Catalytic hydrogenation of nitrile-butadiene copolymers by cationic rhodium complexes

Article abstract of DOI:10.1016/S1381-1169(97)00309-9

The hydrogenation of nitrile-butadiene copolymers (NBR), catalyzed by a series of cationic rhodium complexes [Rh(diene)(L₂)]⁺ (diene = norbornadiene (NBD) and 1,5-cyclooctadiene (COD); L₂ = (PPh₃)₂, Ph₂P(CH₂),PPh₂ (n = 2, 3 and 4); Cy₂P(CH₂)₂PCy₂), was studied under 3.42 MPa hydrogen pressures at 140°C in monochlorobenzene. The general activity patterns found include: (1) the activity of complexes [Rh(diene)(PPh₃)₂]⁺ is similar to that of Wilkinson catalyst RhCl(PPh₃)₃; (2) with chelating diphosphines, the activity of complexes [Rh(diene)(L₂)]⁺ increases with the increase of the chain length of the diphosphine; (3) complexes [Rh(diene)(PPh₃)₂]⁺ are more active catalysts than those containing chelating phosphines. These results demonstrate that there are a number of similarities and remarkable differences between the hydrogenation of NBR and the hydrogenation of simple olefins reported, which is mainly attributed to the steric effect of olefins. In contrast to simple olefin hydrogenation where the coordination of olefin to a metal center is typically facile, this step becomes rate-determining in the case of the hydrogenation of large olefins such as NBR. In order to determine the effects of the factors on the reaction, such as catalyst concentration, polymer concentration, hydrogen pressure and temperature, a detailed kinetic study on the reaction catalyzed by complex [Rh(NBD)(PPh₃)₂]⁺ was carried out. On the basis of the kinetic results and the related statistical analysis, both a reaction pathway and a reaction rate law are proposed. The dilute solution viscosities of fully hydrogenated NBR were found to be independent of reaction conditions employed, suggesting that the properties of hydrogenated NBR are uniform.

Full text of DOI:10.1016/S1381-1169(97)00309-9

Ž
.
Journal of Molecular Catalysis A: Chemical 135 1998 121–132
Catalytic hydrogenation of nitrile-butadiene copolymers by
cationic rhodium complexes
)
Tian-Fu Mao, Garry L. Rempel
Department of Chemical Engineering, UniÕersity of Waterloo, Waterloo, Ontario, Canada, N2L 3G1
Received 20 October 1997; accepted 12 December 1997
Abstract
The hydrogenation of nitrile-butadiene copolymers NBR , catalyzed by a series of cationic rhodium complexes
Ž
.
w
Ž
.Ž .xq Ž
Ž
.
Ž
.
Ž
.
Ž
.
Ž
Rh diene L2 dienesnorbornadiene NBD and 1,5-cyclooctadiene COD ; L2s PPh3 2, Ph2P CH2 nPPh2 ns2, 3
.
Ž
.
.
and 4 ; Cy2P CH2 2PCy2 , was studied under 3.42 MPa hydrogen pressures at 1408C in monochlorobenzene. The general
Ž . .Ž
w
Ž
.2xq
is similar to that of Wilkinson catalyst
activity patterns found include: 1 the activity of complexes Rh diene PPh3
Ž .
Ž
.
w
Ž
.Ž .xq
RhCl PPh3 3; 2 with chelating diphosphines, the activity of complexes Rh diene L2
increases with the increase of the
Ž .
3
w
Ž
.Ž
.2xq
complexes Rh diene PPh3 are more active catalysts than those containing
chain length of the diphosphine;
chelating phosphines. These results demonstrate that there are a number of similarities and remarkable differences between
the hydrogenation of NBR and the hydrogenation of simple olefins reported, which is mainly attributed to the steric effect of
olefins. In contrast to simple olefin hydrogenation where the coordination of olefin to a metal center is typically facile, this
step becomes rate-determining in the case of the hydrogenation of large olefins such as NBR. In order to determine the
effects of the factors on the reaction, such as catalyst concentration, polymer concentration, hydrogen pressure and
w
Ž
.Ž
.2xq
temperature, a detailed kinetic study on the reaction catalyzed by complex Rh NBD PPh3 was carried out. On the
basis of the kinetic results and the related statistical analysis, both a reaction pathway and a reaction rate law are proposed.
The dilute solution viscosities of fully hydrogenated NBR were found to be independent of reaction conditions employed,
suggesting that the properties of hydrogenated NBR are uniform. q 1998 Elsevier Science B.V. All rights reserved.
Keywords: Cationic rhodium complexes; Hydrogenation; Nitrile-butadiene rubber; Kinetics; Mechanism
1. Introduction
polymers, the modified polymers usually exhibit
enhanced stability towards thermal and oxida-
tive degradation to meet the commercial de-
mands for robust elastomers that can be used in
Catalytic modifications of butadiene poly-
mers have attracted much attention as they are
among the most important approaches to the
production of specialty elastomers with desir-
able chemical and physical properties which are
generally unattainable by conventional poly-
w x
severe environments 3 .
A most common method employed for poly-
mer modifications is the catalytic hydrogenation
reaction. Many examples have been listed in
w
x
merization methods 1,2 . In contrast to existing
w
x
recent review articles 1,2 . In the case of nitrile
Ž
.
butadiene rubber NBR , there are two types of
unsaturation present, namely, olefin and nitrile
)
Corresponding author. Fax: q1-519-7464979.
1381-1169r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S1381-1169 97 00309-9

