Mechanistic insights into a supramolecular self-assembling catalyst system: Evidence for hydrogen bonding during rhodium-catalyzed hydroformylation

Article abstract of DOI:10.1002/anie.201203768

The structural integrity and flexibility provided by intermolecular hydrogen bonds leads to the outstanding properties of the 6- diphenylphosphinopyridin-(2H)-1-one ligand (see scheme) in the rhodium-catalyzed hydroformylation of terminal alkenes, as demonstrated by the combination of spectroscopic methods and DFT computations. Hydrogen bonds were also detected in a competent intermediate of the catalytic cycle. Copyright

Full text of DOI:10.1002/anie.201203768

Angewandte
Chemie
DOI: 10.1002/anie.201203768
Supramolecular Catalysis
Mechanistic Insights into a Supramolecular Self-Assembling Catalyst
System: Evidence for Hydrogen Bonding during Rhodium-Catalyzed
Hydroformylation**
Urs Gellrich, Wolfgang Seiche, Manfred Keller, and Bernhard Breit*
The hydroformylation of olefins is one of the most important
applications of homogeneous catalysis in industry, annually
producing millions of tons of aldehydes.[1] The selectivity
towards the desired linear aldehyde is usually achieved by
crafting the microenvironment of the catalytically active
metal species through the binding of appropriate ligand
architectures. Amongst the many ligands developed for this
process, bidentate phosphines with a wide bite angle such as
Xantphos (2; Scheme 1) are particularly effective because of
activity and selectivity in the hydroformylation of terminal
alkenes allows reactions to be carried out at room temper-
ature and ambient pressure.
This has enabled the development of tandem processes
such as a tandem hydroformylation/asymmetric organocata-
lytic cross-aldol reaction.[5] More recently, the utility of this
system was demonstrated by the devlopement of a highly
selective tandem hydroformylation/hydrogenation process.[6]
The hydrogen-bonding properties of this ligand system have
been intensively studied in a [Cl2Pt(3)2] complex.[7] However,
evidence that the proposed hydrogen-bonding interaction
occurs during the catalytic reaction was lacking. Herein, we
present a mechanistic investigation that proves the existence
of hydrogen bonding during the catalytic cycle, and demon-
strate its importance for catalyst activity and selectivity. We
were able to characterize an intermediate of the catalytic
cycle, the acyl complex [(COR)Rh(3)2(CO)2], and demon-
strate its catalytic competence.
Our investigation commenced with attempts to prepare
an analogue to the Wilkinson complex. Pleasingly, the
reaction of [Rh(CO)2(acac)] (acac = acetylacetanoate) with
ligand 3 under syngas pressure furnished [HRh(3)3CO] (4) in
80% yield as an orange crystalline compound. Thorough 1H,
31P, 103Rh NMR and IR spectroscopic studies indicated the
presence of hydrogen bonds between one ligand in the
hydroxypyridine form and one ligand in the pyridone form.
Furthermore, detailed analysis of variable-temperature NMR
spectra showed that the rhodium complex 4 exists as a dimer,
the structure of which could be determined by X-ray
diffraction (Figure 1).
Scheme 1. Selectivity and activity observed in the rhodium-catalyzed
hydroformylation of 1-octene with various ligands. Conditions:
Rh/L/substrate=1:20(10):7500, 808C, 10 bar, toluene.
their high regioselectivity for the linear aldehyde (high n/iso
ratio).[2] Previously, we demonstrated that monodentate
phosphorus ligands possessing hydrogen-bond recognition
motifs are able to achieve levels of selectivity previously only
attained with bidentate ligands, together with outstanding
activities.[3,4] The most prominent example of such supra-
molecular self-assembling ligands is the 6-diphenylphosphi-
nopyridin-2-(1H)-one (6-DPPon; 3) system. Its exceptional
By pressurizing complex 4 with CO (4 bar) we were able
to detect the formation of [HRh(3)2(CO)2] (5) by in situ IR
spectroscopy. DFT calculations enabled the vibrations of this
complex to be assigned to equatorial/equatorial (eqeq) and
axial/equatorial (axeq) conformers, and are in good agree-
ment with those observed experimentally. A broad absorption
[*] U. Gellrich, Dr. W. Seiche, Dr. M. Keller, Prof. Dr. B. Breit
Institut fꢀr Organische Chemie und Biochemie
Freiburg Institute for Advanced Studies (FRIAS)
Albert-Ludwigs-Universitꢁt Freiburg
between 2300 cmꢀ1 and 3500 cmꢀ1 corresponds to the O
ꢀ
ꢀ
H···O and N H···N vibrations of the hydrogen-bonding
network. The unusual shape of these vibrations is caused by
a fast double-proton transfer process, which is in accordance
with our previous studies on the [Cl2Pt(3)2] complex.[6]
Complex 5 was also produced by the reaction between
[Rh(CO)2(acac)] and 10 equiv 3 in a CO/H2 atmosphere and
was identified as the resting state of the catalytic trans-
formation through in situ IR spectroscopy.[8] This finding
indicates, in agreement with kinetic studies, which show a first
order dependency of the TOF on the alkene concentration, an
early rate-determining step in the catalytic cycle.[9] To gain
Albertstrasse 21, 79104 Freiburg i. Brsg. (Germany)
E-mail: bernhard.breit@chemie.uni-freiburg.de
[**] Support from the Freiburg Institute for Advanced Studies (FRIAS),
the Deutsche Forschungsgemeinschaft (IRTG 1038), and the Fonds
der chemischen Industrie (PhD fellowship to U.G.) is gratefully
acknowledged. We thank S. Diezel for support with the kinetic
experiments, J. Schꢁfer for help with X-ray structure analysis, and Dr.
M. Cooke and Prof. Dr. D. Plattner for stimulating discussions.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201203768.
Angew. Chem. Int. Ed. 2012, 51, 11033 –11038
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11033

.
Angewandte
Communications
a deeper insight into the mechanism we performed a detailed
DFT study on the catalytic cycle (Figure 2).[10] Several
conformers (e.g. cis/trans and eqeq/axeq) and tautomers
were optimized for each intermediate and transition state
(TS). Propene was used as a model for the 1-octene used
experimentally. Despite the fact that many previous studies
have focused on the TS for hydrometalation,[11] less attention
has been paid to alkene coordination.[12] We were able to
locate and optimize transition states for alkene coordination,
but the hydrometalation remains rate-determining on the
computed DG surface (Figure 2).
Comparison of the DFT-optimized hydrogen-bond geo-
metries of the pentacoordinated [HRh(3)2(CO)2] (5) complex
(P-Rh-P: 1128) and the catalytically active species trans-
[HRh(3)2(CO)] (6; P-Rh-P: 1548) shows an elongation of the
ꢀ
N H···N bond by 0.5 ꢀ in the latter (Figure 3). In contrast,
ꢀ
the O H···O bond length remains unchanged. This observa-
ꢀ
tion indicates that the N H···N bond corresponds to the
flexible part of the hydrogen-bonding network, whereas the
ꢀ
O H···O bond is responsible for the chelating character of the
ligand backbone.[13] This synergism of flexibility and structural
integrity is likely to be a key for understanding the excep-
tional high activity of this ligand compared to covalently
bound bidentate ligands, since it facilitates the adoption of
different coordination geometries and bite angles without
severe energy penalties. In agreement with this assumption,
Figure 1. X-ray structure of the [HRh(6-DPPon)3(CO)] (4) complex.
ꢀ
Selected bond lengths [ꢂ] and angles [8]: N H···O: 2.880(3), 164(3);
ꢀ
ꢀ
O H···O: 2.657(3), 170(4); N H···N: 2.892(3), 169(3); P-Rh-P:
115.23(3).
Figure 2. Free energy surface of the catalytic cycle (M06/6-311+G(2d,p)[C,H,N,O,P]+SDD[Rh] using the PCM model for toluene on B3LYP/6-
31G(d,p)[C,H,N,O,P]+LANL2DZ[Rh]-optimized structures. Thermodynamic corrections were obtained from a frequency calculation at the B3LYP
level).
11034 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11033 –11038

Angewandte
Chemie
Figure 4. B3LYP-optimized structures for the TS of the prolinear hydro-
metalation with hydrogen bonds (left), and a conformer where hydro-
gen bonding is prevented for geometrical reasons (right).
Figure 3. B3LYP-optimized structures and hydrogen-bond lengths of
the resting state (5) and the actually catalytic active species, the
trans-[HRh(3)2(CO)] (6) complex.
and activity, we synthesized O- and N-methylated derivatives
of 3 (12 and 13) so as to prevent the possibility of hydrogen
bonding and then evaluated them in the hydroformylation of
1-octene under identical conditions (Scheme 2).[20] Indeed 12
and 13 showed selectivities comparable to classical mono-
the reported lowest energy pathways for the Xantphos- and
Thixantphos-catalyzed hydroformylation involves cis-coordi-
nated hydride species.[2,14]
Interestingly, the lowest TS for the prolinear hydrometa-
lation and the probranched hydrometalation do not start from
the same alkene complex, which was confirmed by IRC
calculations (Figure 2).[15] The prolinear alkene complex is
higher in energy, and therefore the resulting DGTS value is
smaller; this is an interesting example of the concept of
survival of the weakest.[16] The n/iso ratio, calculated from the
two different DGTS values, is in good agreement with the
experimental observations (Table 1).[17] The computed free
energy surface shows that the prolinear hydrometalation step
is exothermic (DGr = ꢀ6.18 kcalmolꢀ1) whereas the pro-
branched hydrometalation is slightly endothermic (DGr =
1.35 kcalmolꢀ1).
Table 1: Comparison of the computed selectivity and the experimentaly
obtained selectivity.
Scheme 2. Selectivity observed in the rhodium-catalyzed hydroformyla-
tion of 1-octene with various ligands. Conditions: Rh/L/substrate=
1:20:7500, 808C, 10 bar, toluene.
B3LYP[a]
M06[b]
Expt.
DDGTS [kcalmolꢀ1
n/iso
]
1.17
88:12
1.75
95:5
–
96:4
dentate ligands such as PPh3 (1). Furthermore, the reactions
with 12 and 13 were significantly slower than with the 6-
DPPon (3) catalyst, thus highlighting the importance of the
hydrogen bonds for the outstanding activity.
[a] 6-31G(d,p)[C,H,N,O,P]+LANL2DZ[Rh]. [b] 6-311+G(2d,p)-
[C,H,N,O,P]+SDD[Rh] (PCM for toluene).
In agreement with the computed reaction enthalpies,
deuteroformylation experiments reveal that deuterium incor-
poration occurs only in the C1-position of the alkene after
20% conversion, thus indicating that the probranched hydro-
metalation is reversible.[18] However, it is likely that the
reversibility depends on the CO pressure since the dicarbonyl
species 9 is thermodynamically favored, which is reflected in
a lower regioselectivity at higher pressures.[19] Therefore, the
regioselectivity seems to be due to a combination of kinetic
and thermodynamic control. We then focused on the opti-
mization of several transition states for hydrometalation in
which hydrogen bonding is prevented by geometrical con-
straints (Figure 4). In these transition states the pyridone
tautomer is favored, which agrees with previous investiga-
tions.[7] However, M06 and B3LYP calculations predict that
hydrogen bonding stabilizes the rate-determining TS by at
least 7 kcalmolꢀ1 (Figure 4).
To further validate the presence of hydrogen bonds during
the catalytic reaction, we performed in situ IR experiments
starting from isolated 4. After pressurizing this complex with
CO (4 bar) the spectrum of complex [HRh(3)2(CO)2] (5) was
substracted. The formation of an intermediate was observed
upon addition of 1.5 equiv 1-octene under CO pressure
(4 bar). This intermediate shows a strong absorption at
1663 cmꢀ1, which indicates the formation of a rhodium–acyl
complex (Figure 5).[21] DFT computations were performed to
assign the emerging vibrations (Table 2).
After correction for anharmonicity, the computed vibra-
tions were found to be in good agreement with the exper-
imental observations (Table 2). This finding suggests that the
complex formed is a rhodium–acyl complex bearing two 6-
DPPon (3) ligands. To validate that a catalytically active
intermediate was formed, not a thermodynamically stable,
unreactive complex, we then pressurized the solution with
CO/H2 (4 bar), which immediately resulted in the formation
of nonanal, as indicated by the appearance of its CO
To get additional experimental proof for the importance
of the hydrogen-bond network for the outstanding selectivity
Angew. Chem. Int. Ed. 2012, 51, 11033 –11038
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11035

.
Angewandte
Communications
Figure 5. Difference in situ IR spectra obtained after the addition of
1.5 equiv 1-octene under CO pressure to the solution of the
[HRh(3)2(CO)2] complex (10ꢀ2 m in toluene). The orange line marks the
addition of H2.
Figure 6. 13C NMR spectrum of the intermediate (193 K, 500 MHz).
Table 2: Vibrations of the intermediate observed by in situ IR spectros-
copy and computed vibrations for an axeq- and eqeq-coordinated acyl
complex.
Table 3: Comparison of the experimentally observed 13C NMR shifts of
the intermediate and the computed shielding.
Expt. [cmꢀ1 [a]
]
eqeq [cmꢀ1 [b]
]
axeq [cmꢀ1 [b]
]
Expt. [ppm]
eqeq [ppm][a]
axeq [ppm][a]
C3axeq
C2axeq/eqeq
C3eqeq
C1axeq
C1eqeq
195.7
197.6
201.3
230.6
250.1
–
188.9
192.4
–
234.9
–
COeq
COeq-eq
COax
COeq+eq
COacyl
2040
2004
1969
1956
1663
2031 (2113)
–
192.2
196.0
–
1987(2067)
–
1949 (2028)
1704 (1773)
1968 (2048)
–
1667 (1734)
248.5
[a] Si-comp, toluene [10ꢀ2 m]. [b] B3LYP/6-31G(d,p)[C,H,N,O,P]
+LANL2DZ[Rh], the values in italics are obtained by anharmonic
correction.
[a] M06/6-311+G(2d,p)[C,H,N,O,P]+SDD[Rh] on B3LYP/6-31G(d,p)
[C,H,N,O,P]+LANL2DZ[Rh] optimized structures referenced to
[HRh(PPh3)3(CO)] (d=206.8 ppm).[25]
stretching vibration (1734 cmꢀ1, Figure 5).[22,23] To investigate
the presence of hydrogen bonding in this intermediate by
NMR spectroscopy, we performed the experiment described
previously with 13C-labeled CO in [D8]toluene. After the same
IR spectroscopic signature was observed (taking into account
the frequency shift of the CO vibrations because of the
isotopic labeling[8]), a sample was removed for low-temper-
ature NMR experiments.[24] The 13C NMR spectrum shows
signals at 250.1 ppm and 230.6 ppm, which can be assigned to
the acyl carbon atoms of the eqeq-(14) and axeq-coordinated
(15) rhodium–acyl complexes (C1 in Figure 6).
Table 4: Comparison of the experimentally observed 1H NMR spectro-
scopic shifts of the intermediate and the computed shielding.
Expt. [ppm]
eqeq [ppm][a]
axeq [ppm][a]
H2axeq
H1eqeq
H1axeq
H2eqeq
13.79
13.45
13.05
12.90
–
13.71
–
12.91
–
13.33
–
12.71
[a] M06/6-311+G(2d,p)[C,H,N,O,P]+SDD[Rh] on B3LYP/6-31G(d,p)
[C,H,N,O,P]+LANL2DZ[Rh] optimized structures referenced to tetra-
methylsilane (d=0.00 ppm).
The computed chemical shifts for 14 and 15 are in
excellent agreement with the experimental value (Tables 3
and 4). The 31P NMR spectrum recorded at 193 K shows five
signals corresponding to a rhodium-bound ligand and one
signal for the free ligand 3, which is present as one equivalent
of 3 was replaced by CO when starting from 4.[8] As expected,
the signals assigned to the acyl carbon atoms show HMBC
cross-peaks to the neighboring CH2 signals (Figure 7).
If, as assumed, hydrogen bonding occurs in acyl complexes
14 and 15, hydrogen-bond signals (10–15 ppm) should be
observable in a 1:2 ratio relative to the CH2 group. Indeed,
four signals in the hydrogen-bond region could be detected
integration of the CH2 group of the acyl residue suggests
that the axeq complex is slightly preferred. This preference is
in agreement with M06 single-point calculations (DE =
0.9 kcalmolꢀ1) on the B3LYP-optimized structures.
In summary, we have demonstrated that the hydrogen-
bonding network of the 6-DPPon (3) system enhances the
activity and selectivity of the hydroformylation of 1-octene.
To ensure that hydrogen bonding is not only present in the
resting state but throughout the catalytic cycle, a catalytically
competent intermediate (proven by in situ IR studies) was
characterized. The existence of hydrogen bonds was verified
unambiguously through correlation of 1H and 13C NMR
spectroscopy supported by DFT calculations. These results
show that the outstanding activity and selectivity of the 6-
DPPon (3) self-assembly catalyst is a direct consequence of
ꢀ
ꢀ
and assigned by DFT calculations to the N H···N and O
H···O hydrogen bonds of 14 and 15 (Figure 8). The broad
ꢀ
signal at 12.6 ppm corresponds to the N H···O proton of 3,
which forms symmetrical pyridone dimers. Interestingly,
11036 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11033 –11038

Angewandte
Chemie
2001, 1 – 36; c) Rhodium Catalyzed Hydroformylation (Eds.:
P. W. N. M. van Leeuwen, C. Claver), Kluwer Academic, Dor-
drecht, 2000.
[2] a) M. Kranenburg, Y. E. M. van der Burgt, P. C. J. Kamer,
P. W. N. M. van Leeuwen, Organometallics 1995, 14, 3081 –
3089; b) T. J. Devon, G. W. Philips, T. A. Puckette, J. L. Stavi-
noha, J. J. Vanderbilt, US Patent 4 694 109, 1988; c) E. Billig,
A. G. Abatjoglou, D. R. Bryant, US Patent 4 769 498, 1988;
d) G. D. Cunny, S. L. Buchwald, J. Am. Chem. Soc. 1992, 114,
5535 – 5543.
[3] a) W. Seiche, B. Breit, J. Am. Chem. Soc. 2003, 125, 6608 – 6609;
b) W. Seiche, A. Schuschkowski, B. Breit, Adv. Synth. Catal.
2005, 347, 1488 – 1494; c) C. Waloch, J. Wieland, M. Keller, B.
Breit, Angew. Chem. 2007, 119, 3097 – 3099; Angew. Chem. Int.
Ed. 2007, 46, 3037 – 3039.
[4] For other approaches to supramolecular catalysis, see a) P.
Dydio, W. I. Dzik, B. de Bruin, J. N. H. Reek, Angew. Chem.
2011, 123, 416 – 420; Angew. Chem. Int. Ed. 2011, 50, 396 – 400;
b) M. Kuil, T. Soltner, P. W. N. M. van Leeuwen, J. N. H. Reek, J.
Am. Chem. Soc. 2006, 128, 11344 – 11345; c) T. S. Koblenz, J.
Wassenaar, J. N. H. Reek, Chem. Soc. Rev. 2008, 37, 247 – 262;
d) L. Diab, T. Smejkal, J. Geier, B. Breit, Angew. Chem. 2009,
121, 8166 – 8170; Angew. Chem. Int. Ed. 2009, 48, 8022 – 8026;
e) A. W. Kleij, M. Lutz, A. L. Spek, P. W. N. M. van Leeuwen,
J. N. H. Reek, Chem. Commun. 2005, 3661 – 3663; f) V. F. Slagt,
P. C. J. Kamer, P. W. N. M. van Leeuwen, J. N. H. Reek, J. Am.
Chem. Soc. 2004, 126, 1526 – 1536; g) V. F. Slagt, J. N. H. Reek,
P. C. J. Kamer, P. W. N. M. van Leeuwen, Angew. Chem. 2001,
113, 4401 – 4404; Angew. Chem. Int. Ed. 2001, 40, 4271 – 4274;
h) J. M. Takacs, D. S. Reddy, S. A. Moteki, D. Wu, H. Palencia, J.
Am. Chem. Soc. 2004, 126, 4494 – 4495; i) P. W. N. M. van Leeu-
wen, D. Rivillo, M. Raynal, Z. Freixa, J. Am. Chem. Soc. 2011,
133, 18562 – 18565.
Figure 7. HMBC spectrum (253 K, 500 MHz) which shows the relation-
ship between the acyl carbon atoms of 10 and 11 and the neighboring
CH2 groups.
[5] O. Abillard, B. Breit, Adv. Synth. Catal. 2007, 349, 1891 – 1895.
[6] D. Fuchs, G. Rousseau, L. Diab, U. Gellrich, B. Breit, Angew.
Chem. 2012, 124, 2220 – 2224; Angew. Chem. Int. Ed. 2012, 51,
2178 – 2182.
[7] U. Gellrich, J. Huang, W. Seiche, M. Keller, M. Meuwly, B. Breit,
J. Am. Chem. Soc. 2011, 133, 964 – 975.
[8] The spectra are shown and discussed in detail in the Supporting
Information.
[9] Kinetic studies revealed a first order dependency of the alkene
concentration, thus suggesting that either alkene coordination or
hydrometalation is rate-determining. For details, see the Sup-
porting Information.
Figure 8. 1H NMR spectrum of the intermediate (253 K, 500 MHz).
[10] All calculations were performed with Gaussian09: M. J. Frisch
et al., Gauusian09 Rev.B.01, Gaussian, Inc., Wallingford CT,
2010. The complete reference, alternative pathways, and details
of the calculations are reported in the Supporting Information.
[11] See for example: a) J. J. Carbꢁ, F. Maseras, C. Bo, P. W. N. M.
van Leeuwen, J. Am. Chem. Soc. 2001, 123, 7630 – 7637; b) R.
Lazzaroni, R. Settambolo, G. Alagona, C. Ghio, Coord. Chem.
Rev. 2010, 254, 696 – 706.
[12] Energy barriers for the coordination of ethene to model
phosphane-modified systems: D. Gleich, J. Hutter, Chem. Eur.
J. 2004, 10, 2435 – 2444; an energy barrier for the coordination of
ethene to thixantphos-modified rhodium catalysts has been
reported, but only a barrier for coordination to the cis complex
could be located and a barrierless coordination to the trans
complex was reported: E. Zuidema, L. Escorihuela, T. Eichel-
sheim, J. J. Carbꢁ, C. Bo, P. C. J. Kamer, P. W. N. M. van Leeu-
wen, Chem. Eur. J. 2008, 14, 1843 – 1853.
the hydrogen-bonding interaction of the ligand. The hydrogen
bonding provides a synergism of flexibility and structural
integrity which facilitates the adoption of different coordina-
tion geometries without a significant energy penalty whilst
maintaining the regiodiscriminating properties of a chelating
ligand. The findings reported herein should assist in the
rational design of future supramolecular catalysts.
Received: May 16, 2012
Published online: September 28, 2012
Keywords: DFT calculations · hydroformylation ·
.
hydrogen bonds · IR spectroscopy · self-assembly
[13] Additional confomers of 5 and 6 were optimized in which
hydrogen bonding is prevented for geometrical reasons, but
these confomers are predicted to be 8.2 kcalmolꢀ1 and 4.8 kcal
molꢀ1, respectively, higher in energy by the M06//B3LYP level of
theory.
[1] a) K. Weissermel, H.-J. Arpe, Industrial Organic Chemistry,
Wiley-VCH, Weinheim, 2003, pp. 127 – 144; b) B. Breit, Top.
Curr. Chem. 2008, 279, 139 – 172; B. Breit, W. Seiche, Synthesis
Angew. Chem. Int. Ed. 2012, 51, 11033 –11038
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 11037

.
Angewandte
Communications
[14] a) J. Uddin, C. R. Landis, J. Chem. Soc. Dalton Trans. 2002, 729 –
742; b) see Ref. [12].
[15] For a detailed analysis of the different possible alkene complexes
and the resulting transition states see Figure S27 in the
Supporting Information.
[16] J. Wassenaar, E. Jansen, W.-J. van Zeist, F. M. Bickelhaupt,
M. A. Siegler, A. L. Spek, J. N. H. Reek, Nat. Chem. 2010, 2,
417 – 421.
[17] By taking the pre-equilibria between 6 and 7 into account, for
example, regarding a keff = k6!7k7!8/(k6!7+k7!6) relationship for
the two pathways, a selectivity of 89:11 is computed at the M06//
B3LYP level.
[18] To ensure that the same catalytic active species is present we
followed this reaction by in situ IR spectroscopy: a) R. Lazzar-
oni, G. Uccello-Baretta, M. Benetti, Organometallics 1989, 8,
2323 – 2327; b) P. C. Casey, L. M. Petrovich, J. Am. Chem. Soc.
1995, 117, 6007 – 6014; c) S. C. van der Slot, P. C. J. Kamer,
P. W. N. M. van Leuween, J. A. Iggo, B. T. Heaton, Organome-
tallics 2001, 20, 430 – 441.
[21] J. Feng, M. Garland, Organometallics 1999, 18, 417 – 427.
[22] P. C. J. Kamer, A. van Rooy, G. C. Schoemaker, P. W. N. M.
van Leeuwen, Coord. Chem. Rev. 2004, 248, 2409 – 2424.
[23] Recently, we (C. H. Beierlein, R. A. Paz-Schmidt, D. A. Platt-
ner, B. Breit, Organometallics 2010, 29, 2521 – 2532) and others
(J. Meeuwissen, A. J. Sandee, B. de Bruin, M. A. Siegler, A. L.
Spek, J. N. H. Reek, Organometallics 2010, 29, 2413 – 2421)
discussed the possibility of an insertion of the rhodium atom into
a g- or b-NH bond of the ligand, thereby resulting in a rhodium
hydride species that is able to undergo reductive elimination. We
ꢀ
were able to locate a TS for rhodium insertion into a N H bond,
but the computed barrier renders this process unlikely to happen
(see the Supporting Information). This is in agreement with the
in situ IR experiments, which show that by using 1.5 equivalents
of 1-octene the formation of nonanal occurs only after pressur-
izing the solution with H2 (see Figure 5).
[24] For NMR studies of rhodium–acyl complexes, see a) A. G. Kent,
J. M. Brown, J. Chem. Soc. Perkin Trans. 2 1988, 1587 – 1607;
b) see Ref. [18c].
[19] The results of the hydroformylation in terms of dependency on
the CO/H2 pressure can be found in the Supporting Information.
[20] For the synthesis of these ligands, see reference [7].
[25] J. M. Brown, L. R. Canning, A. G. Kent, P. J. Sidebottom, J.
Chem. Soc. Chem. Commun. 1982, 721 – 723.
11038 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 11033 –11038