Synthesis and bonding in carbene complexes of an unsymmetrical dilithio methandiide: A combined experimental and theoretical study

Article abstract of DOI:10.1002/chem.201303115

Herein, we report the preparation of a new unsymmetrical, bis(thiophosphinoyl)-substituted dilithio methandiide and its application for the synthesis of zirconium- and palladium-carbene complexes. These complexes were found to exhibit remarkably shielded ¹³C NMR shifts, which are much more highfield-shifted than those of normal carbene complexes. DFT calculations were performed to determine the origin of these observations and to distinguish the electronic structure of these and related carbene complexes compared with the classical Fischer and Schrock-type complexes. Various methods show that these systems are best described as highly polarized Schrock-type complexes, in which the metal-carbon bond possesses more electrostatic contributions than in the prototype Schrock systems, or even as masked methandiides. As such, geminal dianions represent a kind of extreme Schrock-type ligands favoring the ionic resonance structure M⁺-CR₂^- as often used in textbooks to explain the nucleophilic nature of Schrock complexes.

Full text of DOI:10.1002/chem.201303115

FULL PAPER
DOI: 10.1002/chem.201303115
Synthesis and Bonding in Carbene Complexes of an Unsymmetrical Dilithio
Methandiide: A Combined Experimental and Theoretical Study
Viktoria H. Gessner,*[a] Florian Meier,[b] Diana Uhrich,[a] and Martin Kaupp*[b]
Abstract: Herein, we report the prepa-
ration of a new unsymmetrical, bis-
ACHTUNGTRENNUNG(thiophosphinoyl)-substituted dilithio
and to distinguish the electronic struc-
ture of these and related carbene com-
plexes compared with the classical
Fischer and Schrock-type complexes.
Various methods show that these sys-
tems are best described as highly polar-
ized Schrock-type complexes, in which
the metal–carbon bond possesses more
electrostatic contributions than in the
prototype Schrock systems, or even as
“masked” methandiides. As such,
geminal dianions represent a kind of
“extreme” Schrock-type ligands favor-
methandiide and its application for the
synthesis of zirconium- and palladium-
carbene complexes. These complexes
were found to exhibit remarkably
shielded 13C NMR shifts, which are
much more highfield-shifted than those
of “normal” carbene complexes. DFT
calculations were performed to deter-
mine the origin of these observations
ing the ionic resonance structure
ꢀ
M+ CR2 as often used in textbooks
ꢀ
to explain the nucleophilic nature of
Schrock complexes.
Keywords: carbenes · density func-
tional calculations
·
dianions ·
lithium · transition metals
Introduction
Floch, Mꢀzailles, and co-workers focused on the thiophos-
phinoyl congeners. Hence, a huge variety of different transi-
tion-metal carbene complexes (incorporating early (e.g., Ti,
Zr, Hf) and late transition-metals (e.g., Pd, Pt)) that also
complex with lanthanides and actinides, have been pre-
pared.[4] These complexes revealed interesting, partly differ-
ent reactivities depending on the ligand and the metal
center. As such, metallo-Wittig-like reactivity was observed
in reactions of zirconium and scandium complexes with car-
Transition-metal carbene complexes have experienced huge
interest over the past decades due to their many applica-
tions in both stoichiometric and catalytic organic transfor-
mations. It is well-known that the nature and reactivity of
the carbene complexes greatly depends on the electronic
properties of the substituents bound to the carbine-carbon
atom. Those with hydrogen- or alkyl/aryl substituents are
usually classified as Schrock carbenes (alkylidenes), whereas
those containing electronegative heteroatom substituents,
such as O, N, or S, are designated as Fischer carbenes.[1] In
recent years, bis(phosphonium)-stabilized methandiides
have received increasing interest as precursors for new car-
bene complexes.[2] These complexes were reported to exhibit
unique electronic properties differing from the generally
known Fischer-type carbene and the Schrock-type alkyli-
dene complexes.[3] So far, this research has mainly focused
on the oxidized versions of the commercially available bis-
bonyl compounds,[5] whereas C H bond activation was ob-
ꢀ
served with an yttrium complex.[6]
Although quite a number of carbene complexes derived
from dilithio methandiides has been synthesized over the
past years, the bonding situation is still under debate. Only
recently, Mindiola and Scott highlighted the differences of
theses carbene complexes compared with the typical Fischer
and Schrock complexes.[3] Formally, the bonding situation in
these complexes can be described by a four-electron dona-
tion from the ligand to the metal, leaving a negatively
charged carbon atom bound to electron-withdrawing sub-
stituents (Figure 1).[2] Although these carbenes were often
reported to possess Schrock/alkylidene-like reactivity, no
comparative study of the electronic structure has been un-
dertaken to rank these complexes within the Fischer and
Schrock classification Scheme.[4k] Thus the question arises,
whether carbene complexes derived from dilithio methan-
diides indeed display a new class of carbene complexes as
depicted in Figure 1; or, do they simply offer a new route to
Schrock alkylidenes, which also allows for the syntheses of
complexes with late transition-metals in low oxidation
states.
ACHTUNGTRENNUNG(diphenylphosphino)methane (DPPM) ligand. The groups of
Cavell and Liddle studied the coordination chemistry of the
bis(iminophosphorane) derivative in depth, whereas Le
[a] Dr. V. H. Gessner, D. Uhrich
Institut fꢁr Anorganische Chemie
Julius-Maximilians-Universitꢂt Wꢁrzburg
Am Hubland, 97074 Wꢁrzburg (Germany)
E-mail: vgessner@uni-wuerzburg.de
[b] F. Meier, Prof. Dr. M. Kaupp
Technische Universitꢂt Berlin, Institut fꢁr Chemie
Sekr. C7, Strasse des 17. Juni 135
10623 Berlin (Germany)
E-mail: martin.kaupp@tu-berlin.de
It should be noted that also the terms Fischer- and
Schrock-type complexes have been used differently in litera-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201303115.
Chem. Eur. J. 2013, 19, 16729 – 16739
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
16729

stituted systems.[8,9] So far, almost exclusively symmetric pre-
cursors have been used and applied as ligands in carbene
complexes. Earlier studies on silyl-substituted systems indi-
cated the somewhat limited flexibility in changing the substi-
tution pattern at the central carbon atom due to the re-
quired anion-stabilizing ability.[10,11] Thus, we decided to
retain the bis(phosphonium) framework, which has been
proven as suitable stabilizing moiety, and to modify the sub-
stituents at phosphorus. The starting compound 2 was syn-
thesized from dicyclohexylmethylphosphine sulfide 1, which
was prepared in a one-pot reaction from phosphorus tri-
chloride. Stepwise addition of two equiv of cyclohexylmag-
nesium chloride followed by methylmagnesium bromide and
elemental sulfur afforded 1 as colorless solid in 72% yield
(Scheme 1).[12] Deprotonation of the methyl group of 1 with
Figure 1. Representation of the dominant orbital interactions in a) Fisch-
er-type carbenes, b) Schrock-type alkylidenes, and c) complexes based on
methandiides.
ture. This originates from the fact that classifications accord-
ing to textbook descriptions (Fischer-type: p-donor substitu-
ents, metals in low oxidation states, electrophilic character)
are often ambiguous and the transitions are smooth. As
such, complexes have been reported featuring a metal
center in low oxidation state (Fischer-type), but with nucleo-
philic character and thus Schrock-type reactivity. Alterna-
tively, both types of complexes were distinguished by the
bonding situation. Thus, Fischer-carbenes were described to
be formed by donor–acceptor interactions of singlet frag-
ments, whereas Schrock-alkylidenes bind covalently as trip-
let species.[7] Additionally, distinction solely by means of the
oxidation state of the metal or the type of substituents at
the carbene carbon atom are made. Because of the focus of
this paper on the influence of the methandiide ligand sys-
tems on the bonding situation, we decided to use the terms
Schrock and Fischer complexes depending on the a-substitu-
ents.
Scheme 1.
n-butyllithium was accomplished at room temperature. Sub-
sequent treatment with diphenylchlorophosphine and oxida-
tion with elemental sulfur furnished the bis(thiophosphi-
noyl)methane 2 as colorless crystalline solid in good yields
of 67%. Compound 2 is characterized by a doublet of dou-
1
blet of the methylene protons in the H NMR spectrum and
a set of two doublets in the 31P NMR spectrum (dP =31.4
and 63.9 ppm, 2JPP =18.2 Hz). The molecular structure of
1 is shown in Figure 2. Compound 1 crystallizes in the ortho-
In this publication, we describe the synthesis of an unsym-
metrical dilithio methandiide and its application as a ligand
in transition-metal carbene complexes. We have chosen zir-
conium as the early- and palladium as the late-metal center
to compare their influence on the bonding situation. In the
course of these studies, we observed remarkably shielded
13C NMR shifts, which prompted us to perform detailed
computational studies to understand these observations and
especially to distinguish the bonding situation in these sys-
tems from known Fischer and Schrock complexes. To gain
insight into the electronic structure, electron localization
function (ELF), atoms in molecules (AIM) and natural
bond orbital (NBO) analyses have been taken out. We show
that despite the unique synthesis of these complexes, gemi-
nal dianions can be described as “extreme” Schrock-type li-
gands that polarize the metal–carbon bond towards the
carbon atom.
