Selective dimerization of lewis-acid/base-stabilized phosphanylalanes

Article abstract of DOI:10.1002/chem.201203074

The reaction of [{(CO)₅W}PRH₂] (R=H, Ph) with H ₃Al×NR₃ (R=Et, Me) leads to the formation of four-membered heterocyclic compounds [({(CO)₅W}P(H)AlH× NEt₃)₂] and [({(CO)₅W}PhPAlH× NMe ₃)₂]. Upon dissolving the solid compounds, fast equilibria between the isomers are observed on the NMR timescale. Further insight into the stability and reactivity of the isomers was gained by applying theoretical methods. DFT calculations predict that hydrogen elimination in the case of [({(CO)₅W}PhPAlH×NMe₃)₂] may be reversible. Basic changes: Minor changes in the amine base and the phosphorus substituent of phosphanylalanes significantly affect the reaction outcome and the behavior of the products in solution. A fast equilibrium between different isomers is observed by NMR spectroscopy (see scheme). Copyright

Full text of DOI:10.1002/chem.201203074

FULL PAPER
DOI: 10.1002/chem.201203074
Selective Dimerization of Lewis-Acid/Base-Stabilized Phosphanylalanes
Michael Bodensteiner,[a] Alexey Y. Timoshkin,[b] Eugenia V. Peresypkina,[c]
Ulf Vogel,[a] and Manfred Scheer*[a]
Dedicated to Professor Hartmut Bꢀrnighausen on the occasion of his 80th birthday
Abstract:
The
reaction
of
compounds, fast equilibria between the
isomers are observed on the NMR
timescale. Further insight into the sta-
bility and reactivity of the isomers was
gained by applying theoretical meth-
ods. DFT calculations predict that hy-
drogen elimination in the case of
[({(CO)5W}PhPAlH·NMe3)2] may be
reversible.
[{(CO)5W}PRH2] (R=H, Ph) with
H3Al·NR3 (R=Et, Me) leads to the
formation of four-membered heterocy-
clic compounds [({(CO)5W}P(H)AlH·
Keywords: aluminum
·
density
functional calculations · dimeriza-
tion · isomerization · phosphorus
NEt3)2]
and
[({(CO)5W}PhPAlH·
NMe3)2]. Upon dissolving the solid
Introduction
numerous different structural motifs.[13] To avoid polymeri-
zation and to stabilize the non-existent monomers, we devel-
oped the concept of LA/LB stabilization, thereby enabling
us to synthesize the first LA/LB-stabilized phosphanyl- and
arsanylboranes, as well as phosphanylalanes and -gallanes,[14]
The storage and activation of hydrogen is an active topic in
the chemistry of compounds that contain Group 13 and
Group 15 elements. In particular, owing to its high hydro-
gen-storage capacity, ammonia–borane (H3B·NH3) has at-
tracted much attention.[1,2] Acids,[3] transition-metal cata-
lysts,[4] nanoparticles,[5] and ionic liquids[6] have been success-
fully employed to liberate hydrogen from this system. Re-
cently, spontaneous hydrogen elimination from aromatic
amine boranes, Ph*NH2·BH3, in THF, without any catalyst,
was reported by Manners and co-workers.[7]
by using M(CO)5 (M=Cr, W) or EACTHUNGTRENUNG(C6F5)3 (E=B, Ga) as
LAs and amines or N-heterocyclic carbenes as LBs.[15] Com-
putations predicted that stabilization of compounds of the
À
type H2E E’H2 can be achieved by using either LAs or LBs
but not necessarily both; indeed, we succeeded in synthesiz-
ing the first LB-stabilized, hydrogen-substituted Group 13/
[16]
À
15 compound, H2P BH2·NMe3, by abstraction of the LA.
Phosphorus–boron compounds also show facile hydrogen
activation in frustrated Lewis acid/Lewis base (LA/LB)
For the phosphanylalane analogue, abstraction of the LA is
not possible; the LA/LB-stabilized compounds already
easily lose hydrogen under formation of oligomers, even at
ambient temperatures. Thus, the challenge is to control this
process by applying appropriate conditions, such as solvent
and temperature, and also the features of the LA or LB. We
recently found that, depending on the solvent and the tem-
perature, [{(CO)5W}H2PAlH2·NMe3] trimerized and we were
able to gain mechanistic insight into this reaction.[17] We
have now discovered that the LB has a decisive influence on
the oligomerization process. Thus, the use of the slightly
bulkier base NEt3 results in the exclusive formation of a
dimer. The NMe3 derivative also undergoes selective dimeri-
zation if one hydrogen atom of the phosphine group is sub-
stituted by a phenyl group. Further aggregation of those
four-membered rings would then lead to the tetramer. Theo-
retical investigations on the smaller model LA/LB substitu-
ents BH3 and NH3 revealed a distorted cube-shaped com-
pound, which has already been reported by Cowley et al. for
the non-LA/LB-stabilized derivative (Figure 1).[18] However,
the geometries of these compounds differ significantly be-
cause they were influenced by the steric and the electronic
properties of the substituents.
(C6F5)2.[8] Even non-frus-
pairs, such as (C6H2Me3)2PACHTNUGTRENNUG(C6F4)BCAHTUNGTRENNUGN
trated adducts can undergo hydrogen elimination to form
polymers by using RhI catalysts.[9] Aluminum–phosphorus
ꢀ
systems were recently used for the activation of C C triple
bonds[10] and for chemically binding CO2.[11] Our research fo-
À
cuses on the hydrogen-substituted parent compounds H2E
E’H2 (E=B, Al, Ga ; E’=P, As), which have only been stud-
ied theoretically owing to their instability.[12] These com-
pounds undergo oligomerization and polymerization proc-
esses by intramolecular H2 elimination, thereby resulting in
[a] Dr. M. Bodensteiner, Dr. U. Vogel, Prof. Dr. M. Scheer
Institut fꢀr Anorganische Chemie
Universitꢁt Regensburg, 93040 Regensburg (Germany)
Fax: (+49)941-943-4439
E-mail: manfred.scheer@ur.de
[b] Prof. Dr. A. Y. Timoshkin
Department of Chemistry, St. Petersburg State University
University pr. 26, 198504 Old Peterhoff, St. Petersburg (Russia)
[c] Dr. E. V. Peresypkina
Nikolaev Institute of Inorganic Chemistry SB RAS
Ak. Lavrentiev pr. 3, 630090 Novosibirsk (Russia)
Herein, we investigate the role of the amine and the
phenyl substituent on the oligomerization products.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201203074.
Chem. Eur. J. 2013, 19, 957 – 963
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
957

frequencies (1833 cmÀ1, 1744 cmÀ1) were observed in
CH2Cl2. Despite the absence of the molecular-ion peak in
the mass spectrum (EI), peaks of characteristic fragments
were detected. In the solid state, two molecules of com-
pound 1 arrange through hydride bridges, thereby resulting
in a distorted trigonal bipyramidal geometry at the alumi-
num atom (Figure 2). In comparison to [{(CO)5W}-
H2PAlH2·NMe3], the distance between the molecules is
À
elongated, owing to the larger base (Al Al’ 3.0067(3) ꢃ in
À
compound 1 compared to 2.908(2) ꢃ). Furthermore, the W
À
P Al angle is widened by more than eight degrees
(127.02(10)8 in compound 1, in comparison with 118.44(4)8
in [{(CO)5W}H2PAlH2·NMe3]).
Figure 2. Hydride-bridged arrangement of compound 1 in the solid state;
hydrogen atoms of the triethylamine moiety are omitted for clarity. Se-
À
À
À
lected bond lengths [ꢃ] and angles [8]: W P 2.559(2), P Al 2.387(3), Al
À
À
N 2.066(6); P Al N 105.77(19).
If the reaction mixture of W(CO)5PH3 and AlH3·NEt3 is
not cooled down, or a solution of compound 1 is warmed up
to 48C, yellow needles of [({(CO)5W}HPAlH·NEt3)2] (2) are
obtained. Surprisingly, compound 2 does not undergo trime-
rization, but, instead, undergoes dimerization, in comparison
to the NMe3 derivative.[17] Based on the NMR data and the-
oretical calculations, a four-membered-ringed intermediate,
[({(CO)5W}H2PAlH2·NMe3)2], was also proposed for the tri-
methylamine system. However, it could not be isolated
owing to its further reactivity towards the ladder-shaped
compound [({(CO)5W}HPAlH·NMe3)2{(CO)5WPAl·NMe3}].
Similar additional aggregation was not observed for the
NEt3 derivative at all; the four-membered ring was the only
product of this reaction. The 31P{1H} NMR spectrum of a
solution of compound 2 in CD2Cl2 showed three signals
(Figure 3).
À
Figure 1. a) Theoretically computed [{(H3B)PAl
À À
ACHTUGNTRENUN(NG NH3)}4] (P Al 2.381 ꢃ;
À
À
Al P Al 78.08, P Al P 100.88) and b) experimentally observed
À À
À
[{(Ph3Si)PAl
ACHTUNGTRENNUNG(iBu)}4] heterocubanes (P Al 2.409(4)–2.417(4) ꢃ; Al P Al
[18]
À
À
90.9(2)–92.38, P Al P 87.7(2)–89.0(2)8).
Results and Discussion
The triethylamine-substituted compound [{(CO)5W}-
H2PAlH2·NEt3] (1) can be synthesized by the reaction be-
tween W(CO)5PH3 and AlH3·NEt3 in CH2Cl2, but the solu-
tion has to be cooled immediately to À788C to inhibit fur-
ther elimination of hydrogen and aggregation. Crystals of
compound 1 can be isolated in good yield (89%) at low
temperatures. As a solid, compound 1 is stable at ambient
temperature, but is extremely air-sensitive. Its 31P NMR
spectrum shows the expected triplet at d=À242 ppm (1J-
1
In addition, in the H NMR spectrum, resonances for dif-
(P,H)=286 Hz), which is in good accordance with the data
for the NMe3 derivative. IR spectra could not be acquired
from a solid sample because, even in a glovebox, the com-
pound rapidly decomposed during the preparation of a KBr
pellet. Nevertheless, P H (2322 cm ) and Al H stretching
ferent aminoethyl groups are detected, which indicate the
presence of different isomers in solution. Discounting the
possibility of ring puckering, which is expected to be an
effect of the arrangement of the substituents, the structures
of the possible isomers are depicted in Figure 4.
À1
À
À
958
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 957 – 963

Dimerization of Lewis-Acid/Base-Stabilized Phosphanylalanes
FULL PAPER
Figure 3. 31P NMR (top) and 31P{1H} NMR spectra (bottom) of compound
2 in CH2Cl2 at 300 K.
Figure 5. Puckered four-membered ring of compound 2 in the solid state;
hydrogen atoms on the triethylamine moiety are omitted for clarity. Se-
À
À
À
lected bond lengths [ꢃ] and angles [8]: W P 2.589(3), P Al1 2.382(4), P
À
À
À À
Al2 2.385(4), Al1 N1 2.000(13), Al2 N2 1.990(14); Al1 P Al2
À
À
À
À
84.15(15), P Al1 P 94.62(18), P Al2 P 94.45(18).
sition of some of the substance during solvation. A small
1
signal for free NEt3 is observed in the H NMR spectrum of
compound 2. Such an exchange of one NEt3 ligand leads to
the formation of isomers ii and iii, which give the additional
resonances in the NMR spectra (Scheme 1).
