Rhodium-catalyzed hydroformylation of alkenes using in situ high-pressure IR and polymer matrix techniques

Article abstract of DOI:10.1021/om020854n

This paper exploits an unusual high-pressure cell together with in situ FTIR and polymer matrix techniques to study Rh-catalyzed hydroformylation of 1-octene, 1-butene, propene, and ethene, using either Rh(acac)(CO)₂ or Rh(acac)(CO)(PPh₃) in a polyethylene matrix. The technique requires only micrograms of catalyst. Using Rh(acac)(CO)₂ as the catalyst precursor, we were able to characterize the previously observed acyl rhodium tetracarbonyl intermediates, RC(O)Rh(CO)₄, using IR spectroscopy. For the PPh₃-modified reactions, the analogous acyl rhodium tricarbonyl triphenylphosphine intermediates, RC(O)Rh(CO)₃(PPh₃), have been observed. The IR spectra suggest a trigonal-bipyramidal structure with the acyl ligand and the PPh₃ ligand occupying the axial positions. The acyl rhodium tetracarbonyl intermediates, RC(O)Rh(CO)₄, can easily react with ethene to form acyl rhodium tricarbonyl ethene species RC(O)Rh(CO)₃(C₂H₄), again identified via IR. The resulting hydroformylation products were extracted from the polyethylene film after the reaction and analyzed by GC. The results indicate that, in the PE matrix, the PPh₃-modified hydroformylation gives a higher linear-to-branched ratio for the unsymmetrical alkenes.

Full text of DOI:10.1021/om020854n

1612
Organometallics 2003, 22, 1612-1618
Rh od iu m -Ca ta lyzed Hyd r ofor m yla tion of Alk en es Usin g
in Situ High -P r essu r e IR a n d P olym er Ma tr ix
Tech n iqu es
J ie Zhang, Martyn Poliakoff,* and Michael W. George*
School of Chemistry, University of Nottingham, University Park,
Nottingham, NG7 2RD, United Kingdom
Received October 15, 2002
This paper exploits an unusual high-pressure cell together with in situ FTIR and polymer
matrix techniques to study Rh-catalyzed hydroformylation of 1-octene, 1-butene, propene,
and ethene, using either Rh(acac)(CO)2 or Rh(acac)(CO)(PPh3) in a polyethylene matrix. The
technique requires only micrograms of catalyst. Using Rh(acac)(CO)2 as the catalyst
precursor, we were able to characterize the previously observed acyl rhodium tetracarbonyl
intermediates, RC(O)Rh(CO)4, using IR spectroscopy. For the PPh3-modified reactions, the
analogous acyl rhodium tricarbonyl triphenylphosphine intermediates, RC(O)Rh(CO)3(PPh3),
have been observed. The IR spectra suggest a trigonal-bipyramidal structure with the acyl
ligand and the PPh3 ligand occupying the axial positions. The acyl rhodium tetracarbonyl
intermediates, RC(O)Rh(CO)4, can easily react with ethene to form acyl rhodium tricarbonyl
ethene species RC(O)Rh(CO)3(C2H4), again identified via IR. The resulting hydroformylation
products were extracted from the polyethylene film after the reaction and analyzed by GC.
The results indicate that, in the PE matrix, the PPh3-modified hydroformylation gives a higher
linear-to-branched ratio for the unsymmetrical alkenes.
Sch em e 1. Ca ta lytic Cycle for Rh -Ca ta lyzed
Hyd r ofor m yla tion
In tr od u ction
Rhodium-catalyzed hydroformylation has received a
great deal of attention from both academia and industry.
Since the early 1960s, this type of reaction has been the
subject of numerous mechanistic studies.1-15 Although
the general mechanistic concepts proposed by Wilkinson
and co-workers4 are still accepted as being essentially
correct, many of the intermediates involved in the
catalytic cycle (see Scheme 1) remain elusive. For
unmodified reactions (L ) CO), the acyl rhodium
intermediates (C) have been observed in more than 20
cases using in situ IR spectroscopic techniques.12 The
observation of the ethyl rhodium intermediate (B) has
* Corresponding authors. Fax: +44 (0)115 951 3058. E-mail:
Martyn.Poliakoff@nottingham.ac.uk; Mike.George@nottingham.ac.uk.
(1) Bath, S. S.; Vaska, L. J . Am. Chem. Soc. 1963, 85, 3500.
(2) Osborn, J . A.; Wilkinson, G.; Young, J . F. Chem. Commun.
(London) 1965, 17.
(3) Vaska, L. J . Am. Chem. Soc. 1966, 88, 4100.
(4) Evans, D.; Osborn, J . A.; Wilkinson, G. J . Chem. Soc. A 1968,
3133.
(5) Yagupsky, G.; Brown, C. K.; Wilkinson, G. J . Chem. Soc. A 1970,
1392.
(6) Unruh, J . D.; Christenson, J . R. J . Mol. Catal. 1982, 14, 19.
(7) Moasser, B.; Gladfelter, W. L.; Roe, D. C. Organometallics 1995,
14, 3832.
(8) Schmid, R.; Herrmann, W. A.; Frenking, G. Organometallics
1997, 16, 701.
(9) van Rooy, A.; Kamer, P. C. J .; van Leeuwen, P. W. N. M. J .
Organomet. Chem. 1997, 535, 201.
(10) Matsubara, T.; Koga, N.; Ding, Y.; Musaev, D. G.; Morokuma,
K. Organometallics 1997, 16, 1065.
(11) Feng, J .; Garland, M. Organometallics 1999, 18, 417.
(12) Liu, G.; Volken, R.; Garland, M. Organometallics 1999, 18, 3429.
(13) Bianchini, C.; Lee, H. M.; Meli, A.; Vizza, F. Organometallics
2000, 19, 849.
also been assigned.14,16 The hydride rhodium intermedi-
ate (A) has only very recently been observed under
hydroformylation conditions;17 however its existence had
(14) Liu, G.; Garland, M. J . Organomet. Chem. 2000, 613, 124.
(15) van der Slot, S. C.; Kamer, P. C. J .; van Leeuwen, P. W. N. M.;
Iggo, J . A.; Heaton, B. T. Organometallics 2001, 20, 430.
(16) King, R. B.; King, A. D., J r.; Iqbal, M. Z. J . Am. Chem. Soc.
1979, 101, 4893.
10.1021/om020854n CCC: $25.00 © 2003 American Chemical Society
Publication on Web 03/12/2003

Rhodium-Catalyzed Hydroformylation of Alkenes
Organometallics, Vol. 22, No. 8, 2003 1613
been demonstrated earlier by employing drastic syn-
thetic conditions using high-pressure IR.18 For triph-
enylphosphine-modified reactions (L ) PPh3), the hy-
dride rhodium intermediates (A) have been characterized
fully using in situ IR19 and NMR20 techniques, and the
acyl rhodium intermediates (C) have been observed in
in situ NMR experiments.20-22
PE film and to surround this film by a high pressure of
hydrogen. The reaction then takes place inside the PE
film, and the various intermediates formed during the
reaction can be characterized using IR spectroscopy. The
technique requires only micrograms of chemicals and
greatly facilitates the investigations of organometallic
reactions involving gaseous reactants, for the gaseous
reactants can be easily removed from the reaction
system, or one type of reactant can be easily replaced
by another. The effects of diffusion of gases into and out
of the PE film have been studied in some detail.39 In
this paper, our aim has been to validate the use of our
cell for studying hydroformylation of alkenes catalyzed
by rhodium catalysts. Our strategy has been (i) to use
this technique to obtain the spectra of acyl rhodium
tetracarbonyl intermediates RC(O)Rh(CO)4 in the un-
modified rhodium-catalyzed hydroformylation of 1-octene
and compare our results with the published data, (ii)
to extend these studies to other alkenes especially
gaseous alkenes, (iii) to search for the corresponding acyl
rhodium tricarbonyl triphenylphosphine intermediates
RC(O)Rh(CO)3(PPh3) when the reactions are carried out
in the presence of PPh3, and (iv) to study the rhodium-
catalyzed hydroformylation of ethene with and without
PPh3. We have successfully observed not only the well-
documented acyl rhodium tetracarbonyl intermediates
in the unmodified reactions but also, for the first time,
the acyl rhodium tricarbonyl triphenylphosphine inter-
mediates in the triphenylphosphine-modified reactions.
We also found that the rhodium carbonyl species at
2089, 2038, and 2017 cm-1 formerly observed in the
unmodified rhodium-catalyzed hydroformylation of
ethene under lower CO pressure and assigned to the
ethyl rhodium tetracarbonyl intermediate C2H5Rh(CO)4
is in fact an acyl rhodium tricarbonyl ethene species
C2H5C(O)Rh(CO)3(C2H4). This species is formed by the
reaction of acyl rhodium tetracarbonyl intermediates
with ethene.
The two most frequently used techniques for mecha-
nistic study of hydroformylation are high-pressure NMR
and high-pressure IR spectroscopy. However, both high-
pressure NMR and high-pressure IR have shortcomings,
and the crucial intermediates may still escape direct
observation.19 In high-pressure NMR experiments,7,23-27
high concentrations are required in order for the signal-
to-noise ratios to be satisfactory. At these high concen-
trations the reactions usually deviate considerably from
catalytic experiments. High-pressure IR techniques23,28-33
do not suffer from this limitation. However, while
reflectance high-pressure IR spectroscopy is a real in
situ technique, transmission high-pressure IR spectros-
copy has the drawback that the cell must be connected
to the autoclave via a suitable delivery system. Trans-
portation from the autoclave to the cell to obtain the
spectra can mean that the results are not necessarily
representative of the contents of the autoclave. Fur-
thermore, once an autoclave is involved in the experi-
ment, a considerable quantity of catalyst is always
required no matter which method is used.
For several decades, matrix isolation has been one of
the major techniques for characterizing intermediates
in organometallic reactions.34 Quite recently, we devel-
oped a miniature high-pressure low-temperature copper
cell35,36 which allowed us to use polyethylene (PE)
matrixes in combination with high-pressure gases. In
this way, we have studied intermediates in homoge-
neously catalyzed hydrogenation.37,38 Our strategy has
been to impregnate a substrate and a catalyst into a
(17) Li, C.; Widjaja, E.; Chew, W.; Garland, M. Angew. Chem., Int.
Ed. 2002, 41, 3786.
(18) Vidal, J . L.; Walker, W. E. Inorg. Chem. 1981, 20, 249.
(19) Dieguez, M.; Claver, C.; Masdeu-Bulto, A. M.; Ruiz, A.; van
Leeuwen, P. W. N. M.; Schoemaker, G. C. Organometallics 1999, 18,
2107.
(20) Brown, J . M.; Kent, A. G. J . Chem. Soc., Perkin Trans. 2 1987,
1597.
(21) Yagupsky, M.; Brown, C. K.; Yagupsky, G.; Wilkinson, G. J .
Chem. Soc. A 1970, 937.
(22) Brown, J . M.; Kent, A. G. J . Chem. Soc., Chem. Commun. 1982,
723.
(23) Moasser, B.; Gladfelter, W. L. Inorg. Chim. Acta 1996, 242, 125.
(24) Kegl, T.; Kollar, L.; Radics, L. Inorg. Chim. Acta 1997, 265,
249.
(25) Castellanos-Paez, A.; Castillon, S.; Claver, C.; van Leeuwen,
P. W. N. M.; de Lange, W. G. J . Organometallics 1998, 17, 2543.
(26) Elsevier, C. J . J . Mol. Catal. 1994, 92, 285.
(27) Roe, D. C. J . Magn. Reson. 1985, 63, 388.
(28) Trzeciak, A. M.; Ziolkowski, J . J . J . Mol. Catal. 1986, 34, 213.
(29) Moser, W. R.; Papile, C. J .; Brannon, D. A.; Duwell, R. A.;
Weininger, S. J . J . Mol. Catal. 1987, 41, 271.
(30) Garland, M.; Bor, G. Inorg. Chem. 1989, 28, 410.
(31) J ongsma, T.; Challa, G.; Van Leeuwen, P. W. N. M. J .
Organomet. Chem. 1991, 421, 121.
Exp er im en ta l Section
Gen er a l Str a tegy. Dicarbonylacetylacetonato rhodium(I)
(99%, Strem), ethene (99%, Air Products), propene (99%, BOC
Gases), 1-butene (99%, Aldrich), 1-octene (99%, Aldrich),
triphenylphosphine (99%, Acros), and syngas (1/1, Air Prod-
ucts) were used as received. Hostalen GUR 4150 polyethylene
(HPE) was supplied by Hoechst.40 Our previous investigations
have shown that this form of PE has a very low degree of
unsaturation and have demonstrated that interaction between
olefinic double bonds in the PE and unsaturated photofrag-
ments is negligible.41 The PE film was melting pressed at 155
°C from powders using a constant thickness film-maker
(Specac P/N 15620). The thickness of the PE film was 0.5 mm.
GC analysis was carried out on a Shimadzu GC-17A gas
chromatograph with a RTX-5 column.
P r ep a r a tion of th e Ca ta lyst F ilm . For the unmodified
reactions, the catalyst film was made by impregnating Rh-
(acac)(CO)2 into PE film. The impregnation was achieved by
(32) van Rooy, A.; Kamer, P. C. J .; van Leeuwen, P. W. N. M.;
Goubitz, K.; Fraanje, J .; Veldman, N.; Spek, A. L. Organometallics
1996, 15, 835.
(33) Horvath, I. T. Organometallics 1986, 5, 2333.
(34) See, for example: Almond, M.; Downs, A. J . Adv. Spectrosc.
1989, 17.
(35) Cooper, A. I.; Poliakoff, M. Chem. Phys. Lett. 1993, 212, 611.
(36) Cooper, A. I. Ph.D. Thesis, University of Nottingham, 1994.
(37) Poliakoff, M.; George, M. W. J . Phys. Org. Chem. 1998, 11, 589.
(38) Childs, G. I.; Cooper, A. I.; Nolan, T. F.; Carrott, M. J .; George,
M. W.; Poliakoff, M. J . Am. Chem. Soc. 2001, 123, 6857.
(39) Goff, S. E. J .; Nolan, T. F.; George, M. W.; Poliakoff, M.
Organometallics 1998, 17, 2730.
(40) HPE is made using a Ziegler-Natta catalyst. It has an
extremely high average molecular weight (7.3 × 106), a density of 0.93
g/cm3, 52% crystallinity, a melting temperature of 130-135 °C, and a
remarkable toughness at very low temperatures (Hoechst: Hostalen
GUR-Ultrahigh Molecular Weight Polyethylene (PE-UHMW); Ho-
echst Aktiengesellschaft: Frankfurt, 1993.
(41) Clarke, M. J .; Howdle, S. M.; J obling, M.; Poliakoff, M. J . Am.
Chem. Soc. 1994, 116, 8621.

1614 Organometallics, Vol. 22, No. 8, 2003
Zhang et al.
F igu r e 1. Schematic representation of the high-pressure
Cu cell used in the experiments described here. A gas can
be introduced to (or removed from) the cell at any time.
The gas can freely permeate around the polyethylene
catalyst film (PE) where the reactions take place. The
reactions are monitored through the CaF2 windows (W)
using in situ FTIR.
F igu r e 2. IR spectra of the catalyst precursors in PE
film: (a) Rh(acac)(CO)(PPh3); (b) Rh(acac)(CO)2.
placing a PE film in a cyclohexane solvent saturated with
dicarbonylacetylacetonato rhodium(I) and leaving it in the
solvent for ca. 5 h. For the PPh3-modified reactions, the
catalyst film was made by forming Rh(acac)(CO)(PPh3) com-
pound in situ inside PE film. The formation of Rh(acac)(CO)-
(PPh3) in PE film was achieved by the following steps. First,
a PE film was placed in cyclohexane saturated with tri-
phenylphosphine and left for ca. 4 h. In this way, some
triphenylphosphine was impregnated into the PE film. Second,
the PPh3-impregnated PE film was transferred to a cyclohex-
ane solvent saturated with dicarbonylacetylacetonato rhodium-
(I) and left until Rh(acac)(CO)2 in the film is detectable by
FTIR. In this way, all the impregnated PPh3 in the PE film
was converted into Rh(acac)(CO)(PPh3). Finally, the film was
placed in clean cyclohexane solvent until all of the free Rh-
(acac)(CO)2 was extracted. In this way, all excess Rh(acac)-
(CO)2 was removed, leaving only Rh(acac)(CO)(PPh3) in the
film.
F igu r e 3. In situ IR spectra of unmodified Rh-catalyzed
hydroformylation of 1-octene at 298 K and 1950 psi of
syngas (CO/H2 ) 1/1) using Rh(acac)(CO)2 as the catalyst
precursor. The peaks are labeled as follow: (1) free CO
inside the cell; (2) the acyl rhodium tetracarbonyl inter-
mediate C8H17C(O)Rh(CO)4; (3) Rh(acac)(CO)2; (4) 1-octene;
(5) the hydroformylation products C8H17CHO.
Hyd r ofor m yla tion Exp er im en ts. Hydroformylation reac-
tions were carried out in the high-pressure cell, which has been
described in detail elsewhere (see Figure 1).35,37 Briefly, the
catalyst PE film was clipped into the cell (for hydroformylation
of liquid alkene, the substrate was impregnated into the film
in advance together with the catalyst) and the cell was
evacuated. An appropriate pressure of syngas (CO/H2 ) 1/1)
was then added at room temperature (for hydroformylation
of gaseous alkenes, a mixture of alkene and syngas was used).
The reaction took place inside the PE film at room temperature
and was monitored in situ by IR spectroscopy. After the
required time, the reaction was quenched by venting the gas
mixture, and IR spectroscopy was used to characterize and
monitor the decay of the intermediates. Venting and exchang-
ing gases is very rapid, taking less than 1 min for the entire
procedure.
Ta ble 1. P ea k P osition s (cm -1) of th e IR Sp ectr a of
th e Ca ta lyst P r ecu r sor s
species
ν(CO)
ν(acac)
medium
ref
Rh(acac)(CO)(PPh3) 1980
Rh(acac)(CO)(PPh3) 1982
1568, 1517 Nujol
1569, 1518 PE
42
this work
this work
Rh(acac)(CO)2
2081, 2012 1581, 1525 PE
Un m od ified Rh -Ca ta lyzed Hyd r ofor m yla tion of
1-Octen e, 1-Bu ten e, a n d P r op en e in P E F ilm . The
unmodified rhodium-catalyzed hydroformylation of al-
kenes has been investigated by Garland and co-workers
in detail.11,12,14,30 We investigated this reaction using the
Rh(acac)(CO)2-impregnated PE film as catalyst to both
validate our approach and to capture any possible new
intermediates in the catalytic cycle.
Initially, we used 1-octene as the substrate. Figure 3
shows the in situ IR spectra of this reaction at 298 K
under 1950 psi of syngas (CO/H2 ) 1/1). The absorptions
at 2012, 1581, and 1525 cm-1 (labeled 3) are assigned
to Rh(acac)(CO)2, and the absorptions at 1821 and 1641
cm-1 (labeled 4) are due to 1-octene. The absorption at
1733 cm-1 (labeled 5) indicates the formation of the
hydroformylation product (nonyl aldehyde). The weak
absorption at 2018 cm-1 (labeled 2) is attributed to the
acyl rhodium tetracarbonyl intermediate.12 The CO in
the syngas surrounding the PE film gives a very broad
absorption in the range 2030-2230 cm-1 (labeled 1).
This band obscures most of the absorptions of interest,
but this is not a problem because the intermediates are
trapped in the PE film. We can stop the reaction at
anytime by venting the syngas and then monitor the
Sp ectr oscop y. All IR spectra were recorded on a Nicolet
Avatar 360 FTIR interferometer with a liquid nitrogen cooled
HgCdTe detector interfaced to a PC running OMNIC software.
The resolution was set to 4 cm-1 for all measurements. All
structure assignments are based on IR spectra.
Resu lts a n d Discu ssion
P r ep a r a tion of th e Ca ta lyst F ilm . Rh(acac)(CO)2
was impregnated into PE film and used directly as an
unmodified catalyst. For triphenylphosphine-modified
reaction, Rh(acac)(CO)(PPh3) was used as the catalyst
precursor. Figure 2 shows the IR spectra of Rh(acac)-
(CO)2 and Rh(acac)(CO)(PPh3) in the PE film. The
corresponding peak positions are listed in Table 1. The
presence of Rh(acac)(CO)(PPh3) can be confirmed by
comparison with published results.42
(42) Bonati, F.; Wilkinson, G. J . Chem. Soc. 1964, 3156.

Rhodium-Catalyzed Hydroformylation of Alkenes
Organometallics, Vol. 22, No. 8, 2003 1615
Ta ble 3. Ch em oselectivity of th e
Hyd r ofor m yla tion of Alk en es in P E F ilm
linear-to-branched ratioa
catalyst precursor
propene
1-butene
1-octene
Rh(acac)(CO)2
Rh(acac)(CO)(PPh3)
1.0 ( 0.1
1.7 ( 0.1
1.2 ( 0.1
3.2 ( 0.1
1.1 ( 0.1
2.4 ( 0.1
a
Linear-to-branched ratios are determined by the peak area
in the GC spectra.
F igu r e 4. IR spectra derived from the unmodified Rh-
catalyzed hydroformylation of (a) 1-octene, (b) 1-butene,
and (c) propene at 298 K and 1950 psi of syngas (CO/H2 )
1/1) using Rh(acac)(CO)2 as catalyst precursor. The spectra
were obtained by subtractions of the spectra of the PE film
taken intermittently after the gas mixture has been
removed from the cell. Asterisks (*) indicate bands that
have been incompletely removed by subtraction.
F igu r e 5. Decay of the four bands at 2055 (2), 2000 (b),
1981 (1), and 1681 (9) cm-1 in the PPh3-modified Rh-
catalyzed hydroformylation of 1-butene after the syngas
has been removed from the cell. Note that the absorbance
of the bands has been normalized and offset to emphasize
that all four bands decay at the same rate and are therefore
due to the same species, C4H9C(O)Rh(CO)3(PPh3).
Ta ble 2. P ea k P osition s (cm -1) of th e IR Ba n d s of
th e In ter m ed ia tes RC(O)Rh (CO)4 Der ived fr om th e
Un m od ified Rh -Ca ta lyzed Hyd r ofor m yla tion of
Alk en es
alkene
ν(CO)
ν(CdO)
ref
1-octene
1-octene
1-butene
propene
ethene
2111, 2064, 2038, 2020
2110, 2063, 2037, 2018
2110, 2063, 2037, 2019
2111, 2064, 2037, 2019
2111, 2063, 2038, 2019
1703
1703
1702
1701
1696
12
of rhodium-catalyzed hydroformylation, especially for
the gaseous reactants.
this work
this work
this work
this work
P P h 3-Mod ified Rh -Ca ta lyzed Hyd r ofor m yla tion
of 1-Octen e, 1-Bu ten e, a n d P r op en e in P E F ilm .
Our success in monitoring hydroformylation using the
unmodified catalyst prompted us to investigate the
PPh3-modified reactions by making Rh(acac)(CO)(PPh3)
in the PE film (see above). We then used this catalyst
film to hydroformylate 1-octene, 1-butene, and propene.
In each case, an intermediate was detected with four
IR bands at ca. 2056, 2001, 1981, and 1680 cm-1, which
disappeared after the syngas was vented from the cell.
Figure 5 shows this process for 1-butene. Subtractions
of the spectra of PE catalyst film taken sequentially
during the decay produced good quality IR spectra (see
Figure 6a, b, c). These spectra have patterns very
similar to the spectrum of the known cobalt compound
CH3C(O)Co(CO)3(PPh3) with the PPh3 ligand in the
apical position.43 Thus we tentatively attribute these
spectra to the acyl rhodium tricarbonyl triphenylphos-
phine intermediates RC(O)Rh(CO)3(PPh3) with equa-
tional CO groups (Chart 1). The peak positions are listed
in Table 4.
decay of the intermediates by IR spectroscopy. Subtrac-
tions of these spectra give the bands of the intermedi-
ates (see Figure 4a). The pattern and the peak positions
(see Table 2) of the intermediate spectra are almost
identical to the well-documented acyl rhodium tetra-
carbonyl intermediates C8H17C(O)Rh(CO)4.12
We have extended these studies to the hydroform-
ylation of 1-butene and propene. Subtractions of the
spectra of the PE film taken after the syngas had been
vented from the cell produced IR spectra that have
patterns similar to that of Figure 4a (see Figure 4 and
Table 2). Thus, we have successfully obtained the IR
spectra of C4H9C(O)Rh(CO)4 and C3H7C(O)Rh(CO)4, by
an approach that is simpler and involves less com-
pound than previous methods.
To confirm that the hydroformylation of alkenes does
take place inside the PE film, we carried out solvent
extraction (acetone) of the PE film after the reaction and
analyzed the extracts by GC. The GC analyses indicate
that the hydroformylation reactions were indeed taking
place in the PE film and the linear-to-branched ratios
obtained from these experiments are very reproducible
(see Table 3).
Again, to confirm that hydroformylation was really
taking place inside the PE film, we extracted the
contents of the PE film after the reaction and analyzed
the extracts by GC. As expected, the analysis gave a
higher linear-to-branched ratio than with the unmodi-
fied reaction (see Table 3).
Although we did not detect any new intermediates
in these experiments, the results definitely show the
effectiveness of our approach for the mechanistic studies
(43) Ford, P. C.; Massick, S. Coord. Chem. Rev. 2002, 226, 39.

1616 Organometallics, Vol. 22, No. 8, 2003
Zhang et al.
acyl ligand (linear and branched). However, as in the
case of the acyl rhodium tetracarbonyl intermediate,12
the isomerization of the acyl ligand appears to have
little influence on the IR absorption of other terminally
coordinated CO ligands in the acyl rhodium intermedi-
ate under the experimental conditions.
Rh -Ca ta lyzed Hyd r ofor m yla tion of Eth en e a n d
th e Rea ction of RC(O)Rh (CO)4 In ter m ed ia tes w ith
Eth en e in P E F ilm . The unmodified Rh-catalyzed
hydroformylation of ethene has also received consider-
able attention. In 1979, King and co-workers observed
infrared absorbance at 2115, 2037, and 2019 cm-1
during the hydroformylation of ethene under consider-
able pressure of CO and attributed these new IR bands
to the ethyl rhodium tetracarbonyl intermediate C2H5-
Rh(CO)4.16 Very recently, Garland and co-workers re-
investigated this system in more detail.14 Garland found
that this system is difficult to study using IR spectros-
copy and complex spectra were obtained. At higher CO
partial pressures, the IR spectra correspond exactly to
the well-documented mononuclear acyl rhodium tetra-
carbonyl species (2112, 2068, 2038, and 2020 cm-1). At
lower CO partial pressures, bands at 2089, 2038, and
2017 cm-1 were obtained. These bands were tentatively
attributed to the ethyl rhodium tetracarbonyl interme-
diate C2H5Rh(CO)4. However, it was also noted that
these bands could have been due to the acyl species,
C2H5C(O)Rh(CO)3(C2H4). We have used our technique
in an attempt to distinguish between these possibilities.
First, we investigated hydroformylation of ethene
using Rh(acac)(CO)(PPh3) as the catalyst precursor at
298 K under different pressures of syngas and ethene.
In every case, we observed the same spectrum, at-
tributed to the acyl rhodium tricarbonyl triphenylphos-
phine intermediate, C2H5C(O)Rh(CO)3(PPh3), by com-
parison with the intermediates in other PPh3-modified
hydroformylation. Unlike the higher olefins, the IR
spectrum has a symmetrical absorption around 1680
cm-1 (see Figure 6d and Table 4). This is consistent with
the fact that there is no possibility of acyl ligand
isomerization in the case of C2H5C(O)Rh(CO)3(PPh3).
We have investigated the unmodified hydroformyla-
tion of ethene using Rh(acac)(CO)2 as the catalyst
precursor at 298 K under different pressures of syngas
and ethene. In these reactions, propionaldehyde was
formed (IR and GC), but no ketone or polyketone
products were detectable by GC analysis of the extract
from the PE film. This contrasts with the recent
results44 from Garland and co-workers where ketone/
polyketone were detected in solution reactions involving
ethene. With 50 psi of ethene and 1800 psi of syngas
(CO/H2 ) 1/1), we observed IR bands at ca. 2111, 2088,
2064, 2038, 2017, and 1694 cm-1 (Figure 7a), while 700
psi of ethene and 200 psi of syngas (CO/H2 ) 1/1)
produced IR bands at ca. 2088, 2037, 2015, and 1693
cm-1 (Figure 7b). These results are similar to those of
Garland in the ν(CO) region.14 However, in our experi-
ments, these ν(CO) absorptions appear to be associated
with an acyl carbonyl absorption. After subtracting
spectrum (b) from spectrum (a), bands are found at ca.
F igu r e 6. IR spectra derived from the PPh3-modified Rh-
catalyzed hydroformylation of (a) 1-octene, (b) 1-butene, (c)
propene, and (d) ethene at 298 K and 1950 psi of syngas
(CO/H2 ) 1/1) using Rh(acac)(CO)(PPh3) as catalyst pre-
cursor. The spectra were obtained by subtractions of the
spectra of the PE film recorded while intermediates decay
after the syngas has been removed from the cell.
Ch a r t 1
Ta ble 4. P ea k P osition s (cm -1) of th e IR Ba n d s of
In ter m ed ia tes RC(O)Rh (CO)3(P P h 3) Der ived fr om
th e P P h 3-Mod ified Rh -Ca ta lyzed Hyd r ofor m yla tion
of Alk en es
alkene
ν(CO)
ν(CdO)
1-octene
1-butene
propene
ethene
2056, 2000, 1980
2055, 2000, 1981
2056, 2001, 1981
2056, 2001, 1981
1681
1681
1681
1679
Very recently, van Leeuwen and co-workers detected
the existence of the acyl rhodium diphosphone inter-
mediate RC(O)RhL2(CO)2 (L ) monodentate phosphorus
diamide ligand) using in situ IR and NMR techniques.15
Because several isomers exist for this intermediate,
seven absorptions in the terminal CO region were
observed in IR spectra. In addition, the strong amide
bands of the ligand obscure the expected rhodium-acyl
absorption band around 1600-1700 cm-1. In our case,
because the precursor has only one PPh3 ligand in the
molecule, it is likely that the acyl rhodium carbonyl
intermediates also contain only one PPh3. The relative
intensities of IR bands suggest that the molecule adopts
a configuration with the two bulky groups, acyl ligand
and triphenylphosphine ligand, occupying the two axial
positions with comparatively simple IR spectra in the
terminal CO region. The absorption at ca. 1680 cm-1 is
clearly unsymmetrical in the cases of 1-octene, 1-butene,
and propene. This may be due to the two isomers of the
(44) Li, C.; Guo, L.; Garland, M. The Homogeneous Hydroformyla-
tion of Ethylene Catalysed by Rh4(CO)12. The Application of BTEM to
Identify a New Rhodium Carbonyl Spectrum. The 13th International
Symposium on Homogeneous Catalysis, Tarragona Spain, Sept 3-7,
2002.

Rhodium-Catalyzed Hydroformylation of Alkenes
Organometallics, Vol. 22, No. 8, 2003 1617
F igu r e 8. IR spectra from an experiment where an acyl
rhodium tetracarbonyl intermediate reacts with ethene: (a)
spectrum of the species C2H5C(O)Rh(CO)3(C2H4), repeated
from Figure 7b; (b) obtained by subtraction of the spectra
of the PE film taken after ethene has been removed from
the cell containing C8H17C(O)Rh(CO)4 and ethene at 263
K. Asterisks (*) indicate bands that have been incompletely
removed by subtraction. Notice: The patterns of the bands
in the two spectra are very similar.
F igu r e 7. IR spectra derived from the unmodified Rh-
catalyzed hydroformylation of ethene using Rh(acac)(CO)2
as the catalyst precursor at 298 K and (a) 50 psi of ethene
and 1800 psi of syngas (CO/H2 ) 1/1), (b) 700 psi of ethene
and 200 psi of syngas (CO/H2 ) 1/1). Spectra (a) and (b)
were obtained by subtractions of the spectra of the PE film
taken after the gas mixture was removed from the cell.
Spectra (a - b) were obtained by scaled subtraction of (a)
and (b). Asterisks (*) indicate bands that have been
incompletely removed by subtraction.
Ta ble 5. P ea k P osition s (cm -1) of th e IR Ba n d s of
th e Sp ecies RC(O)Rh (CO)3(C2H4)
species
ν(CO)
ν(CdO)
C2H5C(O)Rh(CO)3(C2H4)
C8H17C(O)Rh(CO)3(C2H4)
2088, 2037, 2015
2087, 2038, 2014
1693
1696
tricarbonyl ethene species C8H17C(O)Rh(CO)3(C2H4).
The absorption at 1996 cm-1 is broader and less sym-
metrical than the corresponding absorption (at 1993
cm-1) in Figure 8a. This is consistent with isomerization
of the nonyl ligand (linear and branched). By contrast
the C2H5C(O) ligand cannot isomerize easily.
Sch em e 2. In ter a ction of RC(O)Rh (CO)4 w ith
Eth en e
Our results suggest that the acyl rhodium tetracar-
bonyl intermediates RC(O)Rh(CO)4 can easily react with
ethene to form acyl rhodium tricarbonyl ethene species
RC(O)Rh(CO)3(C2H4). We believe that reactions of this
type would be difficult if not impossible to carry out in
more conventional experiments with spectroscopic au-
toclaves. The peak positions of the two acyl rhodium
tricarbonyl ethene species observed in this work are
listed in Table 5.
2111, 2063, 2038, 2019, and 1696 cm-1 (Figure
7(a-b)). By comparison with the higher olefins, these
bands can be tentatively assigned to the acyl rhodium
tetracarbonyl intermediate C2H5C(O)Rh(CO)4. Under
higher partial pressure of ethene, C2H5C(O)Rh(CO)4 was
not observed, but instead there is a second intermediate,
spectrum (b). We believe this intermediate to be an acyl
rhodium tricarbonyl ethene species, C2H5C(O)Rh(CO)3-
(C2H4), formed by the reaction of C2H5C(O)Rh(CO)4 with
ethene (Scheme 2). A spectrum very similar to 7b has
just been reported by Garland and co-workers.44
To support this suggestion, we investigated the reac-
tion of another acyl rhodium tetracarbonyl intermediate,
C8H17C(O)Rh(CO)4, with ethene. Our strategy was to
generate C8H17C(O)Rh(CO)4 by hydroformylation of
1-octene at room temperature (as above), cool the
catalyst film to a lower temperature (263 K), and vent
the syngas from the cell. The cell was then filled with
high-pressure ethene (900 psi), and IR spectroscopy was
used to follow the reaction of C8H17C(O)Rh(CO)4 with
ethene over several minutes. Finally, ethene was vented
from the cell and IR spectra were used to monitor the
decay of the resulting species. Subtraction of these
spectra produced an IR spectrum that has peak position
and band intensities very similar to those assigned to
C2H5C(O)Rh(CO)3(C2H4) (see Figure 8b). Spectrum 8b
can be reasonably attributed to the acyl rhodium
Con clu sion s
The rhodium-catalyzed hydroformylation of alkenes
has been carried out within a polyethylene film, and the
labile acyl rhodium intermediates RC(O)Rh(CO)3L (L
) CO, PPh3) in the catalytic cycle have been observed.
The IR spectra suggest that the intermediates RC(O)-
Rh(CO)3(PPh3) may have a trigonal-bipyramidal struc-
ture, with the acyl ligand and the PPh3 ligand occupying
the axial positions, and normally give IR bands at ca.
2056, 2001, 1981, and 1680 (acyl) cm-1. One CO ligand
in intermediates RC(O)Rh(CO)4 can be easily replaced
by an ethene ligand to form RC(O)Rh(CO)3(C2H4). As
expected, the PPh3-modified Rh-catalyzed hydroform-
ylations in PE film give a higher chemoselectivity for
linear aldehyde compared with the unmodified reac-
tions, and the reproducibility is good. We believe that
our approach is a valuable addition to the existing
techniques for studying such procedures. Although the

1618 Organometallics, Vol. 22, No. 8, 2003
Zhang et al.
conditions in our reactions are considerably different
from those of industrial hydroformylation reactions,
hydroformylation does occur and we can detect and
quantify the products from that reaction. It is precisely
because of these differences in conditions that we are
able to detect species that would be too short-lived or
reactive to be observed easily under normal conditions.
Ack n ow led gm en t. We thank A. Cowon, Dr. X.-Z.
Sun, and Dr. J . Hyde for their collaboration in some
experiments, Mr. M. Guyler for his technical support,
and the EPSRC for financial support (Grant No. GR/
N38312). We thank Professor M. Garland for helpful
comments.
OM020854N