Synthetic applications of synergism using catalytic binuclear elimination reactions. Further examples of rhodium-manganese and rhodium-rhenium-catalyzed hydroformylations

Article abstract of DOI:10.1002/adsc.200800008

Synergism has been previously observed in both rhodium-manganese-and rhodium-rhenium-catalyzed hydroformylation. Furthermore, detailed in situ spectroscopic investigations have conclusively shown that the phenomenological origin of this synergistic effect is catalytic binuclear elimination (J. Am. Chem. Soc. 2003, 125, 5540-5548; 2007, 129, 13327-13334). In the present contribution, further substrates are used in the hydroformylation reaction with both rhodium-manganese and rhodium-rhenium. In situ spectroscopic studies show that (i) significant rate enhancements occur in the mixed metal systems with the new substrates and (ii) the organometallics present in the active systems, and their concentration profiles are consistent with those present in the previouslystud ied catalytic binuclear elimination reactions (CBER). It is therefore concluded that catalytic binuclear elimination is a rather general mechanism in mixed metal hydroformylations and is rather independent of the substrates used. Further discussion is given to mechanistic aspects, synthetic efficiency, and the possibility that such synergistic effects might be useful to other classes of organic syntheses.

Full text of DOI:10.1002/adsc.200800008

FULL PAPERS
DOI: 10.1002/adsc.200800008
Synthetic Applications of Synergism using Catalytic Binuclear
Elimination Reactions. Further Examples of Rhodium-Manganese
and Rhodium-Rhenium-Catalyzed Hydroformylations
Chuanzhao Li,a,b Li Chen,b and Marc Garlanda,b,*
a
Department of Chemical and Biomolecular Engineering, National Universityof Singapore, Singapore 117576
Institute of Chemical and Engineering Sciences, A*STAR (Agencyfor Science, Technologyand Research), 1 Pesek Road,
b
Jurong Island, Singapore 627833
Fax : (+65)-6316-6185; e-mail: marc_garland@ices.a-star.edu.sg
Received: January7, 2008; Published online: March 12, 2008
Abstract: Synergism has been previously observed in centration profiles are consistent with those present
both rhodium-manganese- and rhodium-rhenium-cat- in the previouslystudied catalytic binuclear elimina-
alyzed hydroformylation. Furthermore, detailed in tion reactions (CBER). It is therefore concluded that
situ spectroscopic investigations have conclusively catalytic binuclear elimination is a rather general
shown that the phenomenological origin of this syn- mechanism in mixed metal hydroformylations and is
ergistic effect is catalytic binuclear elimination (J. rather independent of the substrates used. Further
Am. Chem. Soc. 2003, 125, 5540–5548; 2007, 129, discussion is given to mechanistic aspects, synthetic
13327–13334). In the present contribution, further efficiency, and the possibility that such synergistic ef-
substrates are used in the hydroformylation reaction fects might be useful to other classes of organic syn-
with both rhodium-manganese and rhodium-rheni- theses.
um. In situ spectroscopic studies show that (i) signifi-
cant rate enhancements occur in the mixed metal Keywords: catalytic binuclear elimination; hydrofor-
systems with the new substrates and (ii) the organo- mylation; rhodium; spectroscopy; synergism; syn-
metallics present in the active systems, and their con- thetic efficiency
Introduction
ent. In this system, ruthenium promotes the abstrac-
tion of iodine from an iridium compound, leading to
Synergism is the term frequently used to describe the coordinative unsaturation on iridium and then eventu-
combined application of more than one metal, leading al elimination of acetic acid from an iridium com-
to regio-, chemo- and stereoselectivities and/or activi- pound. This is the synthetic basis for the well known,
ties which differ significantlyfrom a strictlyadditive
effect.[1,2]Accordingly, synergism holds considerable
promise (perhaps greatlyunderutilized) as an ap-
commercial and widelyused Cativa process.
Another class of multiple-nuclearity catalytic sys-
tems, namelycatalytic binuclear eliminations, has also
proach for improving synthetic strategies and synthet- been proposed as an origin of synergism.[7] In such
ic efficiency.
systems, binuclear elimination between two mononu-
A number of phenomena have been proposed to clear intermediates leads to organic product forma-
explain synergistic observations in homogeneous cat- tion and a dinuclear complex. Thus, these systems
alysis. One of the most frequent has been cluster cat- possess simultaneouslyboth mononuclear and hetero-
alysis,[3–5] i.e., the catalytic system consists of inter- binuclear complexes. Numerous examples exist in the
mediates all having the same hetero-bimetallic or literature for stoichiometric binuclear eliminations
hetero-multimetallic nuclearity. Another explanation leading to organic products,[8] and the catalytic version
has been the simultaneous presence of intermediates has been the focus of considerable work and
with various nuclearities interacting in a complex debate.[9–11] Eq. (1) shows a non-elementarystep in-
manner. A well documented example involves the I/ volving a hetero-bimetallic stoichiometric binuclear
G
Ru/Ir-promoted carbonylation of methanol to acetic elimination, where R1 and R2 are ligands which elimi-
acid,[6] where mononuclear ruthenium carbonyl, mon- nate to form the organic product R1R2, L are other li-
onuclear iridium carbonyl and bimetallic ruthenium-
iridium carbonyl complexes are simultaneously pres-
Adv. Synth. Catal. 2008, 350, 679 – 690
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
679

Chuanzhao Li et al.
FULL PAPERS
gands not involved in elimination, and M1 and M2 are lation systems are re-investigated using additional
different metallic elements. substrates. The immediate purpose is to demonstrate
Recently, the existence of the hetero-bimetallic cat- that hydroformylation CBER systems are not restrict-
alytic binuclear elimination reaction (CBER) has ed to just a few substrates. The broader purpose is to
been conclusivelydemonstrated in the hydroformyla-
highlight the synergistic effects and hence synthetic
tion reaction using rhodium-manganese and rhodium- advantages of mixed-metal CBER. This might pro-
rhenium mixed-metal systems.[12] Detailed in situ FT- mote further consideration within the catalytic com-
IR spectroscopic measurements showed the simulta- munityof (1) other classes of organic syntheses which
neous presence of both mononuclear and dinuclear might be achieved using a similar approach and (2)
intermediates in the active systems. Detailed kinetic other classes of organic syntheses which might be
modelling demonstrated that the rate limiting and more rapidlyor selectivelyachieved using a similar
product forming steps were the attack of HM(CO)5 approach.
{M=Mn,
Re}
on
RCORh(CO)4
{R=C5H9,
CH2CH2C
ACHTREUNG
rhodium-rhenium system, RhRe(CO)9 was spectro- Results
scopicallyobserved and shown to activate molecular
hydrogen at a very high rate. Perhaps most important- Rhodium-Rhenium Systems
ly, the catalytic binuclear elimination mechanism is
synthetically much more efficient than the classic uni- The organometallics Rh4(CO)12 and HRe(CO)5 were
cyclic mechanism which involves the hydrogenolysis used as precursors in this section. Separate hydrofor-
RCORh(CO)4. In the CBER systems, HM(CO)5 is ca. mylation experiments were conducted with the sub-
170–1200 times more reactive toward RCORh(CO)4 strates 3,3-dimethylbut-1-ene, styrene, and methylene
than molecular hydrogen. Scheme 1 shows a com- cyclohexane. In situ spectroscopic FT-IR measure-
bined unicyclic and CBER mechanism for hydrofor- ments were conducted for each system, and the sig-
mylation.
nals were analysed using the BTEM algorithm. The
In the present contribution, the rhodium-manga- experimental and computational issues are presented
nese and rhodium-rhenium mixed-metal hydroformy- in the Experimental Section.
Scheme 1. A combined unicyclic and CBER mechanism for mixed metal-catalyzed hydroformylation. The outer cycle con-
sists of just mononuclear rhodium intermediates and represents the classic hydroformylation mechanism. In this scheme,
CBER is made possible by(1) a step a representing the transformation of a dinuclear complex to mononuclear complexes
and (2) a step b representing binuclear elimination.
680
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 679 – 690

Synthetic Applications of Synergism using Catalytic Binuclear Elimination Reactions
FULL PAPERS
Observable Species Present
The spectroscopic results using 3,3-dimethylbut-1-ene
as substrate were selected to illustrate the methods
used for detailed analysis. A total of 120 spectra were
first subjected to singular value decomposition. The
resulting right singular vectors (basis vectors for the
observable components) are ordered according to
their contribution to the total signal variance and
some of these are shown in Figure 1. Localized signals
with good signal-to-noise ratios were clearlyobserved
in the first ca. 20 right singular vectors. In contrast,
the right singular vectors from the 50th vector onwards
consist of more-or-less randomlydistributed white
noise.
Since most of the useful signals associated with the
observable components are present in the first ca. 20–
25 right singular vectors, these were taken for further
analysis using the BTEM algorithm. The resulting
pure component spectral estimates are shown in
Figure 2. The pure component spectra of the solvent
and dissolved CO as well as the organic reactants 3,3-
dimethylbute-1-ene and 4,4-dimethylpentanal were
easilyrecovered. In addition, good spectral estimates
Figure 2. The BTEM recovered pure component spectra of
the solute species from the Rh-Re-catalyzed hydroformyla-
tion of 3,3-dimethylbut-1-ene.
for four observable organometallics were obtained, The Time-Dependent Concentrations Of
[13]
namely, Rh4(CO)12, HRe(CO)5, RCORh(CO)4 and Organometallics
RhRe(CO)9.[14] All of these organometallic species
were present at concentrations of 80 ppm or less. The The time-dependent concentrations of the reactive
last two species are non-isolatable complexes formed solutes were determined using an algebraic system
under reaction conditions.
identification algorithm (see Experimental Section for
details). The time-dependent concentrations of all
four observable organometallics were obtained and
these are shown in Figure 3a. It is seen that the con-
centrations of the organometallic precursors
Rh4(CO)12 and HRe(CO)5 decrease and the inter-
mediates RCORh(CO)4 and RhRe(CO)9 increase at
initial reaction times. At ca. 80 min, the concentra-
tions of all observable organometallics have reached a
pseudo-steadystate. It is noteworthythat the concen-
tration profiles are quite smooth, even at the verylow
concentrations present. This is due in part to the
high-quality in situ spectroscopic measurements but
also due to the multivariate numerical methods used.
The organometallics present in this active system,
namelyRh 4(CO)12, HRe(CO)5, RCORh(CO)4 and
RhRe(CO)9, are the same as those observed in the
hydroformylation of cyclopentene which was mod-
[12d]
elled previouslyin considerable detail elsewhere.
For comparison purposes, the time-dependent concen-
trations for Rh4(CO)12 and RCORh(CO)4 from a
pure rhodium-catalyzed hydroformylation are also
shown in Figure 3b. Comparison of Figure 3a and b
shows that the presence of HRe(CO)5 accelerates the
transformation of Rh4(CO)12 to RCORh(CO)4. This
phenomenon has been attributed to the attack of
HRe(CO)5 on a transient species {Rh4(CO)14} and this
Figure 1. The right singular vectors from a singular value de-
composition of the in situ FT-IR spectroscopic data from the
Rh-Re catalyzed hydroformylation of 3,3-dimethylbut-1-ene. has been modelled in detail elsewhere.[12d]
Adv. Synth. Catal. 2008, 350, 679 – 690
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
681

Chuanzhao Li et al.
FULL PAPERS
Figure 3. The time-dependent mole fractions of the organometallic solutes. (a) Initial reaction conditions were 12.0 mg
Rh4(CO)12, 4.5 mL HRe(CO)5. (b) Initial reaction conditions were 13.8 mg Rh4(CO)12. Both reactions used 0.5 mL 3,3-dime-
thylbut-1-ene in 50 mL hexane with 2.0 MPa CO and 1.0 MPa H2 at 298.0 K.
Observable Increase in Rate of Product Formation
tion/min (Rh-Re system) and 2.5”10À5 mole fraction/
min (Rh system). This marked rate increase is analo-
The time-dependent concentrations of the organic gous to that observed in the previouslyreported hy-
product 4,4-dimethylpentanal from the rhodium-rhe- droformylation of cyclopentene.[12d]
nium system are shown in Figure 4. In addition, and
for comparison purposes, the time-dependent concen-
trations of 4,4-dimethylpentanal in a pure rhodium Remarks for Styrene and Methylene-Cyclohexane
hydroformylation at comparable conditions are also
shown in this figure. It is clearlyseen that the rhodi-
BTEM analysis of the other two systems, using sty-
um-rhenium system is significantly more active than rene and methylenecyclohexane as substrates, also re-
the pure-rhodium system. Specifically at 120 min the sulted in the observation of Rh4(CO)12, HRe(CO)5,
rates of hydroformylation were 1.2”10À4 mole frac- RCORh(CO)4 and RhRe(CO)9 in the active mixed-
metal systems. In addition, significant hydroformyla-
tion rate increases were observed for both systems.
The corresponding concentrations of the total alde-
hyde products are shown in Figure 5 and Figure 6.
Since styrene and methylenecyclohexane are less re-
active than 3,3-dimethylbut-1-ene, slightly longer re-
action times are shown.
Kinetic Modelling of Product Formation
In mixed rhodium-rhenium hydroformylation, there
are two simultaneous mechanisms for product forma-
tion. Accordingly, there are two dimultaneous product
forming steps. These are (1) the hydrogenolysis of the
rhodium intermediate RCORh(CO)4 and (2) the
attack of HRe(CO)5 on RCORh(CO)4. Since the in-
stantaneous concentrations of these organometallics
Figure 4. A comparison of the time-dependent mole frac-
as well as the organic product can be measured, the
tions of the organic product 4,4-dimethylpentanal in a pure
contributions of each mechanism to total product for-
Rh-catalyzed hydroformylation and a Rh-Re-catalyzed hy-
mation can be evaluated. Eq. (2) describes this total
droformylation. Initial reaction conditions for Rh-Re were
12.0 mg Rh4(CO)12, 4.5 mL HRe(CO)5. Initial reaction condi-
tions for Rh onlywere 13.8 mg Rh (CO)12. Both reactions
4
used 0.5 mL 3,3-dimethylbut-1-ene in 50 mL hexane with
2.0 MPa CO and 1.0 MPa H2 at 298.0 K.
682
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 679 – 690

Synthetic Applications of Synergism using Catalytic Binuclear Elimination Reactions
FULL PAPERS
two parts, as shown in Eq. (3), to reflect the contribu-
tions from the unicyclic rhodium catalysis and the
rhodium-rhenium CBER.
The experimental runs using substrates 3,3-dime-
thylbut-1-ene, styrene, and methylenecyclohexane
were analyzed using Eq. (2). The resulting coefficients
k1 and k2 for the product forming steps are shown in
Table 1. For comparison purposes, the results from
pure rhodium hydroformylations are also included, as
well as the previouslyreported result for the rhodi-
um-rhenium hydroformylation of cyclopentene.
Table 1 clearlyshows that the k values for acyl hy-
1
drogenolysis are consistent for each particular sub-
strate (compare the value for the pure rhodium ex-
periments and the mixed rhodium-rhenium experi-
ments). In addition, it is clearlyseen that the second
term in Eq. (3), i.e., the attack of HRe(CO)5 on
RCORh(CO)4 is the main contribution to product for-
Figure 5. A comparison of the time-dependent mole frac-
tions of the organic product phenylpropanal in a pure Rh-
catalyzed hydroformylation and a Rh-Re-catalyzed hydro-
formylation. Initial reaction conditions for Rh-Re were
36.3 mg Rh4(CO)12, 45 mL HRe(CO)5. Initial reaction condi-
tions for Rh onlywere 91.2 mg Rh (CO)12. Both reactions
4
used 25 mL styrene in 300 mL hexane with 5.0 MPa CO and mation. When interpreting these results, it is instruc-
0.5 MPa H2 at 308.0 K.
tive to recognize that the molar ratio of Re/H2 in the
liquid phase is verylow, that is, on the order of 0.002
in all experiments. Therefore, it can be concluded that
the CBER synthetic pathway utilizes rhodium far
more effectively than the unicyclic synthetic pathway.
Rhodium-Manganese Systems
The organometallics Rh4(CO)12 and a mixture of
HMn(CO)5/Mn2(CO)10 were used as precursors in this
section (see Experimental Section). Separate hydro-
formylation experiments were conducted with the
substrates styrene, and methylenecyclohexane. In situ
spectroscopic FT-IR measurements were conducted
for each system, and the signals were analyzed using
the BTEM algorithm as demonstrated before.
Observable Species Present
Figure 6. A comparison of the time-dependent mole frac-
tions of the organic product 2-cyclohexylacetaldehyde in a
pure Rh-catalyzed hydroformylation and a Rh-Re-catalyzed
hydroformylation. Initial reaction conditions for Rh-Re
were 40.3 mg Rh4(CO)12, 20 mL HRe(CO)5. Initial reaction
condition for Rh onlywas 52.3 mg Rh 4(CO)12. Both reac-
tions used 2.1 mL methylenecyclohexane in 300 mL hexane
with 2.0 MPa CO and 2.0 MPa H2 at 289.7 K.
The spectroscopic results using methylenecyclohexane
were selected to illustrate the methods used for de-
tailed analysis in this section. A total of 90 spectra
were first subjected to singular value decomposition.
Since most of the useful signals associated with the
observable components are present in the first ca. 20–
25 right singular vectors, these were taken for further
analysis using the BTEM algorithm. The resulting
pure component spectral estimates are shown in
rate of formation for aldehydes where the reaction Figure 7. The pure component spectra of the solvent
[12d]
orders have been determined previously.
Finally, and dissolved CO as well as the organic reactants
methylenecyclohexane and 2-cyclohexylacetaldehyde
the total rate can be convenientlypartitioned into
Adv. Synth. Catal. 2008, 350, 679 – 690
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
683

Chuanzhao Li et al.
FULL PAPERS
Table 1. Kinetic parameters for Rh-Re hydroformylations.
Reaction Systems
Initial metal loading
(mol ratio)
k1 (minÀ1) k2 (minÀ1) rateRhÀRe CBER/rateRh T (K) Remarks
HRe(CO)5/Rh4(CO)12
3,3-Dimethylbut-1-ene with pure Rh NA
3,3-Dimethylbut-1-ene with Rh/Re 1.8
Methylenecyclohexane with pure Rh NA
0.48
0.51
0.05
0.05
3.4
NA
241
NA
5.5
NA
243
110
NA
6.1
NA
1.8
NA
9.1
4.2
298.0 This study
298.0 This study
289.7 This study
289.7 This study
308.0 This study
308.0 This study
Methylenecyclohexane with Rh/Re
Styrene with pure Rh
2.4
NA
6.0
2.3
Styrene with Rh/Re
2.5
0.33
Cyclopentene with Rh/Re
289.7
A
precursor Rh4(CO)12 approaches zero at long reaction
times. Again, the concentration profiles are quite
smooth, even at the verylow concentrations present.
The organometallics present in this active system,
namelyRh 4(CO)12, HMn(CO)5, RCORh(CO)4 and
Mn2(CO)10 are the same as those observed in previous
hydroformylations of 3,3-dimethylbut-1-ene and cy-
clopentene which were modelled in detail elsewher-
e.[12a,b] For comparison purposes, the time-dependent
concentrations for Rh4(CO)12 and RCORh(CO)4
from a pure rhodium-catalyzed hydroformylation are
also shown in Figure 8.
Observable Rate Increases
The time-dependent concentrations of the organic
product 2-cyclohexylacetaldehyde from this rhodium-
manganese system are shown in Figure 9. In addition,
and for comparison purposes, the time-dependent
concentrations of 2-cyclohexylacetaldehyde in a pure
rhodium hydroformylation are also shown in this
figure. It is clearlyseen that the rhodium-manganese
system is significantly more active than the pure rho-
dium system. Specifically at ca. 200 min the rates of
Figure 7. The BTEM recovered pure component spectra of
the solute species from the Rh-Mn-catalyzed hydroformyla-
tion of methylenecyclohexane.
were easilyrecovered. In addition, spectral estimates
for four observable organometallics were obtained, hydroformylation were 2.6”10À5 mole fraction/min
namely, Rh4(CO)12, HMn(CO)5, RCORh(CO)4 and (Rh-Mn system) and 1.0”10À5 mole fraction/min (Rh
Mn2(CO)10.
system). This marked rate increase is analogous to
that observed in the previous hydroformylations of
3,3-dimethylbut-1-ene and cyclopentene.[12a,b]
The Time-Dependent Concentrations of
Organometallics
Remarks for Styrene
The time-dependent concentrations of all four observ-
able organometallics were obtained and are shown in BTEM analysis of the styrene system also resulted in
Figure 8. It is seen that the concentrations of the or- the observation of Rh4(CO)12, HMn(CO)5, RCORh(-
ganometallic precursors decrease and the intermedi- CO)4 and Mn2(CO)10 in the active mixed-metal sys-
ate RCORh(CO)4 increases at initial reaction times. tems. In addition, significant hydroformylation rate
Furthermore, it is seen that (i) the concentrations of increases were observed for this system. The corre-
all observable organometallics approach a pseudo- sponding concentrations of the total aldehyde prod-
steadystate veryrapidly(more rapidlythan the Rh-
Re system) and (ii) the concentration of the limiting
ucts are shown in Figure 10.
684
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 679 – 690

Synthetic Applications of Synergism using Catalytic Binuclear Elimination Reactions
FULL PAPERS
Figure 8. The time-dependent mole fractions of the organometallic solutes. (a) Initial reaction conditions were 18.1 mg
Rh4(CO)12, and a mixture of HMn(CO)5/Mn2(CO)10 obtained from 425.3 mg Mn2(CO)10; (b) Initial reaction conditions were
52.3 mg Rh4(CO)12, Both reactions used 2.1 mL methylenecyclohexane in 300 mL hexane with 2.0 MPa CO and 2.0 MPa H2
at 289.7 K.
Figure 10. A comparison of the time-dependent mole frac-
tions of the aldehyde in a pure Rh-catalyzed hydroformyla-
tion and an Rh-Mn-catalyzed hydroformylation. Initial reac-
tion conditions for Rh-Mn were 91.5 mg Rh4(CO)12, and a
mixture of HMn(CO)5/Mn2(CO)10 obtained from 187.5 mg
Mn2(CO)10; Initial reaction conditions for Rh onlywere
91.2 mg Rh4(CO)12, Both reactions used 25 mL styrene in
300 mL hexane with 5.0 MPa CO and 0.5 MPa H2 at
308.0 K.
Figure 9. A comparison of the time-dependent mole frac-
tions of the organic product 2-cyclohexylacetaldehyde in a
pure Rh-catalyzed hydroformylation and an Rh-Mn-cata-
lyzed hydroformylation. Initial reaction conditions for Rh-
Mn were 18.1 mg Rh4(CO)12, and a mixture of HMn(CO)5/
Mn2(CO)10 obtained from 425.3 mg Mn2(CO)10; Initial reac-
tion conditions for Rh onlywere 52.3 mg Rh (CO)12, Both
4
reactions used 2.1 mL methylenecyclohexane in 300 mL
hexane with 2.0 MPa CO and 2.0 MPa H2 at 289.7 K.
Kinetic Modelling of Product Formation
(2) the attack of HMn(CO)5 on RCORh(CO)4. Since
the instantaneous concentrations of these organome-
In mixed rhodium-manganese hydroformylation, tallics as well as the organic product can be measured,
there are two simultaneous mechanisms for product the contributions of each mechanism to total product
formation. Accordingly, there are two simultaneous formation can be evaluated. Eq. (4) describes this
product forming steps. These are (1) the hydrogenoly- total rate of formation for aldehydes where the reac-
[12a,b]
sis of the rhodium intermediate RCORh(CO)4 and tion orders have been determined previously.
Fi-
Adv. Synth. Catal. 2008, 350, 679 – 690
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
685

Chuanzhao Li et al.
FULL PAPERS
Table 2. Kinetic parameters for Rh-Mn hydroformylations.
Reaction Systems
Initial metal loading
(mol ratio)
k1 (minÀ1) k2 (minÀ1) rateRhÀMn CBER/rateRh T (K) Remarks
HMn(CO)5/Rh4(CO)12
Methylenecyclohexane with pure Rh NA
Methylenecyclohexane with Rh/Mn 0.7
0.05
0.04
3.4
3.4
0.45
0.31
NA
26
NA
192
44.5
37
NA
1.3
NA
1.2
1.3
1.1
289.7 This study
289.7 This study
308.0 This study
308.0 This study
Styrene with pure Rh
Styrene with Rh/Mn
NA
0.2
3,3-Dimethylbut-1-ene with Rh/Mn 0.4
Cyclopentene with Rh/Mn 0.5
298.0
289.7
ACHTREUNG
AHCTREUNG
hydroformylation of cyclopentene. In addition, a sig-
nificant synergistic rate increase was measured for all
the present hydroformylations in the presence of Rh-
Re as compared to just Rh. This too is consistent with
nally, the total rate can be conveniently partitioned the previous measurements of the mixed-metal rhodi-
into two parts, as shown in Eq. (5), to reflect the con- um-rhenium hydroformylation of cyclopentene.
Simple regressions of the present total hydroformyla-
tion rates versus the time-dependent concentrations
provided reasonable kinetic parameters. Taken to-
gether, these results support the conclusion that the
tribution from the unicyclic rhodium catalysis and the present systems consist of two simultaneous catalytic
rhodium-manganese CBER.
systems (1) a classic unicyclic rhodium catalytic cycle
The experimental runs using substrates methylene- consisting of onlymononuclear intermediates and (2)
cyclohexane and styrene were analysed using Eq. (4). a mixed metal rhodium-rhenium catalytic binuclear
The resulting coefficients k1 and k2 for the product elimination. The interconnected mechanisms of the
forming steps are shown in Table 2. For comparison Rh-Re system are shown in Scheme 2.
purposes, the results from pure rhodium hydroformy-
lations are also included, as well as the previouslyre-
ported result for the rhodium-manganese hydrofor- Rh-Mn
mylation of 3,3-dimethylbute-1-ene and cyclopentene.
Table 2 clearlyshows that the k values for acyl hy- An analogous situation holds for the present mixed-
1
drogenolysis are consistent for each particular sub- metal
rhodium-manganese
[Rh4(CO)12
and
strate between the pure rhodium experiments and the HMn(CO)5/Mn2(CO)10] hydroformylations of methyl-
mixed rhodium-manganese experiments. In addition, enecyclohexane and styrene. Indeed, the present hy-
it is clearlyseen that the second term in Eq. (5), that
droformylations behave very similarly to the previous
is, the attack of HMn(CO)5 on RCORh(CO)4 is the rhodium-manganese hydroformylations of 3,3-dime-
main contribution to product formation. Again, when thylbut-1-ene and cyclopentene. The expected organ-
interpreting these results, it is instructive to recognize ometallic species were observed in the active systems
that the molar ratio of HMn(CO)5/H2 in the liquid and their concentration profiles were similar, the ex-
phase is on the order of 0.005 in all experiments.
pected rate increases were measured and regression
of the results provided reasonable kinetic parameters.
Accordingly, it is concluded that the present systems
consist of two simultaneous catalytic systems (1) a
classic unicyclic rhodium catalytic cycle consisting of
onlymononuclear intermediates and (2) a mixed
metal rhodium-manganese catalytic binuclear elimina-
Discussion
Rh-Re
The in situ spectroscopic results in the present mixed- tion. The interconnected mechanisms of the Rh-Mn
metal rhodium-rhenium hydroformylations of 3,3-di- systems are shown in Scheme 3.
methylbut-1-ene, methylenecyclohexane and styrene
showed that the active systems contain the observable
organometallics
Rh4(CO)12,
HRe(CO)5, Implications for Organic Synthesis
ACHTREUNG
well as the associated concentration profiles, is consis- The present study shows that synergistic CBER sys-
tent with the previous mixed-metal rhodium-rhenium tems consisting of Rh-Re and Rh-Mn are not limited
686
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 679 – 690

Synthetic Applications of Synergism using Catalytic Binuclear Elimination Reactions
FULL PAPERS
Scheme 2. A combined unicyclic and CBER mechanism for Rh-Re-catalyzed hydroformylation. The outer cycle consists of
just mononuclear rhodium intermediates and represents the classic hydroformylation mechanism.
Scheme 3. A combined unicyclic and CBER mechanism for Rh-Mn-catalyzed hydroformylation. The outer cycle consists of
just mononuclear rhodium intermediates and represents the classic hydroformylation mechanism.
to just a few selected substrates. Therefore, Rh-Re
and Rh-Mn CBER systems clearly have broader syn- reactions provide a number of additional opportuni-
thetic potential for hydroformylations. ties. Theses include (1) the possibilityof dramatically
The increased rates observed in these CBER type
Adv. Synth. Catal. 2008, 350, 679 – 690
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
687

Chuanzhao Li et al.
FULL PAPERS
(99%) were purchased from Fluka. Styrene was distilled
with CaH2 under argon before use and the other two al-
kenes were used as obtained. The Puriss quality n-hexane
(99.6%, Fluka AG) was distilled from sodium-potassium
under argon for ca. 5 h to remove the trace water and
oxygen.
increasing rate using a cheaper base metal (or de-
creasing the loading of the precious metal), and (2)
the possibilityof dramatic process intensification.
The latter issue arises from the bi-linear kinetics in
metal loadings, and hence, the opportunities to de-
crease the volumes of reaction vessels (assuming that
other mass-transfer effects, particularlygas-liquid
mass transfer, do not interfere).
[15]
Apparatus
Opportunities probablyexist for the development
of new CBER type synthetic strategies for other types
of reactions. This might include the identification of
(1) other classes of organic syntheses which might be
achieved using a similar approach and (2) other
classes of organic syntheses which might be more rap-
idlyor selectivelyachieved using a similar approach.
Since one or both metals can be potentiallymodified
in a CBER type catalytic system, a very broad range
of possibilities are available to the experimentalist for
fine-tuning synthetic aspects.
The in situ kinetic studies were performed in two kinds of
high-pressure FT-IR set-ups. For the hydroformylations per-
formed in 300 mL of n-hexane, a 1.5-L stainless steel
(SS316) autoclave (Pmax =22.5 MPa, Buchi-Uster, Swizer-
land) connected with a high pressure infrared cell was used.
The system is the same as that has been used for numerous
rhodium hydroformylations by this group.[18] The autoclave
was equipped with a packed magnetic stirrer with six-bladed
turbines in both the gas and liquid phases (autoclave Engi-
neers, Erie PA) and has a mantle for heating/cooling. The
liquid-phase reaction mixtures were circulated from the au-
toclave to and from the high pressure IR cell with a high
pressure membrane pump (Model DMK 30, Orlita AG,
Geissen Germany). A polyscience cryostat Model 9505 is
used to keep the entire system isothermal (Dtꢀ0.58C) for
the range 289–308 K. The high-pressure infrared cell was sit-
uated in a Perkin–Elmer Spectrum 2000 FT-IR spectrometer
with spectral resolution 4 cmÀ1 and intervals 0.2 cmÀ1 for the
range 1000–2500 cmÀ1. The detailed experimental scheme
can be found in ref.[18d]
For the hydroformylations performed in 50 mL of n-
hexane, an in-house designed 100-mL high pressure SS316
reactor was used, which was connected with an in situ injec-
tion block, and a high pressure magneticallydriven gear
pump (Model GAH-X2, Micro pump, USA). The high-pres-
sure infrared cell was situated in a Bruker Vertex-À710 FT-IR
spectrometer and the spectral resolution was 2 cm with an
interval of 0.12 cmÀ1 for the range 1000–2500 cmÀ1. A de-
tailed description of the reactor, pump and injection block
can be found elsewhere.[19]
Conclusions
Additional mixed-metal Rh-Re and Rh-Mn hydrofor-
mylation systems have been investigated. In situ spec-
troscopy during these hydroformylation, together with
kinetic modelling of product formation, indicate that
these systems consist of two simultaneous mecha-
nisms, namely, (1) a classic unicyclic rhodium catalytic
cycle consisting of only mononuclear intermediates
and (2) a mixed metal catalytic binuclear elimination.
These results indicate the broader applicabilityof
mixed-metal Rh-Re and Rh-Mn catalysis for hydro-
formylation, and suggest that CBER type synthetic
strategies might be applicable to other classes of or-
ganic syntheses.
The high-pressure infrared cell was constructed at ETH
Zurich using SS316 steel and could be heated and cooled.
The CaF2 single crystal window (Korth Monokristalle, Kiel
Germany) has dimensions of 40 mm diameter by 15 mm
thickness. Two sets of Viton and silicone gaskets provide the
necessarysealing, and Teflon spacers are used between the
windows.
Experimental Section
General Information
All solutions preparations and transfers were carried out
under purified argon (99.9995%, Saxol, Singapore) atmos-
phere using standard Schlenk techniques.[16] The argon was
further purified before use bypassing it through a de-oxy
and zeolite column. Purified carbon monoxide (Research
grade, 99.7%, Saxol, Singapore) and purified hydrogen
(99.9995%, Saxol, Singapore) were also further purified
through de-oxyand zeolite columns before theywere used
in the hydroformylation experiments.
In Situ Spectroscopic and Kinetic Studies
For the experiments carried out in the 1.5-L reactor, n-
hexane was transferred under argon to the autoclave. The
total system pressure was raised to the set CO partial pres-
sure, and the stirrer and high-pressure membrane pump
were started. Alkene solutions dissolved in 50 mL of n-
hexane were prepared, transferred to the high-pressure re-
servoir under argon, pressurized with CO and then added to
the autoclave. After equilibration, Rh4(CO)12 dissolved in
50 mL of n-hexane and a solution of Mn2(CO)10/HMn(CO)5
or HRe(CO)5 in n-hexane were prepared, transferred to the
high-pressure reservoir under argon, pressurized with CO
and then added to the autoclave. After equilibration, hydro-
gen was then added to initiate the syntheses.
All precursors Rh4(CO)12 (98%), Mn2(CO)10(98%) and
N
HRe(CO)5 (99%) were purchased from Strem Chemicals
(Newport, MA, USA) and were used as obtained.
MnH(CO)5 was prepared using a modification[12b] of a
known literature preparation.[17] This resulted in a mixture
of Mn2(CO)10 and MnH(CO)5 in n-hexane. The detailed
procedure can be found in ref.[12b] All alkenes, 3,3-dimethyl-
but-1-ene (99%), methylenecyclohexane (98%) and styrene
688
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 679 – 690

Synthetic Applications of Synergism using Catalytic Binuclear Elimination Reactions
FULL PAPERS
Table 3. Experimental design.
Experiment
n-
Second
CO
[MPa]
H2
[MPa]
Alkene
[mL]
Rh4(CO)12
[mg]
Temp.
[K]
Hexane Metal
[mL]
3,3-Dimethylbut-1-ene with Rh 50
3,3-Dimethylbut-1-ene with Rh/ 50
Re
NA
2.0
1.0
1.0
0.5
0.5
13.8
12.0
298.0
298.0
HRe(CO)5 [4.5 mL] 2.0
Methylenecyclohexane with Rh 300
NA
2.0
2.0
2.0
2.0
2.1
2.1
52.3
40.3
289.7
289.7
Methylenecyclohexane with
Rh/Re
300
HRe(CO)5
[20.0 mL]
Mn2(CO)10
[425.3 mg]
NA
HRe(CO)5
[45.0 mL]
Mn2(CO)10
[187.5 mg]
Methylenecyclohexane with
Rh/Mn
300
2.0
2.0
2.1
18.1
289.7
Styrene with Rh
Styrene with Rh/Re
300
300
5.0
5.0
0.5
0.5
25.0
25.0
91.2
36.3
308.0
308.0
Styrene with Rh/Mn
300
5.0
0.5
25.0
91.5
308.0
For the hydroformylations performed in 50 mL of n-
hexane, Rh4(CO)12 was dissolved in 50 mL of n-hexane and
transferred under argon to the reactor. The system was pres-
surized with CO, the stirrer and high-pressure gear pump
were started. Alkenes and HRe(CO) were injected through
the injection valves and the reaction was initiated byadding
H2. The experimental design is listed in Table 3.
imum rate of hydroformylation observed in the 1.5-L reac-
tor was ca. 6”10À7 mole/s and the maximum rate of hydro-
formylation observed in the 100-mL reactor was ca. 8”10À7
mole/s. Calculations confirm that all reactions performed in
this studybelong to the Hatta CategoryH, that is, infinitely
slow reaction compared to gas-liquid mass transfer. Accord-
ingly, the presently reported rates represent the intrinsic re-
action kinetics and not mass-transfer controlled kinetics.
Further details of such mass transfer considerations can be
found elsewhere.[20]
Conversions and Selectivities
In this study, initial rate data are emphasized. Thus conver-
sions of substrates ranged from only ca. 10% (for some pure
rhodium systems) to ca. 50% (for some mixed metal sys-
tems). Due to the rather low substrate loadings and due to
these low conversions, the partial pressures of the gaseous
Spectroscopy Issues
The spectra from each experimental run were consolidated
and analyzed with the spectral reconstruction algorithm
band-target entropyminimization (BTEM) in order to re-
cover the pure component spectra present. BTEM has been
reactants, namelyH and CO typically varied less than 10%
2
in anyrun.
Some selectivityissues arise with the use of the alkenes
3,3-dimethylbut-1-ene, methylenecyclohexane and styrene.
Although hydrogenation in unmodified rhodium-catalyzed
hydroformylation at these low temperatures is virtually un-
detectable, regio-isomers of the resulting aldehydes are
formed. Previous studies have shown that the linear to
branched ratios were typically (i) greater than 95:5 when
3,3-dimethylbut-1-ene was used as the substrate,[18a] (ii)
greater than 95:5 when methylenecyclohexane was used as
the substrate,[18b] and (iii) ca. 85:15 when styrene was used as
the substrate at 308 K.[18d] Accordingly, in this study, prod-
ucts will be reported as 4,4-dimethylpentanal, 2-cyclohexyl-
acetaldehyde, and lump-sum mixture of 2-phenylpropanal
and 3-phenylpropanal.
successfullyused to analyze data from numerous homogene-
[12,21]
ouslycatalyzed reactions.
Details of this algorithm can
be found elsewhere.[22] A review of this approach and its ap-
plication to homogeneous catalysis can be found else-
where.[23]
After BTEM analysis, to obtain the pure component spec-
tra and to emmunerate the number of observable species
present, further calculations were performed. These include
(1) renormalization of the data in order to account for
changes in path length of the high pressure cell as well as
changes in the reaction volume[24] as well as (2) further alge-
braic manipulations to obtain internal calibrations for all
species present and all the instantaneous reagent concentra-
tions.[25] Mole fractions are used throughout this study.
Mass Transfer Issues
Evaluation of Reaction Rates
Both experimental systems have been characterized previ-
ouslywith respect to gas-liquid mass transfer and the overall
mass transfer coefficients were obtained as 0.1 sÀ1 for H2
and 0.6 sÀ1 for CO (1.5-L reactor, 200 rpm), and ca. 0.03 sÀ1
for H2 (25-mL reactor, 600 rpm, the present 100-mL reactor
is geometricallysimilar to the 25-mL reactor).
A central difference scheme was used to evaluate all the ob-
servable reaction rates from the time-dependent concentra-
tion data. This approach has been shown to provide reliable
reaction rates as well as estimates of turn-over-frequen-
cies.[26]
[18,19]
The max-
Adv. Synth. Catal. 2008, 350, 679 – 690
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
689

Chuanzhao Li et al.
FULL PAPERS
j) C. K. Brown, D. Georglou, G. Wilkinson, J. Chem.
Soc. A 1971, 3120–3127; k) J. Schwartz, J. Cannon, J.
Am. Chem. Soc. 1974, 96, 4721; l) J. C. Barborak, K.
Cann, Organometallics, 1982, 1, 1726–1728.
Acknowledgements
Part of this work was supported by the Science and Engineer-
ing Research Council of A*STAR (Agency for Science, Tech-
nology and Research), Singapore.
[10] a) W. D. Jones, R. G. Bergman, J. Am. Chem. Soc.
1979, 101, 5447–5449; b) W. D. Jones, J. Huggins, R. G.
Bergman, J. Am. Chem. Soc. 1981, 103, 4415–4423;
c) J. P. Collman, J. Belmont, J. Brauman, J. Am. Chem.
Soc. 1983, 105, 7288–7794; d) I. P. Beletskaya, A. Z.
Voskoboinikov, G. K. Magomedov, Metalloorg. Khim.
1989, 2, 810–813; e) I. P. Beletskaya, G. K. Magome-
dov, A. Z. Voskoboinikov, J. Organomet. Chem. 1990,
385, 289–295.
[11] J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke,
Principles and Applications of Organotransition Metal
Chemistry, UniversityScience Books, Mill Valley, 1987.
[12] a) C. Li, E. Widjaja, M. Garland, J. Am. Chem. Soc.
2003, 125, 5540–5548; b) C. Li, E. Widjaja, M. Garland,
Organometallics 2004, 23, 4131–4138; c) G. Liu, C. Li,
L. Guo, M. Garland, J. Catal. 2006, 237, 67–78; d) C.
Li, L. Chen, M. Garland, J. Am. Chem. Soc. 2007, 129,
13327–13334.
References
[1] J. R. Kosak, Chem. Ind. 1996, 68, 31–41.
[2] G. Jenner, J. Organomet. Chem. 1988, 346, 237–251.
[3] a) E. L. Muetterties, Bull. Soc. Chim. Belg. 1975, 84,
959–986; b) R. Ugo. Catal. Rev. Sci. Eng. 1975, 11,
225–297; c) J. Lewis, B. F. G. Johnson, Pure Appl.
Chem. 1975, 44, 43–79; d) E. L. Muetterties, Science
1977, 196, 839–848.
[4] a) L. N. Lewis, Chem. Rev. 1993, 93, 2693–2730; b) G.
Suss-Fink, G. Meister, Adv. Organomet. Chem. 1993,
35, 41–134.
[5] R. D. Adams, F. A. Cotton, (Eds.), Catalysis by di- and
polynuclear metal cluster complexes,Wiley, New York,
1998.
[6] a) G. J. Sunley, D. J. Watson, Catal. Today 2000, 58,
293–307; b) R. Whyman, A. P. Wright, J. A. Iggo, B. T.
Heaton, J. Chem. Soc. Dalton Trans. 2002, 771–777;
c) A. P. Wright, R. Whyman, R. J. Watt, G. E. Morris
J. A. Iggo, B. T. Heaton, in: 13th International Symposi-
um on Homogeneous Catalysis (Book of Abstracts),
Tarragona, Spain, 2002, 245.
[7] a) D. S. Breslow, R. F. Heck, Chem. & Ind. 1960, 467;
b) R. F. Heck, D. S. Breslow, J. Am. Chem. Soc. 1961,
83, 4023–4027.
[8] a) I. Kovacs, F. Ungvary, L. Marko, Organometallics
1986, 5, 209–215; b) J. R. Norton, Acc. Chem. Res.
1979, 12, 139–145; c) M. Christina , F. G. A. Stone,
Inorg. Chem. 1972, 11, 102–107; d) P. Renaut, G. Tain-
turier, B. Gautheron, J. Organomet. Chem. 1978, 150,
C9-C10; e) B. Martin, D. K. Warner, J. R. Norton, J.
Am. Chem. Soc. 1986, 108, 33–39; f) R. J. Hoxmeier,
J. R. Blickensderfer, H. D. Kaez, Inorg. Chem. 1979, 18,
3453–3461; g) I. Kovacs, C. D. Hoff, F. Ungvary, L.
Marko, Organometallics 1985, 4, 1347–1350.
[9] a) K. H. von Brandes, H. B. Jonassen, Z. Anorg. Allg.
Chem. 1966, 343 , 215–219; b) F. Ungvary, L. Marko, J.
Organomet. Chem. 1969, 20, 205–209; c) B. H. Byers,
T. L. Brown, J. Am. Chem. Soc. 1977, 99, 2528–2532;
d) F. Ungvary, L. Marko, Organometallics 1983, 2,
1608–1612; e) C. Hoff, F. Ungvary, R. B. King, L.
Marko, J. Am. Chem. Soc. 1985, 107, 666–671; f) J. R.
Norton, W. Carter, J. Kelland, S. Okrasinski, Advances
in Chemistry Series, 1978, 167, 170–180; g) R. T.
Edidin, K. Hennessy, A. Moody, S. Okrasinski, J. R.
Norton, New J. Chem. 1988, 12, 475–477; h) M. J.
Nappa, R. Santi, S. Diefenbach, J. Halpern, J. Am.
Chem. Soc. 1982, 104, 619–621; i) M. J. Nappa , R.
Santi, J. Halpern, Organometallics 1985, 4, 34–41;
[13] M. Garland, G. Bor, Inorg. Chem. 1989, 28, 410–413.
[14] C. Li, L. Guo, M. Garland, Organometallics 2004, 23,
5275–5279.
[15] C. Ramshaw, Green Chem. 1999, 1, G15-G17.
[16] D. F. Shriver, M. A. Drezdzon, The Manipulation of
Air-Sensitive Compounds, Wiley, New York, 1986.
[17] B. H. Byers, T. L. Brown, J. Am. Chem. Soc. 1977, 99,
2527–2532.
[18] a) M. Garland, P. Pino, Organometallics 1990, 10,
1693–1704; b) G. Liu, R. Volken, M. Garland, Organo-
metallics 1999, 18, 3429–3436; c) J. Feng, M. Garland,
Organometallics 1999, 18, 417–427; d) J. Feng, M. Gar-
land, Organometallics 1999, 18, 1542–1546.
[19] F. Gao, K. Ng, C. Li, K. Krummel, A. Allian, M. Gar-
land. J. Catal. 2006, 237, 49–57.
[20] M. Garland, Transport Effects in Homogeneous Cataly-
sis, in: Encyclopedia of Catalysis, (Ed.: I. T. Horvath),
Wiley, New York, 2002.
[21] C. Li, E. Widjaja, W. Chew, M. Garland, Angew. Chem.
2002, 114, 3939–3943; Angew. Chem. Int. Ed. 2002, 41,
3785–3789.
[22] a) W. Chew, E. Widjaja, M. Garland, Organometallics
2002, 21, 1882–1990; b) E. Widjaja, C. Li, M. Garland,
Organometallics 2002, 21, 1991–1997.
[23] M. Garland, in: Mechanisms in Homogeneous Cataly-
sis: A Spectroscopic Approach, (Ed.: B. Heaton),
Wiley, Weinheim, 2005, pp 151–193.
[24] W. Chew, E. Widjaja, M. Garland, Applied Spectrosco-
py, 2007, 61, 734–746.
[25] E. Widjaja, C. Li, M. Garland, J. Catal. 2004, 223, 278–
289.
[26] R. Shirt, M. Garland, D. W. T. Rippin, Anal. Chim.
Acta 1998, 374, 67–91.
690
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 679 – 690