Correlations and Contrasts in Homo- and Heteroleptic Cyclic (Alkyl)(amino)carbene-Containing Pt0 Complexes

Article abstract of DOI:10.1002/chem.201502048

An improved synthetic route to homoleptic complex [Pt(CAAC^Me)₂] (CAAC=cyclic (alkyl)(amino)carbenes) and convenient routes to new heteroleptic complexes of the form [Pt(CAAC^Me)(PR₃)] are presented. Although the homoleptic complex was found to be inert to many reagents, oxidative addition and metal-only Lewis pair (MOLP) formation was observed from one of the heteroleptic complexes. The spectroscopic, structural, and electrochemical properties of the zero-valent complexes were explored in concert with density functional theory (DFT) and time-dependent density functional theory (TD-DFT) calculations. The homoleptic [Pt(CAAC)₂] and heteroleptic [Pt(CAAC)(PR₃)] complexes were found to be similar in their spectroscopic and structural properties, but their electrochemical behavior and reactivity differ greatly. The unusually strong color of the CAAC-containing Pt⁰ complexes was investigated by TD-DFT calculations and attributed to excitations into the LUMOs of the complexes, which are predominantly composed of bonding π interactions between Pt and the CAAC carbon atoms. An improved synthetic route to homoleptic complex [Pt(CAACMe)₂] and convenient routes to new heteroleptic complexes of the form [Pt(CAACMe)(PR₃)] are presented. The reactivity, spectroscopic, structural, and electrochemical properties of the zero-valent complexes were explored in concert with DFT and time-dependent DFT calculations. The homoleptic [Pt(CAAC)₂] and heteroleptic [Pt(CAAC)(PR₃)] complexes were found to be similar in their spectroscopic and structural properties, but their electrochemical behavior and reactivity differ greatly (see figure).

Full text of DOI:10.1002/chem.201502048

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
DOI: 10.1002/chem.201502048
Full Paper
&
Synthetic Methods |Hot Paper|
Correlations and Contrasts in Homo- and Heteroleptic Cyclic
(Alkyl)(amino)carbene-Containing Pt0 Complexes
Nicole Arnold, Holger Braunschweig,* Peter B. Brenner, Mehmet Ali Celik, Rian D. Dewhurst,
Martin Haehnel, Thomas Kramer, Ivo Krummenacher, and Todd B. Marder[a]
Abstract: An improved synthetic route to homoleptic com-
plex [Pt(CAACMe)2] (CAAC=cyclic (alkyl)(amino)carbenes) and
convenient routes to new heteroleptic complexes of the
form [Pt(CAACMe)(PR3)] are presented. Although the homo-
leptic complex was found to be inert to many reagents, oxi-
dative addition and metal-only Lewis pair (MOLP) formation
was observed from one of the heteroleptic complexes. The
spectroscopic, structural, and electrochemical properties of
the zero-valent complexes were explored in concert with
density functional theory (DFT) and time-dependent density
functional theory (TD-DFT) calculations. The homoleptic
[Pt(CAAC)2] and heteroleptic [Pt(CAAC)(PR3)] complexes were
found to be similar in their spectroscopic and structural
properties, but their electrochemical behavior and reactivity
differ greatly. The unusually strong color of the CAAC-con-
taining Pt0 complexes was investigated by TD-DFT calcula-
tions and attributed to excitations into the LUMOs of the
complexes, which are predominantly composed of bonding
p interactions between Pt and the CAAC carbon atoms.