Figure 2. Molecular structure of compound 2. For crystallographic details
see the Supporting Information (thermal ellipsoids at 50% probability
Results and Discussion
ꢀ
ꢀ
level). Selected bond lengths [ꢄ] and angles [8]: C1 P1 1.8272(17), C7
Preparation of an unsymmetrical dilithio methandiide: In
the course of our studies on dilithio methandiides we
became interested in the synthesis of unsymmetrically sub-
ꢀ
ꢀ
ꢀ
P1 1.8179(17), C14 P2 1.8383(17), C20 P2 1.8366(17), P2 S2 1.964(1),
ꢀ
ꢀ
ꢀ
P1 S1 1.951(1), C13 P1 1.841(2), C13 P2 1.841(2), P1-C13-P2 122.6(1),
C13-P2-S2 107.5(1), C13-P1-S1 116.9(1).
16730
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16729 – 16739

Synthesis and Bonding in Carbene Complexes
FULL PAPER
rhombic crystal system, space group P212121. Both thiophos-
phinoyl units point to opposite directions in the crystal, thus
affording a conformational change for coordination of
metals through both sulfur atoms. Bond lengths and angles
are in the expected region.
Dimetalation of the methylene unit was first probed in
a deuteration experiment (Scheme 2). Treatment of a suspen-
sion of the ligand in diethyl ether with methyllithium gave
Scheme 2.
way to a clear yellowish solution under gaseous evolution,
which was treated with D2O after 12 h. NMR spectroscopy
of the crude product showed the quantitative formation of
the bis-deuterated compound, which could be isolated in
97% yield. Compound [D2]2 is characterized by a broad sin-
Figure 3. Molecular structure of dilithio methandiide 2. For crystal
AHCTUNGTRENNUNG
ACHTUNGTRENNUNGlog-
2
glet in the H NMR spectrum at d=3.29 ppm and two mul-
ꢀ
ꢀ
ꢀ
ꢀ
and angles [8]: C1 P2 1.676(3), C1 P1 1.681(3), C1 Li2 2.216(6), C1 Li1
tiplets in the 31P NMR spectrum (d=31.2 and 63.7 ppm).
For isolation of the intermediate dilithio methandiide, a sus-
pension of 2 in diethyl ether was treated with two equiva-
lents of methyllithium at room temperature. After 2 was
completely dissolved, the dilithium salt 2-Li2 starts to crys-
tallize from the reaction mixture as colorless plates and can
be isolated in up to 89% yield. Compound 2-Li2 was charac-
terized by multi-nuclear NMR spectroscopy, elemental anal-
ysis, and single-crystal X-ray diffraction analysis. The precip-
itation of 2-Li2 from the reaction mixture is quite surprising
as the symmetrical tetraphenyl-substituted derivative is still
soluble under these conditions. When introducing the cyclo-
hexyl moieties we expected the contrary tendency. NMR
spectroscopy of the dilithio methandiide showed the inclu-
sion of one equivalent of ether in the compound. The
31P NMR spectrum showed two signals at d=35.3 and
55.4 ppm.
ꢀ
ꢀ
ꢀ
ꢀ
2.230(6), Li1 S2 2.484(3), S2 Li3’ 2.405(5), Li2 S1 2.522(2), Li3 O1
ꢀ
ꢀ
ꢀ
1.879(6), Li3 S1 2.395(5), P1 S1 2.039(1), P2 S2 2.035(1); P2-C1-P1
131.4(2), P2-C1-Li2 131.4(2), P1-C1-Li2 92.5(2), P2-C1-Li1 91.3(2), P1-
C1-Li1 129.8(2), Li2-C1-Li1 67.6(3), S2’-Li1-S2 144.0(4), S1’-Li2-S1
148.0(4), C1’-Li1-C1 111.9(4).
carbon atom possesses two contacts to the lithium atoms,
Li1 and Li2, which lie on the C2 axes of the dimer, thus
forming a planar (CLi)2 ring. The two other lithium atoms
are solely coordinated by the sulfur atoms of the thiophos-
phinoyl moieties and diethyl ether. The four lithium atoms
ꢀ
form a planar Li4 square with Li Li distances of 3.262(7)
ꢀ
and 3.282(7) ꢄ. The Li C distances of 2.216(6) and
2.230(6) ꢄ are in the range of oligomeric organolithium
compounds.[14]
In the past, it was shown that the double deprotonation of
bis(phosphonium)methanes results in a considerable change
in the bond lengths of the SPCPS backbone. These changes
are also obvious in compound 2-Li2. The most remarkable
The molecular structure of 2-Li2 is depicted in Figure 3.
The methandiide crystallizes in the monoclinic crystal
system, space group I2/a as a highly symmetric dimeric
structure, containing only half a molecule in the asymmetric
unit, which is assembled to the aggregate through C2 sym-
metry. In the crystal, the phenyl and cyclohexyl substituents
are disordered over the whole molecule with 50% probabili-
ty for both possible positions. This disorder was found in re-
peated measurements also when crystals were grown under
more dilute conditions and at low temperatures. For refine-
ment, the cyclohexyl substituents were fitted to an ideal
model so that no discussion concerning the organic periph-
ery of the structure is possible. Nevertheless, the most inter-
ꢀ
changes concern the shortening of the P C bond lengths by
ꢀ
approximately 8% and the elongation of the P S bond
lengths by 4%. Furthermore, a widening of the P-C-P angle
from 122.68 in compound 2 to 131.48 in the methandiide is
observed. These changes from the neutral to the dianionic
compound are in accordance with previous studies on sym-
metric bis(phosphonium)methandiides[13] and can be re-
ferred to the electronic structure, which is best described by
a Lewis structure with high charge-separation. Resonance
structures with double-bond character have no contribution
as also has been found in other methandiides. The charge
separation in 2-Li2 results in strong electrostatic interactions
ꢀ ꢀ ꢀ
esting part concerns the bonding properties of the S P C
ꢀ
ꢀ ꢀ
P S (SPCPS) backbone. In total, the structure is similar to
within the P C P backbone and thus in the contraction of
ꢀ
the structure reported for the analogue symmetric methan-
diide.[13] The molecule consists of two dimetalated methan-
diides and two diethyl ether molecules. The metalated
the P C bonds. Further stabilization of the negative charge
at the methanide carbon atom is obtained by negative hy-
ꢀ
perconjugation into antibonding P S orbitals, which results
Chem. Eur. J. 2013, 19, 16729 – 16739
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16731

V. H. Gessner, M. Kaupp et al.
in the lengthening of the respective bonds. This is reflected
by the calculated atomic charges and Wiberg bond indices
obtained from computational studies of the 2-Li2.
Preparation of metal complexes: The applicability of meth-
andiide 2-Li2 as ligand for transition-metal carbene com-
plexes was tested with zirconocene dichloride as the early
transition-metal and (bistriphenylphosphine) palladium
chloride as the late transition-metal (Scheme 3). The prepa-
Figure 4. Top: Molecular structure of zirconocene complex 3. Bottom:
a) Molecular structure of palladium complex 4; b) 13C NMR resonance
for carbenic carbon in 4. For crystallographic details see the Supporting
Scheme 3.
ꢀ
Information. Selected bond lengths [ꢄ] and angles [8]: 3: C1 P1 1.836(2),
ꢀ
ꢀ
ꢀ
ꢀ
ration was accomplished by salt metathesis with one equiva-
C7 P1 1.846(2), C13 P2 1.669(2), C13 P1 1.682(2), C14 P2 1.827(2),
ꢀ
ꢀ
ꢀ
ꢀ
C20 P2 1.819(2), C13 Zr 2.259(2), P1 S1 2.0297(7), P2 S2 2.0105(8),
lent of [ZrCl2Cp2] (Cp=cyclopentadienyl) and [PdCl2-
ꢀ
ꢀ
ꢀ
P2 Zr 3.1508(6), S1 Zr 2.7502(6), S2 Zr 2.7519(7); P2-C13-P1
143.02(12), P2-C13-Zr 105.68(9), P1-C13-Zr 107.87(9), C13-P1-S1
100.40(7), C13-P2-S 103.20(7), P1-S1-Zr1 82.70(2), P2-S2-Zr1 81.23(2). 4:
ACHTUNGTRENNUNG
(PPh3)2], respectively.[15] Monitoring of the reaction by using
NMR spectroscopy showed the clean formation of the de-
sired complexes. After work-up, the zirconium complex 3
could be isolated by crystallization in 73% yield, and the
palladium complex 4 in 63% yield. Both complexes were
characterized by multi-nuclear NMR spectroscopy, elemen-
tal analysis, and X-ray diffraction analysis.
ꢀ
ꢀ
ꢀ
ꢀ
C13 P1 1.682(3), C13 P2 1.692(3), C13 Pd1 2.101(3), C1 P1 1.822(3),
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
C7 P1 1.820(3), C14 P2 1.831(3), C20 P2 1.842(3), P1 S1 2.040(1), P1
ꢀ
ꢀ
ꢀ
ꢀ
Pd1 2.829(1), P2 S(2) 2.051(1), P2 Pd1 2.802(1), P3 Pd1 2.303(1), Pd1
S2 2.374(1), Pd1 S1 2.388(1); P1-C13-P2 141.3(2), S2-Pd1-S1 158.55(3),
C13-Pd1-P3 177.72(7), C13-Pd1-S2 80.13(8), C13-Pd1-S1 80.77(8), P3-
Pd1-S1 98.65(3), P3-Pd1-S2 100.05(3).
ꢀ
The 31P{1H} NMR spectrum of 3 features two doublets at
d=20.1 and 47.8 ppm, which are both upfield-shifted com-
pared with the neutral ligand 2. In comparison with typical
carbene complexes, the carbenic carbon atom resonates at
highfield (d=29.2 ppm). This has been attributed to the re-
maining negative charge at the carbon atom and the low
double-bond character of the metal–carbon bond (see
below).[5b] The coupling constants are considerably larger
carbene complexes derived from dilithio methandiides. Due
to coupling with the adjacent phosphorus atoms the metha-
nide carbon gives rise to an ABX spin system (Figure 4)
with JCP coupling constants of JCP =85.0 and 71.1 Hz.
The molecular structure of the complex 3 and 4 are de-
picted in Figure 4. Crystal data and structure refinement de-
tails are given in the Supporting Information. In accordance
1
1
1
(1JCP =78.7 Hz, JCP =68.8 Hz) than those found for 2, which
ꢀ
can be attributed to an higher s character of the central
carbon atom in the zirconium complex. The 31P{1H} NMR
spectrum of the palladium complex 4 features three signals
of equal intensity. The PPh3 signal resonates at d=20.8 ppm
as a virtual triplet (3JPP =14.7 Hz), whereas the phosphorus
atoms of the carbene ligand resonate as doublets of doublets
with the NMR spectroscopic data, the C Zr (2.259(2) ꢄ) is
ꢀ
rather long and more related to a C Zr single bond. The
methanide carbon atom features a trigonal-planar coordina-
tion environment (sum of angles: 356.6(1)8) with C P dis-
tances of 1.682(2) ꢄ to the PCy2 and 1.669(2) ꢄ to the PPh2
moiety. The slightly longer bonds to the cyclohexyl-substi-
tuted moiety account for the more electron-rich phosphorus
and thus for the less electrostatic attraction with the nega-
ꢀ
3
at d=36.6 (2JPP =33.4 Hz, JPP =13.8 Hz) and 70.1 ppm (dd,
2JPP =33.4 Hz, 3JPP =15.5 Hz). Interestingly, the carbenic
carbon atom appears at even higher field than that of com-
plex 3 (d=ꢀ18.3 ppm), which is, to the best of our knowl-
edge, the most shielded carbon-shift compared with related
ꢀ
tively charged carbon atom. The C Zr bond however, re-
mains unaffected by the PCy2 substitution.[5b] Overall, the
ꢀ
P C bond lengths are comparable to those found in the di-
16732
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16729 – 16739

Synthesis and Bonding in Carbene Complexes
FULL PAPER
Table 1. Comparison of computed and experimental 13C NMR chemical
ACHTUNGTRENNUNGlithio methandiide 2-Li2 and considerably shortened com-
pared with the neutral ligand 2, clearly indicating the exist-
ing electrostatic interactions within the ligand backbone.
shifts of the carbine-carbon atoms in 3 and 4 (in ppm vs. TMS).[a]
d
13C
Exptl
ZORA-SR
ZORA-SO
(C carbene)
ꢀ
This is in line with the elongated P S bond (2.0297(7) and
3
4
29.2
ꢀ18.3
34.7
ꢀ2.2
33.1
ꢀ14.8
2.0105(8) ꢄ) accounting for negative hyperconjugation.
In the molecular structure of 4, the palladium atom fea-
tures a distorted square-planar (sum of angles: 350.68) coor-
dination environment with bond angles between 80.13(8)
[a] B3LYP/TZ2P/ZORA/GIAO results.
ꢀ
and 100.05(3)8. The Pd C distance amounts to 2.101(3) ꢄ,
which is slightly shorter than the one reported for the sym-
metric bis(thiophosphinoyl) ligand (2.113(2) ꢄ), but longer
small for 3. At the two-component level, agreement with ex-
periment may be considered excellent. Already, the impor-
tance of spin-orbit effects for 4 shows that a naꢅve interpre-
tation of the differences in terms of charge is not indicated
(indeed, the natural population analysis (NPA) charge on
the carbene carbon is more negative in 3, see Table 2).
[15]
ꢀ
than real Pd C double bonds.
ligand, complex 4 features a bent coordination mode. There-
by, the PCP backbone arranges almost perpendicularly to
the S1 Pd S2 P3 plane comprising an angle between both
planes of 70.0(2)8. Analogous to the zirconocene complex 3,
compound 4 features the same tendencies within the ligand
As for the symmetric
ꢀ
ꢀ ꢀ
Table 2. Results of the NBO analysis of the carbene complexes A–L.
ꢀ
backbone: the P C bond distances are shorter (1.682(3) and
Charges[a]
Metal–carbon bond
WBI
ꢀ
1.692(3) ꢄ) and the P S bonds longer (2.051(1) and
qC
qM
(occupation)
M
C
2.040(1) ꢄ) than in the free protonated ligand 2. Due to the
unsymmetrical substitution pattern at the carbon atom and
its pyramidalization in 4, the carbon becomes stereogenic.
However, in solution, no splitting of the signals (diasterotop-
ic Ph substituents) or broadening was observed.
[%]
[%]
A
B
C
D
E
F
+0.18
ꢀ0.43
ꢀ0.62
ꢀ0.69
ꢀ1.04
ꢀ1.15
ꢀ1.56
ꢀ1.53
ꢀ1.56
ꢀ1.36
ꢀ1.33
ꢀ1.32
ꢀ0.98
+0.45
+1.30
+0.94
+1.26
+1.34
+1.11
+1.55
+1.18
+0.41
ꢀ0.37
+0.28
s (1.91)
no p bond
s (1.95)
p (1.70)
s (1.97)
p (1.94)
s (1.98)
p (1.87)
s (1.94)
p (1.90)
s (1.96)
p (1.90)
s (1.86)
LPC (1.57)
s (1.81)
LPC (1.57)
s (1.86)
p (1.73)
s (1.87)
LPC (1.50)
LPC (1.45)
LPC (1.40)
s (1.74)
31.52
68.48
0.84
1.70
1.82
1.68
1.82
1.63
0.72
0.67
0.68
0.89
0.63
0.44
42.39
54.27
32.08
40.17
32.15
55.87
25.72
58.96
24.72
44.57
18.95
–
18.04
–
17.40
11.93
26.92
–
57.61
45.73
66.82
56.98
67.75
44.01
74.28
41.04
75.18
55.35
81.05
100
81.96
100
82.60
88.07
73.08
DFT calculations: bonding and NMR chemical shifts: DFT
calculations were performed to gain further insight into the
electronic structures of 3 and 4 and of related metal-carbene
complexes. Structure optimizations were performed on the
real systems employing the BP86 functional. The optimized
structures are in excellent agreement with the experimental
data (see Table S9, the Supporting Information). For com-
parison, also prototype Fischer complex A and Schrock-type
complexes B–F as well as the model systems G–L were cal-
culated (Figure 6).
G
H
I
Let us first examine the 13C NMR shifts observed for the
carbene carbon atoms, as such parameters have been used
previously to discuss donor properties of other carbene li-
gands, particularly of N-heterocyclic carbenes (NHCs).[16] In
the case of carbene complexes derived from dilithio methan-
diides, a wide range of 13C NMR shifts (if observed at all)
have been reported, with 4 (d=ꢀ18.3 ppm) featuring the
most highfield-shifted resonance.[4] Notably, the shifts are in
a considerably more shielded region compared with typical
Schrock carbene complexes, which typically feature shifts of
d=200–400 ppm.[17] To understand the differences in the
13C NMR shifts of 3 and 4 relative to each other, to related
derivatives, and to more typical Schrock complexes, and to
gain further insights into electronic structure, the shifts were
computed and analyzed at different computational levels.
Table 1 shows results obtained at scalar-relativistic (ZORA-
SR) and two-component (ZORA-SO) B3LYP/TZ2P/
ZORA/GIAO level in the ADF code, in which the two-
component results show the effect of spin-orbit (SO) cou-
pling (NMR shifts obtained at other levels are given in
Table S18 in the Supporting Information). SO effects are
significantly shielding (about d=ꢀ12 to ꢀ13 ppm) for 4 but
J
100
100
100
60.99
100
K
L
–
–
39.01
–
LPC (1.70)
[a] Partial charges of the carbene ligand and the metal center.
Complex 3 exhibits a d=68 ppm more shielded value
than its homologue, in which the Cp ligands are replaced by
chlorine.[4] This is confirmed by computations (see scalar rel-
ativistic BP86 results for truncated model complexes G and
I in Table S18, the Supporting Information). These show
also that, interestingly, for the simple Schrock-type carbene
complexes [ZrACHTUNTRGENNUG(CH2)Cp2] (F) and [ZrAHCUTNGTRNEN(UGN CH2)Cl2] (E) matters
are just reversed for Cp versus Cl ligands.
To understand this better, we have carried out a detailed
analysis of the paramagnetic term by breaking it down into
individual couplings between (localized) occupied and (ca-
AHCTUNGTREGnNNUN onical) virtual MOs using the IGLO method for the gauge
problem (Pipek–Mezey localized MOs have been used, as
these are known to provide a separation between s- and p-
Chem. Eur. J. 2013, 19, 16729 – 16739
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16733

V. H. Gessner, M. Kaupp et al.
orbital manifolds). These detailed analyses for the model
complex I, its chlorine analogue G, and the two abovemen-
tioned simple Schrock-type complexes F and E are given
and illustrated in Tables S19–S22 in the Supporting Informa-
tion (the local MAG/ReSpect program has been used for
these analyses, complex L is analyzed in Table S23). Here
we summarize only the main findings. First of all, the para-
magnetic contributions to the isotropic carbene 13C shield-
ings in all four systems are by far dominated by one particu-
larly deshielded component, s11, oriented in the carbene
structure of the Zr and Pd complexes 3 and 4, respectively,
was studied by NBO analyses. Previous studies have shown
that carbene complexes based on dilithio methandiides are
often characterized by a highly ionic character of the metal–
carbon bond. Indeed, NPA charges for 3 and 4 feature a re-
markably negative charge on the carbenic carbon atom,
which is slightly more pronounced for the zirconium com-
plex (Figure 5). In the case of the metal atom, the palladium
ꢀ
plane but perpendicular to the Zr C bond. Differences be-
tween the cyclic complexes G and I on one side and the
classical Schrock complexes E and F on the other side are
also dominated by changes in this component, as are differ-
ences between Cp- versus Cl complexes. We may thus con-
centrate our following discussion on s11.
The latter tends to be dominated by contributions from
localized molecular orbitals (LMOs) corresponding to a) the
ꢀ
ꢀ
Zr C s-bond and b) the Zr C p-bond. More precisely, the
ꢀ
former couples to virtual MOs with essentially p*
N
Figure 5. NPA charges and Wiberg bond indices of the complexes 3 and
4.
character, and the latter to virtual MOs with close to s-anti-
bonding character (Tables S19–S22, the Supporting Informa-
tion). The visualization of the occupied LMOs indicates
both bonds to be evenly distributed over Zr and C for the
simple Schrock complexes, whereas the p-bond in the cyclic
complexes is essentially a lone pair on the carbene carbon
atom. Overall, the much less pronounced deshielding in the
cyclic complexes G and I compared with the classical
Schrock complexes E and F may be attributed both to
smaller matrix elements of the relevant perturbation opera-
tors (external magnetic field and nuclear magnetic
moment), as well as to larger energy denominators.
The dꢁ106 ppm larger s11 component for I compared
with G is largely due to larger energy denominators (Ta-
bles S21 and S22, the Supporting Information). That is, the
better p-donor ligand Cp raises the low-lying unoccupied
MOs on Zr compared with Cl ligands and thereby reduces
the paramagnetic terms. Whereas such increased energy de-
nominators are to some extent also present in F compared
with E, here the trend appears to be dominated by increased
perturbation matrix elements (Tables S19 and S20, the Sup-
atom bears a smaller positive charge compared with the zir-
conocene complex (qZr = +1.14), which can be considered
as a zwitterionic species. This is in line with the calculated
Wiberg bond indices of the metal–carbon bonds. Analogous
to previous studies, the WBIs were found to be considerably
smaller than 1 and thus disagree with a “real”, covalent
M=C double bond. It is noteworthy, that the NPA charges
and Wiberg bond indices are quite similar to those found
for the dilithio methandiide 2-Li2 (see above), consistent
with the high importance of electrostatic interactions in the
ligand backbone.
To better understand the bonding situation in carbene
complexes based on dilithio methandiides, we turned our at-
tention to a comparative study of a series of model com-
plexes including the prototype Fischer complex A and
Schrock-type complexes B–F (Figure 6). In the case of the
dianion-based systems, the complexes G–L were chosen;
they differ in the metal center, the co-ligands, and also the
ligand backbone (SCS (C) vs. NCN (D)). Detailed computa-
tional studies by Frenking et al. have revealed unambiguous
differences in the bonding situation of Fischer and Schrock
complexes in NBO analyses.[18] It was found that the car-
bene ligands in Schrock complexes carry a distinctly nega-
tive partial charge, whereas those in Fischer complexes fea-
ture only a small negative or even positive charge. The same
tendency (but of opposite direction) was observed for the
metal centers. Furthermore it was found, that Schrock com-
plexes always possess metal-carbene s and p bonds, whereas
for Fischer complexes often only the s-bond is observed.
Thereby, the p-electrons in Fischer complexes are polarized
towards the metal center, whereas they are more equally
distributed in Schrock complexes.[7,18,19] These findings are
reflected by the NBO analyses of the prototype Fischer and
Schrock complexes A and B–F (Table 2). Considering the
ꢀ
porting Information), consistent with changes in Zr C cova-
lency (including some involvement of the Cp ligands in the
ꢀ
Zr C s-bonding LMO and thus the shapes of the relevant
orbitals involved. This explains why in this case the chloro
complex exhibits an approximately d=68 ppm lower car-
bene 13C shift. Additionally, we find that the s22 component
obtains shielding paramagnetic contributions pointing to off-
center ring currents for G and I (less so for E and F; see
Tables S19–S22). Overall, these analyses indicate a relatively
complicated interplay of covalency and energy gaps, thus
rendering an understanding without detailed quantum-
chemical analysis virtually impossible. Hence, no clear-cut
classification of the carbene character in these systems can
be made by means of 13C NMR shifts.
Therefore, we turned our attention to a detailed study of
the bonding situation of the complexes. First, the electronic
16734
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16729 – 16739

Synthesis and Bonding in Carbene Complexes
FULL PAPER
charge on the carbene carbon atom than L, in spite of the
fact that the Zr d0 center in the latter may function as a p-
acceptor, whereas the Pd d8 center in L may not. This re-
flects the lower electronegativity of Zr (mainly in the
s bond).[21]
The bonding situation obtained from the NBO analyses is
well reflected by the canonical molecular orbitals of these
compounds. Figure 7 exemplarily depicts the important mo-
Figure 6. Calculated model complexes.
dianion-based complexes G–L, the most remarkable feature
concerns the charge concentration on the carbenic carbon
atom. All compounds, independent of the metal or the co-li-
gands, feature a high negative charge (qC =ꢀ1.32 to ꢀ1.56).
These charges are in general higher than those found in typ-
ical Schrock complexes such as B (qC =ꢀ0.43), but compara-
ble to the hypothetical Schrock complex F. Contrary to the
charge at the carbon atom, the partial charges of the metal
were found to vary strongly, depending on the nature of the
metal and the co-ligands. Analogous to the “real” systems 3
and 4 the Wiberg bond indices are significantly smaller than
1, with the smallest value (WBI=0.44) found for the palla-
dium complex L. These values are comparable to the Fisch-
er carbene complex A thus indicating a more dative or ionic
nature of the metal–carbon bond than normally found in
Schrock complexes. Even Schrock complex F, with the most
Figure 7. Top: Molecular orbitals of Schrock complex B. Bottom: Titani-
um complex I (isosurface value at 0.05).
lecular orbitals for the Schrock-type complex B and the tita-
nium complex J. Overall, these data show that carbene com-
plexes based on dilithio methandiides are formally Schrock-
type ligands, however with a stronger polarization and no
truly covalent nature of the metal–carbon bond. In compari-
son to the prototype Schrock complexes such as B, which
are best described by a classical M=C double bond (reso-
nance structure cov, Figure 7), these ligands favor the more
ionic resonance structure ion with a metal–carbon single
bond and a remaining lone pair at the carbon. It should be
noted that this ionic formulation is often used in textbooks
to explain the nucleophilic nature of Schrock complexes.[22]
This leads us to conclude that geminal dianions are a kind
of “extreme” Schrock-type ligands. However, one has to
keep in mind that the metal–carbon interaction also depends
on the nature of the metal center. Thus, interactions be-
tween the two extreme canonical forms M=C and M2+···C2-
are possible. So far, the most “covalent” interactions are
hitherto observed for uranium complexes of methandiides
with bond indices in the range of archetypal Schrock com-
plexes.[23] We would like to point out that the methandiide-
based systems presented in this manuscript do not fulfill the
criterion of M=C double-bonded systems. In the case of the
palladium complexes 4 or L, the complexes are actually
pronounced charge separation, features
a bond index
(WBI=1.68) that is clearly larger than 1.
Additional information about the metal–carbon bonds in
the complexes based on the methandiide ligands is given by
the NBOs involved in the bonding (we keep in mind that
the NBO Lewis structures do not fully describe the one-par-
ticle density matrix in these transition-metal complexes, in
particular, when Cp ligands are present). Similar to Fischer
carbenes, only a s-bond is observed for most of the exam-
ples G–L. These bonds are clearly polarized towards the
carbon end (e.g., 83% in I). Furthermore, the complexes G–
L generally feature an additional lone pair (LP) at the car-
bene carbon, which is in line with the high negative partial
charge at the carbene ligand and the usually observed nucle-
ophilic character. In contrast to this picture, all calculated
Schrock complexes possess metal-carbene s and p bonds as
reported earlier.[18,20] This is also the case for the more polar-
ized complex F, which is (in terms of charges) the system
best comparable with the complexes based on the methan-
diide ligands. Notably, complex I exhibits a larger negative
Chem. Eur. J. 2013, 19, 16729 – 16739
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16735

V. H. Gessner, M. Kaupp et al.
better described as masked methandiides with the totally
ionic resonance structure M2+···C2-.
plexes. Charge separation in the metal–carbon bond is com-
parable but yet more pronounced than in typical Schrock
systems underlining the overall nucleophilic character of
both classes of compounds. However, bond orders were
found to be significantly lower than in the Schrock com-
plexes and in most of the cases the p-bonds are strongly po-
larized to carbon (or truly absent for the Pd complex). Car-
bene complexes based on dilithio methandiides may thus be
viewed formally as Schrock-type systems, but with a more
pronounced ionic nature of the metal–carbon bond. This
leads to the conclusion that geminal dianions represent
Table S26 in the Supporting Information also provides
a graphical comparison of the electron localization function
(ELF; see the Computational Details) for compounds 3, 4,
and for the Schrock model complex F. The ELF confirms
the relatively localized p-type carbon lone-pair in 3, a more
asymmetrical one in 4, and the involvement of the corre-
ꢀ
sponding orbitals in covalent Zr C p-bonding in the
Schrock complex F. A similar picture is obtained from the
AIM studies.[24] Table 3 gives selected results of the topologi-
a kind of “extreme” Schrock-type ligands often favoring the
ꢀ
ionic resonance structure M+ CR2 as used to in textbooks
ꢀ
ꢀ
Table 3. Bond critical point (BCP) properties for the M C carbene
bond.
to explain the nucleophilic nature of Schrock complexes. In
some cases the totally ionic structure M2+···C2ꢀ is even
a more realistic description (i.e., masked dianion). Never-
theless, geminal dianions provide a unique strategy to access
Schrock-like complexes, also allowing the introduction of
late transition-metals. It will be interesting to see whether
also other methandiides than bis(phosphonium)-substituted
systems can be applied in the synthesis of the corresponding
carbene complexes. By the use of less anion-stabilizing moi-
eties, a more covalent metal–carbon interaction and thus
more “classical” Schrock complexes should be realizable.
2
1BCP
[a.u.]
5 1
e
G
[a.u.]
ACHTUNGTRNE[NUNG a.u.]
E
F
H
I
3
L
4
0.153
0.128
0.095
0.084
0.067
0.108
0.106
0.159
0.173
0.134
0.135
0.131
0.128
0.124
0.62
0.54
0.28
0.21
0.21
0.02
0.03
[a] Electron density (1BCP), Laplacian (521), and ellipticity (e).
cal analysis of some of the complexes. The ellipticity e of
ꢀ
the electron density at the M Ccarbene bond critical point
(BCP) is the most significant property, as it provides a mea-
sure of the double-bond character in the carbene complexes.
Cylindrically symmetrical s bonds are characterized by an
ellipticity value close to zero, whereas the presence of one
p-bond gives rise to nonzero ellipticities. This is seen nicely
for the Schrock complexes E and F (e>0), thus confirming
the p-bonding interaction.[18] For the methandiide-based
complexes significantly smaller ellipticities are calculated,
Experimental Section
General procedures: All experiments were carried out under a dry,
oxygen-free argon atmosphere by using standard Schlenk techniques. In-
volved solvents were dried over sodium or potassium and distilled prior
to use. H2O is distilled water. Organolithium reagents were titrated
against diphenylacetic acid prior use. 1H, 13C, 31P NMR spectra were re-
corded on Avance-500 or AMX-400 Bruker spectrometers at 228C if not
stated otherwise. All values of the chemical shift are in ppm regarding
the d scale. All spin-spin coupling constants (J) are printed in Hertz
(Hz). To display multiplicities and signal forms correctly the following
abbreviations were used: s=singlet, d=doublet, t=triplet, q=quartet,
m=multiplet, br=broad signal. GC/MS analyses were performed on
ꢀ
corroborating the small role of M C p-interactions in these
compounds.[25] This is particularly true for the palladium
complexes L and 4, which feature typical values of purely s-
bonded systems.
a
Varian 320MS-GC/MS. Diphenylchlorophosphine, cyclohexylmagne-
AHCTUNGTERGsNNUN ium chloride, bis(triphenylphosphine)palladium dichloride, zirconocene
dichloride were purchased and used without further purification.
Conclusion
One-pot synthesis of compound 1: In a 500 mL Schlenk flask posphorus
trichloride (4.08 g, 29.7 mmol) was dissolved in THF (80 mL). Under vig-
orous stirring cyclohexylmagnesium chloride ((30 mL, 60.0 mmol; 3m so-
lution in diethyl ether) were slowly added at ꢀ788C resulting in the pre-
cipitation of magnesium chloride. After 5 h at room temperature the sol-
vent was partly removed in vacuo and pentane (60 mL) was added. The
mixture was filtered by cannula transfer and the remaining solid washed
twice with additional pentane (40 mL). Subsequently, THF (40 mL) were
added to the solution followed by the addition of methylmagnesium bro-
mide (12 mL, 23.0 mmol; 3m solution in diethyl ether) at ꢀ788C. The
mixture was allowed to warm to RT and stirred overnight. Then, elemen-
tal sulfur (1.28 g, 40.0 mmol) was added and the mixture stirred for addi-
tional 3 h. The yellow solution was treated with water (50 mL) and the
mixture extracted with diethyl ether. The combined organic layers were
dried over sodium sulfate and the solvent removed in vacuo giving
a yellow solid. Purification by sublimation by using Kugelrohr distillation
(132–1378C, 1ꢆ10ꢀ3 mbar) afforded the product as colorless solid in
72% yield (5.21 g, 21.3 mmol). The spectroscopic data matches the pub-
lished data.[26]
We present the synthesis of a new unsymmetrical geminal
dianion and its application as ligand in zirconium- and palla-
dium-carbene complexes. Both complexes were found to ex-
hibit extremely shielded 13C NMR shifts, which are even
more highfield-shifted than those of related carbene com-
plexes based on dilithio methandiides. DFT analyses indicate
significant spin-orbit effects for the palladium complex. But
mostly the lower shifts of both Zr and Pd complexes com-
pared with standard Schrock carbene complexes may be at-
tributed to a combination of 1) smaller relevant energy de-
ꢀ
nominators and 2) more delocalized and polar Zr C bond-
ing.
The electronic structure of these and related complexes
has been compared to typical Fischer and Schrock com-
16736
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16729 – 16739

Synthesis and Bonding in Carbene Complexes
FULL PAPER
Synthesis of 2: n-butyllithium (4.10 mL, 6.57 mmol, 1.6 m solution in
hexane) was added to a cooled solution of 1.60 g (6.57 mmol) of 1 in
THF (40 mL) giving an orange-colored solution. After stirring for 2 h at
RT the mixture was again cooled to ꢀ788C and slowly added to a solution
of diphenylchlorophosphine (1.68 g, 7.00 mmol) in diethyl ether (60 mL).
The solution was allowed to warm to RT and stirred overnight. Subse-
quently, elemental sulfur (0.22 g, 7.00 mmol) was added and the mixture
stirred for 3 h at RT. After the addition of water (50 mL) the mixture
was extracted with diethyl ether. The combined organic layers were dried
over sodium sulfate and the solvent removed in vacuo giving a yellow oil.
Purification by flash chromatography on silica (eluent: pentane/diethyl-
ether: v/v=2:1) gave the product as a colorless solid in 67% yield (2.03g,
4.40 mmol). Rf =0.6; 1H NMR (500 MHz, CDCl3): d=1.02–1.12 (m, 6H,
CH2,cyclohexyl), 1.38–1.44 (m, 4H; CH2,cyclohexyl), 1.59–1.87 (m, 10H;
removal of the remaining solution (139 mg, 0.20 mmol; 73%). 1H NMR
(400 MHz, C6D6): d=0.93–0.97 (m, 2H, CH2,Cy), 1.09–1.22 (m, 4H,
CH2,Cy), 1.39–1.42 (m, 2H; CH2,Cy), 1.55–1.62 (m, 4H, CH2,Cy), 1.70–1.82
(m, 8H; CH2,Cy), 2.13-2.14 (m, 2H; CHP), 6.31 (d, 10H; CHCp), 7.16–7.21
(m, 6H; CHPh,meta,para), 7.96–8.02 ppm (m, 4H; CHPh,ortho); 13C NMR
(100.6, C6D6): d=26.1 (d, JCP =2.36 Hz; CH2,Cy), 26.2 (d, JCP =1.68 Hz;
CH2,Cy), 26.6 (d, JCP =1.75 Hz; CH2,Cy), 26.9 (d, 2JCP =5.30 Hz CH2,Cy),
27.1 (d, 2JCP =5.87 Hz; CH2,Cy), 29.2 (dd, 1JCP =78.7 Hz, 1JCP =68.8 Hz;
PCP), 42.8 (dd,1JCP =42.0 Hz, 3JCP =1.33 Hz; PCH), 112.8 (CCp), 128.3 (d,
2
3JCP =11.5 Hz; CHmeta), 130.1 (d, 4JCP =2.90 Hz; CHpara), 130.6 (d, JCP
=
11.6 Hz, CHortho), 141.3 ppm (dd, 1JCP =69.9 Hz, 3JCP =1.65 Hz, PCipso);
31P NMR (162.0 MHz, C6D6): d=20.1 (d, 2JPP =3.30 Hz; PPh2), 47.8 ppm
(d, 2JPP =3.30 Hz, PCy2); elemental analysis calcd (%) for C37H48P2S2Zr:
C, 62.58; H, 6.81; S 9.03; found: C, 62.17; H, 6.49; S 8.95.
2
2
CH2,cyclohexyl), 2.40–2.43 (m, 2H; CHP), 3.31 (dd, JHP =13.8 Hz, JHP
=
Synthesis of complex 4: Dianion (90 mg, 0.16 mmol) and bis(triphenyl-
phosphine)palladium dichloride (112 mg, 0.16 mmol) were suspended in
toluene (10 mL) and stirred for 3 d at RT. The obtained red suspension
was filtered through a filter cannula and the solvent removed in vacuo to
afford a red solid. The residue was taken up in toluene (8 mL) and the
solvent reduced to about 2 mL, from which the complex started to pre-
cipitate. The remaining solvent was removed, the solid washed three
times with pentane/diethyl ether (v/v=1:1) and dried in vacuo giving the
product as red solid (83 mg, 0.10 mmol; 63%). 1H NMR (400 MHz,
CD2Cl2): d=0.90-1.18 (m, 6H, CH2,Cy), 1.34–1.54 (m, 6H, CH2,Cy), 1.58–
1.69 (m, 4H; CH2,Cy), 1.70–1.80 (m, 4H, CH2,Cy), 1.90–2.02 (m, 2H;
CHP), 7.28-7.46 (m, 15H; CHPh,meta,para), 7.53–7.58 (m, 4H; CHPPh3,ortho),
7.69–7.75 ppm (m, 4H; CHPPh2,ortho); 13C NMR (125.8, CD2Cl2): d=ꢀ18.3
12.1 Hz, 2H; PCH2P), 7.46-7.52 (m, 6H; CHPh,meta,para), 7.90–7.93 ppm (m,
4
4H; CHPh,ortho); 13C NMR (125.8 MHz, CDCl3): d=25.4 (d, JCP =1.71 Hz,
PCHACHTUNGTRENNUNG(CH2CH2)2CH2), 26.0 (d, JCP =7.18 Hz; CH2), 26.1 (d, JCP =7.51 Hz;
CH2), 26.6 (d, JCP =2.60 Hz; CH2), 26.7 (d, JCP =3.41 Hz; CH2), 32.8 (dd,
1JCP =39.8 Hz, 1JCP =31.1 Hz; PCH2P), 37.8 (d, 1JCP =47.3 Hz; PCH),
128.7 (d, 3JCP =12.5 Hz; CHmeta), 129.9 (d, 2JCP =10.6 Hz, CHortho), 131.8
(d, 4JCP =3.02 Hz; CHpara), 133.4 ppm (dd, 1JCP =82.2 Hz, 3JCP =2.44 Hz,
PCipso); 31P NMR (162.0 MHz, CDCl3): d=31.4 (d, 2JPP =18.2 Hz; PPh2),
63.9 ppm (d, 2JPP =18.2 Hz; PCy2); elemental analysis calcd (%) for
C25H34P2S2: C, 65.19; H, 7.44; S 13.92; found: C, 65.30; H, 7.56; S 13.72.
Synthesis of [D2]2: Compound 2 (200 mg, 0.43 mmol) was suspended in
diethyl ether (10 mL) and cooled to ꢀ308C. Methyllithium (0.56 mL,
0.88 mmol; 1.58M in diethylether) was added upon which the solid fully
dissolved under gaseous evolution. The yellowish solution was stirred for
12 h and subsequently D2O (3 mL) were added. After 2 h stirring at
room temperature, the mixture was extracted with diethyl ether (3ꢆ
10 mL), the combined organic layers dried over sodium sulfate and the
solvent removed in vacuo. The obtained solid was subsequently purified
via column chromatography (pentane/Et2O=2:1) giving the product as
colorless solid (192 mg, 42 mmol; 97%). 1H NMR (500 MHz, CDCl3):
d=1.02–1.12 (m, 6H, CH2,cyclohexyl), 1.38-1.42 (m, 4H; CH2, cyclohex-
yl), 1.58–1.89 (m, 10H; CH2,cyclohexyl), 2.40–2.43 (m, 2H; CHP), 7.45–
7.50 (m, 6H; CHPh,meta,para), 7.90–7.95 ppm (m, 4H; CHPh,ortho); 2H NMR
(76.8 MHz, CDCl3): d=3.29 ppm (br); 13C NMR (125.8, CDCl3): d=25.4
1
1
2
(ddd, JCP =85.0 Hz, JCP =71.1 Hz, JCP =56.1 Hz; PCP), 26.1 (m; CH2,Cy),
26.4 (d, 2JCP =1.66 Hz; CH2,Cy), 26.7 (d, 2JCP =6.66 Hz; CH2,Cy), 26.8 (d,
3
2JCP =7.26 Hz CH2,Cy), 41.6 (dd,1JCP =41.9 Hz, JCP =2.13 Hz; PCH), 128.5
(d, 3JCP =11.4 Hz; CHPPh2,meta), 128.6 (d, 3JCP =11.5 Hz; CHPPh3,meta), 130.0
(d, 2JCP =11.8 Hz, CHPPh2,ortho), 130.2 (d, 4JCP =2.00 Hz; CHPPh3,para), 130.9
(d, 4JCP =3.02 Hz; CHPPh2,para), 133.1 (d, 1JCP =35.1 Hz; PCPPh3,ipso), 134.6
3
(d, 2JCP =12.6 Hz, CHPPh3,ortho), 141.2 ppm (ddd, 1JCP =72.2 Hz, JCP
=
3.04 Hz, 4JCP =1.52 Hz; PCPPh2,ipso); 31P NMR (162.0 MHz, CD2Cl2): d=
20.8 (vt, 3JPP =14.7 Hz; PPh3), 36.6 (dd, 2JPP =33.4 Hz, 3JPP =13.8 Hz;
PPh2), 70.1 ppm (dd, 2JPP =33.4 Hz, 3JPP =15.5 Hz; PCy2); 31P NMR
(162.0 MHz, C6D6): d=35.3 (t, 2JPP =15.9 Hz, 2JPP =30.0 Hz; PPh3), 36.3
(dd, 2JPP =14.2 Hz, 2JPP =38.8 Hz; PPh2), 68.3 ppm (dd, 2JPP =15.8 Hz,
2JPP =38.8 Hz; PCy2); elemental analysis calcd (%) for C43H47P3S2Pd: C,
62.43; H, 5.73; S 7.75; found: C, 61.99; H, 5.71; S 7.42.
(d, 4JCP =1.70 Hz, PCH
(CH2CH2)2CH2), 26.1 (d, JCP =7.60 Hz; PCH
CP =2.57 Hz; PCHCH2), 26.7 (d, CP =3.36 Hz; PCHCH2), 32.4 (br;
A
JCP =7.29 Hz; PCH-
A
ACHTUNGTRENNUNG
J
J
3
2
Single-crystal X-ray structure determination: Single crystals were selected
from a Schlenk flask under an argon atmosphere and covered with inert
oil (perfluoropolyalkylether). Data for compounds 2 and 2-Li2 were col-
lected on a Bruker X8Apex diffractometer, data for 3 and 4 were collect-
ed on Bruker APEX-CCD (D8 three-circle goniometer) (Bruker AXS).
Integration was conducted with SAINT and an empirical absorption cor-
rection (SADABS) was applied. The structures were solved by direct (2,
2-Li2, 3 and 4) methods (SHELXS-97) and refined by full-matrix least-
squares methods against F2 (SHELXl-97).[27] All non-hydrogen atoms
were refined with anisotropic displacement parameters. Hydrogen atoms
were placed in calculated positions and refined using the riding model.
Relevant details about the structure refinements are given Tables S1–S10
(the Supporting Information). CCDC-965733 (2) and CCDC-965734 (2-
Li2) CCDC-965735 (3), and CCDC-965736 (4) contain the supplementary
crystallographic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
PCD2P), 37.8 (m; PCH), 128.7 (d, JCP =12.5 Hz; CHmeta), 129.9 (d, JCP
=
10.6 Hz, CHortho), 131.8 (d, 4JCP =3.03 Hz; CHpara), 133.4 ppm (dd, JCP
=
1
82.1 Hz, 3JCP =2.47 Hz, PCipso); 31P NMR (162.0 MHz, CDCl3): d=31.2
(m; PPh2), 63.7 (m; PCy2); elemental analysis calcd (%) for C25H34P2S2:
C, 64.90; H, 7.84; S 13.86; found: C, 64.83; H, 7.48; S 13.78.
Synthesis of 2-Li2: Compound 2 (100 mg, 0.23 mmol) was suspended in
diethyl ether (5 mL). Methyllithium (0.32 mL, 0.50 mmol; 1.56 m in dieth-
yl ether) was then added giving a yellow solution. Under gaseous evolu-
tion the solid completely dissolved and the product started to crystallize
out of the reaction mixture. After 18 h at RT, the remaining solution was
removed and the crystalline solid dried in vacuo giving the product as
colorless solid in 80–89 % yield. 1H NMR: (400.1 MHz, C6D6): d=1.52–
2.25 (m, 20H; CH2Cy), 3.23–3.28 (m, 2H; CHCy), 7.02–7.21 (m, 6H; CHPH,
meta/para), 8.18–8.23 ppm (m, 4H; CHPh,ortho); 7Li NMR (155.5 MHz, C6D6):
d=1.9; 31P NMR (162.0 MHz, C6D6): d=35.3 (PPh2), 55.4 ppm (PCy2).
Solubility too low in benzene, toluene and THF to obtain 13C NMR data;
elemental analysis calcd (%) for C49H52Li2OP2S2: C, 63.72; H, 7.74; S
11.73; found: C, 63.46; H, 8.03; S 11.50.
Crystal data for 2: C25H34P2S2; Mr =460.58; colorless block; 0.35ꢆ0.28ꢆ
0.24 mm3; orthorhombic; space group P212121; a=9.6725(4), b=
Synthesis of complex 3: Zirconocene dichloride (80 mg, 0.27 mmol) was
dissolved in tetrahydrofurane and added to a suspension of dilithium salt
1 (150 mg, 0.27 mmol) in toluene. The suspension was stirred at room
temperature for 20 h, upon which the solution turned yellow and all
solids completely dissolved. After removal of the solvent in vacuo the
residue was taken up in toluene and filtered to remove lithium chloride.
After reduction of the solvent to about 1 mL, the product started to crys-
tallize out of the reaction mixture and was isolated as yellow solid after
12.4667(5), c=20.2529(7) ꢄ; V=2442.18(16) ꢄ3; Z=4; 1calcd
=
1.253 gcmꢀ3; m=0.359 mmꢀ1; F
wR2 =0.0560; 4280 unique reflections (q<24.99) and 262 parameters.
ACHTUNGTERN(NUGN 000)=984; T=100(2) K; R1 =0.0246 and
Crystal data for 2-Li2: C58H84Li4O2P4S4; Mr =1093.13; colorless plate;
0.21ꢆ0.21ꢆ0.06 mm3; monoclinic; space group I2/a; a=16.9670(12), b=
13.0321(9), c=28.097(2) ꢄ; b=106.860(4)8; V=5945.7(8) ꢄ3; Z=4;
1calcd =1.221 gcmꢀ3; m=0.307 mmꢀ1; F
ACTHNUTRGEN(UNG 000)=2336; T=100(2) K; R1 =
Chem. Eur. J. 2013, 19, 16729 – 16739
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16737

V. H. Gessner, M. Kaupp et al.
0.0492 and wR2 =0.1256; 5236 unique reflections (q<25.00) and 478 pa-
rameters.
versity of Wꢁrzburg for financial support. Work in Berlin was funded by
the DFG excellence cluster on “Unifying concepts in catalysis” (UniCat).
We also thank Rockwood Lithium for the supply of chemicals.
Crystal data for 3: C35H42P2S2Zr; Mr =679.97; yellow block; 0.31ꢆ0.21ꢆ
3
¯
0.17 mm ; triclinic; space group P1; a=10.1685(11), b=11.8680(13), c=
13.4916(15) ꢄ; a=93.889(2), b=96.726(2), g=91.096(2)8; V=
1612.6(3) ꢄ3; Z=2; 1calcd =1.400 gcmꢀ3; m=0.593 mm-1; F
ACTHNUGRTENUNG(000)=708; T=
[1] a) U. Klabunde, E. O. Fischer, J. Am. Chem. Soc. 1967, 89, 7141–
7142; b) R. R. Schrock, J. Am. Chem. Soc. 1974, 96, 6796–6797.
[2] For reviews, see: a) N. D. Jones, R. G. Cavell, J. Organomet. Chem.
2005, 690, 5485–5496; b) T. Cantat, N. Mꢀzailles, A. Auffrant, P. Le
Floch, Dalton Trans. 2008, 1957–1972; c) S. T. Liddle, D. P. Mills, A.
Wooles, J. Organomet. Chem. 2010, 36, 29; d) S. Harder, Coord.
Chem. Rev. 2011, 255, 1252; e) T. K. Panda, P. W. Roesky, Chem.
Soc. Rev. 2009, 38, 2782–2804; f) S. T. Liddle, D. P. Mills, A. J.
Wooles, Chem. Soc. Rev. 2011, 40, 2164–2176; g) G. R. Giesbrecht,
J. C. Gordon, Dalton Trans. 2004, 2387–2393; h) O. T. Summer-
scales, J. C. Gordon, RSC Adv. 2013, 3, 6682–6692.
173(2) K; R1 =0.0275 and wR2 =0.0715; 5664 unique reflections (q<
25.00) and 361 parameters.
Crystal data for 4: C43H47P3PdS2; Mr =827.24; red plate; 0.21ꢆ0.18ꢆ
3
¯
0.09 mm ; triclinic; space group P1; a=9.5630(19), b=14.231(3), c=
14.758(3) ꢄ; a=93.09(3), b=100.56(3), g=98.48(3)8; V=1946.1(7) ꢄ3;
Z=2; 1calcd =1.412 gcmꢀ3; m=0.738 mmꢀ1; F
ACTHNUTRGNEUNG(000)=856; T=173(2) K;
R1 =0.0286 and wR2 =0.0771; 6858 unique reflections (q<25.00) and 442
parameters.
Computational details: All calculations (except for 2-Li2, C2 symmetry)
were performed without symmetry restrictions. Starting coordinates were
obtained directly from the crystal structure analyses. All calculations
were done with the Gaussian 03 (Revision E.01) program package.[28a]
Structure optimizations were performed at the density-functional theory
level using the Becke–Perdew (BP86) functional.[29] The 6-31+G* basis
set was used for hydrogen, carbon, and lithium, the 6-311+G** basis set
for all other atoms (P, S). For palladium and zirconium the
LANL2TZ(f)[30] basis set of triple-z quality was used augmented with an
f polarization function of exponent 1.472 and 0.875, respectively.[31] Har-
monic vibrational frequency analyses were performed at the same levels
of theory to confirm that the structures were indeed minima on the po-
tential energy surface (PES). NBO analyses were carried out on the opti-
mized systems using the NBO 5.G program,[28b] interfaced to the Gaussi-
an 03 program. The wave functions were also analyzed in DGrid[32] pro-
gram by means of the electron localization function (ELF)[33] and the
quantum theory of atoms in molecules (AIM).[24] For this purpose, the
Kohn–Sham orbitals of the single point calculations were transferred to
the DGrid and the examined property was calculated on a grid with 50
points per Bohr. The ELF results were visualized in selected planes using
the standard ELF color scale and the Paraview program.[34]
[3] D. J. Mindiola, J. Scott, Nat. Chem. 2011, 3, 15–17.
[4] For examples, see: a) R. P. K. Babu, R. McDonald, S. A. Decker, M.
Klobukowski, R. G. Cavell, Organometallics 1999, 18, 4226–4229;
b) O. J. Cooper, D. P. Mills, J. McMaster, F. Moro, E. S. Davies, W.
Lewis, A. J. Blake, S. T. Liddle, Angew. Chem. 2011, 123, 2431–
2434; Angew. Chem. Int. Ed. 2011, 50, 2383–2386; c) R. G. Cavell,
R. P. K. Babu, A. Kasani, R. McDonald, J. Am. Chem. Soc. 1999,
121, 5805–5806; d) T. Cantat, L. Ricard, N. Mꢀzailles, P. Le Floch,
Organometallics 2006, 25, 6030–6038; e) R. P. Kamalesh Babu, R.
McDonald, R. G. Cavell, Chem. Commun. 2000, 481–482; f) N. D.
Jones, G. Lin, R. A. Gossage, R. McDonald, R. G. Cavell, Organo-
metallics 2003, 22, 2832–2841; g) G. Lin, N. D. Jones, R. A. Gossage,
R. McDonald, R. G. Cavell, Angew. Chem. 2003, 115, 4188–4191;
Angew. Chem. Int. Ed. 2003, 42, 4054–4057; h) T. Cantat, T. Arli-
guie, A. Noel, P. Thuꢀry, M. Ephritikhine, P. Le Floch, N. Mꢀzailles,
J. Am. Chem. Soc. 2009, 131, 963–972; i) J.-P. Tourneux, J.-C. Ber-
thet, P. Thuꢀry, N. Mꢀzailles, P. Le Floch, M. Ephritikhine, Chem.
Commun. 2010, 46, 2494–2496; j) H. Heuclin, X. F. Le Goff, N. Mꢀ-
zailles, Chem. Eur. J. 2012, 18, 16136–16144; k) M. T. Gamer, M.
Rastꢂtter, P. W. Roesky, Z. Anorg. Allg. Chem. 2002, 628, 2269–
2272; l) D P. Mills, O. J. Cooper, F. Tuna, E. J. L. McInnes, E. S.
Davies, J. McMaster, F. Moro, W. Lewis, A. J. Blake, S. T. Liddle, J.
Am. Chem. Soc. 2012, 134, 10047–10054; m) D. P. Mills, A. J.
Wooles, J. McMaster, W. Lewis, A. J. Blake, S. T. Liddle, Organome-
tallics 2009, 28, 6771–6776.
Quasirelativistic all-electron DFT calculations of the nuclear shieldings
have been performed using the Amsterdam Density Functional (ADF)[35]
program suite, employing Slater-type orbital (STO) basis sets of triple-
zeta doubly polarized (TZ2P) quality and the B3LYP[36] functional. Both
scalar and SO relativistic effects were treated by the two-component
zero-order regular approximation (ZORA).[37] Gauge-including atomic
orbitals (GIAOs)[38] were used. The computed 13C nuclear shieldings
were converted into chemical shifts (d, in ppm) relative to the shielding
of tetramethylsilane (TMS), obtained at the same computational level.
[5] M. Fustier, X. F. Le Goff, P. Le Floch, N. Mꢀzailles, J. Am. Chem.
Soc. 2010, 132, 13108–13110.
[6] a) D. P. Mills, L. Soutar, W. Lewis, A. J. Blake, S. T. Liddle, J. Am.
Chem. Soc. 2010, 132, 14379–14381; b) D. P. Mills, W. Lewis, A. J.
Blake, S. T. Liddle, Organometallics 2013, 32, 1239–1250; c) D. P.
Mills, L. Soutar, O. J. Cooper, W. Lewis, A. J. Blake, S. T. Liddle, Or-
ganometallics 2013, 32, 1252–1264.
Analyses of the nuclear shielding constants in model complexes were car-
ried out separately, in two steps: first the restricted Kohn–Sham orbitals
were generated with the Gaussian 09[39] program package, using the
BP86[40] functional. For Zr and Pd, a quasi-relativistic energy-adjusted
small-core Stuttgart-type pseudopotential (effective-core potential, ECP)
[7] T. E. Taylor, M. B. Hall, J. Am. Chem. Soc. 1984, 106, 1576–1584.
[8] a) A. Kasani, R. P. K. Babu, R. McDonald, R. G. Cavell, Angew.
Chem. 1999, 111, 1580–1582; Angew. Chem. Int. Ed. 1999, 38, 1483–
1484; b) L. Orzechowski, G. Jansen, S. Harder, Angew. Chem. 2009,
121, 3883–3887; Angew. Chem. Int. Ed. 2009, 48, 3825–3829; c) M.
Demange, L. Boubekeur, A. Auffrant, N. Mꢀzailles, L. Ricard, X.
Le Goff, P. Le Floch, New J. Chem. 2006, 30, 1745–1754; d) L. Orze-
chowski, G. Jansen, S. Harder, J. Am. Chem. Soc. 2006, 128, 14676–
14684; e) J. Vollhardt, H.-J. Gais, K. L. Lukas, Angew. Chem. 1985,
97, 607–609; Angew. Chem. Int. Ed. Engl. 1985, 24, 610–611; f) J.
Vollhardt, H.-J. Gais, K. L. Lukas, Angew. Chem. 1985, 97, 695–697;
Angew. Chem. Int. Ed. Engl. 1985, 24, 696–697; g) H.-J. Gais, J.
Vollhardt, J. Am. Chem. Soc. 1988, 110, 978–980; h) J. F. K. Mꢁller,
M. Neuburger, M. Zehnder, Helv. Chim. Acta 1997, 80, 2182–2190;
i) G. Linti, A. Rodig, H. Pritzkow, Angew. Chem. 2002, 114, 4685–
4687; Angew. Chem. Int. Ed. 2002, 41, 4503–4506; j) J.-H. Chen, J.
Guo, Y. Li, C.-W. So, Organometallics 2008, 27, 4617; k) O. J.
Cooper, A. J. Wooles, J. McMaster, W. Lewis, A. J. Blake, S. T.
Liddle, Angew. Chem. 2010, 122, 5702–5705; Angew. Chem. Int. Ed.
2010, 49, 5570–5573.
with corresponding (8s7p6d)/ACHTUNGTRNEUNG
[6s5p3d] GTO valence basis set was used.[41]
Ligand atoms were treated by Huzinaga–Kutzelnigg-type IGLO-III basis
sets.[41] Subsequently, the Kohn–Sham orbitals were transferred to the
MAG-ReSpect[42] property package by interface routines. The nuclear
shielding constants were calculated either by using independent gauges
for localized molecular orbitals (IGLO)[41] in combination with a Pipek-
Mezey[43] localization procedure (together with an excitation analysis as
implemented in MAG-ReSpect program package) or at GIAO level. The
visualization of localized and canonical molecular orbitals was done with
Molekel.[44]
Acknowledgements
V.H.G. acknowledges the Deutsche Forschungsgemeinschaft for an
Emmy-Noether grant (DA-1402/1-1) as well as the Fonds der Chemi-
schen Industrie, the Alexander von Humboldt Foundation and the Uni-
[9] P. Schrçter, V. H. Gessner, Chem. Eur. J. 2012, 18, 11223–11227.
16738
www.chemeurj.org
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 16729 – 16739

Synthesis and Bonding in Carbene Complexes
FULL PAPER
[10] V. H. Gessner, Organometallics 2011, 30, 4228–4231.
[11] H. Heuclin, M. Fustier-Boutignon, S. Ying-Fu Ho, X.-F. Le Goff, S.
Carenco, C.-W. So, N. Mꢀzailles, Organometallics 2013, 32, 498–508.
[12] Compound 1 has been synthesized before through an alternative re-
action route: E. Payet, A. Auffrant, X. F. Le Goff, P. Le Floch, J. Or-
ganomet. Chem. 2010, 695, 1499–1506.
[13] T. Cantat, L. Ricard, P. Le Floch, N. Mꢀzailles, Organometallics
2006, 25, 4965–4976.
[14] a) T. Stey, D. Stalke in The Chemistry of Organolithium Compounds
(Eds.: Z. Rappoport, I. Marek),Wiley, Chichester, 2004, p. 47–120;
b) V. H. Gessner, C. Dꢂschlein, C. Stohmann, Chem. Eur. J. 2009,
15, 3320–3335.
[15] T. Cantat, N. Mꢀzailles, Y. Jean, P. Le Floch, Angew. Chem. 2004,
116, 6542–6545; Angew. Chem. Int. Ed. 2004, 43, 6382–6385.
[16] H. V. Huynh, Y. Han, R. Jothibasu, J. A. Yang, Organometallics
2009, 28, 5395–5404.
[17] Multinuclear NMR (Ed.: J. Mason), Plenum Press, New York, 1987.
[18] S. F. Vyboishchikov, G. Frenking, Chem. Eur. J. 1998, 4, 1428–1438.
[19] For examples of further calculations on Schrock- and Fischer car-
bene complexes, see: a) G. Occhipinti, V. R. Jensen, Organometallics
2011, 30, 3522–3529; b) G. Occhipinti, H.-R. Bjørsvik, V. R. Jensen,
J. Am. Chem. Soc. 2006, 128, 6952.
[20] a) G. Frenking, M. Solꢇ, S. F. Vyboishchikov, J. Organomet. Chem.
2005, 690, 6178–6204; b) M. Cases, G. Frenking, M. Duran, M. Solꢇ,
Organometallics 2002, 21, 4182–4191; c) D. Nemcsok, K. Wichmann,
G. Frenking, Organometallics 2004, 23, 3640–3646.
[21] The palladium–carbene complex differs from all other complexes in
that way that the HOMO displays an antibonding p-type interac-
tion. Because of the filled d orbitals, no free orbital is available to
allow for a bonding interaction.
[22] C. Elschenbroich, Organometallchemie, 4th ed., Teubner, Stuttgart,
2003, p.313.
[23] O. J. Cooper, D. P. Mills, J. McMaster, F. Tuna, E. J. L. McInnes, W.
Lewis, A. J. Blake, S. T. Liddle, Chem. Eur. J. 2013, 19, 7071–7083.
[24] R. F. W. Bader, Atoms in Molecules: A Quantum Theory, Oxford
University Press, Oxford, 1990.
[25] Y. Smurnyy, C. Bibal, M. Pink, K. G. Caulton, Organometallics 2005,
24, 3849–3855.
[26] L. Mayer, Chem. Ber. 1961, 94, 3056.
[27] a) SAINT v7.68A in Bruker APEX v2011.9, Bruker AXS Inst. Inc.,
Madison, WI, USA, 2008; b) G. M. Sheldrick, SADABS 2008/2;
Universitꢂt Gçttingen, Gçttingen, Germany, 2008; c) G. M. Shel-
drick, Acta Cryst., Sect. A 1990, 46, 467–473; d) G. M. Sheldrick,
Acta Cryst., Sect. A 2008, 64, 112–122.
hold, Theoretical Chemistry Institute, University of Wisconsin, Mad-
ison, 2001, http://www.chem.wisc.edu/~nbo5.
[29] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100; b) P. Perdew,
Phys. Rev. B 1986, 33, 8822–8824.
[30] L. E. Roy, P. J. Hay, R. L. Martin, J. Chem. Theory Comput. 2008, 4,
1029–1031.
[31] A. W. Ehlers, M. Bçhme, S. Dapprich, A. Gobbi, A. Hçllwarth, V.
Jonas, K. F. Kçhler, R. Stegman, A. Veldkamp, G. Frenking, Chem.
Phys. Lett. 1993, 208, 111–114.
[32] M. Kohout, DGrid, version 4.6, Radebeul, 2011.
[33] a) A. D. Becke, K. E. Edgecombe, J. Chem. Phys. 1990, 92, 5397–
5403; b) A. Savin, O. Jepsen, J. Flad, O. K. Andersen, H. Preuss,
H. G. Vonschnering, Angew. Chem. 1992, 104, 186–188; Angew.
Chem. Int. Ed. Engl. 1992, 31, 187–188; c) M. Kohout, A. Savin, Int.
J. Quantum Chem. 1996, 60, 875–882.
[34] Paraview, version 3.8, Kitware Inc., Clifton Park, New York, 2010;
available from http://www.paraview.org.
[35] Amsterdam Density Functional (ADF), version 2012.01, SCM, The-
oretical Chemistry, Vrije Univeriteit, Amsterdam, Netherlands,
2012; available from http://www.scm.com.
[36] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) C. T. Lee,
W. T. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[37] a) S. K. Wolff, T. Ziegler, J. Chem. Phys. 1998, 109, 895–905;
b) S. K. Wolff, T. Ziegler, E. van Lenthe, E. J. Baerends, J. Chem.
Phys. 1999, 110, 7689–7698. For a recent survey of relativistic DFT
calculations of NMR parameters, see also: J. Autschbach, S. Zheng,
Annu. Rep. NMR Spectrosc. 2009, 67, 1–95.
[38] K. Wolinski, J. F. Hinton, P. Pulay, J. Am. Chem. Soc. 1990, 112,
8251–8260.
[39] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X.
Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.
Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, ꢈ.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian,
Inc., Wallingford CT, 2009.
[28] a) Gaussian 03, Revision E.01, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Mont-
gomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004; b) NBO 5.G, E. D. Glendening, J. K. Badenhoop,
A. E. Reed, J. E. Carpenter, J. A. Bohmann, C. M. Morales, F. Wein-
[40] a) D. Andrae, U. Haeussermann, M. Dolg, H. Stoll, H. Preuss,
Theor. Chim. Acta 1990, 77, 123; b) J.M.L. Martin, A. Sundermann,
J. Chem. Phys. 2001, 114, 3408.
[41] W. Kutzelnigg, U. Fleischer, M. Schindler, in NMR Basic Principles
and Progress, Vol. 23: Deuterium and Shift Calculation (Eds.: P.
Diehl, E. Fluck, H. Gꢁnther, R. Kosfeld, J. Seelig), The IGLO-
Method: Ab Initio Calculation and Interpretation of NMR Chemical
Shifts and Magnetic Susceptibilities, Springer, Berlin, 1991, pp. 165–
262.
[42] V. G. Malkin, O. L. Malkina, R. Reviakine, A. V. Arbuznikov, M.
Kaupp, B. Schimmelpfennig, I. Malkin, M. Repisky, S. Komorovsky,
P. Hrobarik, E. Malkin, T. Helgaker, K. Ruud, MAG-ReSpect, ver-
sion 2.1, 2005.
[43] J. Pipek, P. G. Mezey, J. Chem. Phys. 1989, 90, 4916.
[44] U. Varetto, Molekel, version 5.4, Swiss National Supercomputing
Center, Manno, Switzerland; available from: http://molekel.cscs.ch.
Received: August 6, 2013
Published online: October 21, 2013
Chem. Eur. J. 2013, 19, 16729 – 16739
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
16739