Why Does the Intramolecular Trimethylene-Bridged Frustrated Lewis Pair Mes2PCH2CH2CH2B(C6F5)2 Not Activate Dihydrogen?

Article abstract of DOI:10.1002/chem.201505200

The methyl labelled C₃-bridged frustrated phosphane borane Lewis pair (P/B FLP) 2 b was prepared by treatment of Mes₂PCl with a methallyl Grignard reagent followed by anti-Markovnikov hydroboration with Piers' borane [HB(C₆

Full text of DOI:10.1002/chem.201505200

DOI: 10.1002/chem.201505200
Full Paper
&
Frustrated Lewis Pairs
Why Does the Intramolecular Trimethylene-Bridged Frustrated
Lewis Pair Mes PCH CH CH B(C F ) Not Activate Dihydrogen?
2
2
2
2
6 5 2
[
a]
[a]
† [a]
† [a]
† [b]
Thomas Özgün, Ke-Yin Ye, Constantin G. Daniliuc , Birgit Wibbeling , Lei Liu ,
†
[b]
[a]
[a]
Stefan Grimme , Gerald Kehr, and Gerhard Erker*
Abstract: The methyl labelled C -bridged frustrated phos-
phane borane Lewis pair (P/B FLP) 2b was prepared by
that this was not due to kinetically hindered P···B dissocia-
tion of 2b. DFT calculations showed that the hydrogen-split-
ting reaction of the parent compound 2a is markedly ender-
3
treatment of Mes PCl with a methallyl Grignard reagent fol-
2
+
À
lowed by anti-Markovnikov hydroboration with Piers’ borane
gonic. The PH /BH H -splitting product of 2b was indirect-
2
+
À
[
HB(C F ) )]. The FLP 2b is inactive toward dihydrogen under
ly synthesized by a sequence of H /H addition. It lost H at
6
5 2
2
typical ambient conditions, in contrast to the C - and C -
ambient conditions and confirmed the result of the DFT
analysis.
2
4
bridged FLP analogues. Dynamic NMR spectroscopy showed
Introduction
sults is the behaviour of the trimethylene-bridged P/B system
2
a, which was shown to be inactive towards dihydrogen
[
9]
Frustrated Lewis pairs (FLPs) have undergone remarkable de-
velopment in recent years. The combination of nonquenched
bulky main group element based Lewis acids and bases has al-
lowed the detection of a variety of cooperative reactions.
Small molecule binding or activation, features usually attribut-
ed to transition metal containing systems, has become a promi-
nent property of many such Lewis acid/Lewis base combina-
under our typical reaction conditions. So far we were not
able to find reaction conditions where 2a was able to react
with H . We have investigated this at-first-sight strange differ-
2
ence of the behaviour of the FLPs 1 to 3 by a combined exper-
imental/theoretical study and were able to arrive at an explan-
ation for this dichotomy. This will be outlined in this article.
[
1]
tions. Intramolecular FLPs have played an important role in
[
2]
[3]
this development.
Scheme 1) and some of its derivatives, were shown to bind
a variety of oxides of the elements carbon, nitrogen and
The ethylene-bridged P/B FLP 1a
(
[
4]
[5]
sulfur or to assist in their specific transformation. The FLP
a has been one of the most active non-metallic dihydrogen
activators. It rapidly cleaves dihydrogen heterolytically and
has served as a catalyst for the hydrogenation reaction of a va-
Scheme 1. Intramolecular P/B frustrated Lewis pairs.
1
[
6]
[7]
riety of organic substrates. We had a look at the homologues
of the FLP 1a. The tetramethylene-bridged system 3a is readily Results and Discussion
available by means of a hydroboration reaction (see below). It
The question of the strength of the P/B interaction
was shown to be an active P/B FLP for the cleavage of dihy-
[8]
drogen under mild conditions. In stark contrast to these re-
We had previously shown that the activation barrier of the P/B
interaction in saturated vicinal P/B FLPs can be determined by
19
[
a] T. Özgün, Dr. K.-Y. Ye, Dr. C. G. Daniliuc, B. Wibbeling, Dr. G. Kehr,
Prof. Dr. G. Erker
dynamic F NMR spectroscopy by using suitably substituted
[
10]
chiral derivatives.
A typical example is compound 1b
Organisch-Chemisches Institut, Westfälische Wilhelms-Universität
Corrensstrasse 40, 48149 Münster (Germany)
E-mail: erker@uni-muenster.de
(
Scheme 2) which contains a pair of carbon chirality centres
with a defined relative configuration. The compound shows
a mutual P···B interaction in the solid state. In solution it fea-
tures a pair of diastereotopic C F ligands at boron at low tem-
[
b] Dr. L. Liu, Prof. Dr. S. Grimme
Mulliken Center for Theoretical Chemistry
Institut für Physikalische und Theoretische Chemie, Universität Bonn
Beringstraße 4, 53115 Bonn (Germany)
6 5
perature which undergo coalescence with increasing monitor-
ing temperature due to rapidly increasing equilibration with
the high lying open intermediate 1b (open) which features
a trigonal-planar boron coordination geometry with homotop-
[
†] C.G.D. and B.W. conducted the X-ray crystal structure analysis. L.L. and S.G.
conducted the computational chemistry study.
Supporting information for this article and ORCID for one of the authors
are available on the WWW under http://dx.doi.org/10.1002/
chem.201505200.
1
9
ic C6F substituents. From the dynamic F NMR spectra
5
°
a Gibbs activation energy of DG (298 K)=12.1(Æ0.3) kcal
Chem. Eur. J. 2016, 22, 5988 – 5995
5988
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
°
Table 1. Gibbs activation energies (DG ) of P/B dissociation in the FLPs
[a]
1
b to 3b.
[
10a]
1
b
2b
3b
°
DG
T
12.1
298
14.5
310
12.8
285
coal [K]
°
À1
19
[
a] DG = Æ0.3 kcalmol ; from dynamic F NMR spectroscopy at Tcoal
.
À1
mol was calculated for the P···B dissociation in compound
[10a]
1
b.
We needed to determine the favoured structure of the tri-
Figure 1. A view of the molecular structure of compound 2b (the R enantio-
mer is depicted; thermal ellipsoids are set at 50% probability). Selected
bond lengths [Š] and angles [8]: P1-C1 1.846(3), P1-B1 2.076(3), B1-C3
methylene-bridged FLP 2. At the same time it would be con-
venient to have a stereochemical label introduced that would
allow determination of the P···B dissociation barrier by the
NMR method. Therefore, we prepared the suitably labelled var-
iant 2b by the following synthetic route (Scheme 3). Dimesi-
tyl(methallyl)phosphane (4a) was prepared by the treatment
1.636(4), C3-C2 1.535(4), C2-C1 1.545(4), C1-P1-B1 94.3(1), C2-C3-B1 108.9(2),
B1-C3-C2-C1 À54.0(3), C3-C2-C1-P1 22.1(2), B1-P1-C1-C2 8.8(2).
of Mes PCl with methallylmagnesium chloride. The phosphane
In solution ([D ]toluene) compound 2b shows a broad
2
8
11
4
a was isolated as a white solid in 83% yield. It was character-
B NMR signal at d=À3.5 ppm at 213 K, typical of tetra-coor-
31
11
ized by C,H-elemental analysis and by NMR spectroscopy ( P:
d=À24.8 ppm, for further details see the Supporting Informa-
tion).
dinated boron. The B NMR chemical shift is practically tem-
perature invariant between 213 and 299 K; the signal only gets
31
sharper with increasing temperature. The P NMR resonance
was found at d=24.0 ppm (at 213 K). It also did not change
much with temperature. In contrast, compound 2b showed
1
temperature-dependent dynamic H (see the Supporting Infor-
19
mation) and F NMR spectra. At low temperature (253 K) we
19
observed four separate o-C6F5 F NMR signals, indicating hin-
dered rotation of both diastereotopic C F groups around the
6
5
B-C vector. In addition, at that temperature we monitored two
equal intensity p-C F resonances and four m-C F signals. In-
6
5
6 5
creasing the temperature overcomes the barrier of rotation
around the CÀB bond and the barrier of P···B dissociation. The
latter leads to pairwise equilibration of the ortho-, the para-
and the meta-C F resonances of the B(C F ) substituent. From
6
5
6 5 2
19
the coalescence of the pair of p-C6F5 F NMR signals we have
estimated the Gibbs activation energy of the P···B dissociation
°
in the trimethylene-bridged P/B FLP 2b as DG (310 K)=
1
4.5(Æ0.3) kcalmolÀ1.
Scheme 3. Formation of the FLP 2b.
Treatment of the methallylphosphane 4a with one molar
[11]
equivalent of Piers’ borane [HB(C F ) ] resulted in a clean
6
5 2
anti-Markovnikov hydroboration of the pendant C=C double
bond to give the P/B product 2b, which was isolated as
a white amorphous solid in 87% yield. It was characterized by
C,H-elemental analysis, by NMR spectroscopy and by X-ray dif-
fraction (single crystals were obtained from toluene/pentane
by the diffusion method). In the solid state compound 2b
shows a five-membered heterocyclic structure with a mutual
P···B interaction (Figure 1). Both the phosphorus and the boron
Scheme 2. Activation barrier of the P–B dissociation in the FLP 1b.
We exposed the FLP 2b to dihydrogen (for details see the
Supporting Information) but could not observe any formation
of the respective dihydrogen splitting product. However, the
system 2b reacted with the strongly s-donating Arduengo-car-
CCC
CCC
atom show tetra-coordination (SP1 =324.3, SB1 =334.9).
The five-membered core is found in a nonplanar distorted
twist conformation. Carbon atom C2 has the extra methyl
group attached.
[
12]
bene 5. Treatment of the P/B FLP 2b with the N-heterocyclic
carbene 5 went to completion within 30 min at ambient tem-
perature in benzene. Workup eventually gave the product 6 as
Chem. Eur. J. 2016, 22, 5988 – 5995
www.chemeurj.org
5989
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
3
1
a white solid in 70% yield. It showed a P NMR signal at d=
11
À25.7 ppm and an B NMR resonance at d=À14.0 ppm. The
1
H NMR spectrum of the [B]-NHC unit shows hindered rotation
around the B-C and N-C vectors which results in observation of
a total of six mesityl methyl signals and four CH methine reso-
1
9
nances of this unit at 299 K. We observe the F NMR signals of
a pair of diastereotopic C F substituents at boron. It shows
6
5
1
the H NMR features of a pair of diastereotopic mesityl groups
1
at phosphorus and the high field H NMR resonance of the
methyl group at the C -bridge.
3
Single crystals of compound 6 suitable for the X-ray crystal
structure analysis were obtained from dichloromethane. The
structure shows that the P···B bond was cleaved and the N-het-
erocyclic carbene ligand attached to the boron atom
Scheme 4. Formation and chemical features of the FLP 3b.
(
Figure 2). The C -bridge in compound 6 has attained a confor-
3
mation that results in a maximal spatial separation of the pair
of heteroatom based functional groups. The boron atom
shows a pseudotetrahedral coordination geometry; the phos-
CCC
phorus coordination geometry is trigonal-pyramidal (SP1
14.7).
=
3
Figure 3. A view of the molecular structure of compound 3b (thermal ellip-
soids are set at 30% probability). Selected bond lengths [Š] and angles [8]:
P1-C1 1.848(1), P1-B1 2.093(1), C1-C2 1.538(2), C2-C3 1.522(2), C3-C4 1.532(2),
C4-B1 1.640(2), C1-P1-B1 102.0(1), C2-C1-P1 122.3(1), C3-C4-B1 116.1(1), B1-
P1-C1-C2 À33.7(2), C1-P1-B1-C4 46.4(1), C1-C2-C3-C4 À45.5(2).
Figure 2. Molecular structure of the N-heterocyclic carbene addition product
(thermal ellipsoids are set at 30% probability). Selected bond lengths [Š]
6
and angles [8]: P1-C1 1.858(4), C1-C2 1.541(6), C2-C3 1.547(5), C3-B1 1.638(6),
C5-B1 1.683(6), C2-C1-P1 110.6(3), C1-C2-C3 108.7(3), C2-C3-B1 120.8(3), C3-
B1-C5 110.3(3), P1-C1-C2-C3 164.3(3), C2-C3-B1-C5 175.3(3).
(Figure 3). Carbon atom C3 bears the methyl substituent. We
note a marked phosphorus–boron interaction; the P1···B1
bond length in both the compounds 2b and 3b are almost
identical. In 3b both the phosphorus and the boron atom
show pseudotetrahedral coordination geometries.
We prepared the tetramethylene-bridged P/B FLP 4b for
comparison and chose a synthesis that resulted in the intro-
duction of a stereochemical label at carbon atom C3 in order
to allow the determination of the rate of P···B rupture by the
dynamic NMR method analogously as described above.
11
In solution (CD Cl ) compound 3b shows an B NMR signal
2
2
31
at d=À5.0 ppm (233 K) and a P NMR resonance at d=
1.7 ppm (233 K). These resonances are practically temperature
invariant in the 233 to 299 K temperature range. Compound
3b shows a similar dynamic behaviour as described for 2b. At
Dimesitylphosphorus chloride was treated with the respec-
tive Grignard reagent to give the alkenylphosphane 4b, isolat-
ed as a crystalline solid in 55% yield (Scheme 4). The com-
pound was characterized by C,H-elemental analysis, by X-ray
diffraction (see the Supporting Information for details, includ-
ing a projection of the molecular structure) and by spectrosco-
19
253 K (and below) compound 3b shows ten separate F NMR
signals of the pair of C F groups at boron, belonging to four
6
5
o-, two p- and four m-C F fluorine substituents (Figure 4). This
6
5
indicates a P···B interacting structure (just as observed by X-ray
diffraction in the solid state) and “frozen” rotation of both the
3
1
1
py ( P NMR: d=À18.6 ppm; H NMR: d=4.70, 4.68 ppm, = C F groups around the B-C vectors at these conditions on the
6
5
19
CH ). Treatment of 4b with HB(C F ) (one molar equiv) in tolu-
F NMR time scale. Increasing the monitoring temperature re-
sulted in coalescence of each the four o-C F signals, the pair
2
6 5 2
ene at r.t. (30 min) gave the hydroboration product 3b that
6
5
was isolated from the workup procedure as a white solid in
of p-C F resonances and the four m-C F signals, eventually
6 5 6 5
19
8
0% yield.
leading to a simple F NMR spectrum of 3b at high tempera-
ture that showed only three coalesced signals in a 2:1:2 inten-
sity ratio, originating from the o-, p- and m-C F fluorine atoms
The X-ray crystal structure analysis shows 1,4-attachment of
the PMes2 and B(C F ) groups at the saturated C -bridge
6
5 2
4
6 5
Chem. Eur. J. 2016, 22, 5988 – 5995
www.chemeurj.org
5990
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
19
2 2
Figure 4. Temperature-dependent dynamic F NMR (470 MHz, CD Cl ) spectra of compound 3b.
(
Figure 4). This behaviour indicated two dynamic processes
almost equally becoming facile with increasing temperature
that resulted in coalescence of the respective NMR signals,
namely onset of C F group rotation around the BÀC bonds
6
5
19
and P···B opening and closing becoming rapid on the F NMR
time scale. From the coalescence of the pair of p-C F resonan-
6
5
ces we estimated the Gibbs-activation energy of the equilibra-
tion of the global minimum structure of 3b with its high lying
°
À1
3
b(open) isomer as DG (280 K)=12.8(Æ0.3) kcalmol .
The C -bridged FLP 3b was exposed to dihydrogen at near
4
to ambient conditions (r.t., 1.5 bar H , 16 h). This resulted in an
2
Scheme 5. Catalytic hydrogenation reactions.
about 50% conversion to the dihydrogen splitting product 7b,
which was characterized spectroscopically from the reaction
1
mixture. It showed the characteristic [P]-H H NMR signal at
1
d=7.60 ppm with the large JPH ꢀ475 Hz coupling constant
under more forcing conditions at the 3b catalyst (5 mol%; for
further details see the Supporting Information).
31
11
(
P NMR: d=À9.2 ppm) and an B NMR resonance at d=
1
À21.7 ppm with J ꢀ86 Hz [in addition to the respective het-
Our study has shown that the three oligomethylene-bridged
Mes P-(CH ) -B(C F ) FLPs react differently with dihydrogen.
BH
31
eroatom signals of the remaining compound 3b ( P: d=
2
2 n
6 5 2
11
1
.8 ppm, B: d=À2.8 ppm in CD Cl at 299 K)].
Compounds 1a,b split dihydrogen very rapidly and have
served as metal-free catalysts for the hydrogenation of a variety
of substrates. The tetramethylene-bridged system 3a is also an
active dihydrogen splitting FLP; however, dihydrogen splitting
by its methyl-substituted derivative 3b is not complete but ar-
rives at about 1:1 equilibrium. The trimethylene-bridged
system 2b does not split dihydrogen under our typical reac-
tion conditions. The dynamic NMR study has shown that this is
probably not due to a total inhibition of the necessary P···B dis-
sociation. The P···B cleavage in 2b shows a by about 2 kcal
2
2
The analogous reaction was carried out with D to give 7b-
2
3
1
D . We monitored a P NMR 1:1:1 intensity triplet of the [P]-D
2
1
11
moiety at d=À9.5 ppm ( J =72.6 Hz) and a broad B NMR
PD
31
resonance at d=À22.0 ppm, again in the presence of the P/
11
2
B NMR resonances of the starting material 3b. The H NMR
spectrum of compound 7b-D showed signals at d=7.62 ppm
2
(
[P]-D) and d=2.76 ppm ([B]-D), respectively.
The FLP 3b turned out to be a reasonably active metal-free
hydrogenation catalyst for some bulky imines and an enam-
[13]
[7]
À1
ine (Scheme 5). Some slow and incomplete catalytic conver-
sion of the imines 8a,b to the respective sec-amines and the
enamine 10 to the corresponding tert-amine could be ach-
mol higher dissociation barrier as compared to its relatives
1b and 3b, respectively. However, this still represents a fast
opening (and closing) situation of the P···B unit. Consequently,
we could trap the 2b (open) form with an N-heterocyclic car-
bene. We conclude, that unfavourable kinetics of the genera-
tion of the active open FLP isomer is not likely to be the
reason for the observed inactivity of the FLPs 2a,b toward di-
ieved at close to normal conditions (1.5 bar H , r.t., 20 h,
2
1
0 mol% of 3b). With 30 bar H pressure complete imine hy-
2
drogenation was achieved at r.t. within 20 h with 5 mol% of
b. Similarly, the enamine was completely hydrogenated
3
Chem. Eur. J. 2016, 22, 5988 – 5995
www.chemeurj.org
5991
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
19]
hydrogen in contrast to 1a,b or 3a,b. There must be another
reason.
A previous DFT study that we had carried out had shown
that the ethylene-bridged P/B FLP 1a equilibrated with a reac-
tive open gauche-form that effected heterolytic cleavage of di-
hydrogen in a concerned fashion with a distorted chair-shaped
DFT analysis of the systems 1, 2 and 3 and their reaction
with dihydrogen
six-membered transition state. The active P/B FLP isomer 1a
open) was located by our DFT calculation about 9 kcalmol
À1
(
We investigated the reaction profile of the trio of the parent
ethylene-, trimethylene- and tetramethylene-bridged P/B FLPs
above the global 1a minimum (Figure 5, left). From our kinetic
19
study (dynamic F NMR, see above) we knew that 1a (open)
°
1
a, 2a, and 3a with density functional theory (DFT) calcula-
was populated by a rapid pre-equilibrium (DG ꢀ12 kcal
À1
tions. All the structures were fully optimized at TPSS level of
mol ), to be followed by rate determining H -cleavage [transi-
2
[
14]
À1
theory, with the BJ-damped variant of the DFT-D3 dispersion
tion state calculated about 9 kcalmol above 1a (open)] to
[
15]
[16]
+
À
correction in conjunction with the def2-TZVP basis set. Ac-
curate electronic energies were obtained from single-point cal-
give the respective zwitterionic Mes PH -CH CH BH (C F )
2 2 2 6 5 2
product. The overall dihydrogen splitting reaction at the FLP
[17]
À1
culations at the PW6B95-D3 level with a basis set of def2-
TZVP, and the solvation Gibbs free energies were computed by
employing COSMO-RS (conductor-like screening model for real
1a is strongly exergonic by À7.6 kcalmol .
The energy profile of the reaction of the tetramethylene-
bridged P/B FLP with dihydrogen looks similar. The rapid pre-
equilibrium generates the intermediate 3a (open) (+10.4 kcal
[18]
solvents) solvation model (for details, see the Supporting In-
formation).
À1
mol ) which then rapidly reacts via transition state TS 3a (+
2 2 3
Figure 5. Left: PW6B95-D3/def2-TZVP//TPSS-D3/def2-TZVP energy profiles of the H -splitting reactions of: a) C -bridged FLP 1a, b) C -bridged FLP 2a, and
c) C
4
-bridged FLP 3a. Notations used: CLO for closed isomer, OPE for open isomer, TS for transition state, and PD for hydrogenated product. Right: fully opti-
activation by FLP 1a to 3a. The hydrogen atoms are omitted for clarity, except H molecules. Colour
mized (TPSS-D3/def2-TZVP) transition states (TS) for H
2
2
legend: P yellow, B pink, F green, C black and H white. For closed isomers (CLO), open isomers (OPE) and hydrogenated products (PD), see Supporting Infor-
mation.
Chem. Eur. J. 2016, 22, 5988 – 5995
www.chemeurj.org
5992
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
À1
1
3.2 kcalmol above 3a (open); Figure 5, right) to give the H -
2
+
À
splitting product Mes PH -(CH ) -BH (C F ) (7a, that is, PD
2
2 4
6 5 2
3
a). In this case the overall reaction is slightly endergonic (+
À1
1
.5 kcalmol ) which is qualitatively in accord with our finding
of a reversible H -splitting reaction of the methyl-substituted
2
analogue 3b (see above).
According to the DFT calculation the overall appearance of
the reaction profile of the 2+H reaction is similar: the rapid
2
pre-equilibrium generates the P/B dissociated isomer 2a
À1
(
open; +14.6 kcalmol ). The calculated barrier of the subse-
À1
quent H -splitting reaction is a little higher [+11.1 kcalmol
2
above 2a (open)] than found for the ethylene-bridged P/B FLP
Figure 6. Molecular structure of the phosphonium/boron triflate compound
1
a, but lower than tetramethylene-bridged P/B FLP 3a. How-
1
2b (thermal ellipsoids are set at 30% probability). Selected bond lengths
Š] and angles [8]: P1-C1 1.811(2), B1-O1 1.603(3), B1-C3 1.607(4), C1-C2
1.545(3), C2-C3 1.546(3), C2-C1-P1 112.4(2), O1-B1-C3 108.5(2), C2-C3-B1
20.4(2), P1-C1-C2-C3 À167.0(2), C1-C2-C3-B1 65.3(3), O1-B1-C3-C2 À76.3(3).
ever, in contrast to the other two examples the formation of
[
+
À
the hydrogen splitting product Mes PH -(CH ) -BH (C F )
6 5 2
2
2 3
1
(
(
11 a; i.e., PD 2a) in this case is markedly endergonic
+8.6 kcalmol ).
À1 [20]
We conclude that our DFT analysis has revealed the reason
for the different behaviour of the trimethylene-bridged P/B
FLP system 2 as opposed to its ethylene- or tetramethylene-
bridged neighbours: this is due to a thermodynamic effect. In
attached at boron and the opened P/B FLP framework
CCC
(SP1 =342.5(1)).
The hydrocarbyl framework of compound 12b shows a con-
formational antiperiplanar (q P1-C1-C2-C3 À167.0(2)8)/gauche
(q C1-C2-C3-B1 65.3(3)8) arrangement.
solution both the H -splitting reactions by the FLPs 1 and 3
2
are overall exergonic or slightly endergonic, but the trimethy-
lene-bridged system 2 here in question, is markedly endergon-
We then exchanged triflate at boron for hydride. This was ef-
fected by treatment of the triflate 12b with chlorodimethylsi-
lane at r.t. The reaction was carried out in a Young NMR tube
and the reaction was followed by NMR spectroscopy over
ic and, therefore, is not observed under our typical H -splitting
2
conditions.
a period of about 2 days. After the addition of Me Si(H)Cl to
2
compound 12b in CD Cl at r.t. we directly observed a gas
2
2
1
Experimental confirmation of the endothermicity of the H2-
splitting product 11 b at the trimethylene-bridged P/B FLP
core
evolution. Nevertheless, H NMR spectroscopy revealed that
+
À
the PH /BH product 11 b had apparently been formed. We
have monitored the signals of 11 b and 2b in a about 1:1 ratio.
3
1
Compound 11 b shows a P NMR doublet in CD Cl2 at d=
Since we could not prepare the H -splitting product 11 b by re-
2
2
1
1
31
À16.2 ppm ( J ꢀ479 Hz) ( H: d=7.83 ppm) [of 2b: P d=
acting the P/B FLP 2b with dihydrogen, we decided to attach
PH
11
1
2
3.6 ppm] and an B signal at d=À22.2 ppm ( J ꢀ86 Hz). We
a proton and a hydride in two consecutive steps. We em-
BH
19
[21]
have monitored the F NMR signals of the pair of diastereo-
ployed a procedure that we had used before.
topic C6F substituents of compound 11 b at boron at d=
Trifluoromethane sulfonic acid was slowly added to the P/B
5
À133.3, À133.8 ppm (o), d=À164.7, À164.9 ppm (p) and d=
À167.0, À167.2 ppm (m). After standing for 2 d at r.t. the dehy-
FLP 2b in dichloromethane at r.t. Workup eventually gave the
phosphonium/borane triflate 12b as an amorphous solid in
+
À
+
drogenation of the PH /BH product 11 b under these condi-
tions was complete and we only observed the characteristic
NMR signals of the P/B FLP 2b (the spectra are depicted in the
Supporting Information). We carried out a second experiment
at 08C and were actually able to isolate compound 11 b rea-
sonably pure. We then observed the cleavage of dihydrogen at
r.t. with formation of 2b (for details see the Supporting Infor-
mation).
8
1% yield. Compound 12b features a set of typical [P]H
1
1
31
resonances ( H NMR: d=7.69 ppm, JPH ꢀ473 Hz; P: d=
1
9
À17.9 ppm) and we have monitored the typical F NMR reso-
nance of the newly introduced -OSO CF group (d=
2
3
À79.0 ppm) at boron (Scheme 6).
The X-ray crystal structure analysis of compound 12b
(
Figure 6, single crystals were obtained from a dichloromethane
solution at À408C) showed the presence of the triflate group
Conclusions
Our study has led us to a likely solution of the initial question.
It seems that we have found an explanation for the difference
of behaviour between the FLPs 2a,b versus 1a,b and 3a,b
toward dihydrogen. We have found that all three types of
compounds show a marked P···B interaction in their global
minimum structures, but all three open and close rapidly to
generate the respective reactive open P/B FLP intermediates in
Scheme 6. Independent synthesis of the zwitterionic product 11 b.
Chem. Eur. J. 2016, 22, 5988 – 5995
www.chemeurj.org
5993
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
low activation barrier pre-equilibrium steps. The activation of
neither of these systems is kinetically problematic. Our DFT
study has revealed that the observed differences in FLP behav-
iour are most likely thermodynamic in origin. It was shown
that the overall hydrogen splitting reaction of the trimethylene
FLP 2a is markedly endergonic, in contrast to the slightly en-
dergonic reaction of the compound 3a and the exergonic re-
action of 1a with dihydrogen. This computational result was
confirmed by an alternative stepwise synthesis of the formal
hydrogen-splitting product 11 b, which was shown to com-
Synthesis of 2b: A solution of phosphane 4a (100.0 mg, 0.3 mmol,
1
.0 equiv) in toluene (1 mL) was added dropwise to a solution of
Piers’ borane [HB(C F ) ] (106.6 mg, 0.3 mmol, 1.0 equiv) in toluene
6
5 2
(
2 mL) and stirred at r.t. for 30 min. The suspension was filtered
through glass fibre filter (Whatman), all volatiles of the filtrate were
removed in vacuo and the residue was washed with pentane (4”
2
mL). After being dried in vacuo for 6 h compound 2b was ob-
tained as a powdery white solid (178.7 mg, 0.3 mmol, 87%). Crys-
tals suitable for the X-ray crystal structure analysis were obtained
by slow diffusion of pentane into a solution of compound 2b in
toluene. Elemental analysis calcd (%) for C H BF P: C 60.92, H
34 30
10
4
.51; found: C 60.24, H 4.36.
pletely lose H at r.t. as expected for an endothermic com-
2
Synthesis of 3b:
A solution of phosphane 4b (300.0 mg,
pound.
0
.9 mmol, 1.0 equiv) in toluene (2 mL) was added dropwise to a so-
This study served to solve a rather specific problem in FLP
chemistry. However, at the same time its result pointed to
a rather general feature of this chemistry and has reminded us
to check for the thermodynamics of any of the reactions of the
many new FLP systems when an apparent lack of reactivity or
lution of Piers’ borane [HB(C F ) ] (306.6 mg, 0.9 mmol, 1.0 equiv) in
toluene (3 mL) and stirred at r.t. for 30 min. Then all volatiles were
removed in vacuo and the obtained residue redissolved in pentane
6
5 2
(
5 mL). The solution was stored at À308C for 16 h. The pentane
was removed with a pipette and the white solid was washed with
pentane (4”2 mL). After drying the solid in vacuo for 4 h com-
pound 3b was obtained as a powdery white solid (488.6 mg,
[22]
some unusual reaction branching is observed. Often the ki-
netic feasibilities do not seem to pose any problems but spe-
cific thermodynamic features might in some cases pose serious
restrictions. We hope that these findings will be helpful in the
current rapid development of frustrated Lewis pair chemistry.
0
.5 mmol, 80%). Crystals suitable for the X-ray crystal structure
analysis were obtained by slow diffusion of pentane into a solution
of compound 3b in toluene at À358C. Elemental analysis calcd (%)
for C H BF P: C 61.42, H 4.71; found: C 61.38, H 4.68.
35
32
10
Synthesis of 6: Compound 2b (32.1 mg, 0.048 mmol, 1.0 equiv)
was treated with 1,3-bis(2,4,6-trimethylphenyl)-imidazol-2-ylidene
(
3
5; 14.6 mg, 0.048 mmol, 1.0 equiv) in benzene (2 mL) at r.t. for
0 min. Then all volatiles were removed in vacuo and the obtained
Experimental Section
Synthesis and characterisation
white solid was washed with pentane (2”2 mL). After being dried
in vacuo, compound 6 was obtained as a powdery white solid
Synthesis of 4a: 2-Methylallylmagnesium chloride (12.00 mL,
(
34.2 mg, 0.035 mmol, 70%). Crystals suitable for the X-ray crystal
6
.00 mmol, 1.25 equiv) was added to a solution of dimesitylchloro-
phosphane (1.46 g, 4.80 mmol, 1.00 equiv) in tetrahydrofuran
20 mL) at 08C. The reaction mixture was stirred at ambient tem-
structure analysis were obtained from a dichloromethane solution
of compound 6. Elemental analysis calcd (%) for C H BF N P: C
55
54
10
2
(
6
7.77, H 5.58, N 2.87; found: C 66.51, H 5.82, N 2.74.
Synthesis of 11 b: A solution of chlorodimethylsilane (210.0 mg,
47.5 mL, 2.2 mmol, 5.0 equiv) in CH Cl (1.0 mL) was added to a so-
perature for 3 h, before all volatiles were removed in vacuo. The
residue was redissolved in pentane (30 mL) and the resulting sus-
pension was filtered via cannula (Whatman glass fibre filter). The
colourless filtrate was dried in vacuo to give compound 4a as
a powdery white solid (1.35 g, 4.2 mmol, 83%). Elemental analysis
calcd (%) for C H P: C 81.44, H 9.01; found: C 81.20, H 9.68.
2
2
2
lution of compound 12b (369.0 mg, 0.4 mmol, 1.0 equiv) in CH Cl
2
2
(
2.0 mL) at À788C. The reaction mixture was warmed to 08C and
stirred for 3 h at that temperature, before pentane (15 mL) was
added at À788C. The solution was stored at À358C for 16 h and
the suspension filtered through a glass fibre filter (Whatman). The
resulting solid was washed with pentane (3”3 mL) at À788C and
then dried in vacuo to give a white powdery solid (247.1 mg,
2
2
29
Synthesis of 4b: 1st Step: Generation of the Grignard-reagent: Pow-
dered magnesium (0.6 g, 25.0 mmol, 1.0 equiv) was suspended in
tetrahydrofuran (50 mL) under an argon atmosphere in a dried
three-necked round-bottom flask with reflux condenser. A solution
of 4-bromo-2-methyl-1-butene (3.7 g, 25.0 mmol, 1.0 equiv) in tet-
rahydrofuran (40 mL) was added dropwise using an addition
funnel. The reaction mixture was refluxed for 4 h and then stirred
for another 16 h at room temperature. A conversion to the
Grignard-reagent of 100% was assumed for the calculation of the
stoichiometry in the next step. 2nd Step: Preparation of dimesityl(3-
methylbut-3-en-1-yl)phosphane: A solution of dimesitylchlorophos-
phane (6.1 g, 20.0 mmol, 0.8 equiv) in tetrahydrofuran (60 mL) was
added dropwise at 08C to the Grignard-solution and stirred for
0
.37 mmol, 82%).
Synthesis of 12b: A solution of compound 2b (200.0 mg,
.30 mmol, 1.0 equiv) in dichloromethane (2.0 mL) was added
dropwise to a solution of triflic acid (44.4 mg, 25.8 mL, 0.30 mmol,
.0 equiv) in dichloromethane (2.0 mL) at room temperature in
0
1
a Schlenk flask. The reaction mixture was stirred for 30 min at
room temperature and then all volatiles were removed in vacuo.
The resulting white solid was washed with pentane (8”2 mL) and
dried in vacuo to give
a white powdery solid (200.4 mg,
0
.24 mmol, 81%). Crystals suitable for the X-ray crystal structure
2
0 h at ambient temperature. All volatiles were removed in vacuo
analysis were obtained from a solution of compound 12b in di-
chloromethane at À408C. Elemental analysis calcd (%) for
C H BF O PS: C 51.24, H 3.81; found: C 51.07, H 3.66.
and pentane (120 mL) was added to the sticky yellow residue. The
pale yellow suspension was filtrated via cannula (Whatman glass
fibre filter) and the obtained filtrate was dried in vacuo to give
a pale yellow oil. The oil was purified by column chromatography
35 31
13
3
(
(
silica: CH Cl :CyH=3:10; R : 0.44) to give a white, crystalline solid
2 2 f
3.6 g, 10.7 mmol, 55%). Crystals suitable for the X-ray crystal struc-
Acknowledgements
ture analysis were obtained from a dichloromethane solution of
compound 4b. Elemental analysis calcd (%) for C H P: C 81.62, H
Financial support from the Deutsche Forschungsgemeinschaft
and the European Research Council is gratefully acknowl-
2
3
31
9
.23; found: C 81.32, H 8.93.
Chem. Eur. J. 2016, 22, 5988 – 5995
www.chemeurj.org
5994
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
edged. The work of S.G. and L.L. was supported by the DFG in
the framework of the “Gottfried-Wilhelm-Leibniz” prize.
[11] a) D. J. Parks, R. E. v. H. Spence, W. E. Piers, Angew. Chem. Int. Ed. Engl.
1995, 34, 809–811; Angew. Chem. 1995, 107, 895–897; b) D. J. Parks,
W. E. Piers, G. P. A. Yap, Organometallics 1998, 17, 5492–5503.
12] a) A. J. Arduengo III, H. V. Rasika Dias, R. L. Harlow, M. Kline, J. Am. Chem.
Soc. 1992, 114, 5530–5534; b) F. E. Hahn, M. C. Jahnke, Angew. Chem.
Int. Ed. 2008, 47, 3122–3172; Angew. Chem. 2008, 120, 3166–3216;
c) M. N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature 2014, 510,
[
[
Keywords: boron
·
density functional calculations
·
dihydrogen activation · frustrated Lewis pair · phosphorus
4
85–496.
13] See for a comparison: a) Z. M. Heiden, D. W. Stephan, Chem. Commun.
011, 47, 5729–5731; b) P. Jochmann, D. W. Stephan, Angew. Chem. Int.
[
1] D. W. Stephan, G. Erker, Angew. Chem. Int. Ed. 2015, 54, 6400–6441;
Angew. Chem. 2015, 127, 6498–6541.
2
[
[
2] G. Kehr, S. Schwendemann, G. Erker, Top. Curr. Chem. 2013, 332, 45–83.
3] P. Spies, G. Erker, G. Kehr, R. Frçhlich, S. Grimme, D. W. Stephan, Chem.
Commun. 2007, 5072–5074.
Ed. 2013, 52, 9831–9835; Angew. Chem. 2013, 125, 10014–10018;
c) J. A. Hatnean, J. Thompson, P. A. Chase, D. W. Stephan, Chem.
Commun. 2014, 50, 301–303.
[
[
4] D. W. Stephan, G. Erker, Chem. Sci. 2014, 5, 2625–2641.
5] See for example, a) M. Sajid, A. Stute, A. J. P. Cardenas, B. J. Culotta,
J. A. M. Hepperle, T. H. Warren, B. Schirmer, S. Grimme, A. Studer, C. G.
Daniliuc, R. Frçhlich, J. L. Petersen, G. Kehr, G. Erker, J. Am. Chem. Soc.
[
[
14] J. Tao, J. P. Perdew, V. N. Staroverov, G. E. Scuseria, Phys. Rev. Lett. 2003,
9
1, 146401.
15] a) S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
1
1
54104; b) S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32,
456–1465.
2
012, 134, 10156–10168; b) M. Sajid, L.-M. Elmer, C. Rosorius, C. G. Dan-
iliuc, S. Grimme, G. Kehr, G. Erker, Angew. Chem. Int. Ed. 2013, 52, 2243–
246; Angew. Chem. 2013, 125, 2299–2302; c) M. Sajid, G. Kehr, C. G.
[16] a) F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–3305;
2
b) F. Weigend, Phys. Chem. Chem. Phys. 2006, 8, 1057–1065.
Daniliuc, G. Erker, Angew. Chem. Int. Ed. 2014, 53, 1118–1121; Angew.
Chem. 2014, 126, 1136–1139; d) J. C. M. Pereira, M. Sajid, G. Kehr, A. M.
Wright, B. Schirmer, Z.-W. Qu, S. Grimme, G. Erker, P. C. Ford, J. Am.
Chem. Soc. 2014, 136, 513–519.
6] S. M. Whittemore, G. Edvenson, D. M. Camaioni, A. Karkamkar, D. Neiner,
K. Parab, T. Autrey, Catal. Today 2015, 251, 28–33.
[
[
17] Y. Zhao, D. G. Truhlar, J. Phys. Chem. A 2005, 109, 5656–5667.
18] a) A. Klamt, G. Schürmann, J. Chem. Soc. Trans. 2 1993, 799–805; b) A.
Klamt, J. Phys. Chem. 1995, 99, 2224–2235.
[
[
19] S. Grimme, H. Kruse, L. Goerigk, G. Erker, Angew. Chem. Int. Ed. 2010, 49,
[
[
1402–1405; Angew. Chem. 2010, 122, 1444–1447.
20] See for a comparison: M. Sajid, A. Lawzer, W. Dong, C. Rosorius, W.
Sander, B. Schirmer, S. Grimme, C. G. Daniliuc, G. Kehr, G. Erker, J. Am.
Chem. Soc. 2013, 135, 18567–18574.
21] J. Yu, G. Kehr, C. G. Daniliuc, C. Bannwarth, S. Grimme, G. Erker, Org.
Biomol. Chem. 2015, 13, 5783–5792.
22] See for a comparison: a) H. Wang, R. Frçhlich, G. Kehr, G. Erker, Chem.
Commun. 2008, 5966–5968; b) L. Greb, P. OÇa-Burgos, B. Schirmer, S.
Grimme, D. W. Stephan, J. Paradies, Angew. Chem. Int. Ed. 2012, 51,
7] a) P. Spies, S. Schwendemann, S. Lange, G. Kehr, R. Frçhlich, G. Erker,
Angew. Chem. Int. Ed. 2008, 47, 7543–7546; Angew. Chem. 2008, 120,
7
654–7657; b) M. Lindqvist, K. Borre, K. Axenov, B. Kotai, M. Nieger, M.
[
[
Leskela, I. Papai, T. Repo, J. Am. Chem. Soc. 2015, 137, 4038–4041; c) A.
Feldmann, G. Kehr, C. G. Daniliuc, C. Mück-Lichtenfeld, G. Erker, Chem.
Eur. J. 2015, 21, 12456–12464.
[
[
8] X. Wang, G. Kehr, C. G. Daniliuc, G. Erker, J. Am. Chem. Soc. 2014, 136,
3
293–3303.
10164–10168; Angew. Chem. 2012, 124, 10311–10315.
9] a) Dissertation P. Spies, Münster, 2008; b) P. Spies, R. Frçhlich, G. Kehr,
G. Erker, S. Grimme, Chem. Eur. J. 2008, 14, 333–343, (corrigendum:
Chem. Eur. J. 2008, 14, 779); c) P. Spies, G. Kehr, K. Bergander, B. Wibbel-
ing, R. Frçhlich, G. Erker, Dalton Trans. 2009, 1534–1541.
[
10] a) K. Axenov, C. M. Mçmming, G. Kehr, R. Frçhlich, G. Erker, Chem. Eur. J.
010, 16, 14069–14073; see also b) S. Schwendemann, R. Frçhlich, G.
Kehr, G. Erker, Chem. Sci. 2011, 2, 1842–1849.
2
Received: December 30, 2015
Published online on March 21, 2016
Chem. Eur. J. 2016, 22, 5988 – 5995
www.chemeurj.org
5995
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim