THE HYDROFORMYLATION OF OLEFINS USING
SUPPORTED FILM CATALYSTS
MICHAEL J. NAUGHTON
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
For my family because their love and support are never ending.
There are numerous people to whom I would like to express my gratitude.
It is through their support and encouragement that I was able to finish this phase of
my career. First I would like to thank my research director Dr. Russell S. Drago.
He has a unique relationship with his students which allows him to be both a mentor
and friend. I would also like to thank his wife Ruth for her hospitality. It is very
comforting to arrive in a new town and be greeted by people such as the Dragos who
make you feel that you are part of their family.
My family has been an important influence through out my graduate career.
I would like to thank them for their support throughout this undertaking.
I would also like to thank the members of my graduate committee, Drs. David
Richardson, Jim Boncella, Eric Enholm and David Silverman for the help they have
offered throughout my career. In addition, thanks go to Dr. Jerry Unruh (Hoechst
Celanese) and Dr. Tambra Dunams (Tennessee Valley Authority) for many
informative and enlightening discussions.
While attending Plattsburgh State, during the early portion of my academic
career, I was unsure what field to pursue. If it wasn't for the guidance and
encouragement of Dr. Rudolph J. Bobka I may not have found what wonders
chemistry has to offer. In addition, I would like to thank my graduate advisor at
Plattsburgh, Dr. Edward J. Miller, who encouraged me to strive for more, never give
up the pursuit, and to "hang in there."
Of course no one could do work without the help of his peers. The Drago
group, past and present, has always been there to help and harass me when I needed
it. I would particularly like to acknowledge the guys that joined the group in the
same year as I, Don Ferris, Steve Petrosius and Steve Showalter. Although we had
some lean and rugged times, we all weathered the storm and made it relatively
unscathed. I would also like to recognize some specific members of the group for
their advice and encouragement: Chris Chronister, Jerry Grunewald, Bobby Taylor,
Alan Goldstein, Tom Cundari, Robert Beer, Doug Patton, John Hage, Todd LaFrenz,
Mike Robbins, Phil Kaufman, David Singh, Karen Frank and Rich Reily.
I have made many friends while attending UF. It is impossible to list them all,
but there are a few that need to be singled out. Matt Ryan, long time roommate and
friend (even though the IGNATS won), was always a source of consistency in a world
of changing scenery at both home and work. John New, Linda Casazza and Don
Berhinger were my roommates. Some of the things that I'll never forget are Ben and
Jerry's Yogurt, John Madden Football, and the seemingly endless supply of chickens.
I would also like to thank the guys of the now defunct Idylwilde Basketball
Association (IBA) for many stress releasing, fat burning mornings and afternoons.
During my time at UF I have been blessed with a special person in my life,
Donna Robie. She has been the source of love and support that has carried me
through many difficult times. We have had many memorable times and I shall
cherish them forever.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ..................................... iii
ABSTRA CT ................................................ vii
INTRODUCTION .................................. .......... 1
BACKGROUND AND MECHANISMS OF THE RHODIUM CATALYZED
HYDROFORMYLATION OF OLEFINS .......................... 3
Introduction ............................................. 3
Background ........................................ 3
Phosphine Modified Rhodium Catalysts ................... 8
HYDROFORMYLATION OF GASEOUS SUBSTRATES
Supported Liquid Phase Catalysts (SLPC) .....
Other Heterogeneous Catalysts .............
Experim ental ................................
Chem icals .............................
Synthesis of Reagents ................
Preparation of Supported Film Catalysts ......
Results and Discussion ............. ..........
Hydroformylation in Thin Films .............
Polymers and Triphenylphosphine/Polymer Films
Alternative Films ........................
Other Vapor Phase Substrates ..............
Hydroformylation of Propylene with Supported
HYDROFORMYLATION OF HIGHER OLEFINS
HYDROPHILIC FILM CATALYSTS .........
Supported Transition Metal Catalysts
Supported Aqueous Phase Catalysts .
Chem icals ....................
Apparatus and Instrumentation ...
Catalytic Run .................
Results and Discussion ................
Statement of the Problem ........
Choice of Catalyst, Ligand and Film .
Hydroformylation of Olefins with
. . .
. . .
.. .... ....
. . .
.. ... .... .
.. ... ... ..
. ... .... ..
. ... ... ...
. . .
. .. .
BIOGRAPHICAL SKETCH .......
........... .. ... ... ... .. 149
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
THE HYDROFORMYLATION OF OLEFINS USING
SUPPORTED FILM CATALYSTS
Michael J. Naughton
Chairperson: Russell S. Drago
Major Department: Chemistry
The conversion of homogeneous to heterogenous catalytic systems is an
intriguing and a formidable endeavor. Novel approaches to the hybridization of
heterogenous and homogeneous catalysts are described and their effectiveness is
illustrated with the hydroformylation of gaseous and liquid olefins to their
corresponding aldehydes. Supported film catalysts (SFC) utilize a non-volatile film
consisting of a soluble catalytically active complex supported on silica gel. The use
of polymers and other non-volatile films as reaction mediums for SFC's has not been
well established. We have chosen rhodium (I) aryl phosphine complexes in the
presence of excess phosphine for our catalyst due their acute sensitivity to the
reaction environment towards hydroformylation. Supported film catalysts have been
used to hydroformylate propylene, 1-butene and 1,3 butadiene in a gas flow reactor.
The hydroformylation of propylene shows selectivities greater than 99% to n-
butyraldehyde under mild conditions.
Extending SFC's to liquid substrates, such as 1-hexene, presents the added
problem of leaching the active catalyst and/or film phase into the solvent during
catalysis. We have developed a system which uses hydrophilic polymers (ie.
polyethylene glycol) and catalysts supported on silica gel to inhibit leaching during the
reaction. The rhodium complex HRh(CO)(TPPTS), (TPPTS = triphenylphosphine
trisulfonate sodium salt) has been selected as the catalyst due to its hydrophilic
nature. Comparisons of the SFC's to supported aqueous phase catalysts (SAPC) and
to biphasic aqueous systems are made. Unusually high activities and moderate
selectivities are observed at temperatures ranging between 75C and 125C and
pressures of 50 psig syn gas.
The focus of this dissertation is the development of a new class of supported
liquid phase catalysts (SLPC). Suppored liquid phase catalysts are heterogeneous
catalysts in which a liquid film containing a soluble catalyst coats a support (generally
inorganic oxides). To date the majority of the liquid phases have consisted of high
boiling liquids. These liquids have some volatility and may evaporate from the
support with time. The catalysts presented herein utilize non-volatile films such as
polymers and molten salts as replacements for the high boiling liquids and will be
given the general term supported Film catalysts (SFC).
The hydroformylation of olefins with rhodium catalysts has been chosen as the
test system for the new catalysts. There are several reasons for this selection. 1.
Rhodium catalysts are the most active catalyst for hydroformylation. This high
activity makes for easy analysis and evaluation of the catalysts. 2. The environment
of the rhodium catalyst influences both the activity and selectivity of the catalyst. 3.
Catalytic conversion of olefins to aldehydes is a vital chemical industry. It is the
largest commercial homogeneous catalytic process.
This study is divided into two areas. The first concerns the hydroformylation
of gaseous substrates such as propylene. The second studies the hydroformylation
of liquid substrates (e.g. 1-hexene) with a subclass of SFC's, supported hydrophilic
film catalysts (SHFC).
BACKGROUND AND MECHANISMS OF THE RHODIUM CATALYZED
HYDROFORMYLATION OF OLEFINS
Hydroformylation or the "oxo" process is the reaction of hydrogen and carbon
monoxide with olefins in the presence of a catalyst to form aldehydes. Although the
reaction is energetically favorable (-28 Kcal/mol) it does not occur to a measurable
extent in the absence of a catalyst.' Generally, with low molecular weight olefins two
major products are formed, the linear and branched aldehydes; economically the
linear aldehyde is preferred. The reaction is depicted in Figure 2-1. Side reactions
such as isomerization and hydrogenation also occur. These will be discussed in
greater detail later.
Hydroformylation was discovered by Otto Roelen in 1938 in Germany while
studying the activity of ethylene under Fischer-Tropsch reaction conditions." Since
that time, this process has found its place in the petrochemicals industry.
Approximately 10 billion pounds of products are made annually via the
hydroformylation process, making it the largest homogeneous catalytic process.45
The largest classification of substrates for this process are gaseous olefins (C,-C4).
Table 2-1 shows some commercial catalytic systems.6
RCH=CH2 N +
Figure 2-1. Hydroformylation reaction of olefis to aldehydes.
o 0o \C O
00 00 0i O\
u u 8
It should be noted that aldehydes themselves are not very useful but are
precursors to other more economically attractive chemicals. Aldehydes can be
hydrogenated to alcohols or oxidized to carboxylic acids. These in turn can be further
processed to plasticizers, metal working fluids, alkyd resins, specialty greases and
RCCHZCH2CHCHO -- RCCH CHOCHOH
RCH2CH'CHCHO N RCH2CH2CH2COOH
Several transition metal carbonyls are known hydroformylation catalysts. In
order of decreasing activities they are Rh, Co, Ru, Mn, Fe, Cr, Mo, W, Ni, Pt and
Cu.8 Rhodium and cobalt carbonyls are the most commonly studied and are the only
catalysts active enough for use as commercial catalysts. Rhodium catalysts are
approximately 3500 times more active than cobalt. In addition, the rhodium catalysts
modified with phosphine ligands are more selective than cobalt catalysts. The
principal disadvantage of rhodium catalysts is cost, currently RhCI,- 3H0O is $145/gr.9
Table 2-2 lists several substrates that can be hydroformylated with rhodium
complexes and their relative rates of reaction. Linear olefins such as 1-hexene and
1-octene are more readily hydroformylated than the branched or internal alkenes,
such as 2-methyl 1-pentene and 2-hexene.1
Hydroformylation Rates of Various Olefins Using
HRh(CO)(PPh,), as the Catalyst at 250C.
Phosphine Modified Rhodium Catalysts
Rhodium is by far the most active metal for the hydroformylation of olefins.
Reaction rates and product distributions depend on a number of reaction variables.
These include the substrate and ligand type, the concentrations of the reactants
(hydrogen, carbon monoxide and olefin), the concentration of the catalyst and ligand,
and the reaction temperature. The difficulties with these catalysts are the undesirable
isomerization of olefins and the instability of the catalysts. The addition of phosphine
ligands to these systems not only improves the selectivity and activity of the catalysts
but also increases catalyst lifetimes and reduces the amount of isomerization in the
Effects of equilibrium positions and mechanisms
It has been shown that in the presence of hydrogen, carbon monoxide and
excess phosphine, most any rhodium(I) complex is converted to the form HRh(CO),Py
(P=phosphine).'1 Depending on the concentration of the reactants, several
carbonyl/phosphine species exist in solution (see Figure 2-2).14.15,16 These
intermediates are thought to be the active catalytic species for hydroformylation.
Both the selectivity and rate of reaction are influenced by the position of the
equilibria that are established with reactant variation.
HRh(C O)(PP h3)3
Figure 2-2. Equilibrium species of hydrido rhodium carbonyl phosphine
The predominant species for hydroformylation under standard conditions
(HJCO of 1:1 and no excess phosphine) is the 18 electron species HRh(CO);(P), (A).
From this intermediate either the associative or dissociative mechanism can be
followed as proposed by Wilkinson (see Figure 2-3)..11'23.l117.l19,M21L2223
The associative route starts when HRh(CO)(PR,)(L)2 (L = PR3 or CO)
dissociates a phosphine to form the 16 electron species HRh(L), (1). This complex
then can add CO rapidly to form the active catalytic species, HRh(CO)(L)3 (2), which
corresponds to A B, and C in Figure 2-2. Olefin can then bind to the rhodium
complex to form HRh(CO)(L)3(RCH=CH,) (3) and then rearranges rapidly to the
18 electron species Rh(CO)(L)3(CHCH2R) (4). It is this rearrangement where the
selectivity takes place. Anti-Markovnikov addition leads to the linear alkyl species
and hence a linear product, while Markovnikov to a branched metal alkyl which
produces branched aldehydes (see Figure 2-4). Bulky phosphines promote anti-
Markovnikov addition. From here CO insertion/alkyl migration yields the acyl
complex 5. Formation of linear products is favored by the associative route. This is
attributed to the more sterically hindered olefin coordination.
Dissociative mechanism occurs when another phosphine dissociates from 2 to
give HRh(CO)(L)2 (2) which then binds the olefin to form 8 and rearrangement
occurs to form 9. Due to the loss of phosphine there is a lack of steric influence and
Markovnikov addition occurs more readily than in the associative path, hence
favoring branched products. Addition of the phosphine and CO insertion/alkyl
migration gives 4, the common intermediate in both pathways.
L i I
L= PR3 or CO
Figure 2-3. Mechanisms for the hydroformylation of olefins.
L.,Rh .. R
L' I 'CO
S-* Linear Aldehydes
- Branched Aldehydes
L=PR3 or CO
Figure 2-4. Anti-Markovnikov and Markovnikov addition
to the rhodium catalyst.
In both mechanisms oxidative addition of H2 to the acyl complex gives the six
coordinate intermediate 6. This is the rate determining step. Alternatively CO could
add to the acyl complex to give Rh(L)3(CO)-acyl which is catalytically inactive.
Reductive elimination then yields an aldehyde and the square planar complex
Phosphine concentration effects the rate of reaction as well as the n/b ratio.
Figure 2-5 shows the effect of the change in phosphine using a 1-hexene feed. As the
concentration of phosphine increases the rate initially increases until the phosphine
to rhodium ratio (P/Rh) is in the 20:1 50:1 and then decreases. The n/b ratio
increases with the P/Rh ratio.7 If the concentration of phosphine is high, the resultant
rhodium catalyst has a high degree of substitution by phosphines and the active
catalyst exists as the 18 electron species HRh(CO)(P), (B), refer back to Figure 2-2.
This follows the associative pathway which allows a high selectivity towards linear
aldehydes but slow rates of reaction. In the case of low phosphine concentrations a
phosphine dissociates from A to form the 16 electron species HRh(CO),(P) (D).
Hence the dissociative route predominates. The result is an increase in the reaction
rate but a loss in selectivity."
At low carbon monoxide concentrations A can dissociate a carbonyl to form
the 16 electron complex, HRh(CO)(P), (E). This intermediate allows for high rates
of reactions but also the less selective dissociative mechanism to predominate. At
high carbon monoxide concentrations the reaction is inhibited.
0.4 0.8 1.2 1.6
PPh3 Concentration, M
Effects of phosphine concentration on the activity
and product distribution of rhodium catalyzed
hydroformylation of 1-hexene.
The changes in reaction rates and product distribution can be explained by the
steric influence of the ligands. Phosphine ligands are more sterically hindering than
carbon monoxide. Upon coordination of a phosphine ligand there is relatively less
area about the metal for olefin coordination. This causes the olefin to favor anti-
Markovnikov addition to the metal, resulting in a higher selectivity towards linear
An undesirable side reaction of hydroformylation is the isomerization of
a-olefins to internal olefins (see Figure 2-6). The undesirability is derived from the
fact that internal olefins are much less reactive than their alpha counterparts (see
Table 2-2), and the resultant aldehydes usually are of no economic importance.
Isomerization is thought to come about by olefin coordination to HRh(CO)(L), (1)
prior to CO addition (see Figure 2-3). Isomerization can be repressed by the
addition of excess phosphine ligand.2z4'
In addition to the degree of substitution, the size and basicity of the ligand
effects catalyst performance (see Table 2-3). Bulkier ligands yield a more selective
catalyst, and as the basicity of the ligand decreases the activity and selectivity
increases." In the same vein less bulky phosphines are more susceptible to
substitution by CO resulting in decreased in selectivity. As L is substituted by CO
poor selectivity is realized.
In addition to ligand effects many variables control the activity and product
distribution of hydroformylation catalysts. These include the concentration of the
substrate, hydrogen, carbon monoxide, rhodium, and ligand as well as temperature,
L ., I
Figure 2-6. Mechanism for the isomerization of olefins.
Dependence of Product Distribution
Hydroformylation of 1-Octene.
on Phosphine Ligands in the
Ligand Percent Reaction Cone Angle v(cm )b
Linear Time (min) (0)
P(OCHp-C1), 93 55 128 33.2
P(OPh), 86 50 128 29.2
P(OC.H44pPh), 85 70 128 28.9
P(OC.HpOMe), 83 270 128 28.0
PPh, 82 35 145 12.8
P(OBu), 81 60 107 20.2
P(OC,H,Ph), 78 52 141 28.0
PBu, 71 225 132 4.2
P(OC,H,o-Ph), 52 95 132 28.9
P(O-2,6-Me,CH,), 47 80 190 27.1
" 112 g 1-octene. 10 g 5% Rh/C, and 0.05 mol L
1:1 HJCO at 90"C. Adapted from reference 1.
b v = m-2056.1 cm'.
(L:Rh = 10:1) with 80-100 psig of
pressure and solvent polarity. Table 2-4 shows intrinsically how the increase of a
variable influences the reaction rate and product distribution of rhodium catalysts.
Solvent polarity and basicity contribute to the reaction selectivity. As a general rule,
selectivity increases as solvent polarity and basicity increase.1?
Table 2-4. Effects of Reaction Variables on the Rate
Rhodium Hvdroformylation Catalysts.
and Selectivities of
Variables Rate n/b Ratio Isomerization
[H,] + + +
[olefin] + 0
Increasing + + +
[Rh] + +
[L] + -
Based on the hydroformylation of 1-hexene in benzene using
HRh(CO)(TPP) as the rhodium source. + indicates an increase and a
decrease. Abbreviations: [H], [CO] olefinn], [Rh] and [L] are the
concentration of hydrogen, carbon monoxide, olefin rhodium, and ligand.
respectively. Adapted from reference 1.
HYDROFORMYLATION OF GASEOUS SUBSTRATES
Homogeneous hydroformylation of olefins has matured into a finely tuned art.
Rhodium catalysts are optimal catalysts for hydroformylation yielding systems which
are unusually active and showing selectivities as high as 10:1 to the linear aldehyde.
However, the drawbacks of these systems, namely the recovery of valuable rhodium
catalysts and the convenient separation of products from the reaction solution, merit
Heterogeneous catalysts have been designed to circumvent the shortcomings
of homogenous systems. Several methods have been used to immobilize homogenous
catalysts. These include 1. chemisorption of the metal complex on a solid support;
2. physical adsorption of a catalytic complex; 3. entrapment of complexes in zeolites;
4. coordination of the metal complex to functionalized polymers or inorganic oxides;
and 5. supported liquid phase catalysts. These techniques have yet to produce a
commercially viable catalyst for hydroformylation. The lack of success can be
attributed to leaching of the metal from the support to the surrounding environment
and to the loss in activity and selectivity associated with immobilization of
homogenous catalysts. This chapter will concentrate on heterogeneous catalysts as
they pertain to the hydroformylation of gaseous substrates, such as propylene and
Supported Liquid Phase Catalysts (SLPC)
Of all the types of heterogenous catalysts, supported liquid phase catalysts
(SLPC) seem to have the most promise in the commercialization of a catalytic
process. SLPCs are true hybrids of homogeneous and heterogeneous catalysts in that
a support is coated with a catalytic film that consists of a homogeneous catalyst and
a high boiling liquid.''7 Several applications of supported liquid phase catalysts
exist, these include the oxidation of SO,,' dimerization of alkenes,2 and Wacker
oxidation.3 However, the largest application of SLPC has come in the field of
A multitude of catalytic supports are known. Some supports (e.g. porous glass
and macroreticular polymer resins) are inert, while others (e.g. silica and alumina)
possess acid/base properties.1" These sites may synergistically enhance the activity
of the catalyst or may act to anchor the catalyst to the support.2 The chemical
alteration of organic polymers and inorganic oxides creates ligands to coordinate
active metal species but they can also change the activity of a catalyst.
The interactions of the metal complexes and/or additives with the support are
important." If the support deactivates a complex or competes for the ligands
rendering them unavailable for coordination, then that material may be a poor
support for SLPC. As previously stated, triphenylphosphine is an important additive
for most rhodium based hydroformylation catalysts. Triphenylphosphine can
chemisorb on alumina (see Figure 3-1). In addition, triphenylphosphine can also
hydrogen bond to the inorganic oxide surface (see Figure 3-2). These phosphines are
bound tightly and are not available for coordination to the rhodium complexes.
Rhodium complexes are known to chemisorb to the surface of alumina (see
Figure 3-3). In the presence of excess phosphine the chemisorption of the rhodium
SLPC for the hydrformylation of gaseous substrates
Supported liquid phase catalysts are active for the hydroformylation of gaseous
alkene substrates and realize long term stability in gas flow reactors.'" Generally,
conditions for reaction range from 80-1250C and 250-1000 psig. Table 3-1 shows the
activity and selectivity of some selected catalysts. Activities varied with the change
in support, liquid phase and transition metal complex. Most systems studied used
phosphine modified rhodium catalysts and high boiling liquids as the catalytic liquid
phase. Catalysts which possess a liquid phase have been shown to be more active
than their surface bound counterparts consisting of the same rhodium sources and
supports.3 It has been observed that in these systems the reaction rate is
A crucial requirement for an efficient catalyst in a gas flow reactor is that
reactants and products must be gases at reaction temperatures. This minimizes
condensation of the substrates and/or products in the pores of the support.
a OH 0
PPh3 OH 0
Figure 3-1. Adsorption of triphenylphosphine on alumina.
O anion vacancy
Figure 3-2. Adsorption of triphenylphosphine on silica.
Figure 3-3. Adsorption of rhodium phosphine complex on an oxide
surface site. L= triphenylphosphine
SIn v0 in ir, Ir,
\O C -I ell e
c e rsM r
g s -
vi/ vi" v/i T -T -T -
N r -
- (-1 3N '- C]
*^- u- i r-~
2* ^ tr N:
ur-> >< c!-r' r^r"
'^ ^ >D m 3C -
r^ .r, s ~
3~~ ~~ 3 r <
ZN ZN ZN ZN ZN -
O 0 C..
_ o a- a-
U U '
g- ^ S =-
.c .c c e c
0 0 0 e~ 0 r
C C u V Cu
e~ c c ^ ~
.E E .E -
C U U U Cu u '9
A catalyst with a liquid phase of butylbenzylphthalate and (PPh,),Rh(CO)Cl
and excess phosphine on an alumina support exhibits an optimal activity at a liquid
loading of 0.5 cm3 per gram of catalyst.
Aldehydes, the products of hydroformylation are soluble in many of the liquid
phases. Three possible scenarios may develop: 1. As the reaction proceeds, the
amount of alkanals condensed in the pores fill the pores so that the degree of pore
filling, 6, is greater than one, this leads to leaching of the transition metal catalyst.'
2. The amount of rhodium complex adsorbed to the surface is influenced. 3. The
amount of alkanals in the liquid phase influences the equilibria between various
dissolved rhodium complexes.
A study of SLPCs which utilize molten triphenylphosphine as the liquid phase
showed that condensed aldehydes are discernibly soluble in the liquid phase of the
catalyst, but the solubility of hydrogen and carbon monoxide in this medium is not
appreciably effected. However, the solubility of the alkene is strongly influenced.3'
A SLPC in which aldehyde was added to the film led to a lowering of the linear to
branched ratio from 20 to 15 but an increase in activity by 10%."7
Degree of pore filling and its relation to activity and selectivity
The degree of pore filling, 6, greatly effects the activity of most SLPCs for the
hydroformylation of olefins.3"9 In an ideal system, the liquid phase covers the
entire surface uniformly then preferentially fills the smallest pores (see Figure 3-4).0
Due to the complexity and non-uniform nature of the pore structure of many
Diagram of an ideal liquid filled macropore. The fraction of total pore
length that is occupied by a liquid plug is denoted by x. Adapted from
supports the description of SLPCs can only be done with statistical parameters. Rony
et al.41 utilized these parameters to describe the optimal loading for an alumina
support with a butylbenzylphthalate liquid phase, they found that the optimal value
of 8 was 0.5 cm3 liquid/cm3 pore (see Figure 3-5). The activity decreases as the pores
become filled. It has been claimed that it is rarely beneficial to pore fill over this
value. This is due to the long liquid diffusion paths as S approaches 1.
Generally for a given system, the activity reaches a maximum and then
decreases with increasing 6. Most of the catalysis occurs at the gas/liquid interface,
and it follows that the increase in activity is due to the increase of the gas/liquid
interface. Low or no pore filling leads to catalysts which bound to the support. On
the other hand, SLPCs with a large 6 contain liquid plugs or pools of solution in the
pores. In these areas the majority of the rhodium catalyst exists in the bulk solution,
which'possesses a high concentration of phosphine and a low concentration of carbon
monoxide and substrate. This gives rise to highly substituted rhodium phosphine
species which favors selectivity to linear products. However, due to the decrease in
the gas/liquid interface, the activity of these systems are greatly reduced.
Other Heterogeneous Catalysts
Hydroformylation in polymer matrices
The deposition of rhodium catalysts in thin polymer films provide marginally
active hydroformylation catalysts for ethylene and propylene. A film of cellulose
acetate and HRh(CO)(PPh,)3 gives an activity of 0.41 TON/min under atmospheric
0. 06 -
0. 00 0.20 .. I I
de l ta
0.60 0.80 1.00
( 1 Lq)
Figure 3-5. Conversion curves for the hydroformylation of propylene as a function
of liquid loading. The liquid phase is butylbenzylphthalate. Adapted
from refence 42.
1 : i ; ; r_ ~ l i ~_ _~_ _
pressures and 80CC.4 Active catalysts for hydroformylation also are formed in
polystyrene and polybutylmethacrylate films.
Silica supports. A strong affinity between silica and HRh(CO)(PPh,), exists
and is evidenced by the decolorization of a HRh(CO)(PPh,), benzene solution in the
presence of silica." Phosphine modified rhodium complexes impregnated on
alumina or carbon have activities >30% and n/b ratios of 2. Good long term stability
over a 300 hour period are seen for these catalysts.
Inorganic oxide supports. Coordinating homogenous catalytic complexes to
functionalized organic polymers or inorganic oxides with ligand derivatives is a
common method of heteroginizing homogenous catalysts. Functionalized silica with
phosphorous, nitrogen and oxygen donor ligands have been prepared.5" In the case
of hydroformylation, rhodium is commonly used as a catalyst. Figure 3-6 shows the
preparation of an immobilized rhodium catalyst on a functionalized silica support.
Functionalized polymers. Polystyrene and divinylbenzene based polymers are
by far the most common choice for functionalized supports." This is because the
aromatic side chains are easily modified to contain phosphine side chains, which are
good ligands for rhodium hydroformylation catalysts (see Figure 3-7). Copolymers
have been designed by cross-linking to create supports which are porous, insoluble
iCH CH ,CHPPh, Rh(acacX)CO)
Preparation of an immobilized catalyst by coordination
to a functionalized silica support.
(PPh ) RhCl
Figure 3-7. Preparation of an immobilized catalyst by coordination to
a functionalized polymer.
The rigidity of the polymer also plays an important role in activity.49 If the
polymer chain has low flexibility the access of reactants to the metal center decreases,
hence a decrease in activity.5'" On the other hand, if the chains are too mobile,
deactivation of the rhodium may occur by the formation of rhodium clusters.""
Due to preferred conformations within the polymer some of the phosphines
may be unavailable for coordination to the metal center. If the number of
coordinated ligands to a rhodium center is considerably lower than in homogeneous
systems depends, it follows that the selectivity for the polymer system is lower.52
Rhodium catalysts have been incorporated into zeolites and used in the vapor
phase hydroformylation of olefins.455 The hydroformylation of substrates shows
activity towards ethylene, propylene and 1-butene at 150C and at a n/b ratio of
2:1.56 These catalyst have demonstrated stability for up to one month.
The following were used without any further purification: RhCl- 3HO (Strem
Chemical Inc.); triphenylphosphine (Aldrich); potassium hydroxide (Fisher); absolute
ethyl alcohol (Florida Distilleries); hexanes (Fisher); chloroform (Fisher); polybutyl-
methacrcrylate (Aldrich); silicon gum rubber (Varian); polybutadiene (Aldrich);
Aliquat 336 (Aldrich); tetrabutylphosphonium chloride (Aldrich); formaldehyde
(Aldrich); propylene carbonate (Fisher); 1-octanol (Fisher); silica gel (Davison, pore
volume 1.1 mL/g); hydrogen (Matheson); carbon monoxide (Matheson); propylene
(Matheson); 1-butene (Matheson); 1,3 butadiene (Matheson).
Synthesis of Reagents
Preparation of Hydridocarbonvltris(triphenvlphosphine)rhodium, HRh(CO)(PPh,,
Synthesis of HRh(CO)(PPh3), was carried out as reported.57 Under an
atmosphere of nitrogen, triphenylphosphine (2.64g, 10 mmol) was added to refluxing
ethyl alcohol (100mL, absolute). To this RhCl,.3H,O (0.26g, 1.0 mmol) in hot ethyl
alcohol (20mL) was added. After a 15 second delay, formaldehyde (lOmL, 37% in
water) followed immediately by potassium hydroxide (0.8g) in hot ethyl alcohol
(20mL) were added. It should be noted that the addition of the reactants in the
indicated order is critical in this synthesis. Upon addition of potassium hydroxide the
solution turns yellow and the formation of a precipitate is evident. The yellow
mixture was allowed to reflux an addition 10 minutes then cool with stirring to room
temperature. The yellow solid was filtered, washed with ethanol (2xlOmL), followed
by H,O (2xl0mL), then again with ethanol (2xlOmL) and finally by hexanes (10ml).
The yellow solid is dried under vacuum at 400C for 18 hours. Infra red bands
vco=1919 (lit. 1918), vRh,.H2037 (lit. 2041).
Preparation of Supported Film Catalysts
The following is a description of the preparation of a typical catalyst.
A solution of chloroform (25mL) and polybutyl-methacrcrylate (0.06g) was
added to a solid mixture containing HRh[P(C6H,)]3(CO) (0.10g), 0.262gm of
triphenylphosphine (0.262g) and silica gel (0.60g). After 2 minutes of stirring the
solvent was removed under reduced pressure at room temperature. The yellow
powder was dried under vacuum overnight with the temperature not to exceed 450C.
A 0.50 g sample of this product was placed in a stainless steel column of a
gas flow reactor (Figure 3-8) which was placed in an preheated constant temperature
oil bath at the specified reaction temperature. A premixed gas of HJCO/propylene,
also at specified pressures and flow rates was fed through the reactor. Analysis was
performed by an on-line gas chromatograph. Table 3-2 is a compilation of the
experiments presented in the Results and Discussion section.
Schematic for a gas flow reactor with an on-line gas chromatograph.
S- r C = = = 1- C l-. = ^ -
^ j1 -~ l i M f
- r -^ -* C4 n k C'A t C4
-88-8--cc -"- -
L A -- c- Q- CL -----------------a- ----L-----
=- =- r-~ 'p = = = = -'"- r- = c- r N- "x -, r
- -- Cl(-
. .x x. ..ss. ..s. .
aqaacico^^0 0^ ^ ^^0;^^00
- 7 --
~- ~-- ---
'Cl C4 C( C1 Cl C Z c '
! ---- ----X- -- --- -
u u u u
.G a. a.C
-t --r =
E E 'i- 5
|= .- .
2 = ^ ^ E
Results and Discussion
The system described herein demonstrates the activity of previously reported
SLPC but uses a non-volatile "liquid" phase such as a polymer or molten salt.
Polymer or molten salt films are non-volatile and should retard the sublimation of the
transition metal catalyst or phosphines in the film.
Several polymers and molten salts were chosen as reaction media for the
supported catalysts, these include polybutylmethacrylate (PBMA), polybutadiene
(PBD), silicone gum rubber (SGR), aliquat 366 and tetrabutylphosphonium bromide.
The transition metal complex, hydridocarbonyltristriphenylphosphine rhodium(I), was
chosen as the catalyst, for its high activity, selectivity, and stability in homogenous
catalytic systems. Silica gel was chosen as the support for its porosity, and high
The activities (a) were calculated in units of TON/min, where TON is the
turnover number defined as the number of moles of products formed (n,) per mole
of catalyst (n,), eq. 3-1. The value of n, can be calculated by multiplying the
a(TON/min)- n (eq. 3-1)
conversion to products, ap, by the moles of substrate, ns, (np and n. are in units of
moles/min). The moles of substrate, n,, can be calculated via the ideal gas law
np-aopn (eq. 3-2)
(equation 3-3.). The pressure of the substrate, P,, is measured at the gas outlet. The
value Vf, is the flow rate, R is the gas law constant and T is the temperature at the
sample loop. It should be noted that the flow rate is incorporated into the activity.
n~ (eq. 3-2)
In hydroformylation the selectivity is the percentage of the products which are
aldehydes. Included in this term is the ratio of normal to branched aldehydes (n/b
ratio). Occasionally the equivalent term linear to branched or 1/b is used. This value
is calculated from the calibration corrected area for the two aldehyde products. In
this work many of the catalysts gave only one isomer product, the linear aldehyde.
Hydroformylation in Thin Films
The question of activity in thin polymeric films was addressed in our labs by
utilization of hydroformylation catalysts immobilized in polymer films.5 In previous
work, the rhodium complexes HRh(CO)(PPh,), and (tfa)Rh(PPh,),
(tfa=trifluroacetylacetonate) were placed in a polymer membrane and tested for the
hydroformylation of propylene at 850C and 15 psig. These systems were moderately
active and the activities and selectivities could be influenced by altering the polymer
media. A polystyrene membrane catalyst yielded a system which produced 1x109 mol
of butyraldehyde in 3 days with a n/b of 0.71. Changing the polymer matrix to the
rubbery polymer polybutylmethacrylate produced a system that increased in activity
to 2.2x10"mol in one day and the selectivity was an amazing 500:1. Figure 3.9 is a
schematic of the reactor used in these experiments.
It was suspected that diffusion of the reactant gases (syn gas and propylene,
1:1:1 gas mixture) through the polymer matrix was the limiting effect in the
membrane reactor. In order to increase the surface area of the catalytic film we
supported these films on inorganic oxides. This forms a thin catalytic film over the
support which should decrease the diffusion time of the reactants through the
polymer. In addition, the resultant catalyst is a dry powder that can be easily handled
in the air for a short period of time.
Polymers and Triphenylphosphine/Polymer Films
A catalyst containing HRh(CO)[P(CH,),],, and polybutylmethacrylate (PBMA)
was dispersed on silica gel. The catalysis was carried out in a gas flow reactor, using
premixed gases of hydrogen, carbon monoxide and propylene. This catalyst was
considerably more active than the unsupported membrane reactor systems. The
supported polymer film catalyst had an activity of 0.016 TON/min (conditions 100C,
20 psig, flow rate 10mL/min, 6 = 0.25) and a n/b of 2.6:1, while the membrane
reactor showed an activity of 4.17x105 TON/min under similar conditions.
Sinlered S.S. Plolt
0-12 cm Ihick
|mlcron port llat
Figure 3-9. Thin film membrane reactor.
A PBMA catalyst, with a degree of pore filling, 6, of 0.59 showed an activity
of 6x10-4 TON/min. The selectivity could not be determined due to the low activity.
This film represents a relatively thick coating on the support surface. The activity is
considerably lower than the other catalyst. Figure 3-10 depicts the activity curves for
these experiments. This effect can be attributed to the rigid nature of PBMA films.
Although the polymer is above its glass transition temperature (Tg=80C) at a
reaction temperature of 100C, the polymer is not liquid, but solid. This restricts free
movement of the catalyst and makes diffusion of the reactant gases rate limiting.
Concentration effects. The effects of rhodium concentration within the
supported film was tested. It was found that increasing the concentration of rhodium
in the film increased the activity and the selectivity of the catalysts despite the
increased degree of pore filling (the effects of pore filling on this system will be
discussed in greater detail vida infra) (see Table 3-3 and Figures 3-11 and 3-12). The
activity increased with increasing rhodium concentration.
It has been long realized that the addition of excess phosphine ligands to the
rhodium catalyst increases the lifetimes and stability of the catalysts. We decided to
add triphenylphosphine to the PBMA film to stabilize the rhodium complex. With
this addition we observed an unexpected increase in activity to 0.0373 TON/min, this
is approximately twice that of the previous catalyst. In addition, the system produced
only n-butyraldehyde, the desired linear product. This increase in activity was
Activity of the PBMA film catalyst for the hydroformylation of
propylene. .-6 = 0.25. + 6 + 0.59.
Catalyst: Rhodium source-HRh(CO)(PPh,),; P/Rh 10:1; Polymer-
PBMA.Conditions: 80C, 15 psig (1:1:1 H,/CO/propylene).
Effects of Change in Concetration of the Rhodium Source.
[Rh]xlO0 Activity (TON/min) n/b
(mmol Rh/mi film)
0.25 .0090 1
1.00 0.023 2.5
2.60 0.058 7.0
Catalyst: HRh(CO)(TPP),; Polymer: PBMA; Additive: None. Conditions: 1000C,
15 psig (1:1:1 H/CO/propylene); Flow rate: l0mL/min.
Effects of changing rhodium concentration on the activity for the
hydroformyaltion of propylene.
Catalyst: Rhodium source-HRh(CO)(PPh,)3; Polymer-PBMA.
Conditions: 80C, 15 psig (1:1:1 H/CO/propylene).
0- I --I
0.OOE+00 1.00E-04 2.00E-04 3.00E-04
Figure 3-12. Effects of changing rhodium concentration on n/b ratio for the
hydroformylation of propylene.
Catalyst: Rhodium source-HRh(CO)(PPh,),; Polymer-PBMA.
Conditions: 80C. 15 psig (1:1:1 HjCO/propylene).
attributed to the loosening of the film by molten triphenylphosphine (melting point
810C). This the film should be more mobile than one consisting of PBMA alone.
Table 3-4 shows the activity and selectivity of several catalysts involving PBMA.
The concentration of rhodium within the film was varied while holding both
6 and the percentage of PBMA relative to triphenylphosphine constant. Increasing
the rhodium concentration from 0.040 gRh/g of film to 0.090 gRh/g of film
considerably increased the activity, while not significantly influencing the selectivity
(see Table 3-5).
Effects of temperature and pressure. As expected, the activity of PBMA/PPh,
systems increase with increasing temperature (see Table 3-6). For systems at 50 psig
of reactant gas (1:1:1 HJCO/propylene) an increase of activity was observed from
0.060 TON/min at 800C to 0.14 TON/min at 1000C. At 20 psig similar effects were
observed (0.0064 TON/min, 80C; 0.024 TON/min, 100C). The energy of activation
for this system was calculated to be 52.1 + 11 kJ/mol at 50 psig and 68.15 + 11
KJ/mol at 20 psig (see Figures 3-13 and 3-14). These values are the same with in the
Similar increases in activity are seen with pressure. At 80C and 20 psig the
activity is 0.0064 TON/min and is totally selective to n-butyraldehyde. Increasing the
pressure to 50 psig increases the activity to 0.055 TON/min, while the selectivity drops
to 24. A further increase to 100C and 100 psig brings the activity to 0.309 TON/min
and a n/b of 11.
Table 3-4. Selected Catalysts for the Hydroformylation of Propylene.
Support Film Activity Selectivity
silica PBMA 1.60x102 2.62
none PBMA 1.1x10-to a
silica PBMA/TPP 3.73x102 b
silica TPP 3.29x10' b
a- Trace amounts of isobutyradehyde were formed.
b- n-Butyradehyde was the only product detected.
Rhodium source: HRh(CO)(TPP),. Conditions:
H,/CO/propylene), Flow rate 10 mL/min.
Effect of Concentration of the Rhodium Source in the
[Rh] 5 Activity' (TON/min)
(g Rh/g film)
0.090 0.64 0.0064
0.040 0.64 0.0023
Both catalyst produced n-butyradehyde exclusively.
Catalyst: HRh(CO)(PPh,),; Polymer: PBMA; P/Rh = 10:1. Conditions: 800C,
20 psig (1:1:0.1 H./CO/propylene), flow rate of 4.8 mL/min.
Effects of Temperature and Pressure on the Hydroformylation of
Propylene using the PBMA/Triphenylphosphine Film Catalyst.
Temperature Pressure Activity n/b Ratio
(C) (psig) (TON/min)
80 20 0.006
80 50 0.055 24
90 20 0.014 55
90 50 0.096 16
100 20 0.024 35
100 50 0.140 15
100 100 0.309 11
115 20 0.009 70
120 20 0.005
a n-Butyraldehyde was the only product detected.
Gas mixture 1/1/0.1 HCO/propylene; Flow rate 4.8 mL/min.
Catalyst: Rhodium source HRh(CO)(P03)3; P/Rh = 10:1;
[Rh] = 0.10 g Rh/g catalyst; 6 = 0.64.
Arrhenius plots for the PBMA catalyst at 20 psig. Temperature range
800C to 1000C.
slope= -8200 1300
Ea= 68 11 klJ/nole
A= 8.58x10^ 7
Arrhenius plots for the PBMA catalyst at 50 psig. Temperature range
800C to 100(C.
slope= -6300 + 1300
Ea= 52 l kJ/niole
A= 2.87x10 ^ 6
Influence of phosphine to rhodium ratio
Dependence of activity and selectivity on the ratio of free phosphine to
rhodium complex was studied. A trend developed that showed selectivity increasing
with increasing P/Rh ratio. This is expected from similar results in homogeneous
systems with these catalysts. The role of excess triphenylphosphine in homogeneous
systems is to enhance the stability and lifetime of the catalyst. With only
triphenylphosphine as the film, the selectivity increases with increasing P/Rh ratio
(Figure 3-15). The addition of PBMA to the film increases the n/b ratio dramatically
at lower P/Rh ratios.
Effects of pore filling
The effect of the degree of pore filling of film was studied. Table 3-7 and
Figure 3-16 shows a drop in catalyst activity with increasing 6, while the n/b ratio
increased. The loss in activity can be easily explained by the decrease in surface area.
Since the film is a solid most of the active species are at the gas/film interface, and
any rhodium complex in the interior of the film would be inactive. Hence a reduction
in surface area equates to fewer active species and a decrease in the overall activity.
Effects of propylene feed
It was noticed that varying the composition of the gas feed greatly effects the
activity of the catalyst (see Table 3-8). In several cases, an increase in propylene feed
increased the overall activity. At a gas mixture of 1/1/4 (HJCO/propylene) the
0 2.5 5 10
w/o polymer polymer
Figure 3-15. Change in selectivity with variing P\Rh ratio.
Catalyst: Rhodium source-HRh(CO)(PPh,),; 6 = 0.64.
Conditions: 80C, 15 psig (1:1:1 Hi/CO/propylene).
Effects of Pore Filling (6) on the
Activity of the PBMA Film
6 Activity n/b Ratio
0.64 0.024 35
0.22 0.022 21
-n-Butyraldehyde was the only product detected.
S-Activity was below the detection limit for isobutyraldehyde.
Catalyst: HRh(CO)(TPP),; Polymer: PBMA:
P/Rh = 10; [Rh] = 0.1 g of Rh per g of catalyst.
Conditions: 100C, 22 psig (1:1:0.1 HCO/propylene),
flow rate of 10mL/min.
0.4 0.6 0.8 1
0.4 0.6 0.8 1
I I I
1.2 1.4 1.6
Effects of filling the pores with catalytic film on the activity of
propylene hydroformylation. Catalyst: Rhodium source -
HRh(CO)(PPh,),; P/Rh=10:1; [Rh] = 0.11; Polymer-PBMA.
Conditions: 800C, 15 psig (1:1:1 HJCO/propylene).
Effect of Pregas Mix/Propylene Concentration on the Activity and
Selectivity of the Hydroformylation of Propylene.
Gas mixture Activity (TON/min) Selectivity
1/1/4 0.450 a
1/1/2 0.057 67
1/1/1 0.033 14
1/1/0.1 0.018 a
a. n-Butyradehyde was the only product detected.
Catalyst: Rhodium source HRh(CO)(PPh,),; Polymer PBMA;
P/Rh = 10:1; 6 = 0.64: [Rh] = 0.11.
Conditions: 100C, 20 psig (l:l:x HJCO/propylene),
flow rate = 4.8 mL.min.
activity was 0.45 TON/min. Mixtures of 1/1/1 gave 0.0325 TON/min. At low
propylene concentration (1/1/0.1) the activity was 0.0175 TON/min. This would
indicate that at low propylene feeds the polymer phosphine films are not being
saturated with enough propylene to get optimal conversion.
Selectivity is also affected by the amount of propylene in the gas mixture. It
was evident that the higher the concentration of propylene in the feed the higher the
Effects of flow rates
Since the concentration of propylene plays an important role in the activity of
a catalyst, the flow rate of the reactant gases should also effect the rates. At a flow
rate of 10 mL/min (800C, 50psig) the activity was 0.257 TON/min, while the selectivity
was 2.2. Upon reduction of the flow rate to 4.8 mL/min the activity drops to 0.055
TON/min and the selectivity increases to 24:1. It should be noted that calculation of
the activity includes flow rate as one of the terms. Table 3-9 and Figure 3-17 shows
the steady increase of activity with flow rate at 1000C and 20 psig.
Other polymer films
Silicone gum rubber (SGR) and polybutadiene (PBD) were tested as films
for STFC. These rubbery polymers will produce a more mobile film than PBMA.
Under similar conditions, the SGR catalyst was less active than its PBMA counterpart
[at 20 psig and 800C (0.006 TON/min and 0.015 TON/min respectively)] (see
Activities of PBMA Film Catalyst for the Hydroformylation of
Propylene at Various Flow Rates.
Flow Rate (mL/min)
1000C. 20 psig.
Rhodium source HRh(CO)(P<3)3; P/Rh 10:1;
[Rh] = 0.10 g Rh per g catalyst.
1- I I I r I I I
0 2 4 6 8 10 12 14 16 1
Flow Rate (mL/min)
Figure 3-17. Effects of flow rates on the activity of the PBMA film catalyst on the
hydroformylation of propylene.
Table 3-10 and Figure 3-18), and was very selective producing only n-butyraldehyde.
Two possibilities exist for the decrease in activity: 1) The non-polar nature of the
SGR may decrease the solubility of hydrogen, carbon monoxide and propylene; 2)
The silicon polymer may simply poison the catalyst.
The PBD system was considerably more active (0.81 TON/min) than either
SGR or PBMA, with no apparent loss in activity over a two hour period. The
catalyst produced predominately n-butyraldehyde in a 17:1 n/b ratio. Polybutadiene
is very rubbery under reaction conditions and possesses no functionalities which
would poison the catalyst. Polybutadiene does, however, posses olefinic
functionalities which may be hydroformylated. Analysis of the film was not done but
if this reaction is occurring the film does not deactivate the catalyst. Reactions at
these internal olefinic sites on the polymer is expected to be much slower than the
rate of reaction for the hydroformylation of propylene. This lower activity coupled
with the large excess of propylene used in the reaction should minimize the possibility
of hydroformylation of the polymer alkene units.
Comparisons of activity with temperature and pressure. Figure 3-18
demonstrates the differences in activity of the three polymer systems at 800C. The
polybutadiene film catalyst was the most active followed by PBMA and SGR.
Table 3-11 and Figure 3-19 shows that at 1000C all the polymers are equivalent in
activity. This could be due to loosening of the films by molten triphenylphosphine.
Triphenylphosphine has a melting point of 800C. At 100C triphenylphosphine will
be molten and contribute significantly to the film. Table 3-12 shows the expected
Comparison of Polymers as Films for the Hydroformylation of
Propylene with SLPC's at 800C.
Film Activity (TON/min) n/b Ratio
PBD 0.085 17
SThe only product detected was n-butyraldehyde.
Conditions: 80C, 15 psig (1:1:1 HJCO/propylene).
Catatalyst: Rhodium source HRh(CO)(P0P),; P/Rh ratio 10:1;
[Rh] = 1.0 x 10' mol Rh per gram catalyst; 6 = 0.64;
mass of catalyst = 0.50 g.
I I II
0 10 20 30 40 50
L No Polyme
Figure 3-18. Activity curves of several supported polymer film catalysts for he
hydroformylation of propylene at 80C. Catalyst: Rhodium source-
HRh(CO)(PPh,),;P/Rh = 10:1: 6 = 0.64.
Conditions: 800C, 15 psig (1:1:1 HjCO/propylene).
Comparison of Polymers as Films for the Hydroformylation of
Propylene with SLPC's at 1000C.
Film Activity (TON/min) n/b Ratio
a The only product detected was n-butyraldehyde.
Conditions: 1000C, 15 psig (1:1:1 HJ/CO/propylene).
Catatalyst: Rhodium source HRh(CO)(P0,),; P/Rh ratio 10:1:
[Rh] = 1.0 x 10-4 mol Rh per gram catalyst; 6 = 0.70;
mass of catalyst = 0.50 g.
20 30 40
50 60 70
] PBDNo Po
Figure 3-19. Activity curves of several supported polymer film catalysts for he
hydroformylation of propylene at 1000C. Catalyst: Rhodium source-
HRh(CO)(PPh,),;P/Rh = 10:1; 6 = 0.64.
Conditions: 100C, 15 psig (1:1:1 H./CO/propylene).
Table 3-12. Effects of Varired Pressure with a SGR Film Catalyst.
Temperature Pressure Activity n/b Ratio
(C) (psig) (TON/min)
80 15 0.006
80 60 0.061 44
80 80 0.670 27
n-Butyraldehyde was the only product.
Gas mixture: 1/1/1 H/CO/propylene; Flow rate = 10 mL/min.
Catalyst: Rhodium Source HRh(CO)(PPh,),; P/Rh = 10:1;
[Rh] = 0.10 g Rh per g CAtalyst; 6 = 0.64.
increase in activity and decrease in n/b ratio for the SGR film catalyst with the
increase in pressure.
Molten salts are candidates as liquid phase media", both aliquat 336 (tri-n-
octyl methyl ammonium chloride) and tetrabutylphosphonium bromide demonstrate
similar activity to their polymer counter parts at loadings and conditions comparable
to the polymer systems (see Table 3-13). These systems produced only n-
butyraldehyde and showed no loss in activity over a three hour period. Both Aliquat
336 and tetrabutylphosphonium chloride are molten at room temperature.
As in the case of polymer films, an increase in reaction temperature does
increase the overall activity of the system. At 800C the activity of the aliquat 366
catalyst is 0.010 TON/min and increases at 1000C to 0.044 TON/min. Both catalysts
only produce n-butyraldehyde.
Supported liquid phase catalysts were tested as control catalysts for
hydroformylation of propylene. The liquids examined were 2-octanol, diethylene
phthalate, and propylene carbonate. These catalysts were run as simulants to Rony's
SLPC and to test if the totally selective systems found with rubbery polymers and
molten salts were real phenomena or a product of the flow rates and pressures of the
Table 3-13. Hydroformylation of Propylene with Molten Salt Film Catalysts.
Film Conditions Activity n/b Ratio
Aliquat 336 80/15 0.010
Aliquat 336' 80/15 0.046
Aliquat 336 100/15 0.044
TBPC1 100/15 0.010
PBMA 100/15 0.038
n-Butyraldehyde was the only product detected.
6 = 1.31, [Rh] = 0.06 g Rh per g catalyst.
Reactant gas 1:1:1 H,/CO/propylene, flow rate = 10 mL/min.
Catalyst: Rhodium source HRh(CO)(PPH,),; P/Rh 10:1; [Rh] = 0.1 g Rh per
gas flow reactor used in the experiment. As expected, all the high boiling liquid
catalysts showed slightly higher catalytic activity toward aldehydes than their
polymeric counterparts. In addition to higher activity, the production of iso-
butyraldehyde was appreciable (see Table 3-14).
Several experiments were carried out with propylene carbonate under our
relatively mild conditions for the purpose of comparison with the polymer systems.
It was determined that increasing temperature and pressure generally increase the
activity while lowering the selectivity (see Table 3-15).
Other Vapor Phase Substrates
In addition to propylene, 1-butene and 1,3 butadiene are of interest as
substrates. We have tested these substrates with the
HRh(CO)(PPh3)'/PPh3/PBMA/SG catalyst at 1000C and 20 psig. But-1-ene shows
both the linear and branched products in a 15.7:1 ratio, and a activity of 0.084
TON/min. Buta-1,3-diene produced four products with conversions approaching
Hydroformylation of Propylene with Supported Hydrophilic Catalysts
The hydroformylation of liquid alkenes such as 1-hexene will be described in
the following chapter. In this section we will describe the hydroformylation of
propylene with the catalyst used for liquid alkenes. The catalyst consists of a water
soluble rhodium catalyst (HRh(CO)(TPPTS),) (TPPTS = P(C,6HSO,Na),) and excess
Table 3-14. Hydroformylation of Propylene with Various SLPC's.
SThe only product was n-butyraldehyde.
Conditions: 80C, 15 psig 1:1:1 HJCO/propylene. Flow rate = 10 mL/min.
Catalyst: Rhodium source HRh(CO)(PPh,),; P/Rh = 10:1;
[Rh] = 0.10 g Rh per g catalyst.
Hydrotormylation of Propylene with Various Propylene Carbonate
Temperature Pressure Gas Mixture Activity n/b Ratio
(C) (psig) (H/CO/CH,) (TON/min)
80 15 1/1/1 0.007
80 15 1/1/0.1 0.013
100 15 1/1/1 0.197 41
100 60 1/1/1 0.342 11
" The only product detected was n-butyraldehyde.
Conditions: Flow rate = 10mL/min.
Catalyst: Rhodiun source HRh(CO)(PPh,),; P/Rh = 10:1:
[Rh] = 0.10 gRh per g catalyst; 6 = 0.64.
ligand (TPPTS) in a water soluble polymer (PEG 600) coating silica gel (see
Table 3.16). The degree of pore filling for this catalyst was 1.40. At 100C and 20
psig of 1:1:0.1 H/CO/propylene this catalyst showed a conversion of 0.114 TON/min
and a selectivity of 14.3. An increase in pressure to 40 psig increased the activity to
0.461 TON/min but lowered the n/b ratio to 6.69. A further increase in pressure
raised the activity to 0.519 and a n/b of 5.83. Increasing the temperature to 120C
decrease the activity to 0.196 and increased the selectivity to 8.74.
This demonstrate the feasibility of the water soluble rhodium catalyst in a
hydrophilic film coating silica for the hydroformylation of vapor phase substrates such
as propylene. The activity of this system is high for the degree pore filling (6=1.4).
Further experiments with lower S's should be done.
We developed a novel approach to supported liquid phase catalyst which uses
a polymer or molten salt as the film phase. These catalysts consist of a rhodium
source (HRh(CO)(PPh,)3 in a polymer or molten salt film supported on silica gel.
Previously reported SLPC used high boiling liquids as the liquid phase, leading to
films that evaporate with time, leaving a surface supported rhodium catalyst. We
have seen that the addition of excess phosphine ligand (triphenylphosphine) to the
catalytic film enhances the activity and the selectivity of the overall catalyst. This
enhancement can be attributed to the loosening of the film by molten
triphenylphosphine in the film, at temperatures above 80C.
Table 3-16. Hydroformylation of Propylene with a Supported Hydrophilic
Temperature Pressure Conversion Activity n/b Ratio
(C) (psig) (%) (TON/min)
100 20 3.1 0.114 14.3
100 40 12.5 0.461 6.69
100 60 14.1 0.519 5.83
120 60 5.3 0.196 8.74
Catalyst: Rhodium source HRh(CO)(TPPTS),; Additive: TPPTS;
Polymer: PEG 600; Support: Silica gel: 6= 1.4.
Pregas mixture: 1/1/0.1 H,/CO/propylene, flow rate 4.8 mL/min.
We have varied a number of the conditions and film properties of the catalytic
system and found that even a slight modification of any of the variables affects the
activity of the system dramatically. They include temperature, pressure, degree of
pore filling and concentration of rhodium in the film. It was shown that increased
temperature and pressure increased activity but decreased selectivity of the catalyst,
while the activity decreased significantly with the increase in pore filling. The
concentration of rhodium in the film also had an effect on the activity; as the
concentration increased so did the activity and the selectivity.
We also varied the film material by using different polymers and molten salts.
All of the films tested were active for the hydroformylation of propylene. The molten
salts tested showed extraordinary selectivity by producing only the linear aldehyde.
Liquid films such as propylene carbonate, 1-octanol and diethylene phthalate
demonstrated higher activities than their polymer counterparts but much lower
The catalyst system described herein has shown to be an effective system for
the hydroformylation of vapor phase substrates under mild conditions. The following
chapter will describe a system for the hydroformylation of liquid substrates in a batch
HYDROFORMYLATION OF HIGHER OLEFINS WITH SUPPORTED
HYDROPHILIC FILM CATALYSTS
The goal of heterogenizing homogenous catalysts is to design a catalyst that
would have many of the advantages inherent in both heterogeneous and homogenous
catalytic systems. Homogeneous catalysts offer high activity and selectivity along with
the ease of modification of the system. Heterogeneous catalysts offer the ease of
separation of the catalyst from the reactants and products along with the easy
recovery of expensive transition metals.
Many heterogeneous catalysts have been developed for hydroformylation.
Most include the coordination of rhodium or cobalt to functionalized organic
polymers or inorganic oxides. The functionalized groups are usually phosphines. In
addition, catalysts have been trapped in zeolites and several biphasic systems have
Supported Transition Metal Catalysts
Organic polymers. Numerous polymer-supported transition metal systems
have been developed and studied for catalysis."6 Organic polymers which
contain pendant phosphines have been prepared. The major drawback to these
catalysts is the elution of rhodium from the polymer surface.`' Pittman et al. have
chemically linked rhodium, cobalt, ruthenium and nickel complexes to polymers via
phosphine coordination. The polymer-bound transition metal catalysts were efficient
catalysts for cyclooligomerization, hydrogenation, and hydroformylation."*5' The
catalysts are depicted in Figure 4-1.
The rhodium phosphine polymer hydroformylated 1-pentene at 250-1000 psi
and 40-60C yielding a n/i ratio of 2.5:1.'" Comparison of the rhodium catalyst to
the homogeneous catalyst, HRh(CO)[P(C6H,)3]3, under similar conditions gave
interesting results. At low temperatures, 400C, the activity of the polymer catalyst was
significantly lower than the homogeneous catalyst. However, when the temperature
was increased to 600C, the activity of the two catalysts were nearly identical. This
phenomena was explained by diffusion of the reactants through the polymer matrix.
At lower temperatures the chains are restricted and immobile. The increase in
temperature decreased the viscosity of the resin which in turn greatly reduced the
diffusion time of the reactants to the rhodium centers.
After extended reaction time the activity of the catalyst decreased but time
while the n/b ratio increased. As the reaction is allowed to proceed the rhodium
center becomes more highly coordinated by anchored phosphines leading to low
activities and high n/b ratios. A later study showed that in catalysts with high loadings
of anchored phosphines a high degree of substitution results and n/b ratios of 12:1
are observed. Low loadings of anchored phosphines gave n/b ratios similar to those
of homogeneous catalysts.
Figure 4-1. Coordination of rhodium to a polymer support via pendant phosphine
Allum et al. developed similar polymer supports shown in Figure 4-2.*
Rhodium catalysts can be made by the addition of Rh(acac)(CO), to the polymer.
These copolymers are efficient hydroformylation catalysts for 1-hexene at 900C and
42 atm (1:1 HCO) but n/b ratios are low (0.81). Treatment of these polymers with
tri-n-butylphosphine increases the ratio to 2.3:1. In a continuous flow reactor (pilot
plant conditions) the activity of this catalyst drops dramatically from 86% to 36%
within 10 hours, but selectivity remains unchanged at 2.5:1. Leaching of the catalyst
into the reaction feed occurs. Trace amounts of oxygen facilitates metal leaching
from the support.
Non-phosphine containing polymer supports coordinate rhodium to produce
active hydroformylation catalysts. The polymer contains a organic sulfide group for
the linkage to the transition metal.70 This catalyst converts 100% of the 1-hexene
over an 8 hour period and has a n:i ratio of 1.
Inorganic oxides. Numerous inorganic oxides can be used as supports for
hydroformylation catalysts (e.g. silica gel, alumina, and porous glass)." Direct
deposition of RhCl3 or (PPh3),Rh(CO)Cl lead to catalysts which show negligible
activity for the hydroformylation of 1-hexene at 1000C and 300 psig.7 At low
pressures these catalysts are efficient isomerization catalysts for the formation of 2-
hexene. Increasing pressure and temperature gives an active catalyst for the
hydroformylation of propylene.
Allum et al supported metal complexes on silica by reacting
HRh(CO)[(EtO),SiCHSiCH2CH,PPh,], with functionalized silica.7 Under pilot plant
Preparation via substituted polystyrenes
Prepation via unsubstituted polystyrenes
Figure 4-2. Preparation of functionalized polymers.
CH OCH Cl
conditions the catalyst leaches into the feed initially but after the first few hours no
further leaching occurs. The leaching can be attributed to the presence of trace
amounts of oxygen in the reaction solution.
Zeolites. Various rhodium complexes can be incorporated into zeolites."73'7
Only a few are useful as heterogeneous catalysts for liquid substrates.75 The main
drawback to zeolite supported catalysts is the leaching of the rhodium complex into
the reaction solution. Never the less, there have been a few catalyst which are active
for the hydroformylation of 1-hexene and do not elute rhodium into the solution.76
Davis and coworkers have described the various types of zeolite catalysts for
hydroformylation in which the rhodium source is cation exchanged into the zeolite.77
There are three fashions in which the rhodium can be bound to the support. For
type I the rhodium species is weakly bound to the surface of the support and easily
elutes into solution. For type II the active species is tightly bound to the zeolite
support. The activities of these sites are low, because the rate is diffusion limited.
Type III has the rhodium complex entrapped within the zeolite cage. The n/b ratio
for type III sites are reported to be 2.6:1.
A variety of water soluble ligands and complexes are used for biphasic
catalysis.7'79 These systems employ the use of immiscible solvents to achieve the
desired separation characteristics of heterogeneous catalysts8" and rely heavily on
the solubility of the substrates in the phase in which the catalyst resides. Sulphonate
phosphines are commonly used as ligands for rhodium complexes in biphasic catalytic
systems." Rhone-Poulenc utilizes the water soluble catalyst, HRh(CO)(TPPTS)3
(TPPTS=trisodium tris(m-sufonatophenyl)phosphine), and the water soluble ligand,
TPPTS, for the hydroformylation of propylene.8"'8' The metal complex and
excess ligand resides in the aqueous phase while the majority of the substrates
(propylene) and products (n-butyraldehyde and isobutyraldehyde) are in the organic
phase. Efficient conversions of the olefin are made with a n:i ratio of 3:1. A similar
system uses a chelating sulfonated phosphine as a ligand for rhodium catalysts. These
systems yield high activities at low P/Rh ratios.87
Other biphasic hydroformylation systems have appeared. Taqui Khan et al.
report a system which exclusively produces n-heptaldehyde from 1-hexene using Ru-
EDTA in a 80/20 ethanol-water mixture at 1300C and 50 atm.8 A rate of 11.8
TON/hr is achieved and the catalyst can be recycled with out loss of activity.
Baird et al have developed a rhodium water soluble catalyst with using the
ammonium phosphine ligand Ph2PCH2CHNMe*NO."9 An active catalyst for
hydroformylation can be made with [(norbornadiene)RhCl]2 in the presence of the
phosphine, (at least a 3:1 P/Rh ratio). The active catalyst is
[(norbornadiene)Rh(Ph,PCH,CHNMe),]3" and hydroformylates 1-hexene which is
dissolved in hydrophobic organic solvents. An n:i ration as high as 4:1 is observed,
the selectivity is dependant upon the choice of solvent. Small amounts of
hydrogenation and isomerization products were observed, and leaching of a small
amount of rhodium into the organic media occurred.
The addition of surfactants such as CTAB or quantinary ammonium salts
increases the efficiency of the biphasic rhodium catalyst (see Table 4-1).'" A
biphasic system that consists of PPh,C6H,COH, and a rhodium source in water
converts 2'% d the 1-hexene substrate in three hours. With the addition of a
surfactant, the conversion increases to 44% in one hour. It was also noticed that
high selectivity occurs with poor mixing.
Supported Aqgeous Phase Catalysts
Recent. supported aqueous phase catalysts (SAPC) have been developed by
Davis et al."91 This system covers an inert porous support, such as porous glass,
with a thin aqueous film containing a water soluble catalyst, HRh(CO)(TPPTS),, and
a water solubk phosphine, TPPTS. Hydroformylation of higher olefins such as 1-
octene and oikyl alcohol can be done at moderate rates and a selectivity of 2.3:1
(conditions:7:7C, 700 psig). It is claimed that the rhodium complex is mobile within
the aqueous fIm and the catalysis occurs at the water/organic interface where the
organic phasecontains the reactants and products.
The rale of hydroformylation is unaffected by the size of the olefin in the
SAPC catalyst, whereas in biphasic systems the rate is highly dependent on this
variable.' Ttis indicates that water evaporates from the surface and leaves a
phosphine-rhetium complex bound to the support via hydrogen bonding of the
hydrated sodint-sulphonate groups to the surface (see Figure 4-3).
Effects of Surfactants on the Biphasic Hydroformylation of Olefins.
Surfactant Substrate Time % Conversion n/i
None 1-Hexene 3 2 7
C1H.,N'Me,Br 1-Hexene 1 44 73
None 1-Dodecene 3 2 6
C,2H,N'MeBr 1-Dodecene 1 78 20
none 1-hexadecene 3 0.5
C,_H,N'Me,Br 1-hexadecene 1 73 22
5.5 atms 1:1 H,:CO
Conditions: [Rh],,=300 ppm, 4-Ph,PC.H,COOH,
C,,H,NMe,Br:Rh=20:1 in pH 10 buffer at 800C,
Figure 4-3. Immobilized TPPTS on a hydrophilic support by
hydrogen bonding of the hydrated sodium-sulphonate
groups to the surface.
Water was purified by nanopore filtration and stored under nitrogen. The
premixed syn gas were prepared using a manifold system. A lecture bottle was
pressurized to 500 psig with carbon monoxide and then pressurized to 1000 psig with
hydrogen. The substrates 1-hexene, 1-octadecene, and 1-octene were purchased from
Aldrich Chemical and filtered through acidic alumina before use. Valeraldehyde was
filtered through acidic alumina and freshly distilled before use. Silica gel (grade 62,
mesh 60-200, surface area 700 m g', pore volume 1.1 mL/g) was purchased from
Davison and was dried at 100C overnight, and sealed under nitrogen. The following
were used without any further purification: Rh(acac)(CO), (Strem Chemical);
RhCL,63HO (Strem Chemical); 2-octanone (Fluka); polyvinyl alcohol (Fluka);
absolute ethanol (Florida Distillery Co.); formamide (Aldrich Chemical); polyethylene
glycol (Aldrich Chemical) polyvinyl pyrrolidinone (Polysciences Inc.).
Apparatus and Instrumentation
Catalysis was run in either a 300 mL Parr mini reactor or in glass pressure
bottles equipped with a pressure head (see Figures 4-4 and 4-5). Nuclear magnetic
resonance spectra were obtained by using a Varian VXR 300MHz or a General
Electric QE300 nuclear magnetic resonance (nmr) spectrometer. Infrared spectra
were obtained from a Nicolet 5PC FTIR spectrometer. Gas chromatographs were
obtained by using a Varian 3300, a Perkin-Elmer 900 and a Hewlett Packard 5890
using a DEGS packed column. GC-MS were obtained by using a Varian 3400
equipped with a Finnagan MAT ion trap detector.
Premixed gas cylinder
- psig )
Figure 4-4. Parr High Pressure Reactor.
premixed gas cylinder
Figure 4-5. Parr pressure bottle batch reactor.
All manipulations were carried out under an atmosphere of nitrogen.
Preparation of TPPTS
'The preparation of TPPTS was carried out similarly to the preparation
reported by Arhancet." Oleum (50 mL, 20%) was added dropwise to
triphenylphosphine (8.0g) in a three neck round bottom flask cooled by an ice bath.
After the addition and complete dissolution of the phosphine, the ice bath was
allowed to warm to room temperature and stirring was continued for at least 18
hours (the longer the phosphine solution is allowed to stir the higher the degree of
sulfonation). The solution was then cooled to 0-100C. Cooled deaerated water (200
mL) was added dropwise to the solution to quench the reaction. The acidic
phosphine was extracted from the aqueous solution by the addition of tri-n-
butylphosphate (50 mL, two times). The organic layers (top) were combined and
placed in an ih bath. Concentrated sodium hydroxide (50g, 100 mL) was added
dropwise to tie acidic phosphine solution until neutral. This resulted in the
formation of a white precipitate, from this a white solid forms. The solid was filtered,
washed with ditthylether (50 mL, 5 times) and air dried. The solid was then taken
up in water (75 mL) and methanol was added to precipitate a white solid (mostly
Na2SO,). The solid was removed by filtration and the filtrate was evaporated under
reduced pressure. Phosphorous ("P) nmr of the solid indicated the presence of
TPPTS (-5.50 plpm, 82%), HTPPTS (1.69 ppm, 16%) and OTPPTS (36.11 ppm, 3%).
Barium nitrate test of the solid was positive for the presence of NaSO,. The solid
was taken up in water (50 mL) and the NaSO, was precipitated by the addition of
absolute ethanol (80 mL). Recrystalization from methanol with ethanol removed all
of the phosphine impurities. If any HTPPTS remains it can be removed by washing
the solid with methanol.
Preparation of HRh(CO)(TPPTS),
A. The ligand TPPTS (2.30g, 4.05 mmol) was added to hot water (40mL).
To this RhCl3*3H,O in water (2mL) was added. After 15 seconds formaldehyde
(4mL, 27% in H,O) immediately followed by KOH (0.35g, 6.24 mmol) in water
(5mL) were added. A yellow solution formed almost immediately. The reaction
solution was allowed to stir for 10 minutes and cool to room temperature. Water was
evaporated under reduced pressure. The resultant yellow solid was taken up with
minimal water and the yellow solid was precipitated with ethanol.
B. The ligand TPPTS (0.52g, 0.92 mmol) was added to water (10mL) and
stirred under nitrogen until completely dissolved. Rh(acac)(CO), (0.050g, 0.19 mmol)
was added to the TPPTS solution. Upon complete dissolution an atmosphere of
HJCO (1:1) was introduced and allowed to flow slowly for 18 hours. The yellow
solution was filtered and evaporated to 1 ml. Ethanol was added to precipitate a
dark yellow solid. The solid (I) was filtered, washed with hexanes, and dried under
vacuum. Additional ethanol was added to the filtrate to precipitate additional solid
(II). The mixture was centrifuged and (II) was dried in the vacuum oven. The liquid