demonstrate that, for all com-
The calculations of the geometric and electronic
plexes in the ground state, the structures in which the
2
structures of heterometallic complexes 2–6 (L = H O,
2
H O and HCN ligands are bonded to the M metal atoms
(Zn, Cu, Ni, Co) rather than to the Pd atom are more
energetically favorable.
HCN) showed (Table 4) that low-spin states are more
II
II
favorable for complexes containing Zn and Cu (sin-
glet for the zinc complex and doublet for the Cu com-
plex) and high-spin states are more favorable for the
NBO analysis for complexes 2–6 showed that they
Ni and Co complexes (triplet and quartet, respec- lack natural orbitals in the Pd–M moiety. The nonexist-
II
II
RUSSIAN JOURNAL OF INORGANIC CHEMISTRY Vol. 54 No. 6 2009

8
90
MARKOV et al.
Table 4. Difference in energy (kcal/mol) between low-and high-spin states (∆E and ∆Ed–q) and between possible isomers
s–t
II II
∆
E(LM–Pd) for the Pd M (µ-OOCMe) L complexes calculated by the B3LYP method with the CEP-121G and 6-31G*/SDD
4
basis sets
II II
II II
Pd M (µ-OOOCMe) L
∆Es–t
∆E(LM–Pd
)
Pd M (µ-OOOCMe) L ∆Ed–q
∆E(LM–Pd
)
Basis set
4
4
M = Zn
–27.0
M = Cu
CEP-121G
–
25.8
–31.4
6-31G*/SDD
CEP-121G
L(Zn) = OH2
L(Zn) = NCH
M = Ni
–19.1
–16.4
L(Cu) = OH2
L(Cu) = NCH
M = Co
–30.0
–25.7
–11.1
–3.5
–8.4
–6.1
–
26.7
6-31G*/SDD
CEP-121G
–32.1
14.1
–
30.5
6-31G*/SDD
CEP-121G
1
0.3
5.6
6-31G*/SDD
CEP-121G
L(Ni) = OH2
L(Ni) = NCH
32.1
36.9
L(Co) = OH2
L(Co) = NCH
12.7
–15.7
–13.1
5
.4
6-31G*/SDD
CEP-121G
–16.0
–
13.2
10.6
6-31G*/SDD
ence of the Pd–M bond is also supported by the anti-
bonding character of the HOMOs. The frontier MOs of
complexes 2, 3, and 6 are shown in Fig. 2.
ing orbital of the δ* type involving the d
2
2
Äés of
x – y
the metals as in the dipalladium complex
Pd (HNCHNH) [10]. HOMO-1 and HOMO-2 also
2
4
Naturally, the position and character of the orbitals correspond to the Cotton scheme. These are antibond-
ing MOs of the π* type. Below is the δ* åé involving
for complex 3 (PdNi) should be closest to the Cotton
scheme. Indeed, the HOMO is a strongly antibonding
orbital of the σ* type while the LUMO is an antibond-
the d AOs of the metals and then their bonding ana-
xy
logues are localized.
In complex 2, the position and symmetry of the
HOMO and four LUMOs coincide with those in 3 but
they mainly involve the AOs of the Pd atom (Fig. 2),
then there are two ligand-based MOs, and below are the
bonding analogues of the five upper MOs.
Table 5. Interatomic distances (Å) in the PdNi(µ-
2
+
OOCX) L complexes and [PdNi(µ-OOCX) ] ions calcu-
4
n
4
lated by the DFT-B3LYP/CEP-121G method
PdNi(µ-OOCX)4L R(Pd–Ni), X = CH R(Pd–Ni), X = H
For complexes 2 and 3, the absence of the M–M
bond is consistent with filling of all antibonding MOs
in the Cotton scheme. The removal of two electrons
from complex 3 should result in the formation of an
ordinary bond, which was predicted by calculations of
3
PdNi(µ-OOCX)4
2.575
2.293
2.458
2.608
2.609
2.599
2.302
2.518
2.635
2.633
PdNi(µ-OOCX) L]2+
[
4
2+
2+
model complexes 3 and 3Cl . Complex 3 has a bond-
2
PdNi(µ-OOCX) Cl2
ing MO and, in addition, the Pd···Ni distance becomes
4
shorter. The calculated Pd···M distances are 2.575 Å in
2+
PdNi(µ-OOCX) NCH
3, 2.293 Å in 3 , and 2.458 Å in 3Cl (Table 5).
4
2
The decrease in the Pd···Ni distance is evidence that
oxidation occurs in the Pd···Ni moiety, which can be
PdNi(µ-OOCX) OH2
4
RUSSIAN JOURNAL OF INORGANIC CHEMISTRY Vol. 54 No. 6 2009

ON THE NATURE OF THE CHEMICAL BOND
891
Fe
Zn
Ni
Pd
Pd
Pd
LUMOs of complexes 2, 3 and 6
Zn
Fe
Ni
Pd
Pd
Pd
HOMOs of complexes 2, 3 and 6
Fig. 2. Frontier MOs of complexes 2, 3, and 6 (B3LYP/CEP-121G calculations).
2
4
2
2
4
assigned the electronic configuration (σ π δ δ* π* ). atom (Fig. 2), whereas their bonding analogues are
After the removal of electrons, the antibonding HOMO mainly composed of the AOs of the Pd atom.
(
of the σ* type) in complex 3 (Fig. 2) becomes the
Thus, our findings allow us to draw a definite con-
clusion that the extraordinary short distance between
the metal atoms in the heterobinuclear complexes
2+
LUMO in complex 3 .
The reason for the nonexistence of the Pd–Fe bond
II
II
2+
Pd (µ-OOCMe) M L is caused only by the tightening
in complex 6, which is isoelectronic with complex 3 ,
is that the AOs of these metals (Pd and Fe) have dispar- effect of the acetate bridges while the direct electronic
ate energies, so that they are involved in different MOs Pd ···M interaction is lacking. Such interaction
and, as a result, the Pd atom does not transfer an excess becomes possible only after removal of electrons from
electron to the Fe atom. The antibonding σ* MO in the antibonding orbitals and formation of oxidized
4
II
II
2+
complex 6, as in 3 , is vacant (LUMO+1). However, complexes of the Pd(µ-OOCMe) ML type containing
4 n
this MO, like the other antibonding MOs (of the δ*- and less than 16 electrons in the MO of the metal moiety
π* types), is mainly composed of the AOs of the Fe Pd···M.
RUSSIAN JOURNAL OF INORGANIC CHEMISTRY Vol. 54 No. 6 2009

8
92
E, eV
MARKOV et al.
2. J. M. Thomas, B. F. G. Johnson, R. Raja, et al., Acc.
Chem. Res. 36, 20 (2003).
3
4
5
6
7
8
9
. N. Yu. Kozitsyna, S. E. Nefedov, Zh. V. Dobrokhotova,
et al., Ros. Nanotekhnol. 3 (3–4), 100 (2008).
–
–
–
4
6
8
. O. P. Tkachenko, A. Yu. Stakheev, L. M. Kustov, et al.,
Catal. Lett. 112, 155 (2006).
. N. Yu. Kozitsyna, S. E. Nefedov, N. V. Cherkashina,
et al., Izv. Akad. Nauk, Ser. Khim., 2149 (2005).
. N. Yu. Kozitsyna, S. E. Nefedov, F. M. Dolgushin, et al.,
Inorg. Chim. Acta 359, 2072 (2006).
. S. E. Nefedov, N. Yu. Kozitsyna, M. N. Vargaftik, and
I. I. Moiseev, Polyhedron 28, 172 (2009).
. S. E. Nefedov, M. N. Vargaftik, and I. I. Moiseev, Inorg.
Chem. Commun. 9, 755 (2006).
. Multiple Bonds between Metal Atoms, Ed. by F. A. Cot-
ton, C. A. Murillo, and R. A. Walton, 3rd ed. (Springer
Science and Business Media, Inc., New York, 2005).
–
10
Pd-Zn
Pd-Ni
Pd-Fe
1
1
1
0. J. F. Berry, E. Bill, E. Bothe, et al., J. Am. Chem. Soc.
29, 1393 (2007).
Fig. 3. Energy levels of the MOs of complexes 2, 3, and
(B3LYP/CEP-121G calculations).
1
6
1. R. Alvarez, A. R. de Lera, J. M. Aurrecoechea, and
A. Durana, Organometallics 26, 2799 (2007).
2. T. A. Stephenson, S. M. Morehouse, A. R. Powell, et al.,
J. Chem. Soc., 3632 (1965).
ACKNOWLEDGMENTS
This work was supported by the Russian Foundation 13. SMART (Control) and SAINT (Integration) Software,
Version 5.0, Bruker AXS, Inc., Madison, WI, 1997.
for Basic Research (project nos. 06-03-32578, 08-03-
0
1063, and 09-03-00514), the Council for Grants of the 14. G. M. Sheldrik, SADABS, Program for Scaling and Cor-
President of the Russian Federation for Support of
Leading Scientific Schools (grant no. NSh-
rection of Area Detector Data, Univ. of Göttingen, Göt-
tingen, Germany, 1997.
1
733.2008.3), and the Presidium of the RAS (the pro- 15. A. D. Becke, J. Chem. Phys. 98, 5648 (1993).
gram “Targeted Synthesis of Inorganic Compounds and 16. P. J. Stephens, F. J. Devlin, C. F. Chabalowski, and
Design of Functional Materials”).
M. J. Frisch, J. Phys. Chem. 98, 11623 (1994).
1
1
7. M. J. Frisch, G. W. Trucks, H. B. Schlegel, et al., Gauss-
ian 03, Revision B.01, Gaussian, Inc., Pittsburg, PA,
2
003.
REFERENCES
8. N.
S. Akhmadullina,
N. Yu. Kozitsyna, et al., Inorg.Chim. Acta. 2009 (in
press, doi:10.1016/j.ica.2008.09.042).
N. V.
Cherkashina,
1
. J. H. Sinfelt, Bimetallic Catalysts. Discoveries, Con-
cepts and Applications (Wiley, New York, 1983).
RUSSIAN JOURNAL OF INORGANIC CHEMISTRY Vol. 54 No. 6 2009