An operando FTIR spectroscopic and kinetic study of carbon monoxide pressure influence on rhodium-catalyzed olefin hydroformylation

Article abstract of DOI:10.1002/chem.201402515

The influence of carbon monoxide concentration on the kinetics of the hydroformylation of 3,3-dimethyl-1-butene with a phosphite-modified rhodium catalyst has been studied for the pressure range p(CO)=0.20-3.83MPa. Highly resolved time-dependent concentration profiles of the organometallic intermediates were derived from IR spectroscopic data collected in situ for the entire olefin-conversion range. The dynamics of the catalyst and organic components are described by enzyme-type kinetics with competitive and uncompetitive inhibition reactions involving carbon monoxide taken into account. Saturation of the alkyl-rhodium intermediates with carbon monoxide as a cosubstrate occurs between 1.5 and 2 MPa of carbon monoxide pressure, which brings about a convergence of aldehyde regioselectivity. Hydrogenolysis of the acyl intermediate is fast at 30'°C and low pressure of p(CO)=0.2 MPa, but is of minus first order with respect to the solution concentration of carbon monoxide. Resting 18-electron hydrido and acyl complexes that correspond to early and late rate-determining states, respectively, coexist as long as the conversion of the substrate is not complete.

Full text of DOI:10.1002/chem.201402515

DOI: 10.1002/chem.201402515
Full Paper
&
Transition-Metal Catalysis
An Operando FTIR Spectroscopic and Kinetic Study of Carbon
Monoxide Pressure Influence on Rhodium-Catalyzed Olefin
Hydroformylation
Christoph Kubis,[a] Mathias Sawall,[b] Axel Block,[a] Klaus Neymeyr,[a, b] Ralf Ludwig,[a, c]
Armin Bçrner,[a, c] and Detlef Selent*[a]
Abstract: The influence of carbon monoxide concentration
on the kinetics of the hydroformylation of 3,3-dimethyl-1-
butene with a phosphite-modified rhodium catalyst has
been studied for the pressure range p(CO)=0.20–3.83 MPa.
Highly resolved time-dependent concentration profiles of
the organometallic intermediates were derived from IR spec-
troscopic data collected in situ for the entire olefin-conver-
sion range. The dynamics of the catalyst and organic compo-
nents are described by enzyme-type kinetics with competi-
tive and uncompetitive inhibition reactions involving carbon
monoxide taken into account. Saturation of the alkyl–rhodi-
um intermediates with carbon monoxide as a cosubstrate
occurs between 1.5 and 2 MPa of carbon monoxide pres-
sure, which brings about a convergence of aldehyde regio-
selectivity. Hydrogenolysis of the acyl intermediate is fast at
308C and low pressure of p(CO)=0.2 MPa, but is of minus
first order with respect to the solution concentration of
carbon monoxide. Resting 18-electron hydrido and acyl
complexes that correspond to early and late rate-determin-
ing states, respectively, coexist as long as the conversion of
the substrate is not complete.
Introduction
utilization of nonclassical metals and renewable olefin feed-
stock and on the immediate transformation of aldehydes into
secondary products, for example, by reduction.[3] However, the
well-established class of phosphite-modified rhodium catalyst
remains in the spotlight of research so far because, in part, of
their potential for technical applications.[4] These catalysts
show outstanding activities under mild reaction conditions and
the capability to fulfill a plurality of requests regarding sub-
strate tolerance and chemo- and aldehyde regioselectivities.[5]
Still, a lot of interesting detail remains to be elucidated with re-
spect to the kinetic and mechanistic specifics of hydroformyla-
tion catalysis. Spectroscopic investigations performed in situ
are especially helpful for this purpose.[6] Thus, with the help of
FTIR spectroscopy, it has recently become possible to detect
an acyl complex that belongs to the class of diphosphite-modi-
fied rhodium catalysts under the conditions of the hydroformy-
lation reaction.[7] For monophosphite rhodium catalysts, the
coexistence of the electronically saturated hydrido–rhodium
complex 1 and the acyl–rhodium complex 7 (Scheme 1) within
a broad hydrogen pressure and olefin-conversion range has
been proven and quantified.[8] Further results from our labora-
tory furnished evidence of an axial coordination of the phos-
phite ligand trans to the acyl group present in 7,[8] a structural
detail not anticipated in recent theoretical investigations.[9]
Complementary to the former work on rhodium catalysts
modified with a bulky monophosphite ligand, we address
herein, in more detail, the effects that occur from the influence
of the carbon monoxide concentration on the reaction kinetics.
A negative order has been reported for unmodified rhodium
Olefin hydroformylation is a long-known catalytic process with
industrial importance and sustained academic activities. Al-
though the reaction was discovered with catalytically active
species containing cobalt as the metal center, it acquired in-
tensified academic interest when the higher activities of rhodi-
um-based catalysts became evident.[1] A mechanistic rationale
has been developed.[2] Recent work concentrates on widening
the substrate scope and ligand development, but also on the
[a] Dr. C. Kubis, A. Block, Prof. Dr. K. Neymeyr, Prof. R. Ludwig,
Prof. Dr. A. Bçrner, Dr. D. Selent
Leibniz-Institut fꢀr Katalyse e.V.
Universitꢁt Rostock, Albert-Einstein-Str. 29a
18059 Rostock (Germany)
Fax: (+49)381-128151169
E-mail: detlef.selent@catalysis.de
[b] Dr. M. Sawall, Prof. Dr. K. Neymeyr
Institut fꢀr Mathematik
Universitꢁt Rostock, Ulmenstr. 69
18057 Rostock (Germany)
[c] Prof. R. Ludwig, Prof. Dr. A. Bçrner
Institut fꢀr Chemie
Universitꢁt Rostock, Albert-Einstein-Str. 3
18059 Rostock (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201402515. It contains concentration pro-
files for the metal-containing complexes and organic components for all
the carbon monoxide pressures applied as well as a detailed description of
the handling of the Michaelis–Menten-type kinetics with competitive and
uncompetitive equilibria, and the Christiansen formalism.
Chem. Eur. J. 2014, 20, 11921 – 11931
11921
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
hydrogenolysis are both of relevance for the product-forma-
tion rate within a broad range of the carbon monoxide partial
pressure.