Figure 4. Possible isomers of compounds 2 (R=Et, R’=H) and 3 (R=
Me, R’=Ph).
The isomer determined by X-ray structure analysis is 2(i)
(Figure 4). The four-membered Al2P2 ring is puckered by
Scheme 1. Isomerization equilibrium of compounds 2 and 3 in CH2Cl2.
À
16.88 along the Al Al axis. DFT calculations also show that
this geometry is the most stable for the hypothetical parent
molecule cis-(HAlPH)2.[19] The P Al bond lengths (2.382(4)
Interestingly, a similar four-membered ring motif is found
in the product of the reaction between the mono-phenyl-
substituted phosphane [(CO)5WPPhH2] and the amine
adduct H3Al·NMe3, even in the presence of the smaller base
NMe3. In contrast to compound 2, isomer v, which features
a trans arrangement of similar substituents, is formed in
[({(CO)5W}PhPAlH·NMe3)2] (3) (Figure 6).
Moreover, the four-membered ring is perfectly planar in
the solid state. The monomeric precursor compound
[{(CO)5W}PhHPAlH2·NMe3] cannot be isolated owing to a
fast dimerization to afford compound 3. However, evidence
for its presence is observed in the 31P NMR spectrum of the
crude reaction mixture, which shows an additional doublet
À
and 2.385(4) ꢃ) are in good agreement with those found in
compound 1. However, the angles within the ring are bent
close to 908 to allow the four-membered ring to be formed
À À
À
À
À
À
(Al1 P Al2 84.15(15), P Al1 P 94.62(18), P Al2 P
94.45(18)8).
Unfortunately, all efforts to isolate other isomers in the
solid state failed. The unit-cell parameters for about 100 dif-
ferent crystals were equal within the standard deviation
range. The seven crystals that possessed the most-deviated
cell parameters were fully processed in X-ray experiments
and all gave the structure of compound 2 (isomer i;
Figure 5).
If these crystals are dissolved in CD2Cl2 to record their
31P NMR spectra, the same three signals are detected with
consistent integration ratios. This observation suggests the
existence of an equilibrium between the isomers that is fast
on the NMR timescale. For such compounds, it is known
that, in presence of a free amine, a base-exchange occurs
through an intermediate that features a trigonal-bipyramidal
aluminum atom with two amine ligands.[20] Even with ex-
tremely careful handling, the appearance of traces of free
NEt3 cannot be completely avoided, owing to the decompo-
at d=À136 ppm (1J
(P,H)=296 Hz).[21] Furthermore, as in
ACHTUNGTRENNUNG
the spectra of compound 2, 31P NMR studies of compound 3
also reveal signals for different isomers at d=À138, À140,
À142, and À144 ppm and the H and 31P NMR spectra show
1
traces of decomposition products [(CO)5WPPhH2] and
NMe3.
Theoretical studies: Further information regarding the be-
havior of these four-membered rings was gained from theo-
retical computations. The structural features of all of the
Chem. Eur. J. 2013, 19, 957 – 963
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
959

M. Scheer et al.
The energy difference between the isomers of the NEt3
and NMe3 derivatives is very small (for comparison, data for
the analogous [({(CO)5W}PHAlH·NMe3)2] isomers are given
in parentheses in Table 1).
In contrast, the substitution of a hydrogen atom at a phos-
phorus site by a phenyl group leads to larger energy differ-
ences (up to 25 kJmolÀ1) between the five possible isomers
of compound 3, with isomer 3(v) being 9 kJmolÀ1 more
stable than isomer 3(iv). The Gibbs energy values for the
isomerization reactions in the gas phase suggest that isomer
3(v) is the dominant form at standard temperature. In agree-
ment with theoretical predictions, crystal-structure analysis
revealed that isomer 3(v) featured a planar four-membered
Al2P2 ring (Figure 6). In contrast, the optimized structures
of isomer 2(v) and its NMe3 analogue are asymmetric, with
a puckered Al2P2 ring. This result suggests that the phenyl
groups on the phosphorus atoms induce the planarity of the
Al2P2 ring in isomer 3(v). The unique feature of isomer 3(v)
Figure 6. Planar four-membered ring of compound 3 in the solid state;
hydrogen atoms at the phenyl substituents are omitted for clarity. Select-
À
À
À
ed bond lengths [ꢃ] and angles [8]: W P 2.6256(19), P Al 2.418(3), Al
À À
À À
N 1.978(6); Al P Al 82.81(9), P Al P 97.20(9).
À
À
theoretically considered isomers are summarized in the Sup-
porting Information.[21] Theoretical computations of differ-
ent isomers at the B3LYP/6-31G* (ECP on W) level of
theory reveal that all five of the possible isomers of com-
pound 2 (Figure 3) are close in energy (the maximum
energy difference was only 11 kJmolÀ1) and that isomer 2(ii)
is predicted to be the lowest in energy (Table 1). However,
the three isomers 2(i), 2(ii), and 2(v) are very close in
energy (within 4 kJmolÀ1), which fits well with the experi-
mentally observed equilibrium between the isomers in solu-
tion. The negative value of the isomerization Gibbs energy
indicates that isomer 2(i) should be dominant in the gas
phase at 298 K (assuming that an equilibrium between all of
the isomers is achieved). In our opinion, the fact that only
isomer 2(i) could be isolated in the solid state is due to fa-
vorable packing in the crystal. Isomer 2(i) possesses a rela-
tively large dipole moment, which should increase the crys-
tal lattice energy and facilitate its crystallization.
is the presence of short intramolecular Al H···H C distan-
ces (2.217 ꢃ) between the negatively charged hydridic
hydrogen atom at the aluminum center and the partially
positively charged hydrogen atom on the phenyl group
(Mulliken partial charges are À0.17 and +0.13, respectively).