(
)
122
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
Ž
.
group Scheme 1 . The challenge of this process
is, therefore, to selectively hydrogenate olefin
with the nitrile group remaining intact as the
reduction of nitrile group would reduce the oil
resistant properties of the material. Many transi-
tion metal complexes, generally containing a
group VIII transition metal from the second and
third rows of the periodic table, are used as
homogeneous catalysts for this process
another closely related class of hydrogenation
catalysts of the form, Rh diene L2 q, where
w
Ž
.Ž
.x
Ž
.
diene is norbornadiene NBD or 1,5-cyclooc-
Ž
.
tadiene COD and L is a triphenylphosphine
ligand or L2 is a chelating phosphine ligand. As
part of our ongoing studies on NBR modifica-
tions, it was of interest to investigate the cat-
alytic hydrogenation reactions by these cationic
rhodium complexes, which have also received
considerable attention for simple olefin hydro-
w
x
1,2,4,13 . As for the hydrogenation of olefins,
the Wilkinson catalyst, Rh PPh3 3Cl, is the most
popular catalyst employed for this process
1,2,5–7 . This is due to the fact that it is a
highly active catalyst capable of achieving com-
plete hydrogenation of olefin without any reduc-
tion of nitrile group 7 . In contrast, and to the
best of our knowledge, there are no reports in
the literature on the hydrogenation of NBR by
Ž
.
w
x
genation 8–11 .
Although complexes of the type Rh-
xq
are excellent catalysts for small
w
w
x
Ž
.Ž
.
diene L2
w
x
olefin hydrogenation 8–11 , their activity may
not be directly extrapolated to the hydrogena-
tion of larger olefins such as the carbon–carbon
double bonds within diene based polymers. The
rate of hydrogenation of carbon–carbon double
w x
Ž
.
Scheme 1. Structures of NBR and hydrogenated NBR HNBR .

(
)
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
123
bonds within such polymers is generally much
slower than that within the small olefin analogs
due to the steric inaccessibility of olefin to the
metal coordination sphere. Thus, we have at-
tempted to determine the catalytic activity of a
series of complexes using screening experi-
All reagent grade solvents were used as re-
ceived and were thoroughly degassed prior to
use. NBD and COD were obtained from Aldrich
Chemical and distilled before use. The ligands,
Ž
.
triphenylphosphine, bis diphenylphosphino
Ž
.
Ž
.
ethane dppe , bis diphenylphosphino propane
w
Ž
Ž
Ž
. Ž .
ments. This series of complexes include Rh di-
dppp , and bis diphenylphosphino butane
.Ž
.
xq
. Ž
dppb were purchased from Aldrich. Bis di-
ene L2
where dienesNBD, COD; Ls
Ž
.
Ž
.
.
Ž
.
PPh3 or L 2 s Ph2 P CH 2 2 PPh2 dppe ,
cyclohexylphosphino ethane dcpe was ob-
Ž
.
Ž
.
Ž
.
Ph2 P CH2 3PPh2 dppp , Ph2 P CH2 4PPh2
tained from Strem Chemical. Oxygen-free hy-
Ž
.
Ž
.
Ž
.
Ž
.
dppb ; Cy2 P CH2 2 PCy2 dcpe . Moreover, a
drogen gas 99.99% was obtained from Linde-
Union Carbide Canada.
detailed kinetic study on the catalyst,
Rh NBD PPh3 2 q, was carried out with a
w
Ž
.Ž
. x
2.3. Synthesis of Rh COD dcpe qBF4y
)(
view to further appreciate the intimate catalytic
(
)
process.
Ž
Fluoboric acid 48% in H2O, 29 ml, 0.2
.
mmol was added to a yellow solution of 1,5-
cyclooctadienerhodiumacetylacetonate 69 mg,
0.2 mmol in 2 ml tetrahydrofuran at 08C. The
2. Experimental
Ž
.
2.1. General
reaction mixture was stirred vigorously and al-
All organometallic complexes were prepared
under an atmosphere of nitrogen using standard
Schlenk techniques. Infrared spectra were
recorded on a Nicolet 520 FT–IR spectro-
photometer over the range 400–4000 cmy1. The
1H NMR spectra were recorded on a Bruker 200
MHz or a Bruker 250 MHz spectrometer and
lowed to warm to room temperature. A solution
Ž
.
Ž
of bis dicyclohexylphosphino ethane 88 mg,
.
0.2 mmol in 3 ml tetrahydrofuran was then
added, resulting in rapid formation of an orange
solution. The mixture was stirred for 2 h at
room temperature. The solvent was then re-
moved in vacuo. The residue was recrystallized
31
1
Ä
4
P H NMR spectra were recorded on a Bruker
Ž
.
from CH2Cl2rether to give 97 mg 67% of
1
1
200 MHz spectrometer. The H NMR chemical
Ž
.
light orange powder. H NMR CDCl3 : d 5.31
br s, 4 H, olefin , 1.15–2.40 56 H, 4C6H11,
2CH2, 4CH2 COD ; P H NMR CDCl3 : d
shifts are reported in ppm relative to tetrameth-
Ž
.
Ž
31
1
31
1
Ž
.
Ä
4
ylsilane TMS while P H NMR chemical
shifts are reported relative to external 85%
H3PO4.
Ž
..
Ä
4
Ž
.
Ž
.
67.9 d, JRhP s145 Hz .
2.4. Hydrogenation reactions
2.2. Materials and reagents
Rhodium trichloride hydrate was purchased
from Engelhard. The acrylonitrile-butadiene
The hydrogenation reactions were carried out
in a batch autoclave reactor. In a typical experi-
Ž
.
Ž
.
copolymer NBR was obtained from Bayer.
ment, a polymer 1.55 g was dissolved in 75 ml
monochlorobenzene. This solution was then
transferred into a glass liner, which was placed
inside the autoclave reactor. After the reactor
was assembled, the system was degassed by
charging the reactor with hydrogen to about
3.42 MPa, stirring vigorously under hydrogen
for a few minutes, and venting. After this cycle
Ž
.
This rubber Krynac 38.50 had a molecular
weight of 200 000 with 38% nitrile by weight.
The following compounds were prepared by
published procedures: Rh diene PPh3 2 qPF6y
w
Ž
.Ž
. x
Ž
Ž
.
w
x
w
Ž
.
diene s NBD, COD
12 , Rh diene -
L2 qBF4 diene s NBD, COD; L2 s dppe,
.
x
yŽ
.
dppp, dppb .