Introduction
The isolation of stable singlet carbenes, credited to the groups
of Bertrand and Arduengo, has dramatically broadened the
range of strongly donating ligands available to researchers,
and has led to stunning advances in inorganic, organometallic,
and main-group chemistry, as well as the fields of homogene-
ous and organocatalysis.[1] The field of stable carbenes is per-
haps best represented by their most known sub-class, the so-
called N-heterocyclic carbenes (NHCs, I; Figure 1). However,
many other types exist. Of these, cyclic (alkyl)(amino)carbenes
(CAACs, II), a type of stable carbene pioneered by the group of
Bertrand a decade ago, are arguably the class currently experi-
encing the most rapid growth in interest.[2]
Although the utility of NHCs is based mostly on their strong,
consistent, and usually predictable electron-donating ability,
CAACs have found a fascinating niche as donors that can
adapt their bonding to suit their environment. This bonding
can lie anywhere on the continuum from pure donor to full
multiple (quinoidal) bonding, depending on the extent of p ac-
ceptance by the carbene carbon’s empty p orbital. This adapta-
bility has been exploited very recently in the synthesis of com-
plexes of the form [M(CAAC)2]n+ (M=Group 7–12 metals; III;
[a] N. Arnold, Prof. Dr. H. Braunschweig, Dr. P. B. Brenner, Dr. M. A. Celik,
Dr. R. D. Dewhurst, Dr. M. Haehnel, Dr. T. Kramer, Dr. I. Krummenacher,
Prof. Dr. T. B. Marder
Institut für Anorganische Chemie
Julius-Maximilians-Universität Würzburg
Am Hubland, 97074 Würzburg (Germany)
E-mail: h.braunschweig@uni-wuerzburg.de
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201502048.
Figure 1. Top: Generalized NHC and CAAC bases. Bottom: Isolated homolep-
tic transition-metal complexes of the form [M(CAAC)2]n+
.
Chem. Eur. J. 2015, 21, 12357 – 12362
12357
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Full Paper
Figure 1) by a number of groups working both separately and
collaboratively, those of Bertrand, Roesky, Frenking, Dittrich,
Kaim, Parameswaran, Neidig, Deng, Meyer, Ungur, Stalke, Dalal,
Peters, and Ackermann.[3–8] The sheer quantity of work on com-
plexes of the form [M(CAAC)2]n+ published since the beginning
of 2013 (to say nothing of the equally unusual chemistry of
main-group CAAC adducts[2f]) is astounding and reflective of
the unique abilities of these ligands.
In these studies, CAAC ligands have allowed the surprising
synthesis of complexes, which at first glance contain transition
metals with uncommon oxidation states (e.g., Cu0, Ag0, Au0).
These feats can largely be attributed to two key properties of
the [M(CAAC)2]n+ motif: 1) the bulky CAAC ligands have the
necessary steric bulk to protect reactive metal centers; and
2) the two relatively low-energy empty carbon-based p orbi-
tals, which together can accept one or two residual electrons
from high-energy orbitals on the central metal. In other words,
otherwise unfavorable metal electron configurations can be
stabilized by taking advantage of the ability of CAAC ligands’
p orbitals to act as electron sinks.
Although the unusual electronic structures of these
[M(CAAC)2]n+ complexes are grounds enough to warrant fur-
ther exploration, their reactivity should not be overlooked.
These complexes have already shown, for example, electro-
chemical and catalytic activity, and the ability to activate small
molecules, such as dihydrogen and carbon dioxide.[3–8] Our
entry into this field was motivated by a desire to prepare alter-
natives to the very useful and strongly Lewis basic complexes
[Pt(PR3)2] (R=Et, iPr, Cy) for various purposes, such as catalytic
dehydrocoupling,[9] oxidative addition of EÀX bonds,[10] and the
formation of metal-only Lewis pairs (MOLPs).[10e,11] Herein, we
describe the synthesis of both homoleptic [Pt(CAAC)2] and het-
eroleptic zero-valent complexes of platinum and their reactivity
towards Lewis acidic main-group compounds.
Figure 2. Synthesis of homo- and heteroleptic CAAC complexes of Pt0, and
reactivity of heteroleptic complex 2a.
zene, and showed statistically insignificant PtÀC bond-length
variation from the published molecular structure of 1·0.5THF
(1·C6H6: 1.965(2), 1.969(2) Š; 1·0.5THF: CÀPt 1.964(5),
1.967(5) Š). The CAAC ligands in both structures were found to
coordinate in an approximately linear fashion across the Pt
center (C-Pt-C angle: 1·C6H6: 171.13(7)8; 1·0.5THF: 170.9(2)8).
However, the two structures differ markedly in their degree of
alignment of the CAAC ligands, with the NCC planes in our
structure of 1·C6H6 being oriented nearly orthogonally (N-C-C-
N torsion: 73.18), but those of 1·0.5THF being between syn
and anti-arrangements (N-C-C-N torsion: 48.78). If we consider
the angle between the two least-squares planes of the CAAC
ligands, the differences are likewise very large (1·C6H6 70.08;
1·0.5THF 30.38). The alignment of the CAAC ligands in
1·0.5THF was discussed by Stalke, Frenking, and Roesky[6c] and
attributed to the stronger p backdonation of the CAAC li-
gands.[2e] That our structure shows a very different degree of
alignment, but apparently equivalent PtÀC distances, suggests
instead that the alignment has little bearing on the ability of
the Pt center to undergo p backbonding with the ligands, and
is perhaps not as diagnostic as the authors of that paper[6c]
suggest.
Results and Discussion
Synthesis of CAAC-containing Pt0 complexes
During the production of this manuscript, the groups of Stalke,
Frenking, and Roesky presented the synthesis of six homolep-
tic Pd0 and Pt0 CAAC complexes by the treatment of [M(PPh3)4]
(M=Pd, Pt) with three different CAAC ligands (Figure 1).[6c] Re-
cently, we developed a synthesis of one of these complexes,
[Pt(CAACMe)2] (1, Figure 2), by using the convenient triolefin Pt0
complex [Pt(nbe)3] (nbe=norbornene). The synthesis requires
no heating or fractional crystallization, because the nbe ligands
and solvent pentane can be removed under vacuum. This pro-
duced an orange solid, without the co-crystallized THF mole-
cule reported in the published procedure, in 96% yield, signifi-
cantly higher than the published result.[6c] The recently pub-
lished synthesis of 1 also involves the liberation of four molar
equivalents of triphenylphosphine. Our synthesis thus repre-
sents a marked improvement in atom efficiency and simplicity.
The NMR spectroscopic data of our material was found to
be essentially the same as the published data. Single crystals
of 1·C6H6 suitable for a diffraction study were grown from ben-
Complex 1 was prepared in an attempt to create a superior
transition-metal Lewis base to bis(phosphine) Pt0 complexes
frequently used in our laboratories. However, despite its pre-
sumably higher electron density, 1 was found to be inert to-
wards a range of haloborane Lewis acids including BBr2Ph,
Chem. Eur. J. 2015, 21, 12357 – 12362
www.chemeurj.org
12358
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Full Paper
1
1
BBr2Fc (Fc=ferrocenyl), BBr2Dur (Dur=2,3,5,6-tetramethyl-
phenyl), B2Br2(NMe2)2 and B2Cl2Dur2, even at elevated tempera-
tures (up to 608C). When 1 was treated with BH2Dur, BH2Mes,
or GaCl3, slow and unselective reaction of the starting materials
was observed from which we were unable to isolate or identify
any products. With BBr3, complex 1 decomposes to form
a complex mixture from which we were able to isolate only
the imidazolium salt [CAACMeH][BBr4]. The high steric bulk of
1 presumably precludes its metal-centered Lewis basicity, and
over time, allows slower ligand-cleavage reactions to predomi-
nate (Figure 3).
lites (2: d=69.4 ppm, JPPt =3148; 3: d=57.7 ppm, JPPt =3110),
similar to those found for related heteroleptic complexes
[Pt(NHC)(PCy3)] (NHC=ItBu=:CN2(tBu)2C2H2 d=59.7 ppm,
1JPPt =3986; NHC=IMes=:CN2(Mes)2C2H2 d=58.4 ppm, JPPt
=
1
3640), but with significantly smaller 31P–195Pt coupling con-
stants.[12] Single-crystal X-ray crystallographic studies of 2 and
3
showed considerably shorter PtÀCcarbene distances (2:
1.973(3); 3: 1.86(3) Š) than the NHC analogues [Pt(NHC)(PCy3)]
(NHC=ItBu: 2.027(3); NHC=IMes: 1.991(2) Š), but longer PtÀP
distances in the CAAC complexes (2: 2.2393(7); 3: 2.239(1) Š)
than the NHC complexes (ItBu: 2.211(1); IMes: 2.228(1) Š).
These effects are presumably due to the superior s donation
and p-accepting properties of the CAAC ligands. However, it
should be noted that disorder of the CAAC ligand of 3 over
two orientations limits the accuracy of the distances in this
structure.