This interaction might be responsible for the stabilization of
isomer 3(v) in the solid state.
The planar rings of Group 13/15 elements are quite
common for structures of dimeric imino compounds.[12,13]
For four-membered rings of the heavier Group 15 elements,
both puckered and planar structures have been observed
and the planar structures are usually enforced by very bulky
substituents. For example, the structurally characterized
donor-only stabilized cyclic aluminum–phosphorus com-
pound [(iPr3SiPAlClPy)2] has a planar Al2P2 ring.[22] Struc-
tural and thermodynamic features of model Ga and In com-
pounds have been theoretically explored in
a recent
report.[23]
The experimentally observed equilibrium between the iso-
mers of compound 2 in solution indicates a fast ligand-ex-
change process. There are several plausible mechanisms to
account for this exchange. Our computational data (Table 2)
reveal that the dimers are quite strongly bound with respect
to the dissociation of the Al2P2 ring into monomers. For iso-
mers 2(ii) and 3(v), such a dissociation process is endother-
mic by 190–196 kJmolÀ1 (per mole of dimer). These results
suggest that a dissociation process cannot be responsible for
the fast isomerization of compound 2 at room temperature.
To gain an insight into the exchange mechanism, we per-
formed additional computational studies for the two most
probable mechanisms, based on SN1 and SN2 amine-ex-
change reactions (Figure 7).
Table 1. E80, standard isomerization Gibbs energies (DG8298), dipole mo-
ments (m), and the Debye and dihedral PAlPAl angles (qPAlPAl) for the dif-
ferent isomers of compounds 2 and 3.[a]
Isomer
E80 [kJmolÀ1
]
DG8298 [kJmolÀ1
]
m [8]
qPAlPAl [8]
2(i)
3.9
À3.3
5.7
(6.1)
29.9
ACHTUNGTRNE(NUNG 28.0)
G
E
G
2(ii)
2.5
4.0
3.2
6.2
17.3
2.8
G
G
(1.9)
N
2ACHTUNGTRENNUNG(iii)
G
E
ACHTUNGTNER(NUNG 1.5)
2(iv)
2(v)
21.7
15.8
R
(24.1)
A
E
(4.2)
ACHTUNGTRNE(NUNG 11.4)
3(i)
3(ii)
3ACHTUNGTRENNUNG(iii)
3(iv)
3(v)
25.7
10.8
15.6
9.1
27.7
16.9
15.8
7.9
13.5
14.2
6.5
4.4
0.0
18.6
17.8
24.9
18.3
0.0
We found that the dissociation of NEt3 from compound 2
is endothermic by 107–132 kJmolÀ1 (depending on the
isomer) and the dissociation of NMe3 from compound 3 is
endothermic by 95–135 kJmolÀ1. In both cases, the most-
stable [{(CO)5WPRAlH}2NR’3] product features the
W(CO)5 group in a bridging position.[21] The reaction ener-
gies of the amine-base-dissociation processes are the lower
0.0
0.0
[a] Values
in
parenthesis
correspond
to
the
isomers
of
[({(CO)5W}PHAlH·NMe3)2]. Calculations were performed at the B3LYP/
6-31G* (ECP on W) level of theory.
960
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 957 – 963

Dimerization of Lewis-Acid/Base-Stabilized Phosphanylalanes
FULL PAPER
Table 2. Predicted standard enthalpies and standard Gibbs energies for the gas-phase reactions.[a]
vorable. The elimination of the
second hydrogen molecule with
the formation of monomers
Process
DH8298 [kJmolÀ1
]
DG8298 [kJmolÀ1
]
1a AlH3·NEt3+(CO)5WPH3 =1+H2
À39.1
À24.9
G
(À27.5)
[{(CO)5W}PHAlH·NEt3]
and
1b AlH3·NMe3+(CO)5WP(Ph)H2 =[(CO)5WP(Ph)HAlH2·NMe3]+H2
2a 1=[(CO)5WPHAlH·NEt3]+H2
À33.4
77.0
À24.3
49.6
[{(CO)5W}P(Ph)AlH·NMe3]
(Table 2, processes 2a and 2b)
is unfavorable, but is compen-
sated by the subsequent dimeri-
zation energies (Table 2, pro-
AHCTUNGTERGcNNUN ess 3). The overall reaction
(Table 2, processes 5a and 5b)
is exothermic and thermody-
2b [(CO)5WP(Ph)HAlH2·NMe3]=[(CO)5WP(Ph)AlH·NMe3]+H2
3a [(CO)5WPHAlH·NEt3]=0.52
3b [(CO)5WP(Ph)AlH·NMe3]=0.53
4a 1=0.52+H2
4b [(CO)5WP(Ph)HAlH2·NMe3]=0.53+H2
5a AlH3·NEt3+[(CO)5WPH3]=0.52+2H2
99.9
78.9
À97.9
À95.0
À20.9
4.9
À75.2
À68.8
À25.5
10.1
À60.0
À50.4
(À58.8)
(À48.1)
5b AlH3·NMe3+(CO)5WP(Ph)H2 =0.53+2H2
À28.5
À14.2
namically allowed for both
compounds 2 and 3.
[a] Values in parenthesis correspond to reactions with NMe3 instead of NEt3, see reference [2]. Calculations
were performed at the B3LYP/6-31G*(ECP on W) level of theory.
AHCTUNGTRENNUNG
A comparison of processes
4a and 4b shows that com-
pound 1 is thermodynamically
unstable towards H2 evolution, whilst an analogous H2 elimi-
nation from the [{(CO)5W}P(Ph)HAlH2·NMe3] monomer is
endergonic. The experimental isolation of the more-reactive
compound 1 in the solid state may be attributed to addition-
al hydrogen-bridge stabilization (Figure 1). A similar assem-
bly of the phenyl-substituted derivative appears to be less
favorable, owing to steric hindrance. From our theoretical
computations, it appears that, in the case of the formation
of compound 3, the second hydrogen-evolution step
(Table 2, process 4b) may be reversible, which is of potential
interest for applications in hydrogen-storage/release. These
predictions require additional experimental studies at high
hydrogen pressures. Thermodynamic parameters for the re-
actions that lead to compounds 2 and 3 from the starting
materials are summarized in Table 2, which presents results
for the most-stable isomers, 2(ii) and 3(v). Thus, the substi-
tution of NEt3 by NMe3 only leads to minimal changes in
the thermodynamics of the dehydrogenation reactions.
Figure 7. Proposed SN1 (top) and SN2 (bottom) isomerization pathways
(2: R=Et, R’=H; 3: R=Me, R’=Ph).
limits for the activation energies of the SN1 mechanism.
These values (95–135 kJmolÀ1) appear to be too high to ac-
count for the observed fast isomerization at room tempera-
ture. Thus, an isomerization through an SN1 mechanism with
amine dissociation can be ruled out as being too energetical-
ly demanding.
The alternative SN2 pathway includes an addition of the
amine to the dimeric ring (Figure 7, bottom). Our attempts
to optimize the intermediate structures for the SN2 mecha-
nism (by the addition of the amine to compounds 2 and 3)
failed. Proposed intermediates with pentacoordinate trigo-
nal-pyramidal aluminum atoms were unstable; such struc-
tures lost one amine group upon optimization. This result in-
dicates easy removal of an amine moiety from the inter-
mediate but does not allow us to obtain the activation
energy for the first step of the SN2 reaction pathway. The
presence of free amines in solution (broad signals in the
1H NMR spectra) provides a further hint at a dynamic pro-
Conclusion
In summary, these results have shown that minor changes in
the amine moiety from NMe3 to NEt3 in Lewis-acid/base-
stabilized phosphanylalanes [{(CO)5W}H2PAlH2·NR3] leads
to significantly different reactivity. In the case of the larger
base, smaller four-membered [({(CO)5W}HPAlH·NEt3)2]
rings are formed, which show dynamic behavior in solution.
Similarly, four-membered [({(CO)5W}PhPAlH·NMe3)2] rings
are formed in the case of the phenyl-substituted derivative.
Upon dissolving the solid products, fast equilibria between
different ring isomers are observed. The stability of these
isomers and their interconversion pathways were calculated
by using DFT methods.
ACHTUNGTRENNUNGcess and is in agreement with a fast amine-base-exchange
process through an SN2 mechanism for both compounds 2
and 3 (Figure 7). The origin of the free base in solution is
most probably due to minor decomposition of the com-
pounds owing to side reactions (compounds 2 and 3 are un-
stable at room temperature in solution).
Experimental Section
The first hydrogen-elimination process from the starting
materials (Table 2, processes 1a and 1b) is energetically fa-
All procedures were either performed under an argon atmosphere in a
glovebox or by using standard Schlenk techniques. Solvents were de-
Chem. Eur. J. 2013, 19, 957 – 963
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
961

M. Scheer et al.
gassed and dried over appropriate drying agents. [({CO)5W}PH3],[24]
À144 ppm (br s; PPh); MS (EI, 14 eV): m/z (%): 866 (0.7)
[27]
[{(CO)5W}PH2Ph],[25] H3Al·NEt3,[26] and H3Al·NMe3 were synthesized
[MÀ2PhÀMeÀ2H]+,
433
(67)
[(CO)5WPHPh)]+,
377
(51)
according to literature procedures. All NMR spectra were acquired on a
BRUKER AVANCE 400 spectrometer (1H: 400.13 MHz, 27Al:
104.27 MHz, 31P: 161.98 MHz) and chemical shifts (d) are given in ppm.
IR spectroscopy was performed on a VARIAN FTS800 spectrometer and
absorption maxima are given in cmÀ1. MS (EI) spectra were obtained on
a FINNIGAN MAT SSQ 710A.
[(CO)3WPHPh)]+, 348 (100) [(CO)2WPPh)]+, 320 (44) [(CO)WPPh)]+,
292 (64) [WPPh]+, 242 (16) [(CO)WP]+, 214 (7) [WP]+; IR (CH2Cl2): n˜ =
2076 (s; CO), 1994 (s; CO), 1942 (vs; CO), 1751 cmÀ1 (w; AlH).
NMR data of [{(CO)5W}P(H)(Ph)AlH2·NMe3]: 1H NMR (400 MHz,
CD2Cl2, 300 K, TMS): d=2.17 (s; NMe3), 4.0 (br s; AlH2), 4.24 (d,
J(P,H)=296 Hz; PPhH), 7.2–7.7 ppm (m; Ph); 31P NMR (162 MHz,
CD2Cl2, 300 K, 85% H3PO4): d=À136 ppm (br d,
JACHTNGUTERN(UNG P,H)=296 Hz;
Synthesis of compound 1: [{(CO)5W}PH3] (179 mg, 0.5 mmol) and
H3Al·NEt3 (65 mg, 0.5 mmol) were stirred in CH2Cl2 (3 mL) at RT until
the evolution of gas ceased (20 min) and the reaction mixture was imme-
diately cooled to À788C. Colorless crystals of compound 1 could already
be obtained during the cooling process. Yield: 217 mg (0.445 mmol,
89%).
PPhH); 31P{1H} NMR (162 MHz, CD2Cl2, 300 K, 85% H3PO4): d=
À136 ppm (br s; PPhH).
Crystals of compounds 1, 2, and 3 were taken from a cooled (À308C)
Schlenk flask under a stream of argon. The crystals were immediately
covered with perfluorinated polyethers (Fomblin, Aldrich) and the
chosen single crystal was transferred onto a nylon loop together with
some oil and directly attached onto the goniometer under a stream of
cold nitrogen.[28] All of the crystals were processed on an Oxford Diffrac-
tion Gemini R Ultra CCD diffractometer by using CuKa radiation (l=
1.54178 ꢃ). For compound 1, an analytical absorption correction was ap-
plied[29] and a multi-scan correction from equivalents was carried out for
compounds 2 and 3.[30] The structures were solved by using direct meth-
ods with SIR-92[31] and full-matrix least-squares refinement on F2 was
performed with SHELXL-97.[32] Hydrogen atoms at the phosphorus and
aluminum atoms were refined isotropically with free distances; the
Analytical data for compound 1: 1H NMR (400 MHz, CD2Cl2, 300 K,
TMS): d=1.27 (t, J
2.94 (q,
(H,H)=7 Hz; CH2), 3.8 ppm (br s; AlH2); 1H
(400 MHz, CD2Cl2, 300 K, TMS): d=1.27 (t, J(H,H)=7 Hz; CH3), 2.19
(br s; PH2), 2.94 (q, (H,H)=7 Hz; CH2), 3.8 ppm (br s; AlH2);
27Al NMR (104 MHz, CD2Cl2, 300 K, 1.5m Al
(NO3)3): d=141 ppm (br s,
w1/2 =1500 Hz; AlH2); 31P NMR (162 MHz, CD2Cl2, 300 K, 85% H3PO4):
(P,H)=286 Hz; PH2); 31P{1H} NMR (162 MHz,
ACHTUNGTRENNUNG(H,H)=7 Hz; CH3), 2.19 (br d, JCAHTUNGTRENNNUG
J
N
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
JACHTUNGTRENNUNG
AHCTUNGTRENNUNG
d=À242 ppm (br t,
JACHTUNGTRENNUNG
CD2Cl2, 300 K, 85% H3PO4): d=À242 ppm (br s; PH2); MS (EI, 70 eV):
m/z (%): 429 (0.4) [(CO)5W(PH2AlH2·NEt)]+, 386 (0.5) [(CO)5W-
ACHTUNGTRENNUNG
(PH2AlH2)]+, 358 (66) [(CO)5WPH3]+, 328 (13) [(CO)4WPH]+, 300 (71)
carbon-bound hydrogen atoms were calculated geometrically and
a
[(CO)3WPH]+, 272 (62) [(CO)2WPH]+, 243 (41) [(CO)WP]+, 215 (23)
[WP]+, 184 (5) [W]+, 101 (19) [NEt3]+, 86 (100) [NEt2CH2]+; IR
(CH2Cl2): n˜ =2322 (w; PH), 2068 (s; CO), 1978 (s; CO), 1929 (vs; CO),
1833 (m; AlH), 1774 cmÀ1 (m; AlH).
riding model was used for refinement. The ethyl groups in compound 2
were disordered over two positions; thus, several restraints were used for
refinement. Residual density was located close to the tungsten atoms.
The crystals of compounds 2 and 3 were very sensitive and some decom-
position from the surface could not be avoided. In addition, the thin nee-
dles only provided very weak scattering power.
Synthesis of compound 2: Method A. A solution of [{(CO)5W}PH3]
(179 mg, 0.5 mmol) in CH2Cl2 (5 mL) was added to H3Al·NEt3 (65 mg,
0.5 mmol) and the mixture was stirred for 80 min at RT. At 48C, pale-
yellow crystals were obtained after 2 days. Yield: 377 mg (0.39 mmol,
78%). Method B. A 0.5 mmol solution of compound 1 was stored at 48C;
Crystal data for compound 1: C11H19AlNO5PW; M=487.06; triclinic;
¯
space group P1 (no. 2); a=7.1463(5), b=10.8385(7), c=12.0833(7) ꢃ;
a=86.767(5), b=75.650(6), g=76.089(6)8; V=880.12(10) ꢃ3; Z=2; m=
13.656 mmÀ1; F
ACTHNUTRGNE(NUG 000)=468; T=123(1) K; 6034 reflections measured; 2689
after 2 days, crystals of compound
2 start to grow. Yield: 302 mg
unique reflections (Rint =0.0486); R1 =0.0478 and wR2 =0.1267 for
I>2s(I).
(0.31 mmol, 70% from compound 1, 62% from the starting materials).
Analytical data for compound 2: 1H NMR (400 MHz, CD2Cl2, 300 K,
Crystal data for compound 2·CH2Cl2: C23H36Al2Cl2N2O10P2W2; M=
1037.07; orthorhombic; space group Pnma (no. 62); a=29.201(4), b=
TMS): d=À8.4 (br s; AlH), À8.1 (br s; AlH), À6.6 (br s; AlH), 0.52 (dt,
J
U
ACHTUNGTRENNUNG(H,H)=5 Hz; PH), 0.73 (dm, JACHTUGNTREN(NGUN P,H)=234 Hz; PH),
17.283(4), c=7.1718(5) ꢃ; V=3619.5(10) ꢃ3; Z=4; m=14.671 mmÀ1
;
E
ACHTUNGTRENUN(NG H,H)=7 Hz; CH3), 3.13 ppm
F(000)=2024; T=104(1) K; 8677 reflections measured; 2007 unique
reflections (Rint =0.0365); R1 =0.0418 and wR2 =0.1122 for I>2s(I).
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
Crystal data for compound 3: C28H30Al2N2O10P2W2; M=1038.12; mono-
clinic; space group P21/c (no. 14); a=10.307(3), b=18.925(4), c=
ACHTUNGTRENNUNG
9.143(2) ꢃ; b=97.87(2)8; V=1766.6(7) ꢃ3; Z=2; m=13.667 mmÀ1
;
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
F(000)=992; T=104(1) K; 10142 reflections measured; 1894 unique
reflections (Rint =0.0347); R1 =0.0248 and wR2 =0.0540 for I>2s(I).
AHCTUNGTRENNUNG
CD2Cl2, 300 K, 85% H3PO4): d=À276 (br s; PH), À283 ppm (br d,
J(P,P)=111 Hz; PH); MS (EI, 14 eV): m/z (%): 356 (28) [(CO)5WPH]+,
328 (2) [(CO)4WPH]+, 299 (30) [(CO)3WPH]+, 270 (18) [(CO)2WP]+, 241
(29) [(CO)WP]+, 214 (8) [WP]+, 184 3) [W]+, 101 (100) [NEt3]+; IR
(CH2Cl2): n˜ =2295 (w; PH), 2068 (s; CO), 2060 (m; CO), 1977 (s; CO),
1921 (vs; CO), 1830 (m; AlH), 1774 (m; AlH).
CCDC-897904 (1), CCDC-897905 (2), and CCDC-897906 (3) contain the
supplementary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
Synthesis of compound 3: A solution of [{(CO)5W}PH2Ph] (217 mg,
0.5 mmol) in CH2Cl2 (5 mL) was stirred with H3Al·NMe3 (45 mg,
0.5 mmol) at RT. After the evolution of gas ceased (1 h), the yellow solu-
tion was stored at 48C for 5 days to obtain yellow crystals of compound
3. Yield: 314 mg (0.30 mmol, 61%). The NMR data were taken from the
crude reaction mixture after 30 min and showed additional signals for
[{(CO)5W}P(H)(Ph)AlH2·NMe3] as an intermediate product that could
not be isolated.
Acknowledgements
This work was comprehensively supported by the Deutsche Forschungs-
gemeinschaft and the Alexander von Humboldt Foundation through the
reinvitation program (Prof. A. Y. Timoshkin). The COST action CM0802
PhoSciNet is gratefully acknowledged. A.Y.T. is grateful for a St. Peters-
burg State University research grant (12.37.139.2011).
Analytical data of compound 3: 1H NMR (400 MHz, CD2Cl2, 300 K,
TMS): d=2.46 (s; NMe3), 2.77 (s; NMe3), 4.0 (br s; AlH), 7.2–7.7 ppm
(m; Ph); 27Al NMR (104 MHz, CD2Cl2, 300 K, 1.5m Al
ACTHNUGTRNEGNU(NO3)3): d=
147 ppm (br s, w1/2 =1600 Hz; AlH); 31P NMR (162 MHz, CD2Cl2, 300 K,
85% H3PO4): d=À138 (br s; PPh), À140 (br s; PPh), À142 (br s; PPh),
À144 ppm (br s; PPh); 31P{1H} NMR (162 MHz, CD2Cl2, 300 K, 85%
H3PO4): d=À138 (br s; PPh), À140 (br s; PPh), À142 (br s; PPh),
[1] a) T. B. Marder, Angew. Chem. 2007, 119, 8262–8264; Angew.
Chem. Int. Ed. 2007, 46, 8116–8118; b) A. Staubitz, A. P. M. Robert-
son, I. Manners, Chem. Rev. 2010, 110, 4079–4124.
[2] P. Wang, Dalton Trans. 2012, 41, 4296–4302.
962
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 957 – 963

Dimerization of Lewis-Acid/Base-Stabilized Phosphanylalanes
FULL PAPER
[3] F. H. Stephens, R. T. Baker, M. H. Matus, D. J. Grant, D. A. Dixon,
Angew. Chem. 2007, 119, 760–763; Angew. Chem. Int. Ed. 2007, 46,
746–749.
[4] a) S.-K. Kim, W.-S. Han, T.-J. Kim, T.-Y. Kim, S. W. Nam, M. Mitor-
aj, L. Pieko, A. Michalak, S.-J. Hwang, S. O. Kang, J. Am. Chem.
Soc. 2010, 132, 9954–9955; b) B. L. Conley, T. J. Williams, Chem.
Commun. 2010, 46, 4815–4817; c) P. M. Zimmerman, A. Paul, C. B.
Musgrave, Inorg. Chem. 2009, 48, 5418–5433; d) A. Staubitz, A. P.
Soto, I. Manners, Angew. Chem. 2008, 120, 6308–6311; Angew.
Chem. Int. Ed. 2008, 47, 6212–6215.
Timoshkin, Coord. Chem. Rev. 2005, 249, 2094–2131; d) T. J. Clark,
K. Lee, I. Manners, Chem. Eur. J. 2006, 12, 8634–8648.
[14] a) U. Vogel, P. Hoemensch, K.-Ch. Schwan, A. Y. Timoshkin, M.
Scheer, Chem. Eur. J. 2003, 9, 515–519; b) U. Vogel, A. Y. Timosh-
kin, M. Scheer, Angew. Chem. 2001, 113, 4541–4544; Angew. Chem.
Int. Ed. 2001, 40, 4409–4412.
[15] a) U. Vogel, A. Y. Timoshkin, K.-Ch. Schwan, M. Bodensteiner, M.
Scheer, J. Organomet. Chem. 2006, 691, 4556–4564; b) A. Adolf, U.
Vogel, M. Zabel, A. Y. Timoshkin, M. Scheer, Eur. J. Inorg. Chem.
2008, 22, 3482–3492.
[5] a) ꢄ. Metin, V. Mazumder, S. ꢄzkar, S. Sun, J. Am. Chem. Soc.
2010, 132, 1468–1469; b) J.-M. Yan, X.-B. Zhang, S. Han, H. Shioya-
ma, Q. Xu, Angew. Chem. 2008, 120, 2319–2321; Angew. Chem. Int.
Ed. 2008, 47, 2287–2289.
[6] D. W. Himmelberger, L. R. Alden, M. E. Bluhm, L. G. Sneddon,
Inorg. Chem. 2009, 48, 9883–9889.
[7] H. Helten, A. P. M. Robertson, A. Staubitz, J. R. Vancel, M. F.
Haddow, I. Manners, Chem. Eur. J. 2012, 18, 4665–4680.
[8] a) G. C. Welch, R. R. San Juan, J. D. Masuda, D. W. Stephan, Science
2006, 314, 1124–1126; b) D. W. Stephan, G. Erker, Angew. Chem.
2010, 122, 50–81; Angew. Chem. Int. Ed. 2010, 49, 46–76.
[9] a) H. Dorn, R. A. Singh, J. A. Massey, A. J. Lough, I. Manners,
Angew. Chem. 1999, 111, 3540–3543; Angew. Chem. Int. Ed. 1999,
38, 3321–3323; b) H. Dorn, R. A. Singh, J. A. Massey, J. M. Nelson,
C. A. Jaska, A. J. Lough, I. Manners, J. Am. Chem. Soc. 2000, 122,
6669–6678.
[16] K.-Ch. Schwan, A. Y. Timoshkin, M. Zabel, M. Scheer, Chem. Eur.
J. 2006, 12, 4900–4908.
[17] M. Bodensteiner, U. Vogel, A. Y. Timoshkin, M. Scheer, Angew.
Chem. 2009, 121, 4700–4704; Angew. Chem. Int. Ed. 2009, 48, 4629–
4633.
[18] a) A. Y. Timoshkin, Inorg. Chem. 2009, 48, 8145–8153; b) A. H.
Cowley, R. A. Jones, M. A. Mardones, J. L. Atwood, S. G. Bott,
Angew. Chem. 1990, 102, 1504–1505; Angew. Chem. Int. Ed. Engl.
1990, 29, 1409–1410.
[19] R. D. Davy, H. F. Schaefer, J. Phys. Chem. A 1997, 101, 5707–5711.
[20] U. Vogel, K.-C. Schwan, M. Scheer, Eur. J. Inorg. Chem. 2004,
2062–2065.
[21] cf. supporting information.
[22] C. von Hꢁnisch, Z. Anorg. Allg. Chem. 2003, 629, 1496–1500.
[23] A. Y. Timoshkin, I. V. Kazakov, C. von Hꢁnisch, Russ. J. Gen. Chem.
2009, 79, 1067–1073.
[10] H. Westenberg, J. C. Slootweg, A. Hepp, J. Kçsters, S. Roters, A. W.
Ehlers, K. Lammertsma, W. Uhl, Organometallics 2010, 29, 1323–
1330.
[24] U. Vogel, M. Scheer, Z. Anorg. Allg. Chem. 2001, 627, 1593–1598.
[25] H. C. E. McFarlene, W. McFarlane, D. S. Rycroft, J. Chem. Soc.
Dalton Trans. 1976, 1616–1622.
[11] T. Habereder, H. Nçth, R. T. Paine, Eur. J. Inorg. Chem. 2007,
4298–4305.
[26] J. S. Cha, H. C. Brown, J. Org. Chem. 1993, 58, 3974–3979.
[27] J. K. Ruff, Inorg. Synth. 1967, 9, 30–37.
[12] a) T. L. Allen, W. H. Fink, Inorg. Chem. 1992, 31, 1703–1705;
b) T. L. Allen, A. C. Scheiner, H. F. Schaefer III, Inorg. Chem. 1990,
29, 1930–1936; c) M. B. Coolidge, W. T. Borden, J. Am. Chem. Soc.
1990, 112, 1704–1706; d) H.-J. Himmel, Dalton Trans. 2003, 3639–
3649; e) H.-J. Himmel, Eur. J. Inorg. Chem. 2003, 2153–2163;
f) A. Y. Timoshkin, Phosphorus, Sulfur, Silicon Relat. Elem. 2001,
168, 275–280.
[28] J. Cosier, A. M. Glazer, Appl. Cryst. 1986, 19, 105–107.
[29] R. C. Clark, J. S. Reid, Acta Cryst. 1995, A51, 887–897.
[30] Agilent Technologies, CrysAlisPro 2006–2012, Oxford, UK.
[31] A. Altomare, G. L. Cascarano, C. Giacovazzo, A. Gualardi, J. Appl.
Crystallogr. 1993, 26, 343–350.
[32] G. M. Sheldrick, Acta Cryst. 2008, A64, 112–122.
[13] a) S. Schulz, Adv. Organomet. Chem. 2003, 49, 225–317; b) B. Neu-
mꢀller, E. Iravani, Coord. Chem. Rev. 2004, 248, 817–834; c) A. Y.
Received: August 29, 2012
Published online: November 23, 2012
Chem. Eur. J. 2013, 19, 957 – 963
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
963