(
)
124
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
was repeated three times, the system was further
degassed by bubbling hydrogen through the so-
lution for 30 min. The reactor was heated to
1408C. After equilibrium was reached, a solu-
at 358C in a Ubellohde capillary viscometer.
The data reported are relative to pure solvent.
Ž
.
tion of catalyst 0.02 mmol in 25 ml
monochlorobenzene was added into the reactor
to initiate the hydrogenation reaction. The reac-
tion was maintained at 1408C and the pressure
of the system was adjusted to 3.42 MPa with
hydrogen. In order to maintain constant reaction
pressure of the system throughout the hydro-
genation, hydrogen gas was introduced when
required. The reaction was monitored by period-
ically withdrawing samples from the reactor and
recording their IR spectra. After reaction, the
system was quickly cooled to room temperature,
followed by venting hydrogen. The hydro-
genated polymer was precipitated with methanol
and dried in vacuum.
3. Results and discussion
3.1. Synthesis and characterization of
Rh COD dcpe qBF4y
(
)(
)
The effect of phosphine on the hydrogenation
activity of transition metal complexes is a well
w
x
known phenomenon 17,18 . In order to expand
the scope of complexes Rh diene L2 qBF4y
w
Ž
.Ž
.x
Ž
.
dienesNBD, COD; L2 sdppe, dppp, dppb ,
we attempted to prepare Rh COD dcpe qBF4y.
The synthesis of this complex was initially at-
tempted in a similar way to those of
w
Ž
.Ž
.x
Rh diene L2 qBF4y dienesNBD, COD; L2
w
Ž
.Ž
.
x
Ž
.
w x
sdppe, dppp, dppb 19 . Unexpectedly, addi-
Ž
.Ž
.
tion of dcpe to a mixture of Rh COD acac and
Ž
.
excess of HBF4 3.8 equivalents in THF re-
sulted in the formation of two complexes, 1 and
2, in a ratio of 3:2.
2.5. Kinetic studies
The kinetic experiments were carried out in a
batch reactor, which was interfaced to a com-
puter system. Hydrogen gas consumption was
continuously monitored and the pressure in the
reactor was maintained constant throughout the
hydrogenation. A detailed description of the
w
x
experimental apparatus is given elsewhere 14 .
The experimental procedures are similar to
w
x
those described previously 15 . The polymer
was accurately weighed out and dissolved in
150 ml monochlorobenzene. When a known
amount of catalyst was added into the polymer
solution, the computer program recording reac-
tion time, hydrogen gas uptake, and reaction
temperature, was then started. The experiment
was stopped when hydrogen gas uptake was no
longer noticeable. Finally, the degree of hydro-
genation was calculated based on an IR spec-
trum of the hydrogenated polymer, which was
obtained from a thin film of the polymer on a
w
x
NaCl plate 16 . The viscosity measurements of
Ž
fully hydrogenated NBR solutions 0.250 g of
Both compound 1 and 2 were characterized
31
HNBRr25 ml of 2-butanone were performed
by 1H and P H NMR spectroscopy. The
1
.
Ä 4

(
)
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
125
31
1
Ä
4
P H NMR spectrum of 2 in CD2Cl2 showed
as PPh3 are the most active catalysts. In the
case of diphosphines including dppe, dppp,
dppb, the catalytic activities increase with
the increase of the backbone length of phos-
Ž
.
a doublet at d 99.4 JRhP s137 Hz , indicating
that two phosphines are equivalent. In the corre-
sponding 1H NMR spectrum a doublet of triplets
Ž
.
w
at dy22.4 JRhH s28.7 Hz, JPH s22.6 Hz
phines, namely, in the following order: Rh-
Ž
Ž
.Ž
.
xq
w
Ž
.Ž
.
xq
w
) Rh-
was observed, which is assigned to the hydride.
The possibility of a trigonal–bipyramidal struc-
ture, where one of the phosphines occupies an
in-plane equatorial position and the other takes
up an axial position, can be ruled out since it
would generate two unequivalent phosphines.
diene dppb
) Rh diene dppp
diene dppe q. This trend is consistent with
.Ž
.x
w
x
that found for simple olefins 18 . Since the
electronic properties of dppe, dppp, and dppb
are similar, their activity differences are at-
tributed to the steric effect of the ligands. In
contrast to a short chain phosphine, a phosphine
having a longer chain length is more flexible to
accommodate structure changes during the cat-
alytic process thereby facilitating the reaction.
Ø With the same phosphine ligand, a com-
plex containing a NBD ligand is more reactive
at the early stage of the reaction than that
containing a COD ligand. This is attributed to
the fact that the reduction of norbornadiene is
faster than that of 1,5-cyclooctadiene to gener-
1
31
1
Ä
4
Thus, the H and P H NMR spectroscopic
data are in accord with the structure as shown.
1
31
1
Ä
4
The H and P H NMR spectra of complex 1
w
Ž
.Ž
.
xq
are very similar to those of Rh COD dppe
ClO4y 12 and therefore an analogous structure
is assigned to it.
w
x
Although the mechanism of formation of
complex 2 was not investigated, one can envi-
sion that species 1 is a very possible intermedi-
ate. Since Rh COD dppe qClO4y is resistant
to protonation, the ease of protonation of com-
plex 1 is due to the fact that a dcpe ligand is a
much better donor than a dppe ligand, which
stabilizes cationic species 2. Based on this
premise, the method for the preparation of com-
pound 1 is modified and only one equivalent of
HBF4 was used. Indeed, as described in the
experimental section, the complex 1 was pre-
pared in good yield.
w
Ž
.Ž
.x
w
x
ate a catalytic active species 20 .
The catalytic activity of Rh diene -
xq
w
Ž
.
.
Ø
Ž
.
Ž
PPh3 2 is similar to that of Rh PPh3 3Cl.
Ø The Rh COD dppe q is more active than
xq
w
.Ž
Ž
.Ž
.x
w
Ž
.
Rh COD dcpe
Ø In contrast to simple olefin hydrogenation
xq
catalyst.
w
Ž
.Ž
.
Ž
where the activities of Rh diene L2
L2 s
.
dppe, dppp, dppb are higher than those ob-
served for Rh diene PPh3 2 q, the complexes
xq
w
Ž
.Ž
. x
w
Ž
.Ž
.
of type Rh diene PPh3 2
.
than complexes of the type Rh diene L2
are more active
w
Ž
.Ž
xq
3.2. Catalytic hydrogenation
Ž
.
L2 sdppe, dppp, dppb .
As described above, the hydrogenation of
w Ž .
xq
NBR using Rh diene L2 exhibit some simi-
w
Ž
The catalytic activity of a series of Rh di-
ene L2 q complexes was studied at 1408C and
a hydrogen pressure of 3.42 MPa. The results
are shown in Table 1. Consistent with the hy-
drogenation of simple olefins, all complexes are
active catalysts for NBR hydrogenation. The
general features of the results are summarized
as follows.
larities to that observed for small olefins 18 .
The catalyst activity increases as the chain length
of the diphosphine increases. The activity of
.Ž
.
x
w x
w
Ž
.Ž
.
xq
Ž
.
Rh diene PPh3 2 is similar to RhCl PPh3 3
Thus the hydrogenation processes of NBR with
Rh diene L2 appear to be similar to those of
.
w
Ž
.
xq
w
x
small olefins. Halpern 18 reported that, in the
Ø The catalytic activity of complexes of the
type Rh diene L2
the nature of phosphine. Interestingly, the com-
hydrogenation of small olefins, the cationic
w
Ž
.Ž
.
xq
w
Ž
.
xq
rhodium complexes Rh diene L2 containing
is mainly dependent on
a chelating diphosphine have much higher activ-
plexes containing monophosphine ligands such
ity than complexes containing monophosphines

(
)
126
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
Table 1
The hydrogenation of NBR with cationic rhodium complexesa
Catalyst
% Hydrogenationb
Turnover ratec
y1
mol mol Rh
hy1
.
Ž
Ž
.
Rh COD dppe qBF4y
.Ž
20
22
7
35
40
91
91
97
97
99
178
196
62
312
356
810
810
863
863
881
Ž
Ž
Ž
Ž
Ž
Ž
Ž
Ž
.
.
.
.
.
.
.
.
Rh NBD dppe qBF4y
.Ž
.Ž
Rh COD dcpe qBF4y
.Ž
Rh COD dppp qBF4y
Rh NBD dppp qBF4y
.Ž
.Ž
Rh COD dppb qBF4y
Rh NBD dppb qBF4y
.Ž
Rh COD PPh3 qPF6y
.Ž
2
Rh NBD PPh3 qPF6y
Ž
.Ž
.
2
Ž
.
RhCl PPh3
3
a
y2
Reaction conditions: Catalysts2=10y5 mol, olefin in NBR Krynac 38.50 s1.78=10
mol, solvents100 ml monochlorobenzene,
Ž
.
temperatures1408C, PH s3.42 MPa.
2
bReported for 1 h reaction period.
c Turnover rate were calculated based on the number of moles of olefin reduced per mole of rhodium in the first 1-h interval of the
experiment.
such as Rh diene PPh3 2 q. However, the rate
xq
an olefin is coordinated to a rhodium I com-
2 xq
w
Ž
.Ž
.
x
Ž .
w
Ž
.
w
Ž
.Ž
.
of NBR hydrogenation with Rh diene L2
plex Rh L2 NC
nitrile group present in NBR. With complexes
xq
where NC stands for the
Ž
.
L2 sdcpe, dppe, dppp, dppb is significantly
slower than that with Rh diene PPh3 2 q cata-
lysts. In the case of small olefin hydrogenation,
the rate determining step is usually either the
oxidative addition of H2 or the reductive elimi-
nation of the final products. In contrast, with
macromolecules such as NBR, the step of the
coordination of olefin to the metal center be-
Rh diene PPh3 2
coordinated to a rhodium III complex
as catalysts, an olefin is
w
Ž
.Ž
.
x
w
Ž
.Ž
.
Ž
.
Rh H PPh3 2 NC q as required by a hydride
w
Ž . Ž
. Ž
.x
2
pathway. As an olefin is prone to attack the
more electrophilic rhodium III center instead
of attacking the rhodium I center, the catalytic
Ž
.
Ž .
w
Ž
.Ž
.
xq
reactions with Rh diene PPh3 2
are ex-
w
x
comes rate determining 5,7,21,22 . Thus the
following discussion will focus on the rate de-
termining step of the reaction and show that the
olefin is coordinated to the two types of metal
centers.
pected to be faster than those carried out with
diphosphine complexes. This is also consistent
with the observations that the complex
w
Ž
.Ž
.
xq
Rh COD dcpe is a less efficient catalyst
compared to Rh COD dppe q. Since the dppe
w
Ž
.Ž
.x
The catalytic hydrogenation of simple olefins
ligand is a poorer s-donor than the dcpe ligand,
with Rh diene L2 q complexes have been ex-
w
Ž
.Ž
.
w
x
w
Ž
.Ž
.
2 xq
the Rh metal in the complex Rh dppe NC
x
tensively studied 11,23 . It is generally ac-
cepted that there are two distinct reaction path-
ways associated with this type of catalysts,
which are dependent on the nature of phosphine
ligands. In the case of a monophosphine such as
triphenylphosphine, the reaction follows a hy-
dride pathway. With diphosphines, the reaction
proceeds by an olefin route. Consequently, in
derived from the latter complex is a better elec-
trophilic center than that from the former com-
plex thereby speeding up the reaction. Although
the steric effect in these catalytic processes is
not obvious, a steric bulky dcpe ligand could
prohibit the coordination of a bulky olefin in
NBR thereby retarding the reaction.
w
x
w
Halpern et al. 24 reported that Rh-
w
Ž
.
Ž
.Ž
.
xq
the case of complexes of the form Rh diene -
NBD diphos
reacts with hydrogen to give
q Ž
.
.x
L2
L2 sdcpe, dppe, dppp, dppb as catalysts,
the species Rh diphos q. This species can un-
w Ž
x

(
)
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
127
2
q
w
Ž
.2 x
dergo dimerization to form Rh2 diphos .
pete with olefins for vacant sites at the metal
w x
9 .
More importantly, it is found to react with
arenes to form adducts such as:
K
w
xq
w
xq
RhŽdiphos. qarene | RhŽdiphos.Žarene.
3.3. Kinetics of the hydrogenation
The parameter K was determined with a
w
x
variety of arenes and olefins in methanol 24 .
A series of experiments was carried out to
investigate the effects of variables, such as,
catalyst concentration, carbon–carbon unsatura-
tion, temperature, hydrogen pressure, on the rate
y
1
Ž
.
Ž
They include benzene 18 M ; toluene 97
My ; o-, m-, or p-xylene ;500 M ; 1-
1
y1
.
Ž
.
y
1
y1
Ž
.
Ž
.
hexene 2 M ; methyl acrylate 3 M
.
w Ž .
Clearly, arenes are much better ligands to form
the adducts than olefins, indicating that in the
case of NBR hydrogenation in monochloroben-
of NBR hydrogenation with Rh NBD -
PPh3 2 q. The reason that this catalyst was
Ž
. x
chosen is that it showed the highest activity of
hydrogenation among the complexes described
earlier and resulted in near quantitative olefin
hydrogenation, which made it suitable for mea-
suring hydrogen gas uptake accurately using a
Ž
.
zene MCB , MCB is an effective ligand to
compete with olefins for the vacant sites of
w
Ž
.
xq
Rh diphos thereby retarding the catalytic
hydrogenation of NBR. In contrast, analogous
w
x
arene adducts were not observed in the hydro-
computer controlled gas uptake apparatus 14 .
A summary of kinetic results is given in Table
2. A representative gas uptake profile is shown
w
Ž
.
genation of olefins with complex Rh COD -
PPh3 2 q, suggesting that MCB does not com-
Ž
. x
Table 2
The summary of kinetic data for the hydrogenation of NBR Krynac 38.50 catalyzed by Rh NBD PPh3 2 PF6 in monochlorobenzene
Ž
.
Ž
.Ž
.
k=104,
hrel
w
x
w
x
w
x
w
x
Rh
C5C
CN
PH
H2
Temperature
Ž .
K
T
.
o
2
mM
mM
mM
MPa
mM
sy1
Ž
Ž
.
Ž
.
Ž
.
Ž
.
80.1
80.1
153.0
153.0
152.9
153.0
153.0
152.9
153.0
153.0
76.5
107.3
137.6
137.8
191.2
229.6
275.4
153.1
153.0
153.0
153.0
153.0
153.0
152.9
153.0
153.0
153.0
153.0
95.5
95.5
95.5
95.5
95.5
95.5
95.5
95.5
47.8
67.0
85.9
86.0
119.4
143.4
172.0
95.6
95.6
95.6
95.5
95.5
95.5
95.5
95.6
95.6
95.6
95.6
2.46
2.46
2.46
2.46
2.46
2.46
2.46
2.46
2.46
2.46
2.46
2.46
2.46
2.46
2.46
1.36
3.72
4.86
5.97
5.97
6.96
7.01
2.46
2.46
2.46
2.46
107
107
107
107
107
107
107
107
107
107
107
107
107
107
107
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
388.2
373.2
378.2
383.2
393.2
4.51
4.39
8.12
9.25
12.6
13.4
15.5
16.3
19.8
14.3
13.6
15.1
11.7
10.8
8.79
6.56
17.1
20.8
24.1
23.5
29.9
31.0
5.94
8.17
9.26
16.6
160.2
160.2
240.4
239.8
319.9
320.3
239.8
240.3
240.0
239.9
239.9
240.2
240.4
240.1
240.2
240.0
240.0
240.1
240.1
240.4
240.0
239.8
240.3
239.8
4.32
3.87
4.13
3.89
3.88
4.05
61.7
159
205
250
250
284
285
107
107
107
107
4.54

(
)
128
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
w x w x
Fig. 3. Effect of CN on the rate of hydrogenation. Rh T s240
Fig. 1. Representative olefin conversion profiles for the hydro-
w
x
w
x
Ž
.
Ž
.
—
genation of NBR in monochlorobenzene. Rh T s240 mM; CN
mM; PH s2.46 MPa; T s388.2 K.
Ø
Observed data;
2
s95.6 mM; PH s1.36 MPa; T s388.2 K.
model.
2
in Fig. 1. The catalytic reaction follows a first-
order regression model indicating that there is
no preference for the hydrogenation of
vinylrcisrtrans butadiene isomerization of
NBR under the reaction conditions. This is in
accord with the observations found for
Ž
.
RhH PPh3 4 systems where the viscosities are
w x
in the range from 3.35 to 3.93 7 . The random
viscosities indicate that the product properties
are independent of the reaction conditions em-
ployed.
Ž
.
Ž
.
Rh PPh3 3Cl and RhH PPh3 4 systems investi-
3.3.1. Effect of catalyst concentration
w x
gated under similar conditions 7 . The relative
viscosities hrel reported in Table 2 are in the
range from 3.87 to 4.54, which are somewhat
higher than those found for Rh PPh3 3Cl and
To determine the dependence of the rate of
hydrogenation on catalyst concentration, a se-
ries of experiments was carried out in which the
catalyst concentration was varied from 80 mM
Ž
.
Ž
.
w
x
w
x
Fig. 2. Effect of Rh on the rate of hydrogenation. CN s95.5
Fig. 4. Effect of hydrogen pressure on the rate of hydrogenation.
T
Ž
.
Ž
.
w
x
w
x
Ž .
mM; PH s2.46 MPa; T s388.2 K.
Ø
Observed data;
—
Rh T s240 mM; CN s95.5 mM; T s388.2 K. Ø Observed
2
Ž
.
model.
model.
data;
—

(
)
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
129
w
Ž
.Ž
.2 xq
Scheme 2. Proposed mechanism for the hydrogenation of NBR with Rh NBD PPh3 .
to 320 mM while the polymer concentration
effect of NBR concentration on the rate of the
reaction can be studied by varying NBR concen-
tration. Fig. 3 shows that the reaction rate de-
creases with increase of NBR concentration.
This is due to the presence of nitrile group in
NBR, which can compete with olefin for the
metal coordination sites during the catalytic pro-
Ž
.
Ž
.
153 mM , hydrogen pressure 2.46 MPa , and
Ž
.
temperature 388.2 K were maintained un-
changed. Fig. 2 shows the resulting rate con-
stants and their regression value. The linear
relationship between the reaction rates and the
catalyst concentrations indicates that the reac-
tion is first order in catalyst concentration.
w x
cess. Schrock and Osborn 8 reported that
acetonitrile strongly inhibits the hydrogenation
3.3.2. Effect of polymer concentration
of simple olefins, such as 1-hexene, by
When the catalyst concentration 240 mM ,
Rh NBD PPh3 2 q. This is attributed to
Ž
.
w
Ž
.Ž
. x
q
Ž
.
w
Ž
.
.
reaction temperature 388.2 K , and hydrogen
the facile reaction of RhH2 PPh3 2S2
Ž
.
pressure 2.46 MPa were held constant, the
Table 4
Estimates of model parameters
Table 3
² :
95%
Parameter
Estimate
ASE
Model analysis of variance results
Source
Sum of squares
df
Mean square
Lower
Upper
y1
6.400=10y5
3.314=10y7
6.132=10y5
1.148=10y5
4
1.600=10y5
1.742=10y8
1.34
3.47
0.397
1.04
0.505
1.29
2.17
5.65
Ž .
k2 mM s
Regression
Residual
Total
Ž .
18
22
21
K1 mM
y1
y1
K3 mM
1.67=10y2 3.73=10y3 8.87=10y3 2.45=10y2
1.85=10y3 7.23=10y5 1.70=10y3 2.00=10y3
Ž
.
.
Ž
Corrected
K4 mM

(
)
130
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
with H2 in coordinated solvents such as
methanol and acetonitrile to generate the sol-
xq
w
Ž . Ž
.
w
x
20,24 .
vated species Rh H PPh3 2S2
2
These complexes also react readily with H2 in
the solvents of low coordinating power such as
CH2Cl2 and benzene to generate catalytically
w x
active species 9 . The exact nature of this
species is not known as it has defied identifica-
1
tion by H NMR spectroscopy at low tempera-
ture, presumably the concentrations of these
w x
active species are too low to be observed 9 .
Ž .
Thus, as illustrated in Eq. 1 , we propose that
w
Ž
.Ž
.
xq
the reaction of Rh NBD PPh3 2 with H2 in
X
X
X
Ž
.
the presence of NBR and chlorobenzene gener-
Fig. 5. Residual plot of kactual ykmodel versus k .
w
Ž . Ž
. Ž
.
2 xq
ates a putative species Rh H PPh3 2 NC ,
2
Ž
.
Ssacetone or ethanol , which is a catalytic
where NC stands for the nitrile group present in
w
Ž . Ž
. Ž
.
2 xq
active species, with acetonitrile to generate
NBR. Complex Rh H PPh3 2 NC
subse-
quently dissociates into a sixteen-electron
xq
2
RhH2 PPh3 2 NCCH3 2 q. That nitrile inhibits
the hydrogenation of NBR with Rh PPh3 3Cl
w
Ž
. Ž
. x
Ž
.
w
Ž . Ž
. Ž
.
w
Ž .x
Eq. 2 .
species Rh H PPh3 2 NC
2
w
x
was also observed from this laboratory 5,7 .
3.3.3. Effect of hydrogen pressure
The rate of NBR hydrogenation with
Ž
.
Ž
.Ž .Ž
.
Rh PPh3 3Cl and OsHCl CO O2 PCy3 2 cata-
lysts is a function of the hydrogen pressure, and
the former one exhibits the first to zero order
dependence on the hydrogen pressure while a
second order dependence on the hydrogen pres-
sure is observed in the case of osmium catalyst
w
x
7,25 . Thus, to determine the dependence of the
hydrogenation rate on the hydrogen pressure, a
series of experiments was carried out where the
hydrogen pressure was varied over the range of
1.36–7.01 MPa. The initial polymer concentra-
tion, catalyst concentration and reaction temper-
ature were kept constant. The hydrogen concen-
tration in chlorobenzene was calculated based
on the hydrogen solubility in chlorobenzene
w
Ž . Ž
. Ž
.
xq
The Rh H PPh3 2 NC species then reacts
2
w
Ž .
with olefin to form compound Rh H -
2
w
x
Ž
.Ž
. Ž
.
xq
w
Eq. 3 . This complex
Ž .x
reported by Parent and Rempel 26 . The results
shown in Fig. 4 demonstrate that the rate of
hydrogenation is pseudo-first order with respect
to the hydrogen pressure.
C5C PPh3 2 NC
can undergo further intramolecular addition of a
hydride to the coordinated olefinic bonds and
the reductive elimination of hydrogenated NBR
to give species Rh PPh3 2 NC q, which subse-
quently reacts with nitrile and H2 to regenerate
w
Ž
. Ž
.x
3.3.4. Reaction mechanism and rate law
w
Ž
.
w
Ž . Ž
. Ž
.
2 xq
Ž
Scheme
It is well known that complexes Rh diene -
the species, Rh H PPh3 2 NC
.
2 .
2
Ž
.
xq
PPh3 2
as catalyst precursors readily react

(
)
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
131
step in this case and the corresponding rate law
Ž .
is given in Eq. 4 .
d C5C
skX C5C
dt
q
sk Rh H PPh NC
.
C5C
Ž
.2Ž
3.Ž
2
4
Ž .
w
x
w x
Halpern 23 and Landis and Halpern 27
have demonstrated that, in the catalytic hydro-
genation of simple olefins using cationic
rhodium complexes, the rate-determining step
depends on the reaction conditions. When the
catalytic reaction is carried out at low hydrogen
Ž . Ž .
In combination of the following Eqs. 5 – 8 ,
qx
w
Ž
.Ž
. Ž
.
the concentration of Rh H2 PPh3 2 NC
can
be converted to the total concentration of
w
x
w
x
rhodium Rh , nitrile concentration CN , and
T
w
x
hydrogen concentration H2 and, by substitut-
Ž
.
pressure 1 atm and room temperature, the
oxidative addition of H2 is rate determining.
However, the reductive elimination of final
product becomes rate determining at tempera-
ture lower than y408C. In contrast to small
olefin hydrogenation, Rempel and coworkers
Ž .
ing it into Eq. 4 , the pseudo-first-order rate
X
Ž
.
Ž
Ž ..
constant k is obtained Eq. 9 .
q
2
Rh s Rh H PPh
NC
Ž
.2Ž
3.2 Ž
.
T
q
q Rh H PPh
NC
Ž
.2Ž
3.2 Ž
.
q
w
x
5,7,21,22,28 have demonstrated that, in the
hydrogenation of diene polymers at elevated
temperatures and hydrogen pressures by
q Rh PPh
NC
NC
Ž
3.2 Ž
3.2 Ž
.
.
q
2
q Rh PPh
5
Ž .
Ž
Ž
.
Rh PPh3 3Cl, the olefin coordination step is
typically rate determining, which is attributed to
the steric demands of the polymer that inhibits
an olefin double bond from accessing the metal
center. Thus, the olefin coordination step shown
q
Rh H PPh
NC
NC
Ž
.2Ž
3.2 Ž
.
K1 s
6
Ž .
q
2
Rh H PPh
NC
Ž
.2Ž
3.2 Ž
.
q
2
Rh PPh
NC
Ž
3.2 Ž
.
K3 s
7
Ž .
q
Ž .
in Eq. 3 is assumed as the rate-determining
Rh PPh
NC
NC
Ž
3.2 Ž
.
q
2
Rh H PPh
NC
Ž
.2Ž
3.2 Ž
.
K4 s
8
Ž .
q
Rh PPh
NC
H2
Ž
3.2 Ž
.
2
w
w
xw x
x
k2 K1 K3 K4 H2 Rh
T
KX s
9
Ž .
w
x
w
xw x
1qK3 CN qK1 K3 K4 H2 qK3 K4 CN H2
The Gauss–Newton method for nonlinear
least-squares regression is used for the statistical
analysis of variance to verify the validity of the
w
x
model 29 . The results obtained are listed in
Table 3 and the reaction parameters K1, K3,
K4, and k2 generated are given in Table 4 along
with the 95% confidence ranges. The values of
parameters K1, K3, K4 and k2 are comparable
to the similar parameters found for the NBR
Ž
.
w x
hydrogenation with RhCl PPh3 3 catalyst 7 .
The residual plot is given in Fig. 5. Both a low
value for the mean sum of squares of residues
ww
x
w
x
Fig. 6. Arrhenius plot. Rh T s240 mM; CN s95.5 mM; PH
2
.
Ž
. Ž
Ø Observed data; —
s2.46 MPa; T s373.2 Ky393.2 K.
regression value.

(
)
132
T.-F. Mao, G.L. RempelrJournal of Molecular Catalysis A: Chemical 135 1998 121–132
w x
3
K. Hashimoto, N. Watanabe, A. Yoshioka, Rubber World
shown in Table 3 and the random scattering of
errors shown in Fig. 5 indicates that the ob-
served kinetic data are consistent with the model.
Ž
.
190 1984 32.
w x
4
G.L. Rempel, N.T. McManus, J.S. Parent, U.S. Patent
Ž
.
5561197 1996 .
w x
5
Ž
.
N.A. Mohammadi, G.L. Rempel, Macromolecules 20 1987
2362.
3.3.5. Temperature dependence
A series of experiments were carried out over
the temperature range of 100–1208C, where the
w x
6
S. Bhattacharjee, A.K. Bhowmick, B.N. Avasthi, Ind. Eng.
Ž
.
Chem. Res. 30 1991 1086.
w x
7
J.S. Parent, N.T. McManus, G.L. Rempel, Ind. Eng. Chem.
Ž
.
Res. 35 1996 4417.
Ž
.
catalyst concentration 0.240 mM , the polymer
w x
8
Ž
.
R.R. Schrock, J.A. Osborn, J. Am. Chem. Soc. 98 1976
2134.
Ž
.
concentration 153 mM , and hydrogen pressure
2.46 MPa remained constant. Fig. 6 shows an
Ž
.
w x
9
R.H. Crabtree, A. Gautier, G. Giordano, T. Khan, J.
Ž
.
Organomet. Chem. 141 1977 113.
Arrhenius plot, from which the activation en-
w
w
x
10 G.W. Parshall, S.D. Ittel, Homogeneous Catalysis, Wiley-In-
1
ergy EA of 60.7 kJ moly was obtained. This
terscience, New York, 1992, pp. 25–39.
x
value is in good agreement with those found for
11 R.S. Dickson, Homogeneous Catalysis with Compounds of
y
1
Rhodium and Iridium, D. Reidel Publishing, Dordrecht, 1985,
pp. 62–67.
Ž
.
Ž
.
Ž
.
Rh PPh3 3Cl 73.5 kJ mol
and RhH PPh3 4
y
1
Ž
.
57.4 kJ mol , indicating that the reaction is
w
x
Ž
.
12 R.R. Schrock, J.A. Osborn, J. Am. Chem. Soc. 93 1971
under chemical control and the diffusion of the
reactants is not a rate determining factor under
these reaction conditions. Based on the corre-
sponding Eyring plot, the apparent activation
2397.
w
w
x
Ž
.
13 D.C. Mudalige, G.L. Rempel, J. Mol. Catal. 116 1997 309.
x
14 N.A. Mohammadi, G.L. Rempel, Comput. Chem. Eng. 11
Ž
.
1987 27.
w
w
w
x
15 P. Martin, N.T. McManus, G.L. Rempel, J. Mol. Catal. A
1
enthalpy and entropy obtained are 57.5 kJ moly
Ž
.
126 1997 115.
1
1
and y154.5 J moly Ky
.
x
16 A.J. Marshall, I.R. Jobe, T. Dee, C. Taylor, Rubber Chem.
Ž
.
Technol. 63 1990 244.
x
Ž
.
17 R. Ugo, in: C. Manfredi Ed. , Aspects of Homogeneous
Catalysis, Milano, 1970, pp. 3–75.
w
w
x
Ž
.
18 J. Halpern, Phosphorus Sulfur 18 1983 307.
x
Acknowledgements
19 J.M. Brown, P.A. Chaloner, A.G. Kent, B.A. Murrer, P.N.
Nicholson, D. Parker, P.J. Sidebottom, J. Organomet. Chem.
Ž
.
216 1981 263.
We gratefully acknowledge the support of the
Natural Sciences and Engineering Research
Council of Canada. We thank Dr. Neil T. Mc-
Manus and Mr. Stuart Kehl for all their assis-
tance in carrying out some of the initial kinetic
experiments and in maintaining the equipment.
w
x
20 J.R. Shapley, R.R. Schrock, J.A. Osborn, J. Am. Chem. Soc.
91 1971 2816.
21 X.Y. Guo, G.L. Rempel, J. Mol. Catal. 63 1990 279.
Ž
.
w
w
x
x
Ž
.
Ž
.
22 N.A. Mohammadi, G.L. Rempel, J. Mol. Catal. 50 1989
259.
w
w
x
x
Ž
.
23 J. Halpern, Inorg. Chim. Acta 50 1981 11.
24 J. Halpern, D.P. Riley, A.S.C. Chan, J.J. Pluth, J. Am. Chem.
Ž
.
Soc. 99 1977 8055.
w
w
w
x
25 J.S. Parent, PhD Thesis, University of Waterloo, 1996.
x
Ž
.
Ž
26 J.S. Parent, G.L. Rempel, J. Chem. Eng. Data 41 1996 192.
References
x
.
27 C.R. Landis, J. Halpern, J. Am. Chem. Soc. 109 1987
1746.
w x
1
N.T. McManus, G.L. Rempel, J. Macromol. Sci. Rev.
Macromol. Chem. Phys. C 35 1995 239.
w
w
x
Ž
.
28 X.Y. Guo, P.J. Scott, G.L. Rempel, J. Mol. Catal. 72 1992
Ž
.
193.
w x
2
M.P. McGrath, E.D. Sall, S.J. Tremont, Chem. Rev. 95
x
Ž
.
29 SAS Statistical Analysis System User’s Guide, SAS Insti-
tute, Cary, NC, 1985.
Ž
.
1995 381.