Given these results, we sought a compromise that combined
the metal accessibility of bis(phosphine) complexes [Pt(PR3)2]
(R=Et, iPr, Cy) with the presumed higher electron density of
[Pt(CAAC)2] complexes. The clearest choice was the monophos-
phine, mono(CAAC) scaffold [Pt(CAAC)(PR3)]. In 2010, we devel-
oped a synthetic route to related monophosphine, mono(NHC)
complexes of Pd0 and Pt0 by the simple replacement of one
phosphine ligand of [M(PR3)2] with the carbene.[12] This strategy
was hereby extended to CAAC ligands by using the Pt0 com-
plexes [Pt(PR3)2] (R=iPr, Cy), resulting in the heteroleptic com-
plexes [Pt(CAACMe)(PR3)] (R=iPr, 2; R=Cy, 3; Figure 2). The
complexes each showed single 31P resonances with 195Pt satel-
The monophosphine [Pt(nbe)2(PCy3)], prepared by addition
of tricyclohexylphosphine to [Pt(nbe)3] (see the Supporting In-
formation for synthesis, characterization and single-crystal X-
ray structure), was also found to be a convenient precursor to
3, giving the complex in slightly higher yield (72%) than the
route using [Pt(PCy3)2].
Reactivity of CAAC-containing
Pt0 complexes
Unlike complex 1, the heterolep-
tic complex 2 was found to un-
dergo much more facile reactivi-
ty. Treatment of 2 with either tri-
bromoborane or dibromophe-
nylborane (the latter with heat-
ing to 608C) led to the square-
planar PtII boryl complexes 4
and 5, products of oxidative ad-
dition of one BÀBr bond to the
metal center. The 31P NMR sig-
nals of the products were found
to be significantly shifted to
lower frequencies from those of
their precursor 2, with drastically
reduced coupling constants (4:
1
d=22.1 ppm, JPPt =2397; 5: d=
21.7 ppm, 1JPPt =2415). Broad
singlet 11B NMR resonances were
also detected for the complexes
(4: d=54.0 ppm;
5: d=
73.6 ppm), effectively identical
to those of their most compara-
ble published complexes, the
bis(phosphine) PtII boryl com-
plexes trans-[PtBr(BBr2)(PCy3)2]
Figure 3. Crystallographically derived molecular structures of 1–6. Thermal ellipsoids drawn at the 50% probability
level. Solvent molecules (one molecule of benzene for 1, one molecule of toluene for 5), and some ellipsoids have
been omitted for clarity. Selected bond lengths [Š] and angles [8] for 1: PtÀC1 1.965(2), PtÀC21 1.969(2), N1ÀC1
1.329(2), N2ÀC21 1.326(2); C1-Pt-C21 171.13(7). For 2: PtÀC1 1.973(3), PtÀP 2.2393(7), NÀC1 1.320(3); C1-Pt-P
172.55(7). For 3: PtÀC1 1.86(3), PtÀP 2.239(1), N1ÀC1 1.25(4); C1-Pt-P 173.8(10). For 4: PtÀB 1.960(4), PtÀC1
2.049(3), PtÀP 2.3491(8), PtÀBr1 2.5968(3), C1ÀN 1.306(4); C1-Pt-P 166.12(9), B-Pt-Br1 170.1(1). For 5: PtÀB 1.986(4),
PtÀC1 2.038(4), PtÀP 2.340(1), PtÀBr1 2.6190(4), C1ÀN 1.309(5); C1-Pt-P 166.4(1), B-Pt-Br1 169.80(14). For 6: PtÀGa1
2.3962(3), PtÀC1 2.021(3), PtÀP 2.3227(7), NÀC1 1.323(4); C1-Pt-P 169.83(7), C1-Pt-Ga 91.32(7), P-Pt-Ga 98.44(2).
(d=54 ppm)
and
trans-
(d=
[PtBr{BBr(o-Tol)}(PCy3)2]
73 ppm),[13] suggesting that the
CAAC ligand has little electronic
effect on the boryl ligand itself.
Chem. Eur. J. 2015, 21, 12357 – 12362
www.chemeurj.org
12359
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Full Paper
Likewise, the PtÀB distances of 4 and 5 (4: 1.960(4); 5:
1.986(4) Š) are identical to those of their corresponding
bis(phosphine) counterparts trans-[PtBr(BBr2)(PCy3)2] (1.963(6) Š)
and trans-[PtBr{BBr(o-Tol)}(PCy3)2] (1.982(3) Š) within experi-
mental uncertainty. However, two points of distinction in the
boryl–CAAC derivatives 4 and 5 are their significantly bent C-
Pt-P angles (4: 166.12(9); 5: 166.4(1)8) and somewhat distorted,
non-planar CAAC heterocycles. Given their apparent electronic
similarities to comparable bis(phosphine) derivatives, these dis-
tortions are most likely due to increased steric congestion
around the metal from the bulky CAAC ligand.
In contrast to its oxidative addition reactions with bromo-
boranes, treatment of 2 with GaCl3 gave instead the orange
metal-only Lewis pair [(PiPr3)(CAACMe)Pt!GaCl3] (6), which
1
showed a 31P NMR resonance (d=53.1 ppm, JPPt =2425) close
to that of its bis(phosphine) congener [(Cy3P)2Pt!GaCl3] (d=
49.3 ppm, 1JPPt =2798).[14] Single-crystal X-ray crystallographic
study of 6 indicated that the complex has a shorter PtÀGa dis-
tance (2.3962(3) Š) than that of [(Cy3P)2Pt!GaCl3] (2.4019(2) Š)
by a statistically significant but negligible amount (ca. 0.2%).
In addition, the GaCl3 unit of 6 (S(aCl-Ga-Cl): 309.48) is slightly
more pyramidalized than that of [(Cy3P)2Pt!GaCl3] (S(aCl-Ga-
Cl): 314.38). Overall, the spectroscopic and structural features
of 6 suggest that the Lewis base strength of [Pt(CAACMe)(PiPr3)]
(2) is equivalent to or marginally greater than that of
[Pt(PCy3)], and, by extension, slightly weaker than
[Pt(NHC)(PCy3)] complexes as was determined in an earlier
study.[12]
Electrochemical measurements
Cyclic voltammetry measurements were performed on the
three Pt0 complexes prepared herein (1–3), in addition to
[Pt(PCy3)2] and [Pt(PiPr3)2] for comparison (see the Supporting
Information for further details). For complex 1, we observed
similar behavior to that reported by Stalke, Frenking, Roesky
and co-workers, that is, a quasi-reversible oxidation wave in
THF (our work: E1/2 =À0.30 V vs. [Cp2Fe]/[Cp2Fe]+; published:
E
1/2 =À0.07 V vs. [Cp*2Fe]/[Cp*2Fe]+, or ca. À0.50 V vs. [Cp2Fe]/
[Cp2Fe]+ after correction).[6c] This event was attributed by the
authors to a one-electron oxidation on the basis of DFT calcu-
lations, and also seems consistent with the unique stability of
the [M(CAAC)2]n+ system in oxidation states between those
usually expected for the metals.[3–8] In contrast, the remaining
complexes all showed irreversible oxidation waves (2: Epa =
À0.09 V; 3: Epa =À0.11 V; [Pt(PCy3)2]: Epa =0.00 V; [Pt(PiPr3)2]:
Epa = +0.12 V). If we set aside the unique behavior of 1, we
can infer slightly stronger reductive ability of compounds 2
and 3 compared to the two homoleptic bis(phosphine) com-
plexes. However, as was spectroscopically and crystallographi-
cally demonstrated above, higher electron density does not
always translate faithfully into stronger Lewis basicity.
Figure 4. Top: Experimental UV/Vis spectra of 1–3. Middle and bottom: Cal-
culated UV spectra of 1 and 3 at the PCM-CAM-B3LYP/def2-SVP//PCM-
B3LYP/def2-SVP level.
Table 1. UV/Vis peak list for 1–3, [Pt(PiPr3)2] and [Pt(PCy3)2]. Spectra were
recorded in benzene.
Compd.
l [nm]
1
285.5, 317.0, 365.0, 449.5
2
3
243.5, 301.5, 429.5, 479.5, 516.0
303.5, 431.0, 473.5
[Pt(PiPr3)2]
[Pt(PCy3)2]
247.0, 258.5, 315.5, 342.0, 395.0
247.5, 258.5, 316.0, 339.5, 393.0
UV/Vis spectroscopy
contain multiple excitations tailing off in the region l=430–
530 nm. In contrast, the longest wavelength bands of
[Pt(PiPr3)2] and [Pt(PCy3)2] can be found at l=393 and 395 nm,
The origin of the orange color of 1–3 is evident from their UV/
Vis spectra (recorded in benzene, Figure 4, top; Table 1), which
Chem. Eur. J. 2015, 21, 12357 – 12362
www.chemeurj.org
12360
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim