Reactive separation of dilute ethylene by hydroformylation using slurried rhodium catalysts on phosphinated resins and silica

Article abstract of DOI:10.1016/j.catcom.2011.05.001

A series of slurried rhodium catalysts on phosphinated resins and silica were tested for the reactive separation of dilute ethylene by hydroformylation to propanal. The amount of rhodium leaching was determined for most of the systems investigated. Leaching ranged from <2 to 11% for catalytic runs for resin supported systems, but was found to be much more significant for the phosphinated silica system studied, 19%. The level of rhodium leaching correlated well with the loss in rate of propanal formation between repeated experiments using recycled catalysts for the resin based systems, but the loss in rate was much more significant for the silica based system. This level of leaching is very high for a commercial process for basic chemical synthesis, for which leaching is typically at the ppb level. Rates of propanal formation for resin supported catalyst systems correlated well with the cone angle of the free phosphine, with rates as high as 0.35 mol/l/h. However, while smaller cone angles were found to be directly proportional to rate, supported systems were less active than homogeneous counterparts.

Full text of DOI:10.1016/j.catcom.2011.05.001

Catalysis Communications 12 (2011) 1323–1327
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Reactive separation of dilute ethylene by hydroformylation using slurried rhodium
catalysts on phosphinated resins and silica
William J. Tenn III, Russell C. Singley, Brandon R. Rodriguez ⁎, Jennifer C. DellaMea
Hydrocarbons R&D, The Dow Chemical Company, 2301 N. Brazosport Blvd., Freeport, TX 77541, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of slurried rhodium catalysts on phosphinated resins and silica were tested for the reactive separation
of dilute ethylene by hydroformylation to propanal. The amount of rhodium leaching was determined for
most of the systems investigated. Leaching ranged from b2 to 11% for catalytic runs for resin supported
systems, but was found to be much more significant for the phosphinated silica system studied, 19%. The
level of rhodium leaching correlated well with the loss in rate of propanal formation between repeated
experiments using recycled catalysts for the resin based systems, but the loss in rate was much more
significant for the silica based system. This level of leaching is very high for a commercial process for basic
chemical synthesis, for which leaching is typically at the ppb level. Rates of propanal formation for resin
supported catalyst systems correlated well with the cone angle of the free phosphine, with rates as high as
0.35 mol/l/h. However, while smaller cone angles were found to be directly proportional to rate, supported
systems were less active than homogeneous counterparts.
Received 23 March 2011
Received in revised form 27 April 2011
Accepted 4 May 2011
Available online 14 May 2011
Keywords:
Hydroformylation
Supported catalysts
Propanol
Propanal
Phosphine
Rhodium
©
2011 Elsevier B.V. All rights reserved.
1
. Introduction
Global demand for propylene growth has been outpacing that
for ethylene for more than a decade [1]. Historically, propylene has
been produced as a by-product of ethylene production in steam
crackers, and gasoline production in refinery FCC operations. There
has traditionally been very little “on-purpose” propylene production.
The recent addition of Middle-East ethane crackers is exacerbating
this problem. Thus, there has been a recent trend toward develop-
ment of alternative sources of propylene, without adversely affecting
ethylene production.
One technology being explored is the conversion of ethylene in low-
value, dilute ethylene streams into n-propanol, a shippable intermedi-
ate, which can be subsequently converted to propylene via dehydration.
To address this growing demand a number of on-purpose pro-
pylene technology offerings are being explored, including tech-
nologies such as metathesis, catalytic cracking of higher-olefin rich
feeds, methanol-to-olefins/methanol-to-propylene, and propane
dehydrogenation.
Refinery off-gas, particularly that produced by fluid catalytic
cracker (FCC) units, typically contains fairly high quantities of olefin
components which are not recovered economically by conventional
cryogenic separations, and instead are burned as fuel. The concen-
tration of these olefins can be quite substantial, often up to 20% in
ethylene [2].
The first step can be accomplished either as a 1-step reductive-
hydroformylation of ethylene to propanol, or as a 2-step process of
hydroformylation of ethylene to propanal, followed by a reduction of
the propanal to propanol. Either of which can be essentially viewed as
a reactive separation of ethylene from a dilute gas stream.
In this work we present results of an investigation of slurried
rhodium catalysts ligated/supported by phosphine-functionalized
polymers and a phosphinated silica for the hydroformylation of dilute
ethylene. Each of the supported-phosphine/Rh complexes was gen-
erated in situ and tested for ethylene hydroformylation. The goals of
this study were to establish the feasibility of using these supported
catalysts for the aforementioned reaction, by the comparison of their
reaction rates, lifetime, and selectivity to the non-supported coun-
terparts. The major obstacle with the use of resin-supported hydro-
formylation catalysts, which has largely prevented their commercial
⁎
Corresponding author. Tel.: +1 979 238 1389; fax: +1 979 238 0028.
E-mail address: barodriguez@dow.com (B.R. Rodriguez).
1566-7367/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2011.05.001

1324
W.J. Tenn III et al. / Catalysis Communications 12 (2011) 1323–1327
application, [3] and in particular, their application in flow reactors, is
the leaching of the active material from the support. Thus, an impor-
tant point to be determined is the amount of catalyst that is leached
from the support under reaction conditions.
variable speed stirrer motor was adapted with a larger pulley to permit
stirring speeds in excess of 1000 rpm for optimal gas entrainment.
The reactor had facilities for both gas and liquid sampling while in
operation. In a typical experiment, the reactor was charged with the
catalyst precursor dicarbonylacetylacetonato rhodium(I) (~12 mg), the
supported ligand (~100 mg), and solvent (100 mL) inside a glove-box
under nitrogen. The reactor was then sealed, removed from the glove-
box, mounted to heating/agitation system inside a fume hood, pressure
tested with inert gas, and then heated to the desired temperature
(typically 100 °C). From a 1:1:1 cylinder of ethylene, hydrogen, carbon
monoxide the pressure was increased to the desired level (either 10 atm
or 2 atm) and then nitrogen is added to the system to increase the total
pressure to 34 atm. The 1:1:1 feed cylinder was connected to a Brooks
mass-flow-controller operating as a flow-meter which maintained
the pressure at 34 atm by introducing make-up feed gas to the reactor.
Analysis of the headspace of the reactor was performed before and after
each run to quantitate the amount of ethane formed. The liquid phase of
the reactor was sampled at 0, 15, 30, 60, 90, and 120 min for analysis
with 1 mL samples. Analysis of the liquid phase products of each reac-
tion was carried-out on a Hewlett-Packard 6890 GC equipped with a
methyl-silicone gum capillary column, a flame-ionization detector,
and quantitated using a calibration curve. Typically, only propanal
and the tetraglyme solvent were observed in analyses. The rate of
propanal formation was calculated using the GC analysis of the liquid
phase, and/or the rate of feed consumption through the flow-meter.
The rhodium content of the solutions after the catalytic reactions
was determined by flame atomic absorption spectroscopy. Swelling
volume was determined in a graduated cylinder by recording the
volume of a given quantity of dry resin and then measuring the
expansion of the resin in an excess of toluene (or tetraglyme), after 3
hours at ambient temperature.
The type of phosphinated supports used in this investiga-
tion are lightly cross-linked, macroporous, phosphine-functionalized
poly(styrene-co-divinylbenzene), Merrifield-type resins, and a phos-
phinated silica gel. These resins are insoluble, but able to swell sig-
nificantly in many organic solvent media. The phosphine loading
levels, surface area, pore volume, mesh size, amount of cross-linking,
and swelling volumes for the phosphinated supports are reported in
Table 1. Rhodium hydroformylation catalysts supported on functiona-
lized polymers have been studied since the early sixties [4]. These
resins represent the most widely investigated class of catalyst supports
used in studies of supported hydroformylation catalysts. They are com-
mercially available, are inert toward reaction conditions, and offer
ease of modification. Immobilization of a homogeneous catalyst on a
support can increase the catalyst lifetime, or range of operation of the
catalytic system. By binding the active component at particular sites
and preventing aggregation to larger metal units the system activity
may be maintained [5,6].
2
. Experimental
All of the phosphinated resins and silica utilized in this study were
purchased from commercial sources: triphenylphosphine, polymer-
bound (Aldrich), dicyclohexylphenylphosphine polymer-bound (Al-
drich), di(n-butyl)phenylphosphine, polymer-bound (EMD Chemicals),
benzyldiphenylphosphine, polymer-bound (Fluka), and 2-diphenyl-
phosphinoethyl-functionalized silica gel (Aldrich). The (acac)Rh(CO)
2
Rates are expressed in moles of aldehyde formed per mol of
catalyst per liquid volume within the reactor.
3
. Results
Each of the catalyst systems were evaluated at two sets of feed
partial pressures. In the first case, 3.4 atm partial pressures of each
ethylene, hydrogen, and carbon monoxide were charged to the
reactor, which accounted for 10.2 atm of the total 34 atm pressure on
the reactor. In the second, 0.7 atm each of ethylene, hydrogen, and
carbon monoxide accounting for 2.0 atm of the total 34 atm on the
reactor. The results of experiments at 3.4 atm and various process
conditions are presented in Fig. 1 and Table 2.
was purchased from Strem Chemical, and the solvents (tetraglyme
and propanal) from Aldrich. All reagents were stored inside a nitrogen
purged glove-box and used without further purification. The solvents
were stored over 4 Å molecular sieves.
These experiments were conducted in semi-batch mode in a
00 cm , high-pressure, stirred autoclave, from Parr Instrument Com-
pany (Model 4561). The reactor was equipped with baffles, a hollow
shaft/gas-entrainment impeller and was stirred at 1300 rpm to insure
thorough gas/liquid mixing during the reaction. The standard 1/8 hp
All of the catalysts were found to exhibit stable rates of propanal
formation over the course of each experiment (2–6 h). However,
studies to test for deactivation over longer periods were precluded
due to the semi-batch mode of operation. The rate was found to
be independent of ligand-to-rhodium ratios over the range studied
3
3
(
P/Rh~2–6). Although an excess of the supported ligand should not
Table 1
Properties of the phosphine-modified supports.
P-loading Mesh
mmol/g)
Cross link Pore
Surf. areab Swelling
2
a
(
(% DVB)
size (Å) (m /g)
factor
PS–Ph–PPh
PS–Ph–P(n–Bu)
PS–Ph–PCy
PS–Bz–PPh
2
3.2
100–200
100–200
50–100
200–400
200–400
2
1
1
2
–
–
–
2.0
3.3
2.9
1.2
–
2
0.66
1–2
2.5
–
–
2
–
–
2
–
–
Si–Et–PPh
2
0.7
60
500
a
The swelling factor is the volume of the swollen resin divided by the volume of the
equivalent amount of the dry resin. Determined in toluene after 3 h; no swelling in
tetraglyme was observed after 3 h.
b
Measurement of the dry-state surface area of the resins was attempted by BET, but
the results were not quantifiable due to the ultra-low surface areas, as expected with
the gel-type morphology [7].
Fig. 1. Formation of propanal with time using rhodium with supported ligands: 34 atm
of feed, with 3.4 atm of each CO, H , and C . T=100 °C, stirring rate=1300 rpm.
2
2 4
H

W.J. Tenn III et al. / Catalysis Communications 12 (2011) 1323–1327
1325
Table 2
Results at 3.4 atm each C
1
, and CO.a
2
2
4
H , H
No. Ligand
Rh (mmol) P:Rh Rate (mol/Lh) % Ethane T (°C)
0.8
0.6
0.4
1
2
3
4
5
6
7
8
9
PS–Ph–PPh
PS–Ph–PPh
PS–Ph–PPh
PS–Ph–P(nBu)
PS–Ph–P(nBu)
PS–Ph–PCy
PS–Ph–PCy
PS–Bz–PPh
Si–Et–PPh
2
2
2
0.05
0.05
0.05
0.05
0.10
0.05
0.06
0.05
0.05
6
6
6
6
6
6
3
6
6
0.440± 0.1
0.597
0.481
0.875± 0.2
1.417
0.251± 0.09
0.185
4.393
5.439
5.175
0.220
0.000
0.002
0.687
0.081
0.204
100
120
140
100
100
100
100
100
100
b
b
2
2
2
0.2
2
2
0.081
0.791
0
2
0
1
2
3
4
a
Total pressure was maintained at 34 atm; solvent=tetraglyme, 100 mL; T=100 °C;
time (h)
stirring rate=1300 rpm; reaction time=2 h.
b
Experiments 2 and 3 were performed sequentially after the experiment at 100 °C.
Fig. 3. Rate of propanal formation at two different agitation speeds using PS–Ph–PPh
2
/Rh
2 2 4
with 0.7 atm each CO, H , and C H . Blue diamonds are at 1300 rpm and red triangles are
at 1500 rpm.
be required, typically an excess of phosphine was used with the aim of
suppression of metal leaching. Only the phosphinated silica catalyst
system, Si–Et–PPh
of propanal formation appeared to be stable over the period that
the Si–Et–PPh catalyst was studied, the rate of the second trial was
significantly lower than if that condition had been run initially [8].
2
exhibited an unstable performance. While the rate
times the rate using the analogous supported system, PS–Ph–PPh
Approximately twice as much ethane was generated (0.29% using
2
/Rh.
2
2% PPh
analog.
3 2
vs. 0.7% using PS–Ph–PPh ) by the corresponding supported
Formation of n-propanol was only observed using the Si–Et–PPh
system, and then only in trace amounts.
2
The rate was also examined in propanal solvent using the PS–Ph–
PPh /Rh system and was found to be ~4 times slower (~0.1 mol/Lh at
2
To insure that the system was not under any gas–liquid/liquid–solid
mass transfer limitations, the effect of the rhodium pre-catalyst con-
centration on the rate of propanal formation was measured. The rate
of propanal formation was found to be first order in rhodium; doubling
the rhodium precursor concentration led to approximately double the
observed rate, suggesting gas–liquid resistance is not affecting the rate
as shown in Fig. 2.
In addition to varying the rhodium loading in the system, the affect
of changing the agitation rate was tested. To do so, the rate of propanal
formation was measured at two different agitation speeds as shown
in Fig. 3. No change in the rate of propanal formation occurred when
the agitation rate was adjusted from 1300 to 1500 rpm. Thus, gas–
liquid resistance and solid–liquid mass transfer are not affecting the
rate. A rate of approximately 1000 rpm was necessary in previous
tests for good mixing using the gas-entrainment impeller in-service in
the system.
3.4 atm each of the feeds).
Ethylene hydrogenation was accounted by quantification of the
total ethane accumulated in the headspace of the reactor over the
course of the run, Fig. 5. Hydrogenation was only significant for the
2
PS–Ph–PPh ligated system, the only sample where visual evidence of
rhodium metal was observed (green solution, and grey material). It
would be expected that if there were agglomerated metal in solution
that it would be an excellent ethylene hydrogenation catalyst and
thus the higher levels of ethane observed when using PS–Ph–PPh is
3
consistent.
It was observed that for the three ligand-modified resins wherein
the phosphine was bound to the support through a phenyl linker
(PS–Ph–PR
formation was directly proportional to the cone angle of the free
phosphine analog (PR ), with the rate of propanal formation increasing
2
, where R=Ph, Cy, n-Bu) that the observed rate of propanal
3
with decreasing steric demand. The cone angle (θ) of a monodentate
ligand is defined as the apex angle of a cylindrical cone (for phos-
phines, centered at 2.28 Å from the center of the phosphorous atom),
which touches the outermost atoms of the model [9]. In the commonly
accepted dissociative mechanism for hydroformylation, the rhodium
complex formed with triphenylphosphine generates an active species
Reduction of the partial pressures of the feed to 0.7 atm each
resulted in approximately a 3-fold decrease in the rate of propanal
formation, which is in accordance with previous studies. Our results of
dilute ethylene hydroformylation with 0.7 atm each of CO, H
and 32 atm of nitrogen are presented in Fig. 4 and Table 3.
2
, C
2
H
4
,
For comparison, (acac)Rh(CO)
2
run in the same reactor using 2%
3 3
such as (PPh ) Rh(CO)H. If triphenylphosphine is substituted by a
triphenylphosphine in tetraglyme, at 100 °C, 0.7 atm of each CO, H
C H , with 32 atm of nitrogen, an overall rate of propanal formation
2 4
of 0.6 mol/Lh after 2 h was obtained, which is approximately three-
2
,
more sterically demanding phosphine, the increased steric hindrance
around the rhodium center can inhibit ethylene association and may
result in a decreased overall rate (Fig. 6).
3
2
1
0
.5
2
y = 1.4174x + 0.1136
y = 0.6513x - 0.0057
.5
1
.5
0
0
0.5
1
1.5
2
time (h)
Fig. 2. Rate of propanal formation using double the amount of catalyst typical (200 ppm
Rh and 0.2% ligand), squares, PS–Ph–PPh
2
/Rh with 3.4 atm each CO, H
2
, and C
H
2 4,
,
Fig. 4. Formation of propanal with time using rhodium with various supported ligands:
triangles.
2 2 4
34 atm total, with 0.7 atm of each CO, H , and C H . T=100 °C, stirring rate=1300 rpm.

1326
W.J. Tenn III et al. / Catalysis Communications 12 (2011) 1323–1327
Table 3
Results at 0.7 atm each C
Table 4
2
H , H
4 2
, and CO.a
Rate loss for recycled catalyst and rhodium leaching.
No.
Ligand
Ligand
mmol P)
Rh
(mmol)
P:Rh
Rate
(mol/Lh)
% Ethane
No.
Ligand
Rate loss (%)a
Rhsoln (ppm)b
(
1
PS–Ph–PPh
2
9
5
9.9
5
b 2
11
19
b 2
1
2
3
4
5
6
7
PS–Ph–PPh
PS–Ph–P(nBu)
PS–Ph–PCy
PS–Ph–PCy
PS–Bz–PPh
PS–Bz–PPh
2
0.3
0.3
0.3
0.2
0.3
0.1
0.3
0.05
0.05
0.05
0.06
0.05
0.05
0.05
6
6
6
3
6
2
6
0.17± 0.04
0.23± 0.1
0.05± 0.004
0.05
0.03
0.02
0.71
0.04
0.00
0.19
0.00
0.03
0.08
2
3
4
5
6
PS–Ph–P(n–Bu)
PS–Ph–PCy
PS–Bz–PPh
2
2
2
–
–
2
2
2
Si–Et–PPh
2
78
–
c
2
Blank
2
a
Difference in rate before and after discharging reactor contents and reusing the
Si–Et–PPh
2
0.18
catalyst in a fresh reaction; based on rates at 3.4 atm feed partial pressures.
a
b
Total pressure was maintained at 34 atm. T=100 °C, stirring rate=1300 rpm,
reaction time=2 h, solvent=tetraglyme.
Amount of rhodium in solution after the run by flame A.A.
c
Solvent sample after heating (acac)Rh(CO)
prior to reaction.
2 2
with PS–Ph–P(nBu) in tetraglyme
To determine the rhodium loss by leaching from the catalyst
during the reaction, after the reaction was complete the liquids and
gasses were removed from the reactor by venting and decanting, and
then fresh solvent was added and the reaction was repeated using the
used catalyst from the previous run. The difference in rate between
the two runs was determined. The solvent from the first run was
analyzed by atomic absorption spectroscopy to determine the amount
of free rhodium present in the liquid phase. Significantly, no free
since they were less active. However, this level of leaching (~5%) is
very high for a commercial process for basic chemical synthesis, for
which leaching is typically at the ppb level.
4. Conclusions
Rhodium ligated by phosphine-modified resin and silica supports
have been investigated for the hydroformylation of ethylene in a
dilute gas stream. Rates of propanal formation were determined to be
directly proportional to catalyst concentration. While challenges will
likely present themselves as one moves from a dilute ethylene stream
in inert gas to a more complicated feed stream, these experiments
begin to elucidate likely candidates moving forward and can provide
some baseline economic analyses. Furthermore, using an array of
several phosphine ligands, it was determined that the ligand cone
angle has an indirect correlation with aldehyde formation, with a
maximum rate of 0.85 mol/L/h. Ethane formation varied significan-
tly from ligand to ligand. The amount of rhodium leaching was
determined for most of the systems investigated and ranged from
b2 to 11% for catalytic runs for resin supported systems. The level of
rhodium leaching correlated well with the loss in rate of propanal
formation between repeated experiments using recycled catalysts
for the resin based systems, but the loss in rate was much more
significant for the silica based system. As mentioned previously, the
rates of supported systems are lower than the homogeneous ana-
logues. However, should the metal leaching be reduced allowing
for more complete catalyst recyclability, this would provide a cheap,
easily recollected catalyst capable of hydroformylation of dilute
streams.
rhodium was detected in the solution after heating (acac)Rh(CO)
with PS–Ph–P(n-Bu) prior to initiating hydroformylation. For the PS–
Ph–PPh and PS–Ph–P(n-Bu) ligated systems, the two most efficient
2
2
2
2
systems reported in this study, these two values were found to
correlate well with each other, see Table 4. Experiments using
recycled PS–Bz–PPh /Rh and PS–Ph–PCy /Rh were not performed
2 2
6
5
4
3
2
1
0
PS-Ph-PPh2 PS-Ph-PPh2 PS-Ph-PBu2 PS-Ph-PCy2 Si-Et-PPh2 Ps-Bz-PPh2
recycle)
(
Fig. 5. Mole % of ethane in the headspace of the reactor after the reaction using rhodium
on various phosphinated supports.
Acknowledgements
We gratefully acknowledge Luis Bollmann, Susan Domke, Rick
Wehmeyer, Kurt Olson, Howard Clark, and Mark Kaminsky for helpful
suggestions.
1.2
1
References
0
0
0
0
.8
.6
.4
.2
0
[1] SRI 5.3% annual growth, PP at 7.3%/year.
[
2] (a) L. Xu, J. Liu, Q. Wang, S. Liu, W. Xin, Y. Xu, Appl. Catal. A 258 (2004) 47–53;
b) S. Wang, B. Shen, L. Qu, Catal. Today 98 (2004) 339–342.
(
[
3] (a) N. Yoneda, T. Minami, J. Weiszmann, B. Spehlmann, The Chiyoda/UOP
Acetica™ Process: A Novel Acetic Acid Technology, in: H. Hattori, K. Otsuka
(Eds.), Surface Science and Catalysis, Vol. 121, Kodansha, Tokyo, 1999,
pp. 93–98;(b) Yoneda, N.; Shiroto, Y.; Hamato, K.; Asaoka, S.; Maekima, T.
U.S. Patent 5,334,755, 1994.
(
b) N. Yoneda, T. Minami, K. Hamato, Y. Shiroto, Y. Hosono, J. Jpn. Petrol. Inst. 46
2003) 229 plant design and feasibility study of process for the
hydroformylation of 1-octene to nonanal utilizing rhodium catalyst on
(
A
a
a
120
130
140
150
160
170
180
immobilized on phosphinated silica has been reported: H. van den Berg and
A. G. J. van der Ham, University of Twente, Department of Science and Technology,
Cone Angle (theta)
2004anddiscussed inReek, J.N. H.; vanLeeuwen, P. W. N. M.;van derHam, A.G. J.;
De Haan, A. B. “Immobilizationof Tailor-madeHomogeneousCatalysts” inCatalyst
Separation, Recovery, and Recycling; Cole-Hamilton, D. J.; Tooze, R. P., Eds.; 2006,
Springer, The Netherlands.A rhodium catalyst immobilized on a phosphinated
Fig. 6. Relationship between the rate of propanal formation and the steric demand of
the ligand.

W.J. Tenn III et al. / Catalysis Communications 12 (2011) 1323–1327
1327
resin is utilized in the commercial Acetica Process developed by Chiyoda/UOP for
the carbonylation of methanol.
[6] J.P. Collman, L.S. Hegedus, M.P. Cooke, J.R. Norton, G. Dolcetti, D.N. Marquardt,
J. Am. Chem. Soc. 94 (1972) 1789.
[
[
4] (a) Tjeerd Jongsma “Polymer-Bound Rhodium Hydroformylation Catalysts” Thesis,
Rijksuniversiteit Groningen, 1992.
[7] D.C. Sherrington, A. Lanver, H.-G. Schmalz, B. Wilson, X. Ni, S. Yuan, Angew. Chem.
Int. Ed. 41 (2002) 3656. Commonly b 5 m2g-1, see:
[8] For example, when the system was run first at 3.4 atm of each of the feeds and then
switched to 0.7 atm of each of the feeds, then the rate at 0.7 atm was much less
than when the system was started at 0.7 atm each, and vice versa.
[9] C.A. Tolman, Chem. Rev. 77 (1977) 313.
(b) P.W.N.M. van Leeuwen, C. Claver (Eds.), Rhodium Catalyzed hydroformylation,
Kluwer Academic Publishers, Dordecht, The Netherlands, 2000, Chapter 10 in.
5] R.H. Grubbs, C. Gibbons, L.C. Kroll, W.D. Bonds, C.H. Brubaker, J. Am. Chem. Soc. 95
1973) 2373.
(