Results and Discussion
Kinetic parameters, selectivity, and concentration profiles
The hydroformylation reaction was carried out at 308C in the
range p(CO)=0.20–3.83 MPa with a constant hydrogen pres-
sure of 1.0 MPa applied. 3,3-Dimethyl-1-butene was chosen as
a substrate that is incapable of double-bond isomerization,
thus preventing undesired alteration of the kinetics.[12] Only
4,4-dimethylpentanal and 2,3,3-trimethylbutanal can be
formed as hydroformylation products (Scheme 2). The rhodium
concentration was set to 0.3 mm and the initial concentration
of the substrate was 0.45m. The low temperature is suitable
for a detailed kinetic analysis because the catalyst system is ex-
tremely active at higher temperatures, thus leading to difficul-
ties in controlling the reaction conditions. However, previous
studies have shown that the basic kinetic features are similar
for applied temperatures of 30 and 708C.[7]
Scheme 1. Dissociative mechanism of the bulky monodentate phosphite
ligand (L)-modified rhodium-catalyzed olefin hydroformylation. The pathway
to the normal aldehyde, with only one ligand coordinated to the rhodium
center, is depicted.
catalysis and monophosphite-modified systems, which were
based on initial reaction rates determined at carbon monoxide
pressures equal to or higher than 1 MPa.[10]
At lower pressures, positive reaction orders have been re-
ported.[11] A plausible explanation is the depopulation of the
saturated formally inactive complexes 1 and 7 to the benefit
of reactive intermediates 2 and 6, respectively. For 2, the olefin
competes with carbon monoxide, thus giving intermediates 3
and 4. The formation of rhodium–alkyl compound 5 and also
a subsequent carbonyl insertion toward unsaturated acyl com-
plex 6 benefit from an enhanced concentration of carbon
monoxide. Increasing the p(CO) affords a decrease of the pop-
ulation of 6 by shifting the equilibrium to species 7. This inhib-
iting effect is uncompetitive with respect to the olefin but di-
rectly influences the rate of hydrogenolytic product formation.
We present herein new results for the tris(2,4-di-tert-butyl-
phenyl)phosphite-modified rhodium catalyst applied to the hy-
droformylation reaction of 3,3-dimethyl-1-butene, an olefin
considered to be a suitable model substrate for kinetic and
mechanistic investigations. This work represents a continuation
of a preceding report that dealt with a comprehensive spectro-
scopic characterization of respective rhodium complexes and
kinetic analysis of the influence of the olefinic substrate and
the cosubstrate dihydrogen.[8] Our current contribution focuses
on the effect of carbon monoxide on the reaction kinetics,
molar fractions of the catalytic species, and regioselectivity.
Time-dependent concentration profiles are the experimental
basis for this study. Such profiles were determined for the full
olefin-conversion range for the organic components and the
rhodium–hydride and acyl complexes that are present in sub-
millimolar concentrations by FTIR spectroscopic analysis per-
formed in situ. The changes of the individual concentrations
with time are in accord with a model analogue of enzyme-type
kinetics, in which catalyst saturation with carbon monoxide
occurs and inhibition of olefin coordination and rhodium–acyl
Scheme 2. Hydroformylation of 3,3-dimethyl-1-butene.
At the beginning of every batch, the catalyst precursor [Rh-
(acac)(CO)2] (acac=acetylacetonate) was converted in the pres-
ence of 20 equivalents of monophosphite ligand into a mixture
of the known hydrido complex e-[HRh(CO)3(phosphite)]
(n˜(CO)=2015, 2043, 2093 cmꢀ1) and e,e-[HRh(CO)2(phosphite)2]
(n˜(CO)=2028, 2068 cmꢀ1) under the intended reaction condi-
tions. The presence, identity, and basic geometric features of
the latter hydride have been clearly evidenced recently by
combining experimental FTIR spectroscopic data and DFT cal-
culations.[8] After a complete precursor-to-catalyst transforma-
tion, the olefin was added, which was taken as the reaction
start.
The full sequence of each batch, including the transforma-
tion of the precursor into the rhodium–hydride complexes, the
olefin addition, and the following reaction progress, was fol-
lowed by FTIR spectroscopic analysis. The experimental setup
consisted of a stainless-steel autoclave with a constant pres-
sure controller. The autoclave was connected with a high-pres-
sure transmission IR cell and an automatic sampling device to
take liquid samples for GC analysis. Because of the very low
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11922
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
catalyst concentration applied, a wedged spacer was placed
between the ZnS windows of the transmission IR cell to pre-
vent the disturbing interference fringes that otherwise oc-
curred in the IR spectra.[13] Further experimental details are
given in the Experimental Section.
Table 1. Kinetic values for the hydroformylation of 3,3-dimethyl-1-butene
with the Rh/tris(2,4-di-tert-butylphenyl)phosphite catalyst at 308C (p(H2)=
1 MPa).
p(CO) [CO][a]
[mPa] [mol dmꢀ3
V
k’obs
Kombs
[mol dmꢀ3
Selectivity[c]
obs
sat
[b]
]
[mol dmꢀ3 minꢀ1
]
[minꢀ1
]
]
[n/iso]
The hydroformylation reactions performed showed high che-
moselectivities toward the aldehydes; only trace amounts (<
0.5%) of alcohols were found. For the batch performed at
1 MPa CO/1 MPa H2, the aldehyde products were formed with
constant regioselectivity, as verified by online sampling and GC
analysis (see the Supporting Information). By assuming the
same kinetics for both products with any other p(CO) value ap-
plied, we used the sum of the normal and iso aldehyde con-
centrations for the following kinetic analysis of all the batches.
The effect of the variation of the carbon monoxide pressure
on the product formation is shown in Figure 1. At a pressure
of 0.2 MPa of CO, a turnover frequency of 1065 hꢀ1, based on
0.20
0.35
0.50
0.75
1.00
0.99
1.49
1.99
2.98
3.83
0.022
0.039
0.057
0.084
0.113
0.112
0.169
0.228
0.347
0.451
7.59ꢁ10ꢀ3
3.89ꢁ10ꢀ3
2.96ꢁ10ꢀ3
1.79ꢁ10ꢀ3
1.39ꢁ10ꢀ3
1.31ꢁ10ꢀ3
0.89ꢁ10ꢀ3
0.67ꢁ10ꢀ3
0.44ꢁ10ꢀ3
0.34ꢁ10ꢀ3
25.5
13.1
9.2
5.9
4.6
4.5
3.0
2.3
1.5
1.2
0.196
0.155
0.143
0.137
0.139
0.138
0.130
0.132
0.129
0.130
0.95:0.05
0.93:0.07
0.92:0.08
0.91:0.09
0.90:0.10
0.90:0.10
0.89:0.11
0.88:0.12
0.88:0.12
0.88:0.12
[a] Solution concentrations of CO refer to the amount of pure solvent
used for the experiments and have been calculated from data given in: R.
Koelliker, H. Thies, J. Chem. Eng. Data 1993, 38, 437–440. [b] Ratio of
Vosabts/[Rh]. [c] Derived from GC data.
eters and the respective initial concentration of the olefinic
substrate [Eq. (2)].
obs
sat
obs
sat
d½Pꢁ
dt
V
½Sꢁ
V
ð½Sꢁ0ꢀ½PꢁÞ
V ¼
¼
¼
ð1Þ
ð2Þ
Kombs þ ½Sꢁ Kombs þ ð½Sꢁ0ꢀ½PꢁÞ
obs
sat
V
½Sꢁ0
V0 ¼
Kombs þ ½Sꢁ0
Initially, the increase of p(CO) affords higher amounts of the
iso product. Such an effect is known[10a,14] and has been dis-
cussed to originate from a reduced tendency to undergo b-hy-
dride elimination and a favored carbon monoxide addition/in-
sertion reaction of the isoalkyl analogue of intermediate 4
(Scheme 1), provided that the normal and iso acyl complexes
undergo subsequent hydrogenolysis at the same rate.[15] But,
our results also verify a constancy of selectivity if p(CO)=
1.5 MPa or higher, a result that is not deduced easily from the
accepted mechanism. The change of selectivity depicted in
Figure 2 roughly parallels the known equilibrium shift between
saturated hydrides 1 and [HRh(CO)2(phosphite)2],[8,16] thus sug-
gesting that the latter represents the more selective catalyst.
However, there is no evidence that this bisphosphite complex
contributes to product formation at all. Also, the relevant acyl
complex remains unknown. With a quick equilibrium that
occurs between both hydrides taken into account, 1 may
mainly act as the catalyst. Then, the convergence of selectivity
probably is due to catalyst saturation with carbon monoxide
as a cosubstrate. The respective results and discussion are
given in Figure 9.
Figure 1. Influence of carbon monoxide pressure on aldehyde product for-
mation (P=normal+iso isomers) in the hydroformylation of 3,3-dimethyl-1-
butene in n-hexane with a Rh/tris(2,4-di-tert-butylphenyl)phosphite catalyst
(p(CO)=0.20, 0.35, 0.50, 0.75, 0.99, 1.49, 1.99, 2.98, 3.83 MPa (plots follow
the same sequence); [Rh]=0.3 mm; [olefin]0 =0.45m). Experimental data
from FTIR spectroscopic analysis were acquired after mixing for 270 s. Time
intervals between the IR spectroscopic measurements were 37 s for the first
120 min of each batch and 67 s after that period. See the Supporting Infor-
mation for a comparison of the data obtained by IR spectroscopic analysis
with that from GC analysis.
the initial rate, was determined. This result means that the re-
action is still fast despite the uncommonly low temperature
applied. By progressively increasing the carbon monoxide pres-
sure, the reaction decelerates; consequently, for the batches
run at 2.98 and 3.83 MPa, full conversion was not achieved
even after more than 30 hours of reaction time.
A fit of the numerically integrated Michaelis–Menten equa-
tion [Eq. (1)] was performed to all of the product concentration
profiles from Figure 1. Table 1 summarizes the kinetic values
obtained from this fit together with data for the product selec-
The evolution in the metal carbonyl region of the FTIR spec-
tra with ongoing reaction time is shown for two batches per-
formed at the extremes of the carbon monoxide pressure ap-
plied (i.e., 0.20 and 3.83 MPa; Figure 3). The acyl complex
[C6H13C(O)Rh(CO)3(phosphite)] dominates directly after olefin
addition for both batches. This complex most probably is a mix-
ture of normal and iso regioisomers that could not be resolved
tivity. The procedure of using the Michaelis–Menten equation
obs
sat
for determining the saturation rate V and the Michaelis con-
stant Kombs from the product-concentration profiles that cover
the entire conversion range has been validated.[8] The regres-
sion results of this work are given in the Supporting Informa-
tion. The initial rates can be calculated from the kinetic param-
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11923
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ure 3a). For better insight into the superposition and dynamics
of the experimental spectra, chemometrical treatment based
on factor analysis was performed. In analogy to our former
work, the pure component decomposition (PCD) algorithm
was applied.[17,8] Interestingly, the individual spectra of 1 and
the related bisphosphite complex could not be resolved. This
result indicates that an equilibrium between both hydrides
occurs, which is fast with respect to hydroformylation kinetics.
As a consequence for the kinetic analysis, it is appropriate to
apply the quasiequilibrium approximation with the hydrido
complexes treated as a pseudocomponent. However, it has
been
possible
to
determine
the
ratio
of
1/[HRh(CO)2(phosphite)2], which depends on the partial pres-
sure of carbon monoxide. In the absence of the olefin, we esti-
mated values of 2.3 and 32.3 for p(CO)=0.20 and 3.83 MPa at
an molar ratio P/Rh=20:1 from reported data.[8] In contrast,
the IR spectrum of the acyl complex shows no dependency on
carbon monoxide pressure (Figure 4a,b). For all the carbon
monoxide pressures, the spectrum fits to that assigned to
a structure with the phosphite moiety coordinated in an axial
position trans to the acyl moiety, with three carbon monoxide
ligands occupying equatorial positions in the trigonal-bipyra-
midal complex geometry.[8] Therefore, this structure turns out
to be significantly more stable than any substitution product
Figure 2. Influence of carbon monoxide concentration on the normal(n) se-
lectivity for the hydroformylation of 3,3-dimethyl-1-butene in n-hexane with
a Rh/tris(2,4-di-tert-butylphenyl)phosphite catalyst at 308C and 1 MPa H2
([Rh]=0.3 mm; [ligand]/[Rh]=20:1; [olefin]0 =0.45m). The experimental data
were obtained from GC analysis.
Figure 3. FTIR spectra for selected carbon monoxide pressures in the IR
metal carbonyl region during the hydroformylation of 3,3-dimethyl-1-butene
in n-hexane with a Rh/tris(2,4-di-tert-butylphenyl)phosphite catalyst
([Rh]=0.3 mm; [ligand]/[Rh]=20:1; [olefin]0 =0.45m). Asterisks indicate the
acyl complex formed after the addition of the olefin at t=0 min
(n˜(CO)=1995, 2019, 2072 and 2079 cmꢀ1). [HRh(CO)3(phosphite)] (main com-
ponent): n˜(CO)=2015, 2043, 2093 cmꢀ1; [HRh(CO)2(phosphite)2] (minor com-
ponent): n˜(CO)=2028, 2068 cmꢀ1. Please note the different reaction times.
Figure 4. FTIR spectra of mixtures of [HRh(CO)3(phosphite)] and
[HRh(CO)2(phosphite)2] (hydrido complexes) and
[C6H13C(O)Rh(CO)3(phosphite)] (acyl complex). The spectra were obtained
from catalytic batches with p(CO)=0.20 and 3.83 MPa (see Figure 3 for fur-
ther experimental conditions).
by IR spectroscopic analysis in the shown spectral range. The
complex depletes much more quickly at 0.20 MPa, thus leading
to a mixture of the above-mentioned hydrido complexes (Fig-
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11924
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
quasiequilibrium. Seven additional sets of concentration pro-
files for the other carbon monoxide partial pressures applied
are given in the Supporting Information.
d½Sꢁ
ð3aÞ
ð3bÞ
ð3cÞ
ð3dÞ
¼ ꢀk1½Sꢁ½Catꢁ þ k ½CatSꢁ
ꢀ1
dt
d½Catꢁ
dt
¼ ꢀk1½Sꢁ½Catꢁ þ k ½CatSꢁ þ k2½CatSꢁ
ꢀ1
d½CatSꢁ
dt
¼ k1½Sꢁ½Catꢁꢀk ½CatSꢁꢀk2½CatSꢁ
ꢀ1
d½Pꢁ
¼ k2½CatSꢁ
dt
It can be seen from Figure 5 that all the experimental data
are reconstructed conveniently by the kinetic model, which
holds not only for the concentration profile of 3,3-dimethyl-1-
butene and the aldehyde products, but also for the observable
rhodium complexes. Notably, it was verified once more that
the hydrides and acyl complex [C6H13C(O)Rh(CO)3(phosphite)]
(7) coexist over the full conversion range. At the reaction start,
with a high olefin concentration given, the latter dominates.
