Reactions of a quintuply bonded chromium dimer with alkynes

Article abstract of DOI:10.1039/c1cc15218a

Quintuply bonded [^HL^iPrCr]₂ reacts with alkynes RCCR (R = Me, Et, Ph, CF₃) to form exclusively 1:1 adducts [^HL^iPrCr]₂(RCCR). All products feature relatively short Cr-Cr distances (1.919-1.962 A) and elongated C-C bonds (1.315-1.436 A), consistent with [2+2] cycloaddition reactions. The hydrocarbon adducts are 4-membered metallacycles, the bridging alkynes of which are progressively skewed with respect to the Cr-Cr axis. In contrast, perfluoroalkyne adds across the metal ligand moiety. The Royal Society of Chemistry 2011.

Full text of DOI:10.1039/c1cc15218a

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 12191–12193
www.rsc.org/chemcomm
COMMUNICATION
Reactions of a quintuply bonded chromium dimer with alkynesw
Jingmei Shen,a Glenn P. A. Yap,a Jan-Philipp Wernerb and Klaus H. Theopold*a
Received 22nd August 2011, Accepted 28th September 2011
DOI: 10.1039/c1cc15218a
Quintuply bonded [HLiPrCr]2 reacts with alkynes RC CR
R
(R = Me, Et, Ph, CF3) to form exclusively 1 : 1 adducts
[HLiPrCr]2(RCCR). All products feature relatively short Cr–Cr
˚
distances (1.919–1.962 A) and elongated C–C bonds
˚
(1.315–1.436 A), consistent with [2+2] cycloaddition reactions.
Scheme 1 Synthesis of [HLiPrCr]2(RCCR) (R = Me (2a), Et (2b),
The hydrocarbon adducts are 4-membered metallacycles, the
bridging alkynes of which are progressively skewed with respect
to the Cr–Cr axis. In contrast, perfluoroalkyne adds across the
metal ligand moiety.
Ph (2c).
characterized by the X-ray diffraction; the molecular structure
of representative 2a is depicted in Fig. 1, along with selected
interatomic distances and angles.
Inspection of the structure reveals that 2a (and similarly
2b and 2c, see ESIw) does not adopt the typical perpendicular
(i.e. m2-Z2:Z2) bonding mode of the alkyne, which is generally
thought to maximize alkyne–metal interaction and is favored
by Kempe’s analogues.11 Rather, the structure more closely
Element–element multiple bonds are functional groups of
heightened reactivity, and multiple bonds between metals are
no exception.1 Until fairly recently, the latter were for all
intents and purposes restricted to bond orders ranging from
2–4,2 but the discovery by Power et al. of the first isolable
molecule featuring a formal quintuple bond between two
chromium atoms, i.e., Ar0CrCrAr0,3 was followed by several
reports of dinuclear chromium compounds sporting ever
shorter metal–metal distances and laying claim to quintuple
bonds between two d5 ions.4 Whereas the race toward ‘fusing’
two metal atoms has seemingly abated for now (the current
record appears to be 1.73 A),5,6 there remains the elucidation
of the reactivity of this new kind of bond, which is only now
beginning.7–9 Herein we describe a series of compounds resulting
from pericyclic reactions between quintuply bonded [HLiPrCr]2
(1, where HLiPr = Ar–NQC(H)–(H)CQN–Ar, with Ar =
2,6-diisopropylphenyl)10 and internal alkynes.
approximates
a binuclear metallacycle—i.e., a coplanar
m2-Z1:Z1-C2R2 geometry such as would result from a formal
[2+2] cycloaddition between a quintuple and a triple bond,
generating a polyunsaturated four-membered ring. The rele-
vant bond lengths are in accord with that view; thus the
C–C bond length of 1.326(5) A is exactly what might be
Treatment of Et2O solutions of 1 with excess of 2-butyne,
3-hexyne, or diphenylacetylene at room temperature caused
instant color changes from green to deep purple. Standard
work-up of the reactions yielded the simple adducts
[HLiPrCr]2(RCCR) (R = Me (2a), Et (2b), Ph (2c), see
Scheme 1). No further reaction of 2a–c with excess alkyne
was observed. The air-sensitive complexes 2a–c have been
a Department of Chemistry and Biochemistry, Center for Catalytic
Science and Technology, University of Delaware, Newark,
DE 19711, USA. E-mail: theopold@udel.edu;
Fig. 1 The molecular structure of 2a with thermal ellipsoids at the 30%
probability level; H-atoms and ligand isopropyl groups have been omitted
for clarity. Selected distances [A] and angles [1]: Cr1–Cr2 1.9248(7),
Cr1–N1 1.927(2), Cr1–N3 1.933(2), Cr2–N2 1.901(2), Cr2–N4 1.903(2),
Cr1–C55 1.965(3), Cr2–C54 2.039(3), C55–C54 1.326(5), N1–C1 1.366(4),
C1–C2 1.351(4), N2–C2 1.372(4), N3–C27 1.372(4), C27–C28 1.356(4),
N4–C28 1.368(4), N1–Cr1–Cr2 104.15(7), N1–Cr1–N3 128.395(10),
Cr2–Cr1–N3 103.67(7), N1–Cr1–C55 118.453(12), Cr2–Cr1–C55
83.94(10), N3–Cr1–C55 106.98(12), N2–Cr2–N4 139.21(10), N2–Cr2–Cr1
105.57(8), N4–Cr2–Cr1 103.24(7), N2–Cr2–C54 109.52(12), N4–Cr2–C54
106.75(12), Cr1–Cr2–C54 71.16(9).
Fax: +1 302 831-6335; Tel: +1 302 831-1546
b Institut fur Anorganische und Angewandte Chemie,
Universitat Hamburg, Hamburg, Germany.
¨
¨
E-mail: jan-philipp.werner@chemie.uni-hamburg.de;
Tel: +49 (0)40 42838 3112
w Electronic supplementary information (ESI) available: Synthesis
and characterization of 2a–c and 3; crystal structures of 2b and 2c;
details of the calculations. CCDC 839343–839347. For ESI and
crystallographic data in CIF or other electronic format see DOI:
10.1039/c1cc15218a
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 12191–12193 12191

View Online
expected of a carbon–carbon double bond (cf. 1.339(1) A in
ethylene)12 and the Cr–Cr distance of 1.9248(7) falls into the
range of ‘supershort’ chromium–chromium quadruple bonds
(Cr–Cr o 2.0 A).2 Indeed, this elongation of the two multiple
bonds in return for the formation of two new Cr–C single
bonds (at 1.965(3) and 2.039(3) A) provides independent
chemical evidence for the presence of a quintuple metal–
metal bond in 1 (Cr–Cr = 1.8028(9) A).10 We note that the
backbone bond lengths of the diimine ligands in 1 and 2a do
not appreciably differ; thus the two electrons used to reduce
the alkyne to an alkenediyl ligand must have come from the
chromium atoms rather than from the redox non-innocent
ligands; the latter maintain radical anion character throughout
the reaction.13,14 If the formal oxidation state of chromium in
2a is Cr(II), as seems reasonable, it must have been Cr(I) in 1,
consistent with our original assignment of a quintuple bond
between two d5-ions.
remains one singlet down to ꢀ90 1C. This too may simply
be an accidental degeneracy; alternatively, it would suggest
that rotation of the 2-butyne ligand is a separate, very fast
process with a much lower barrier, which cannot be frozen out
even at ꢀ90 1C. Control experiments established that neither
alkyne dissociation, nor alkyne exchange occurs in solution on
a chemical time scale.
The electronic structures of model complexes 2a0–c0 were
examined by DFT calculations.w All structure optimizations
were performed at the BLYP/def2-SVP (LANL08(f) for the
metal atoms) level using molecules in which the 2,6-diisopropyl-
phenyl substituents had been replaced by hydrogen atoms.
Geometry optimizations on 1 gave bond distances that were
in good agreement with the X-ray structures. All reported
structural parameters refer to singlet ground state structures.
Bonding analyses were performed by means of natural bond
orbital (NBO) analysis and natural population analysis
(NPA).17,18
On closer inspection, the structures of 2a–c reveal some
additional detail. The alkyne centroids are slightly closer to
one chromium, being displaced some way toward a putative
m2-Z1:Z2-C2R2 geometry. To wit, the intraannular distances in
2a are: Cr1–C54, 2.308(3) A; Cr2–C55, 2.603(4) A, placing the
alkyne somewhat closer to Cr1. Furthermore, the Cr2C2 cores
of the molecules are not perfectly planar. For example, the
CQC bond of 2a is twisted from coplanarity with the Cr–Cr
bond by 23.71. This twist angle increases with the steric bulk of
the alkyne substituents (2b: 35.81 and 2c: 41.71); thus steric
interactions with the ligand isopropyl groups are at least
partially responsible for this torsion. There may, however,
also be an electronic contribution; a second order Jahn–Teller
distortion, resulting in mixing of metal–metal bonding MOs
with an alkyne p*-orbital, has been proposed to account for
this type of skew.15,16 In the solid state, molecules of 2a possess
no symmetry whatsoever; however, the deviation from C2v
symmetry is slight and may be anticipated to give rise to
fluxional behavior.
The calculated bond orders for the chromium–chromium
bonds of the alkyne adducts are 3.43 (2a0), 3.22 (2b0) and 3.56
(2c0), to be compared with the value of 4.59 for precursor 10.
Concomitantly, the erstwhile CRC triple bonds now have
bond orders of 1.98, 1.90 and 1.90, respectively. These reduced
values are consistent with a formal oxidative addition of the
CRC triple bonds to the Cr–Cr quintuple bond, resulting in
the formation of Cr–Cr quadruple bonds and CQC double
bonds. The natural population analysis for 2a0 indicates that
the sum of partial charges on the chromium atoms increases
from +1.022 to +1.231, whereas the bound MeCQCMe
fragment picks up a negative charge of ꢀ0.412. This indicates
the charge transfer from the Cr–Cr multiple bonds to CQC
bonds. We also note diminution of the total negative charge
on the ligand N atoms from ꢀ3.241 to ꢀ2.965. There is thus
some charge transfer from the a-dimine ligands to the alkyne.
It was of some interest to explore whether a significant
increase in the driving force of the reaction might change its
course. Given the net electron transfer from the chromium
complex to the alkynes, a more electron deficient alkyne would
probe this question. Accordingly, 1 was exposed to an excess
of hexafluoro-2-butyne; upon addition of CF3CRCCF3 at
room temperature, the color of the solution immediately
changed from green to deep purple again, but turned to red
after about one hour. 1H NMR spectroscopy of the purple
intermediate 3 showed it to be diamagnetic. To isolate 3, the
reaction was allowed to proceed for only 20 minutes, where-
upon single crystals were grown from toluene at ꢀ30 1C.
To our surprise, the structure determination of 3 (see Fig. 2)
revealed a connectivity much different from 2a–c. The fluori-
nated alkyne has been added across a metal–ligand moiety,
resulting in the formation of both carbon–carbon and carbon–
chromium bonds. The functionalized ligand has been
transformed into a localized iminoamide. While the Cr–Cr
distance of 3 (1.9615(6) A) is only marginally longer than those
of 2a–c, the C–C bond of the alkyne has been elongated to
1.436(3) A, indicating much more extensive reduction. In
transition metal chemistry, the compounds most closely
The 1H NMR spectra of 2b and 2c in C6D6 are those
expected of diamagnetic complexes, featuring a multitude of
sharp resonances between 0 and 8 ppm. However, the room
temperature spectrum of 2a exhibited broad resonances, which
is presumably caused by a low barrier, fluxional process. Due
to the asymmetric structure of its N4Cr2C2 core, 2a is expected
to exhibit resonances of eight unique iPr–CH3 and four unique
iPr–CH (while still allowing for fast rotation about the N–CAr
1
bonds). Accordingly, variable temperature H NMR spectra
in toluene-d8 between ꢀ40 1C and 30 1C showed a coalescence
phenomenon resulting in six isopropyl methyl resonances
(in a 1 : 1 : 1 : 1 : 2 : 2 intensity ratio) and three methine signals
(in a 1 : 1 : 1 ratio) at low temperature. Presumably, two pairs
of the closely spaced iPr–Me resonances are accidentally
degenerate and the missing fourth methine resonance may be
obscured by the residual toluene resonance (at 2.09 ppm). In
other words, the low temperature spectrum is consistent with
the solid-state structure of 2a. However, the activation barrier
for the fluxional process—via a twist of the propeller shaped
molecule about the Cr–Cr bond, combined with a subtle slide
of the alkyne—is rather low. The coalescence temperature of
related to 3 are Fruhauf’s products of 1,3 dipolar additions
¨
across M–XQC bonds (X = O, S, N).19–25 In main group
chemistry, the reversible addition of alkynes to dinuclear
ca. 20 1C is commensurate with DGz E 15 kcal molꢀ1
.
Curiously, we note that the 2-butyne methyl resonance
c
12192 Chem. Commun., 2011, 47, 12191–12193
This journal is The Royal Society of Chemistry 2011

View Online
In summary, we have explored the alkyne chemistry of a
quintuply bonded Cr dimer, [HLiPrCr]2 (1). With electron rich
internal alkynes forms [2+2] cycloaddition products
1
[HLiPrCr]2(m2-Z1:Z1-C2R2), i.e., 4-membered dimetallacycles
retaining a Cr–Cr quadruple bond. In contrast, electron poor
CF3CRCCF3 adds across the metal ligand bond of 1. This
dichotomy reflects the unique nature of HOMO of the starting
material, which is ligand based. We are currently investigating
the reactivity of 1 with other unsaturated organic molecules.
This work was supported by a grant from the US National
Science Foundation (CHE-0911081 to KHT)
Notes and references
1 P. P. Power, Acc. Chem. Res., 2011, 44, 627–637.
2 F. A. Cotton, Multiple Bonds Between Metal Atoms, Wiley,
New York, 1982.
3 T. Nguyen, A. D. Sutton, S. Brynda, J. C. Fettinger, G. J. Long
and P. P. Power, Science, 2005, 310, 844–847.
Fig. 2 The molecular structure of 3 with thermal ellipsoids at the
30% probability level; H-atoms, solvent molecules and ligand isopropyl
groups have been omitted for clarity. Selected distances [A] and
angles [1]: Cr1–Cr2 1.9615(6), Cr1–N1 1.9157(19), Cr1–N4 2.0883(18),
Cr2–N2 1.901(2), Cr2–N3 1.898(2), Cr1–C54 2.008(2), Cr1–C55 2.112(2),
4 F. R. Wagner, A. Noor and R. Kempe, Nat. Chem., 2009, 1,
529–536.
5 A. Noor and R. Kempe, Chem. Rec., 2010, 10, 413–416.
6 M. Kaupp, A. Noor, G. Glatz, R. Muller, S. Demeshko and
R. Kempe, Z. Anorg. Allg. Chem., 2009, 635, 1149–1152.
7 A. Noor, G. Glatz, R. Muller, M. Kaupp, S. Demeshko and
R. Kempe, Nat. Chem., 2009, 1, 322–325.
Cr2–C54
1.987(2),
C55–C54
1.436(3),
C27–C55
1.553(3),
N1–C1 1.372(3), C1–C2 1.353(3), N2–C2 1.371(3), N3–C27 1.476(3),
C27–C28 1.501(3), N4–C28 1.282(3), N1–Cr1–Cr2 106.99(6),
N1–Cr1–C54 114.75(9), Cr2–Cr1–C54 60.07(6), N1–Cr1–N4 128.55(7),
Cr2–Cr1–N4 103.36(5), C54–Cr1–N4 116.28(8), N1–Cr1–C55 146.10(8),
Cr2–Cr1–C55 81.44(6), C54–Cr1–C55 40.69(9), N4–Cr1–C55 78.37(8),
N3–Cr2–N2 156.62(8), N3–Cr2–Cr1 99.44(6), N2–Cr2–Cr1 102.88(6),
N3–Cr2–C54 84.70(9), N2–Cr2–C54 112.21(9), Cr1–Cr2–C54 61.14(6).
8 P. P. Power, C. B. Ni, B. D. Ellis and G. J. Long, Chem. Commun.,
2009, 2332–2334.
9 R. Kempe, C. Schwarzmaier, A. Noor, G. Glatz, M. Zabel,
A. Y. Timoshkin, B. M. Cossairt, C. C. Cummins and
M. Scheer, Angew. Chem., Int. Ed., 2011, 50, 7283–7286.
10 K. A. Kreisel, G. P. A. Yap, O. Dmitrenko, C. R. Landis and
K. H. Theopold, J. Am. Chem. Soc., 2007, 129, 14162–14163.
11 R. Kempe, A. Noor, E. S. Tamne, S. Qayyum and T. Bauer,
Chem.–Eur. J., 2011, 17, 6900–6903.
12 J. L. Duncan, D. Vanlerbe and I. J. Wright, J. Mol. Spectrosc.,
1972, 42, 463–477.
13 K. A. Kreisel, G. P. A. Yap and K. H. Theopold, Inorg. Chem.,
2008, 47, 5293–5303.
14 M. Ghosh, S. Sproules, T. Weyhermuller and K. Wieghardt, Inorg.
Chem., 2008, 47, 5963–5970.
15 M. J. Calhorda and R. Hoffmann, Organometallics, 1986, 5,
2181–2187.
16 M. H. Chisholm and M. A. Lynn, J. Organomet. Chem., 1998, 550,
141–150.
17 F. Weinhold and E. D. Glendening, J. Comput. Chem., 1998, 19,
593–609.
gallium complexes of a chelating diamide provides a recent
analogue.26 However, what makes 3 unique is the apparent
retention of its metal–metal multiple bond.
DFT calculation on 30 confirmed that its Cr–Cr bond order
is 3.17, similar to those in compounds 2a–c0. However, the
bond order of the alkyne derived C–C bond is 1.12, much
reduced from the earlier examples and close to a C–C single
bond, in accord with the longer bond distance. In the same
vein, the CF3CRCCF3 fragment accumulates a negative
partial charge of ꢀ0.604 at the central carbons, which significantly
exceeds that of MeCRCMe in 2a0. Interestingly, the sum of
NPA charge of Cr (1.036) hardly differs from that of 1; instead
the negative partial charge of the ligand N atoms decreases
from ꢀ3.241 (in 10) to ꢀ2.883 (in 3). These numbers support
the view that 3 features a disproportionate transfer of charge
from the ligands to the alkyne, leaving the chromium little
changed.
18 F. Weinhold and E. D. Glendening, J. Comput. Chem., 1998, 19,
610–627.
19 R. Siebenlist, H. W. Fruhauf, K. Vrieze, W. J. J. Smeets and
A. L. Spek, Organometallics, 2002, 21, 5628–5641.
20 R. Siebenlist, H. W. Fruhauf, H. Kooijman, N. Veldman,
A. L. Spek, K. Goubitz and J. Fraanje, Inorg. Chim. Acta, 2002,
327, 66–89.
Calculations were also used to address the change in product.
Full optimizations of 2a and 3 and their reaction products
were performed at the BLYP/def2-SVP/LAN2L08(f) level
using the actual, non-curtailed molecules. In both cases, the
[2+2] cycloaddition product is more stable than the product
of the ligand functionalization. However, for 2-butyne the
energy difference is 13.4 kcal molꢀ1, whereas it decreases to a
scant 4.8 kcal molꢀ1 for hexafluoro-2-butyne. We suggest that
the greater exothermicity of the latter reaction leads to an
earlier transition state, and that frontier orbital control of the
reaction path (the HOMO of 1 is entirely ligand based) asserts
itself.
21 R. Siebenlist, M. de Beurs, N. Feiken, H. W. Fruhauf, K. Vrieze,
H. Kooijman, N. Veldman, M. T. Lakin and A. L. Spek, Organo-
metallics, 2000, 19, 3032–3053.
22 N. Feiken, P. Schreuder, R. Siebenlist, H. W. Fruhauf, K. Vrieze,
H. Kooijman, N. Veldman, A. L. Spek, J. Fraanje and K. Goubitz,
Organometallics, 1996, 15, 2148–2169.
23 M. Vanwijnkoop, R. Siebenlist, J. M. Ernsting, P. P. M. Delange,
H. W. Fruhauf, E. Horn and A. L. Spek, J. Organomet. Chem.,
1994, 482, 99–109.
24 M. Vanwijnkoop, R. Siebenlist, P. P. M. Delange, H. W. Fruhauf,
K. Vrieze, W. J. J. Smeets and A. L. Spek, Organometallics, 1993,
12, 4172–4181.
25 H. W. Fruhauf, Coord. Chem. Rev., 2002, 230, 79–96.
¨
26 I. L. Fedushkin, A. S. Nikipelov and K. A. Lyssenko, J. Am.
Chem. Soc., 2010, 132, 7874.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 12191–12193 12193