A versatile ligand platform that supports lewis acid promoted migratory insertion

Article abstract of DOI:10.1002/anie.201203264

A helping hand: Incorporation of Group 2 Lewis acids into a macrocycle appended to a phosphine ligand attached to a rhenium carbonyl complex promotes otherwise unfavorable transformations of coordinated CO (see scheme; M=Ca, Sr). These Lewis acids form relatively weak M-O bonds, thereby enabling release of organic products from the metal center. Copyright

Full text of DOI:10.1002/anie.201203264

Angewandte
Chemie
DOI: 10.1002/anie.201203264
Alkyl Migration
AVersatile Ligand Platform that Supports Lewis Acid Promoted
Migratory Insertion**
Amaruka Hazari, Jay A. Labinger,* and John E. Bercaw*
In light of increasing crude oil prices and global energy
demands, routes for obtaining fuels and chemical precursors
from methane, coal, or biomass by using synthesis gas, which
is a mixture of CO and H2, are emerging as viable
processes.[1,2] The main large-scale processes for utilization
of syngas are methanol synthesis, hydrogen production, and
the Fischer–Tropsch reaction. A severe limitation of the last
of these, which affords complex mixtures of hydrocarbons and
oxygenates, is the inability to control the hydrocarbon
product distribution and to select for a particular product.[3]
Although the prospect of developing homogeneous catalysts
for selective conversion of synthesis gas to hydrocarbons or
oxygenates has generated extensive research for over
40 years, no practical syngas-to-C2+ conversion has yet been
À
À
realized; formation of the first C H bond as well as C C
bonds appear to be the major hurdles that need to be
overcome.[4–7]
À
In previous work, we succeeded in facilitating both C H
À
and C C bond forming reactions by appending Lewis acids to
ligands in Re and Mn carbonyl complexes; however, those
systems have not proven viable for catalytic CO reduction, as
bonds between the Lewis acid (mostly boron) and oxygen in
the resulting products are too robust, preventing non-
destructive release of any organic product.[8–10] A versatile
ligand platform that can support a wide variety of Lewis acidic
À
metals in close proximity to a M CO reaction site would aid
Scheme 1. Synthesis of complexes 2–6.
in the effort to balance reactivity with sufficient lability for
closing a catalytic cycle. We report herein on a rhenium
carbonyl complex of a phosphine-functionalized aza crown
ether ligand,[11,12] which upon coordination of Group 2 Lewis
acids undergoes facile migratory insertions to form Lewis
acid-stabilized acyl complexes that react with H2O and
Brønsted acids to liberate C2 organic compounds.
ring,[14] suggesting that the N of the aza crown ether has
occupied the vacant coordination site on Re. Reaction of 4
with ZnMe2 in CH2Cl2 resulted in both methylation at Re and
insertion of Zn into the macrocycle to give 5; the zinc-free
methyl complex 6 was obtained by passing a THF solution of
5 over alumina. The proposed structures of both 4 and 5 were
confirmed by X-ray crystallography (Figure 1).[15,23]
Conversion of the commercially available aza crown ether
1 to phosphino-aza crown ether 2 followed by reaction with
[Re(CO)4(m-Br)]2 generated phosphine complex
3
(Scheme 1).[13] Treatment of 3 with AgBF4 in CH2Cl2 resulted
in cationic complex 4, the large upfield 31P NMR shift (d À41)
of which is characteristic of phosphorus in a four-membered
Both Ca2+ and Sr2+ were rapidly taken up into the
macrocycle by simple addition of [CaI2(thf)4] or [SrI2(thf)5] to
a CD2Cl2 solution of 6 to form 7a and 7b, respectively,
À
followed by slow (ca. 1 day) insertion of CO into the Re Me
bond to give acyl complexes 8a and 8b (Scheme 2). These
transformations were conveniently followed by 1H and
31P NMR spectroscopy.[13] Upon addition of Ca2+, the
31P NMR signal for 6 at d 2 shifted upfield to give a single
broad signal at d À11.3 ppm; subsequently the 1H signal
[*] Dr. A. Hazari, Dr. J. A. Labinger, Prof. J. E. Bercaw
Arnold and Mabel Beckman Laboratories of Chemical Synthesis,
California Institute of Technology
Pasadena, CA 91125 (USA)
E-mail: jal@caltech.edu
bercaw@caltech.edu
[**] This work was supported by BP through the XC2 program. We thank
À
corresponding to the Re Me protons gradually disappeared
and was replaced by a new singlet at d 2.74, which is consistent
with the formation of an acyl complex. The 13C NMR
spectrum included a diagnostic acyl peak at d 299 ppm (d,
Larry Henling and Michael Day for assistance with crystallography.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201203264.
J
C-P = 8.8 Hz), suggesting formation of the Lewis acid-stabi-
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
A bridging iodide occupies the site on Re vacated by
migratory insertion, forming a five-membered ring that
incorporates both Ca and Re.
The analogous Sr-containing acyl 8b was insoluble in
CD2Cl2 and consequently precipitated from solution during
the course of the reaction, but was sufficiently soluble in
1
CD3CN to obtain H and 13C NMR spectra. The signals of
these spectra are somewhat broadened at room temperature,
but characteristic signals for the acyl carbon (d = 298) and
protons (d = 2.75) were readily identified. Furthermore, the
signal for the methylene protons linking P and N appears as
two distinct sets of doublets of doublets as a result of coupling
both to phosphorus and to each other; the chemical non-
equivalence is expected for the proposed iodide-bridged
structure.[13]
Uptake of Li+ and In3+ into the macrocycle could also be
effected by treatment of 6 with LiBF4 or [InCl3(thf)3], as
evidenced by rapid appearance of new, broad 31P NMR
signals upfield (by 9–12 ppm) from that of 6, but no further
changes indicating insertion were observed. Addition of
[MgX2(thf)] (X = Cl, Br or I) to 6 did result in NMR signals
that suggest formation of migratory insertion products, but
these reactions did not proceed to completion, even after
extended reaction times.[13] These differences in reactivity
may be a consequence of the relative sizes of the Lewis acids.
Ca2+ and Sr2+, the largest ions, are likely
Figure 1. XRD structural representation of a) 4·0.6Et2O and
b) 5·0.5C6H6; solvent molecules are omitted for clarity. Selected bond
lengths [ꢀ] and angles [8]: 4: Re–N 2.2959(18), Re–P 2.4211(6), N–C5
1.511(3), P–C5 1.839(2); N-Re-P 66.96(5), N-C5-P 101.57(14). 5: Zn–
O8 2.0675(16), Zn–O6 2.1862(18), Zn–N 2.2450(17), Zn–O7
2.3723(17); N-Zn-O7 140.63(7), O6-Zn-O8 88.08(7).
to be situated the furthest outside the
macrocycle (the Ca sits 1.529 ꢀ above the
mean plane in 8a) and thus closest to the
À
Re CO reaction site, whereas the smaller
Li+, Mg2+, and In3+ would be expected to
better fit inside the macrocycle.
The five-membered Re-acyl-Ca-I ring
in 8a is closely related to the early
examples of acyl complexes formed by
Scheme 2. Ca2+ and Sr2+ mediated methyl migration.
promotion with AlX3,[18–21] but whereas
the latter undergo facile displacement of
lized acyl complex 8a; the IR spectrum was consistent with
a fac-tricarbonyl complex;[16] and an XRD study confirmed
the proposed structure (Figure 2).[17,23] The Ca2+ center in 8a
is seven-coordinate, interacting with the acyl oxygen and both
iodides as well as the O and N donor atoms of the macrocycle.
the Al-X-M bridges by CO, no reaction was observed when
complex 8a was exposed to an atmosphere of CO or treated
with PMe3. However, addition of [12]crown-4 to a solution of
8a and PMe3 did afford Ca-free acyl complex 9 (Scheme 3),
which was characterized by NMR and IR spectroscopy
(including an acyl stretch at 1567 cmÀ1);[13] it decomposed in
solution to a complex mixture of products over a period of
12 h, precluding its isolation. On treating 8a with [12]crown-4
in the absence of added PMe3, sequestration of CaI2 resulted
in the rapid reformation of methyl complex 6 (the reverse of
migratory insertion).
The observations of Scheme 3 suggest that the role of the
Re-I-M bridge in the kinetic acceleration and/or thermody-
namic stabilization of migratory insertion products 8 is of
importance comparable to that of the Lewis acid–acyl
interaction. Indeed, no observable reaction took place upon
treatment of complex 6 with a halide-free Lewis acid,
Ca(OTf)2, in acetonitrile over a period of several days. (The
use of acetonitrile for this reaction was necessitated by the
insolubility of Ca(OTf)2 in CH2Cl2. This better-coordinating
solvent might well interfere with the desired Lewis acid–
oxygen interactions; however, the reaction of 6 with [CaI2-
Figure 2. XRD structural representation of 8a. The structure is disor-
dered: 8a co-crystallizes with 20% of a second species in which the
bridging acyl is replaced by a second bridging iodine. Selected bond
lengths [ꢀ] and angles [8]: Re–C4 2.25(10), C4–O4 1.24(3), Ca–O5
2.416(10), Ca–O6 2.475(9), Ca–O7 2.426(9), Ca–N 2.576(10), Ca–O4
2.237(9); Re-I-Ca 92.23(5), Re-C4-O4 119.2(17).
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
attack by iodide at Re to release acetaldehyde
(pathway A). We suggest that coordination of water
to the Ca2+ center is required to generate sufficient
acid strength for the protonation of 8a to take place.
With larger amounts of added water (in the absence
of deliberately added acid) the effective pH is
apparently too high (a consequence of the leveling
effect), and pathway B, aquation of Ca2+ followed
by its removal and deinsertion, is followed instead.
Complex 8a is unaffected by the addition of
methanol, but treatment of 8a with HOTf and an
excess of MeOH leads to the formation of 1,1-
dimethoxyethane along with some methyl iodide
(Scheme 5). Again, hydroxycarbene complex 10 is
the probable first intermediate; it reacts with
methanol to give methoxycarbene complex 14,
which undergoes addition of methanol to afford 15
Scheme 3. Sequestration of Ca2+ from the macrocycle.
(thf)4] in acetonitrile did lead to the formation of an acyl
complex after several days.) Likewise, in situ generation of
Ca(OTf)2 in CD2Cl2, by addition of AgOTf to a solution of 6
and [CaI2(thf)4], led to no insertion.
Release of a C2 organic product generated by Lewis acid-
À
mediated C C coupling from the Re center was readily
achieved. Addition of 5–20 equivalents of water to a CD2Cl2
solution of acyl complex 8a or 8b resulted in rapid liberation
1
of acetaldehyde, which was identified by its H NMR signals
À
(d 2.17 d; d 9.76 q), and formation of Re I complex 12, the
Figure 3. XRD structural representation of 13; the iodide counteranion
is not shown. Selected bond lengths [ꢀ] and angles [8]: Re–C4
2.345(10), C4–O4 1.176(10), Ca–N 2.803(8), Ca–O4 2.282(7), Ca–O7
2.432(7), Ca–O8 2.537(7), Ca–O9 2.453(7); Re-I-Ca 94.47(4), Re-C4-O4
116.4(7).
identity of which was confirmed by comparison of spectro-
scopic data to those of an independently synthesized
sample.[13] Larger amounts of water led to a different reaction:
addition of 50 equivalents of water to a CD2Cl2 solution of 8a
gave new complex 13, which
underwent deinsertion to
6
over a period of several hours,
and no acetaldehyde was
observed. The crystal structure
of 13 (Figure 3) reveals that the
terminal iodide on Ca in 8a has
been replaced by two water
molecules to give an eight-coor-
dinate Ca2+ center.[22,23]
In contrast, addition of
50 equivalents
of
acidified
water (2% HOTf) to 8a does
afford acetaldehyde, as does
protonation of 8a with HOTf
in the absence of added water.
In each case,
a single (as
assessed by 31P NMR) as yet
unidentified metal species is
formed. These observations sug-
gest the sequence of steps shown
in Scheme 4, in which the acyl
oxygen of 8a is protonated to
afford hydroxycarbene inter-
mediate 10, followed by a 1,2-
Scheme 4. Reactivity of 8a with H2O. Pathway A is accompanied by the formation of an unidentified
hydride shift and nucleophilic precipitate; presumably this side reaction provides the additional equivalent of CO required to form 12.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[11] S. J. McLain, J. Am. Chem.
Soc. 1983, 105, 6355.
[12] S. J. McLain, Inorg. Chem.
1986, 25, 3124.
[13] See the Supporting Infor-
mation for full experimen-
tal
and
spectroscopic
details.
[14] P. E. Garrou, Chem. Rev.
1981, 81, 229.
[15] Crystal
[C25H28NO7PRe]+-
[BF4]À·0.6(C4H10O),
data
for
4:
tri-
¯
clinic, P1, a = 9.2994(5),
b = 9.9376(4),
17.5007(8) ꢀ,
87.062(2),
c =
a =
Scheme 5. Reactivity of 8a with MeOH/HOTf.
b = 84.077(3),
g = 75.087(3)8, V= 1553.97(13) ꢀ3, Z = 2, 1calcd = 1.712 MgmÀ3
,
along with an unidentified metal species. Competing nucle-
F(000) = 792, m = 4.031 mmÀ1, reflections measured: 101244,
independent reflections: 12609 [Rint = 0.0432], R1 = 0.0305 [I >
2s(I)], wR2 = 0.051 (all data), residual electron density 2.438/
À2.718 eꢀÀ3. For 5: [C27H31NO7PReZn]+[BF4]À·0.5(C6H6), tri-
À
ophilic attack at the O Me bond of 14 by iodide leads to the
formation of methyl iodide.
In conclusion, we have shown that Ca2+ and Sr2+ cations
can be readily incorporated into a macrocycle appended to
a phosphine ligand attached to a rhenium carbonyl complex,
and thus be suitably positioned to act as Lewis acids that
promote otherwise unfavorable transformations of coordi-
nated CO, under mild conditions. These Lewis acids appear to
¯
clinic, P1, a = 11.5843(6), b = 17.2038(8), c = 17.2154(9) ꢀ, a =
88.971(3), b = 77.007(3), g = 89.071(3)8, V= 3342.3(3) ꢀ3, Z = 4,
1calcd = 1.775 MgmÀ3, F(000) = 1764, m = 4.452 mmÀ1, reflections
measured: 338688, independent reflections: 49331 [Rint
0.0454], R1 = 0.0587 [I > 2s(I)], wR2 = 0.0907 (all data), residual
=
electron density 8.977/À4.817 eꢀÀ3
.
À
form weaker intermediate M O bonds and exhibit greater
[16] An acyl CO stretching peak was not observed in the IR
spectrum. Presumably, strong bonding to the Lewis acid lowers
the energy of this peak to a point where it is obscured by the
peaks that are due to phenyl vibrations.
substitutional lability than the boron-based systems studied
previously, thereby enabling the facile release of organic
products from the metal center, although that release has not
yet been accomplished nondestructively, in a manner com-
patible with a closed catalytic cycle. To achieve that goal, we
are continuing to investigate the behavior of a wide range of
Lewis acids, using this and other ligand architectures.
[17] Crystal data for 8a:
0.8(C26H31NO7PCaI2Re)·0.2(C24H28NO6PCaI3Re), monoclinic,
P21, a = 11.8734(4), b = 14.0152(5), c = 12.0094(5) ꢀ, b =
109.605(2), V= 1882.61(12) ꢀ3, Z = 2, 1calcd = 1.821 MgmÀ3, F-
(000) = 979, m = 5.252 mmÀ1, reflections measured: 40053, inde-
pendent reflections: 9988 [Rint = 0.0455], R1 = 0.0602 [I > 2s(I)],
wR2 = 0.1322 (all data), residual electron density 3.262/
Received: April 27, 2012
Published online: && &&, &&&&
À2.835 eꢀÀ3
.
[18] E. Lindner, G. von Au, Angew. Chem. 1980, 92, 843; Angew.
Chem. Int. Ed. Engl. 1980, 19, 824.
Keywords: carbonylation · homogeneous catalysis · insertion ·
.
[19] S. B. Butts, S. H. Strauss, E. M. Holt, R. E. Stimson, N. W.
Alcock, D. F. Shriver, J. Am. Chem. Soc. 1980, 102, 5093.
[20] S. B. Butts, E. M. Holt, S. H. Strauss, N. W. Alcock, R. E.
Stimson, D. F. Shriver, J. Am. Chem. Soc. 1979, 101, 5864.
[21] T. G. Richmond, F. Basolo, D. F. Shriver, Inorg. Chem. 1982, 21,
1272.
Lewis acids · rhenium
[1] A. Y. Khodakov, W. Chu, P. Fongarland, Chem. Rev. 2007, 107,
1692.
[2] C. K. Rofer-DePoorter, Chem. Rev. 1981, 81, 447.
[3] United Nations Development Program. World Energy Assess-
ment Report: Energy and the Challenge of Sustainability, United
Nations, New York, 2000.
[4] G. C. Demitras, E. L. Muetterties, J. Am. Chem. Soc. 1977, 99,
2796.
[5] A. R. Cutler, P. K. Hanna, J. C. Vites, Chem. Rev. 1988, 88, 1363.
[6] B. D. Dombek, Adv. Catal. 1983, 32, 325.
[7] N. M. West, A. J. M. Miller, J. A. Labinger, J. E. Bercaw, Coord.
Chem. Rev. 2011, 255, 881.
[8] A. J. M. Miller, J. A. Labinger, J. E. Bercaw, J. Am. Chem. Soc.
2008, 130, 11874.
[22] Crystal data for 13: [C26H31NO9PCaIRe]+[I]À, monoclinic, C2/c,
a = 50.446(6), b = 8.8199(11), c = 17.567(2) ꢀ, b = 93.862(7), V=
7798.2(17) ꢀ3, Z = 8, 1calcd = 1.969 MgmÀ3, F(000) = 4424, m =
5.127 mmÀ1, reflections measured: 116605, independent reflec-
tions: 9367 [Rint = 0.0436], R1 = 0.0818 [I > 2s(I)], wR2 = 0.1320
(all data), residual electron density 4.437/À3.910 eꢀÀ3
.
[23] Data were collected at 100 K using a Bruker KAPPA APEX II
diffractometer with MoKa radiation (g = 0.71073 ꢀ). Structure
solution and refinements were performed with the SHELXS-97
program. Refinement of F2 against all reflections. CCDC 850303
(4), 822258 (5), CCDC 854071 (8a), and CCDC 854702 (13)
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.
[9] A. J. M. Miller, J. A. Labinger, J. E. Bercaw, Organometallics
2010, 29, 4499.
[10] N. M. West, J. A. Labinger, J. E. Bercaw, Organometallics 2011,
30, 2690.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Alkyl Migration
A. Hazari, J. A. Labinger,*
J. E. Bercaw*
&&&& — &&&&
A Versatile Ligand Platform that Supports
Lewis Acid Promoted Migratory Insertion
A helping hand: Incorporation of Group 2
Lewis acids into a macrocycle appended
to a phosphine ligand attached to a rhe-
nium carbonyl complex promotes other-
wise unfavorable transformations of
coordinated CO (see scheme; M=Ca,
Sr). These Lewis acids form relatively
À
weak M O bonds, thereby enabling
release of organic products from the
metal center.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!