With an ongoing reaction, the fraction of the hydrides increas-
es. Interestingly, the initial molar fraction of 7 is only slightly in-
fluenced by the carbon monoxide pressure, thus rising from
0.7 to 0.77 at p(CO)=0.2 and 3.83 MPa, respectively (Fig-
ure 6a). This outcome is in accord with the uncompetitive and
competitive inhibition reactions that occur in this catalytic
cycle (see below). From the data given in Figure 5, further
quantitative information can be derived. The concentration of
the rhodium–acyl complex, shown for nine different CO partial
pressures in Figure 6a, is linearly correlated with the rate of al-
dehyde formation (Figure 6b). A negative order with respect
to p(CO) for acyl hydrogenolysis is thus proven.[19] These rela-
tionships hold for the full carbon monoxide pressure range ap-
plied, even below 1 MPa, and for the full concentration range
for which the acyl complex is observable.
Figure 5. Experimental concentration profiles of the olefin and aldehydes for
a) p(CO)=0.2 and b) 3.83 MPa, as obtained conventionally by FTIR spectro-
scopic analysis and profiles of the organometallic components from PCD.
(a) Least-squares fit by using the Michaelis–Menten kinetic model.[20] The
concentration range of the rhodium complexes is given by the right-hand
axis. Hydroformylation of 3,3-dimethyl-1-butene in n-hexane with a Rh/
tris(2,4-di-tert-butylphenyl)phosphite catalyst ([Rh]=0.3 mm; [ligand]/
[Rh]=20:1; [olefin]0 =0.45m; p(H2) = 1.0 MPa).
in which the phosphorus ligand has entered the equatorial
plane.[18]
The consequence of the variation of [CO] on the reaction ki-
netics is illustrated by the concentration profiles that include
pure organic components and also the observable rhodium
complexes. Such profile sets were determined for the entire
olefin-conversion range for all the experiments. Figure 5 shows
two selected sets of profiles ob-
The experimental results presented above are in accord with
the established interpretation of the hydroformylation mecha-
nism. Changes of the equilibrium between 6 and 7 are likely
to occur upon pressure variation, thus enhancing the fraction
of the undetected, but catalytically essential intermediate 6 at
low carbon monoxide pressures.
tained
at
p(CO)=0.2
and
3.83 MPa, respectively. Each indi-
vidual profile is accompanied by
the result of a simultaneous inte-
gration/regression of the set of or-
dinary differential Equations (3a–
d) derived from the Michaelis–
Menten mechanism depicted in
Scheme 3, where (Cat) and (CatS)
Scheme 3. Michaelis–Menten
mechanism. Cat=catalyst: hy-
drido–rhodium complexes;
CatS=catalyst–substrate com-
plex: acyl–rhodium com-
plexes; S=olefinic substrate;
P=product aldehyde. The
pseudocomponent approach,
which also includes bisphos-
phite hydrido complexes, ap-
plies.
We want to emphasize that the analysis of the reaction rates
with the help of the acyl complex, as performed here, is a chal-
lenge and not very common in homogeneous transition-metal
catalysis. The determination of a direct correlation of rate with
the mass balance of catalytic intermediates can be hampered
due to chemical reasons, for example, a very low concentration
of the catalyst–substrate complex and also by various meth-
odological aspects.
represent
pseudocomponents,
which consist of the 16- and 18-
electron hydrido and acyl–rhodi-
um complexes, respectively, at
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11925
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
kobs½Catꢁ0½Sꢁ
ꢀ
ꢁꢀ
ꢁ
ꢀ
ꢁ
V ¼
ð4Þ
ð5Þ
½COꢁ
J
½COꢁ
D,uc
1 þ K
Km
þ
þ
1 þ K
½Sꢁ
½COꢁ
D,c
ꢂ
ꢀ
ꢁꢃ
½COꢁ
kobs½Catꢁ0= 1 þ
½Sꢁ
KD,uc
ꢂꢀ
ꢁꢀ
ꢁ ꢀ
ꢁꢃ
V¼
½COꢁ
J
½COꢁ
D,uc
1 þ K
Km
þ
= 1 þ K
þ ½Sꢁ
½COꢁ
D,c
obs
sat
V
½Sꢁ
¼
Kombs þ ½Sꢁ
Equations (6) and (7) result when [S]@Kombs. With the help of
the experimental values of k’obs, an estimate of kobs and KD,uc or
of 1/KD,uc =KInh,uc, respectively, should be possible by plotting 1/
k’obs versus [CO] (such a plot is given in Figure 7a). A line
through the origin was obtained, thus indicating a hydrogenol-
ysis of minus first order with respect to the carbon monoxide
concentration within the whole pressure range applied. A
linear regression of ln(k’obs) versus ln([CO]) gives an order of re-
action with respect to the carbon monoxide concentration of
ꢀ1.02. The fact that a line through the origin was found identi-
fies a very large value for KInh,uc and that the inequation 1/
Figure 6. Dependency from carbon monoxide pressure. a) Concentration
profiles of the rhodium–acyl complex. b) Plots of the rate of product forma-
tion versus the concentration of the acyl complex. Hydroformylation of 3,3-
dimethyl-1-butene in n-hexane with a Rh/tris(2,4-di-tert-butylphenyl)phos-
phite catalyst ([Rh]=0.3 mm: [olefin]0 =0.45m; [ligand]/[Rh]=20:1).
Uncompetitive and competitive inhibition, the Michaelis
constant, and catalyst saturation with CO
Our treatment of the hydroformylation reaction, with special
emphasis on the dependency of the rate of product formation
from the carbon monoxide concentration, follows the Christi-
ansen formalism.[21] The competitive and uncompetitive equili-
bria assigned in Scheme 1, which involve the cosubstrate
carbon monoxide, complexes 2 and 1, and complexes 6 and 7,
respectively, have been incorporated. The general expression
obtained is represented by Equation (4), where kobs is the pseu-
doconstant of irreversible acyl complex hydrogenolysis. The
constants KD,c and KD,uc represent the dissociation constants of
competitive and uncompetitive inhibition. The J value results
from the fact that carbon monoxide acts as a cosubstrate and
is composed of distinct rate constants and the concentration
of the substrate hydrogen (see the Supporting Information for
more details). Equation (5) is derived from Equation (4), it rep-
Figure 7. a) Linear regression: reciprocal observed rate constants k’obs versus
carbon monoxide concentration. b) Comparison of experimental values of
k’obs (open circles) with that obtained by using the ratio KInh,uc/kobs. Hydrofor-
mylation of 3,3-dimethyl-1-butene in n-hexane with a Rh/tris(2,4-di-tert-bu-
tylphenyl) phosphite catalyst ([Rh]=0.3 mm; [olefin]0 =0.45m; [ligand]/
[Rh]=20:1).
resents the Michaelis–Menten equation with composed kinetic
obs
parameters, and was used to determine the values of V and
sat
Kombs depicted in Table 1.
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11926
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
KD,uc ꢁ[CO]=KInh,uc ꢁ[CO]@1 is valid already at very low CO
pressures. This result can be considered to be the reason for
the spectroscopic observation of the coordinatively saturated
18-electron acyl complex only, with the consequence that the
corresponding unsaturated complex is present only in unde-
tectable concentrations. As a result, no distinct numerical
values can be determined for kobs and KInh,uc, but the ratio
K
Inh,uc/kobs is accessible. This ratio is part of Equation (8), which
allows for the calculation of the composed rate constant k’obs
.
The comparison between the calculated and experimental
values of this constant (depicted in Figure 7b) shows a good
agreement.
Figure 8. Partial order of reaction with respect to the concentration of
obs
sat
V
kobs
carbon monoxide for selected substrate concentrations. The calculation is
0
ꢀ
ꢁ
¼
¼ k obs
obs
based on the experimentally determined parameters V and Kombs and Equa-
½Catꢁ0
ð6Þ
sat
½COꢁ
1 þ K
tion (2).
D,uc
1
1
1
¼
þ
½COꢁ
ð7Þ
ð8Þ
k0obs kobs kobsKD,uc
The origin for the dependency depicted in Figure 8 is found
in the influence of [CO] on the Michaelis constant Kombs. This
constant is defined by Equation (9), which represents a part of
Equation (5). It consists of factors a, a’, and the term b, which
account for the competitive and uncompetitive inhibition by
carbon monoxide and the role of carbon monoxide as a cosub-
strate itself. Equation (10) results from 1/KD,uc ꢁ[CO]@1 and 1/
1
k0obs
KInh,uc
kobs
¼
½COꢁ
It would be interesting to see if additional information on
the influence of carbon monoxide partial pressure arises when
the analysis is based on the initial reaction rate. Our finding
K
D,c ꢁ[CO]@1. Thus, this equation respects the fact that only
the saturated hydrido and acyl complexes, respectively, can be
detected spectroscopically. Both the uncompetitive and com-
petitive inhibition equilibria are shifted largely to the side of
the 18-electron complexes.
with respect to the uncompetitive inhibition with carbon mon-
obs
sat
oxide as an inhibitor was based on the V value of the satura-
tion rate. This rate is the highest possible product-formation
rate in which all of the catalyst is complexed with the olefin
(S=reference substrate); namely, [CatS]=[Cat]0. This rate can
only be approached if the relation [S]@Kombs is valid. The initial
rate represents the reaction rate at the initial olefin concentra-
tion [S]0 applied [Eq. (2)]. The initial rate can be close to that at
catalyst saturation or smaller, and it directly corresponds to the
relation between the [S]0 and Kombs values; because carbon
monoxide acts as a substrate as well, this circumstance should
result in a positive influence on the product-formation rate.
ꢀ
ꢁꢀ
ꢁ
½COꢁ
J
1 þ
Km
þ
KD,c
½COꢁ
aðKm þ bÞ
Kombs
¼
¼
ð9Þ
ꢀ
ꢁ
a0
½COꢁ
1 þ K
D,uc
KD,uc
KD,c
K
KD,c
1
1
Kombs ¼ Km
þ
D,uc J
¼ A þ B
ð10Þ
½COꢁ
½COꢁ
To check this hypothesis, initial rates were calculated by
The dependency of the Kmobs values on the carbon monoxide
concentration is characterized by the constants A and B, which
were derived from a regression of the plot of the experimental
values of the observed Michaelis constant given in Table 1
versus the carbon monoxide concentration (Figure 9). It can be
seen that the Kombs value first decreases and then approaches
a plateau with [CO]@B. This dependency shows very nicely
that the Kombs constants deliver information on the role of
carbon monoxide as a cosubstrate by having a slightly positive
effect on the product-formation rate. Considering the Michae-
lis–Menten equation [Eq. (1)], the Michaelis constant is a term
in the denominator, so a decrease of the Kombs constants leads
to an increase of the reaction rate. However, this positive
effect is small because any increase in carbon monoxide con-
centration enforces uncompetitive inhibition and promptly re-
duces the rate of acyl hydrogenolysis.
obs
using Equation (2) with the values V and Kombs determined for
sat
selected initial substrate concentrations. The slope of a double
logarithmic plot of ln(V0) versus ln([CO]) gave the value for the
order of reaction with respect to [CO] for each experiment. A
plot of these reaction orders versus olefin concentration is
shown in Figure 8.
As a first result obtained for [S]0 =0.45m, no significant dif-
ference is seen relative to the result derived from the satura-
tion rate at the initial olefin concentration, which was applied
for the experimental batches. A reaction order of ꢀ0.99 with
respect to [CO] was calculated, but the order of reaction in-
creases and partial orders in [CO] result with decreasing sub-
strate concentration. Such partial orders based on initial rates
have been reported for other hydroformylation catalysts.[11] At
least for the reaction system under study here, we deduced
a shift in the relation between the [S]0 and Kombs values. The
reason for this shift is detailed below.
Mechanistically, the influence of [CO] on the Kombs constant
can be understood in terms of a Michaelis–Menten-type cata-
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11927
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
undetected 16-electron alkyl intermediates. Whereas at the
higher temperatures applied (i.e., 1008C), the formation of
linear and branched rhodium–alkyl complexes was still reversi-
ble at 2 MPa of carbon monoxide applied, and the regioselec-
tivity was mainly determined by the entropy change that
occurs during the formation of the rhodium–alkyl complex for
the same reaction system,[4d] saturation phenomena that occur
with carbon monoxide will strongly interfere when the reac-
tions are performed at ambient temperature.
The fact that the experimentally determined Michaelis con-
stants can be described by Equation (10) can be regarded as
indirect proof for the existence of competitive inhibition. In
the case that inhibition does not play any role, this equation
will have another form because of the a factor is then absent
in the foregoing Equation (9). The Michaelis constant should
further decrease with higher carbon monoxide concentrations;
such a dependency was not observed. The occurrence of com-
petitive inhibition is also in accord with the influence of [CO]
on the initial molar fractions of the acyl complex and the hydri-
do complex already mentioned. These molar fractions were
not changed for carbon monoxide concentrations greater than
0.2m.[22] The reason for this outcome lies in the relationship be-
tween the Kombs constant at a given initial substrate concentra-
tion and the molar fractions of the catalyst and catalyst–sub-
strate complex [Eq. (12)] and the dependency of the Michaelis
constant on [CO] as depicted in Figure 9.
Figure 9. Plot of the observed Michaelis constant versus carbon monoxide
concentration and regression according to Equation (10). Hydroformylation
of 3,3-dimethyl-1-butene in n-hexane with a Rh/tris(2,4-di-tert-butylphenyl)-
phosphite catalyst ([Rh]=0.3 mm; [olefin]0 =0.45m; [ligand]/[Rh]=20:1).
lyst saturation with the cosubstrate carbon monoxide. The rel-
evant part in the reaction mechanism is the coordination of
carbon monoxide to the unsaturated 16-electron alkyl complex
4, thus yielding the respective saturated 18-electron complex
5, which subsequently undergoes CO insertion to give the 16-
electron acyl complex 6. This consideration is represented by
Equation (11), which arises from Equation (10), and is structur-
ally identical to the Michaelis–Menten equation. The saturation
regime related to carbon monoxide is reached already at
[CO]=0.2m, thus representing a ratio of [CO]/[Rh]=667:1 for
our experiments. No significant change of the Kombs constant is
observed with a further increase of the CO pressure (or CO/Rh
ratio).
½Sꢁ0
½CatSꢁ
¼
ð12Þ
Kombs þ ½Sꢁ0
½CatSꢁ þ ½Catꢁ
We tried to get quantitative information with regard to the
obs
competitive inhibition reaction. Dividing Kombs by V eliminates
sat
1
Kombs
½COꢁ
ð1=AÞ½COꢁ
the a’ factor, which accounts for the uncompetitive inhibition
in Equation (9). Equation (13) is obtained based on the as-
sumption that KInh,c ꢁ[CO]@1.[23]
¼
¼
ð11Þ
B þ A½COꢁ ðB=AÞ þ ½COꢁ
With the analysis of the Michaelis constant given herein, the
small positive influence of carbon monoxide on the product-
formation rate is disclosed elegantly. To the best of our knowl-
edge, a similar study has not been reported.
Kombs KmKInh,c
KInh,c
kobs
J
ð13Þ
¼
½COꢁ þ
k0obs
kobs
The respective plot of Kombs/k’obs versus [CO] is given in
By comparing Figures 2 and 9, one can see a similar trend
for the effect of [CO] not only on the Michaelis constant but
also on the n-selectivity. Both dependencies are described by
an equation of the type y=A+B/[CO], which may indicate
that regioselectivity is also influenced by a saturation process
that occurs with carbon monoxide. Such a suggestion is plausi-
ble, that is, that with a large excess of carbon monoxide pres-
ent, the unsaturated normal and iso alkyl complexes 4 become
saturated with carbon monoxide and b-hydride elimination is
suppressed. But, probably the respective equilibria remain
highly mobile and fast with respect to product formation. Con-
sequently, a plateau is reached for regioselectivity at high con-
centrations of carbon monoxide ([CO]@B). Thus, the changes
of selectivity observed will not originate from an equilibrium
shift between the 18-electron tricarbonyl complex 1 and
[HRh(CO)2(phosphite)2], which occurs upon pressure variations
and can easily be followed by IR spectroscopic analysis, but
rather from variations of the concentration of the up-to-now
Figure 10. The linear dependency observed is in agreement
Figure 10. Plot of Kombs/k’obs versus [CO]. Hydroformylation of 3,3-dimethyl-1-
butene in n-hexane with a Rh/tris(2,4-di-tert-butylphenyl)phosphite catalyst
([Rh]=0.3 mm; [olefin]0 =0.45m; [ligand]/[Rh]=20:1).
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11928
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
with the occurrence of competitive inhibition. However, no dis-
tinct values can be derived from the data for KInh,c, thus indicat-
ing a high preference for the electronically saturated hydrido
complexes. The result found herein is analogous to that found
for the inhibition involving the acyl complexes. Obviously, it is
not possible at the moment to describe the uncompetitive
and competitive equilibria comprehensively. The detection and
quantification of the respective coordinatively unsaturated
complexes involved will be required.
Experimental Section
Materials
3,3-Dimethyl-1-butene (>99%, GC) was purchased from Sigma–Al-
drich (95%) and distilled in an argon atmosphere over sodium. The
solvent n-hexane from Sigma–Aldrich (>99%) was distilled in an
argon atmosphere over Sicapent (Merck). The internal GC standard
dodecane (>99%, Sigma–Aldrich) was stored over Sicapent for
drying and distilled under vacuum. The catalyst precursor
[(acac)Rh(CO)2] (39.46% Rh) from Umicore was used without fur-
ther purification. The monophosphite ligand tris(2,4-di-tert-butyl-
phenyl)phosphite (TDTBPP; 98%, Sigma–Aldrich) was used as ob-
tained. The following gases were used: synthesis gas (CO/H2 =1:1,
from 99.997% CO and 99.999% H2, Linde), carbon monoxide
(99.997%, Linde), hydrogen (99.9993%, Linde), and argon
(99.999%, Linde). Bis(2,2,6,6-tetramethyl-4-piperidyl)sebaceate (Ti-
nuvin 770 DF; 100%, Ciba) was used as a phosphite stabilizer and
added in equimolar amounts with respect to the phosphite ligand.
Conclusion
Precise time-resolved insight into the dynamics of the hydro-
formylation reaction of 3,3-dimethyl-1-butene has been provid-
ed by coupling GC, FTIR spectroscopic analysis performed in
situ, and advanced chemometric treatment of IR spectra. Thus,
detailed concentration profiles became accessible not only for
the organic components, but also for the relevant rhodium
complexes. These profiles have allowed a qualitative interpre-
tation and a conclusive quantitative description of the relations
between the concentrations of the rhodium complexes and
the rate of product formation and on the dependency of these
relations on the concentration of carbon monoxide.
Devices and procedures
For the HP FTIR spectroscopic measurements, the experimental
setup consisted of a stainless-steel autoclave (200 mL) with a gas-
entrainment impeller and an oil-bath thermostat (premex reactor
AG; Leimen, Germany) equipped with a heatable transmission
flow-through IR cell (Dr. Bastian Feinwerktechnik GmbH; Wupper-
tal, Germany) and an automatic sampling device (amplius GmbH;
Rostock, Germany) to collect samples during the reaction for GC
analysis (Figure 11).
Hydride and acyl complexes that correspond to early and
late rate-determining states, respectively,[24] coexist for nearly
the entire conversion range. The rate control shifts with con-
version for the whole carbon monoxide pressure range ap-
plied. The two different hydrido–rhodium complexes observed
by employing IR spectroscopy in situ cannot be distinguished
kinetically, thus pointing to very fast equilibria for ligand ex-
change between these species. The acyl–rhodium complex,
which can be regarded to be equipollent to the catalyst–sub-
strate complex, with the olefin representing the reference sub-
strate, could also be tracked quantitatively in a time-resolved
manner. The rate of product formation from this compound
shows a minus first-order dependency with respect to the
carbon monoxide concentration.
For the circulation of the reaction solution between the reactor
and the IR cell, a microgear pump (mzr-7255, HNP Mikrosysteme
GmbH; Parchim, Germany) was used. Olefin injections were per-
formed with the help of a six-port valve and a respective loop
(Knauer GmbH; Berlin, Germany). FTIR spectra were collected on
a Bruker Tensor 27 FTIR spectrometer with a MCT-A detector. The
MCT detector was featured with a dewar, with a hold time of 24 h
with liquid nitrogen as a cooling agent. Variations in band intensi-
ties that arise from a nonconstant detector signal due to N2-level
alteration and the refilling required for longer reaction times were
The overall dynamics can be interpreted in terms of Michae-
lis–Menten kinetics involving competitive and uncompetitive
side reactions, in which saturation of the catalyst occurs also
with the cosubstrate carbon monoxide. An inhibiting influence
of carbon monoxide on the reaction rate is found already at
headspace partial pressures below 1 MPa.
In completion to earlier work, it can now be concluded that
for all the substrates, that is, 3,3-dimethyl-1-butene, dihydro-
gen, and carbon monoxide, the same kinetic approach is appli-
cable to describe their influence on the monophosphite-con-
taining catalytic system, which is a rhodium-catalyst represen-
tative for the most active catalyts known for olefin hydrofor-
mylation. Similar concepts of enzyme-type kinetics have al-
ready earlier been found to be useful for homogeneous
catalysis.[25] However, a comprehensive result, as provided
herein, is not thus widespread, especially for the case in which
the kinetic evaluation benefits from the direct observation of
the catalyst–substrate complex.[26]
Figure 11. Experimental setup for catalytic batches, high-pressure FTIR spec-
troscopic measurements collected in situ, and automated sampling for GC
analyses.
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11929
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
M. Becker, D. Vogt, Catal. Sci. Technol. 2014, 4, 524–530; d) S. Gꢃven,
M. M. L. Nieuwenhuizen, B. Hamers, R. Franke, M. Priske, M. Becker, D.
Vogt, ChemCatChem 2014, 6, 603–610.
eliminated by the baseline-correction procedure. As a window ma-
terial, ZnS (Korth Kristalle GmbH; Kiel, Germany) was used. The
mean optical-path length was 0.48 mm, which resulted from
a wedged spacer. The apparatus provided synthesis gas (CO/H2 =
1:1), carbon monoxide, and hydrogen for pressurization. A con-
stant-pressure controller (Brooks Instrument; Hatfield, PA, USA) was
used for regulation of the consumed gas (CO/H2 =1:1).
[5] R. Franke, D. Selent, A. Bçrner, Chem. Rev. 2012, 112, 5675–5732.
[6] a) Mechanisms in Homogeneous Catalysis (Ed.: B. Heaton), Wiley-VCH,
Weinheim, 2005; selected recent publications: b) M. Garland, G. Li, L.
Guo, ACS Catal. 2012, 2, 2327–2334; c) O. Diebolt, P. W. N. M. van Leeu-
wen, P. C. J. Kamer, ACS Catal. 2012, 2, 2357–2370; d) E. R. Nelsen, C. R.
Landis, J. Am. Chem. Soc. 2013, 135, 9636–9639; e) C. J. Koeken, L. J. P.
van den Broeke, B.-J. Deelman, J. T. F. Keurentjes, J. Mol. Catal. A 2011,
346, 1–11.
[7] C. Kubis, R. Ludwig, M. Sawall, K. Neymeyr, A. Bçrner, K.-D. Wiese, D.
Hess, R. Franke, D. Selent, ChemCatChem 2010, 2, 287–295.
[8] C. Kubis, D. Selent, M. Sawall, R. Ludwig, K. Neymeyr, W. Baumann, R.
Franke, A. Bçrner, Chem. Eur. J. 2012, 18, 8780–8794.
Stock solutions were prepared from the solids by using Schlenk
techniques. All the liquid components, except the olefin were
transferred into the reactor. The stirrer speed was set to 1500 rpm
and the speed of the microgear pump was set to 2333 rpm (dis-
placement volume=48 mL). At the reaction temperature of 308C,
the system was pressurized with the respective partial pressure of
carbon monoxide and 1 MPa of hydrogen. After completion of the
preformation of the hydrido complexes, the hydroformylation reac-
tion was started with an olefin injection.
[9] M. Sparta, K. J. Børve, V. R. Jensen, J. Am. Chem. Soc. 2007, 129, 8487–
8499.
[10] a) A. Van Rooy, E. N. Orij, P. C. J. Kamer, P. W. N. M. Van Leuuwen, Organo-
metallics 1995, 14, 34–43; b) A. A. Dabbawala, H. C. Bajaj, R. V. Jasra, J.
Mol. Catal. A 2009, 302, 97–106; c) A. Van Rooy, E. N. Orij, P. C. J. Kamer,
F. Van den Aardweg, P. W. N. M. Van Leuuwen, J. Chem. Soc. Chem.
Commun. 1991, 1096–1097; d) M. Garland, P. Pino, Organometallics
1991, 10, 1693–1704; e) C. Fyhr, M. Garland, Organometallics 1993, 12,
1753–1764; f) J. Feng, M. Garland, Organometallics 1999, 18, 417–427.
[11] a) S. S. Divekar, R. M. Deshpande, R. V. Chaudhari, Catal. Lett. 1993, 21,
191–200; b) B. M. Bhanage, S. S. Divekar, R. M. Deshpande, R. V. Chaud-
hari, J. Mol. Catal. A 1997, 115, 247–257; c) B. Heil, L. Markꢂ, Chem. Ber.
1968, 101, 2209–2214; d) M. Rosales, A. Gonzꢄlez, Y. Guerrero, I. Pache-
co, R. A. Sꢄnchez-Delgado, J. Mol. Catal. A 2007, 270, 241–249.
[12] For other studies that have used 3,3-dimethyl-1-butene as a substrate,
see, for example: a) M. Garland, G. Bor, Inorg. Chem. 1989, 28, 410–413;
b) M. Garland, Organometallics 1993, 12, 535–543; c) C. Li, E. Widjaja, M.
Garland, J. Catal. 2003, 213, 126–134; d) B. Moasser, W. L. Gladfelter,
D. C. Roe, Organometallics 1995, 14, 3832–3838 and ref. [10d].
[13] M. F. Faggin, M. A. Hines, Rev. Sci. Instrum. 2004, 75, 4547–4553.
[14] A. Van Rooy, P. C. J. Kamer, P. W. N. M. Van Leeuwen, K. Goubitz, J.
Franje, N. Veldman, A. L. Spek, Organometallics 1996, 15, 835–847.
[15] The hydrogenolysis of normal and iso acyl–rhodium intermediates can
proceed at different rates, even if the individual reaction orders with re-
spect to [CO] are equal; see ref. [10 f].
FTIR spectra were recorded between n˜ =3950 and 700 cmꢀ1, and
the spectral resolution was set at 2 cmꢀ1. For one FTIR spectrum,
ten scans were collected (double-sided, forward–backward) with
the mirror speed set at 40 kHz. Time intervals between two meas-
urements were 37 s for the first 120 min and 67 s after that period.
For the GC analyses, a 7890A GC System from Agilent Technologies
with a Petrocol DH 150 column (Supelco, Inc.) was used.
Acknowledgements
We thank the state of Mecklenburg–Western Pomerania and
the German Federal Ministry for Education and Research for
funding this work.
Keywords: hydroformylation · IR spectroscopy · kinetics ·
operando · rhodium
[16] R. Crous, M. Datt, D. Foster, L. Bennie, C. Steenkamp, J. Huyser, L. Kirs-
ten, G. Steyl, A. Roodt, J. Chem. Soc. Dalton Trans. 2005, 1108–1116.
[17] a) K. Neymeyr, M. Sawall, D. Hess, J. Chemometrics 2010, 24, 67–74;
b) M. Sawall, A. Bçrner, C. Kubis, D. Selent, R. Ludwig, K. Neymeyr, J.
Chemometrics 2012, 26, 538–548.
[18] An axial coordination of the phosphorus ligand has been documented
for corresponding triphenylphosphane complexes of rhodium and
cobalt; see, for example: a) J. Zhang, M. Poliakoff, M. W. George, Orga-
nometallics 2003, 22, 1612–1618; b) P. C. Ford, S. M. Massik, Coord.
Chem. Rev. 2002, 226, 39–49; c) S. M. Massick, J. G. Rabor, S. Elbers, J.
Marhenke, S. Bernhard, J. J. Schoonover, P. C. Ford, Inorg. Chem. 2000,
39, 3098–3106; d) J. Somlyai-Haꢄz, F. Haꢄz, V. Galamb, A. Benedetti, C.
Zucci, G. Pꢄlyi, T. Krꢃmmling, B. Happ, T. Bartik, J. Organomet. Chem.
1991, 419, 205–217; e) I. Kovꢄcs, F. Ungvꢄry, Coord. Chem. Rev. 1997,
161, 1–32.
[1] a) O. Roelen, DE 849548, (1938/1952); b) J. F. Young, J. A. Osborn, F. A.
Jardine, G. Wilkinson, J. Chem. Soc., Chem. Commun. 1965, 131–132;
c) B. Cornils, in New Syntheses with Carbon Monoxide (Ed.: J. Falbe),
Springer, Berlin, 1980; d) B. Cornils, W. A. Herrmann, M. Rasch, Angew.
Chem. 1994, 106, 2219–2238; Angew. Chem. Int. Ed. Engl. 1994, 33,
2144–2163; e) Rhodium Catalyzed Hydroformylation (Eds.: P. W. N. M.
Van Leeuwen, C. Claver), Kluwer, Dordrecht, 2000; f) C. D. Frohning,
C. W. Kohlpaintner, H.-W. Bohnen, in Applied Homogeneous Catalysis
with Organometallic Compounds (Eds.: B. Cornils, W. A. Herrmann),
2ednd edWiley-VCH, Weinheim, 2002.
[2] a) R. F. Heck, D. S. Breslow, J. Am. Chem. Soc. 1961, 83, 4023–4027;
b) R. F. Heck, D. S. Breslow, J. Am. Chem. Soc. 1962, 84, 2499–2502;
c) R. F. Heck, Acc. Chem. Res. 1969, 2, 10–16; d) D. Evans, J. A. Osborn,
G. Wilkinson, J. Chem. Soc. A 1968, 3133–3142; e) C. K. Brown, G. Wil-
kinson, J. Chem. Soc. A 1970, 2753–2764.
[19] Such relationship is also known from unmodified hydroformylation of
structurally different alkenes; see: G. Liu, R. Volken, M. Garland, Organo-
metallics 1999, 18, 3429–3426 and refs. [10d,f].
[20] See the Supporting Information for the full set of the concentration
profiles, which were determined for all the different carbon monoxide
pressures applied.
[21] See the Supporting Information for the derivation of the kinetic model
and F. G. Helfferich in Comprehensive Chemical Kinetics, Vol. 38 (Ed.:
N. J. B. Green), Elsevier Science Amsterdam, 2004.
[22] See the Supporting Information for a comparison of the entire set of
the concentration profiles of the rhodium complexes obtained at the
different carbon monoxide pressures applied.
[3] a) V. Agabekov, W. Seiche, B. Breit, Chem. Sci. 2013, 4, 2418–2422;
b) A. R. Almeida, A. F. Peixoto, M. J. F. Calvete, P. M. P. Gꢂis, M. M. Pereira,
Curr. Org. Synth. 2011, 8, 764–775; c) I. Piras, R. Jennerjahn, R. Jackstell,
A. Spannenberg, R. Franke, M. Beller, Angew. Chem. 2011, 123, 294–
298; Angew. Chem. Int. Ed. 2011, 50, 280–284; d) I. Fleischer, L. Wu, I.
Profir, R. Jackstell, R. Franke, M. Beller, Chem. Eur. J. 2013, 19, 10589–
10594; e) L. C. Matsinha, P. Malatji, A. T. Hutton, G. A. Venter, S. F. Mapo-
lie, G. S. Smith, Eur. J. Inorg. Chem. 2013, 4318–4328; f) J. Pospech, I.
Fleischer, R. Franke, S. Buchholz, M. Beller, Angew. Chem. 2013, 125,
2922–2944; Angew. Chem. Int. Ed. 2013, 52, 2852–2872; g) K. J. L.
Chang, P. D. Nichols, S. Blackburn, Lipid Technol. 2013, 25, 199–203; h)
O. Diebolt, C. Mueller, D. Vogt, Catal. Sci. Technol. 2012, 2, 773–777; i) S.
Li, K. Huang, J. Zhang, W. Wu, X. Zhang, Org. Lett. 2013, 15, 3078–3081.
[4] a) O. Diebolt, H. Tricas, Z. Freixa, P. W. N. M. van Leeuwen, ACS Catal.
2013, 3, 128–137; b) H. Tricas, O. Diebolt, P. W. N. M. van Leeuwen, J.
Catal. 2013, 298, 198–205; c) S. Gꢃven, B. Hamers, R. Franke, M. Priske,
[23] A. Cornish-Bowden, Fundamentals of Enzyme Kinetics, Portland Press,
London, 3edrd ed2004.
[24] a) S. Kozuch, J. M. Martin, ChemPhysChem 2011, 12, 1413–1418; the tra-
ditional concept of individual, early, and late rate-determining elemen-
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11930
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
tary steps, which is still used in transition-metal catalysis, needs to be
modified; see, for example: b) E. Zuidema, L. Escorihuela, T. Eichelsheim,
J. J. Carbꢂ, C. Bo, P. C. J. Kamer, P. W. N. M. van Leeuwen, Chem. Eur. J.
2008, 14, 1843–1853.
peatedly by Garland and co-workers; see, for example, the references
cited in [15] and [19]; furthermore, by the same authors, the semibatch
experimental design was shown to be of strong benefit to gain reliable
kinetic information and identify the transition-metal carbonyl com-
plexes present in very low concentrations: a) E. Widjaja, C. Li, M. Gar-
land, Organometallics 2002, 21, 1991–1997; b) C. Li, E. Widjaja, W.
Chew, M. Garland, Angew. Chem. 2002, 114, 3939–3943; Angew. Chem.
Int. Ed. 2002, 41, 3785–3789.
[25] a) C. Bergounhou, D. Neibecker, R. Mathieu, J. Mol. Catal. A 2004, 220,
167–182; b) D. G. Blackmond, Angew. Chem. 2005, 117, 4374–4393;
Angew. Chem. Int. Ed. 2005, 44, 4302–4320; c) H.-J. Drexler, A. Preetz, T.
Schmidt, D. Heller in The Handbook of Homogeneous Hydrogenation
(Eds.: J. G. de Vries, C. J. Elsevier), Wiley-VCH, Weinheim, 2007, pp. 257–
293.
[26] The detailed spectroscopic identification of metal carbonyl complexes
related to nonmodified rhodium-catalyzed homogeneous hydroformy-
lation and the respective kinetic evaluation, has been exemplified re-
Received: March 7, 2014
Published online on July 31, 2014
Chem. Eur. J. 2014, 20, 11921 – 11931
www.chemeurj.org
11931
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim