Rational design of efficient steric catalyst for isomerization of 2-methyl-3-butenenitrile

Article abstract of DOI:10.1016/j.mcat.2020.111259

The catalytic isomerization of 2-methyl-3-butenenitrile (2M3BN), a model reaction in the DuPont process, has been performed using NiL₄ (L=tri-O-p-tolyl phosphite) as a catalyst. The lowered catalytic activity in the isomerization with coexistence of 2-pentenenitrile (2PN) and 2-methyl-2-butenenitrile (2M2BN) indicates that both 2PN and 2M2BN are the catalyst inhibitors, and the quantitative relationship between the conversion of 2M3BN and the content of 2M2BN and 2PN is provided. DFT calculation results suggest that the inhibition effect is attributed to the generation of dead-end intermediates (2PN)NiL₂ and (2M2BN)NiL₂, both of which take nickel atom out of the catalytic cycle in the isomerization process. To suppress the inhibition effect, new catalytic intermediates are rationally designed based on their computational %V_bur. An efficient method that adding extra ligand 1, 5-bis(diphenylphosphino)pentane (dppp5) to the NiL₄ catalyst is selected experimentally. Compared to the results obtained with NiL₄ as catalyst, the (dppp5)NiL₂ increases the conversion of 2M3BN from 74.5 % to 93.4 % at 3 h of reaction and provides a high selectivity to 3PN (> 98 %) at optimal conditions.

Full text of DOI:10.1016/j.mcat.2020.111259

Molecular Catalysis 498 (2020) 111259
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.journals.elsevier.com/molecular-catalysis
Rational design of efficient steric catalyst for isomerization of
2
-methyl-3-butenenitrile
Kaikai Liu, Tiefeng Wang*, Minghan Han*
Department of Chemical Engineering, Tsinghua University, Beijing, 100084, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
The catalytic isomerization of 2-methyl-3-butenenitrile (2M3BN), a model reaction in the DuPont process, has
Isomerization
been performed using NiL
4
(L=tri-O-p-tolyl phosphite) as a catalyst. The lowered catalytic activity in the
2
2
-Pentenenitrile
isomerization with coexistence of 2-pentenenitrile (2PN) and 2-methyl-2-butenenitrile (2M2BN) indicates that
both 2PN and 2M2BN are the catalyst inhibitors, and the quantitative relationship between the conversion of
-Methyl-2-butenenitrile
Density functional theory calculatoins
Steric effect
2
M3BN and the content of 2M2BN and 2PN is provided. DFT calculation results suggest that the inhibition effect
is attributed to the generation of dead-end intermediates (2PN)NiL and (2M2BN)NiL , both of which take nickel
2
2
atom out of the catalytic cycle in the isomerization process. To suppress the inhibition effect, new catalytic
intermediates are rationally designed based on their computational %Vbur. An efficient method that adding extra
ligand 1, 5-bis(diphenylphosphino)pentane (dppp5) to the NiL
the results obtained with NiL as catalyst, the (dppp5)NiL increases the conversion of 2M3BN from 74.5 % to
3.4 % at 3 h of reaction and provides a high selectivity to 3PN (> 98 %) at optimal conditions.
4
catalyst is selected experimentally. Compared to
4
2
9
1
. Introduction
The synthesis of adiponitrile (AdN) from 1, 3-butadiene and
Due to its importance of the isomerization reaction for the DuPont
process, it has been extensively studied [1,3]. The ligand effects [13] are
well-understood thanks to the systematic and meticulous work of Tol-
man et al., such as the electronic factor [14] and steric factor [15]. These
two factors are powerful approaches for explanation of the influence of
phosphorus (III) ligands on the catalytic performance, and they are also
considered as the basis for rational design of better catalysts. Typically,
the catalyst performance depends on the ligand used, and extensive
work has been done to optimize the ligands for the isomerization reac-
hydrogen cyanide (HCN) catalyzed by Ni(0)-phosphite complex, which
is known as the DuPont process, has received extensive attention [1–3].
In general, the DuPont process includes three steps : the initial hydro-
cyanation of 1, 3-butadiene to form a mixture of 2-methyl-3-buteneni-
trile (2M3BN) and 3-pentenenitrile (3PN), the isomerization of
–
–
2
M3BN to 3PN, and the subsequent C
second hydrocyanation to product AdN. Among these steps, the isom-
erization of 2M3BN to 3PN via a C CN bond breaking and forming
C double bond migration and a
tion [1,3]. Lewis acids such as ZnCl
[16]. Solvents have a significant effect on the selectivity to 3PN, and
nonpolar solvents favor C CN cleavage in 2M3BN [17]. The key
2
are found to accelerate this reaction
–
reaction is a key step because it governs the economic profit of the
DuPont process. In addition, there is a close relationship between the
hydrocyanation and isomerization reactions on the typical catalysts
–
development in the isomerization is to investigate the bidentate ligands,
such as diphosphines and diphosphites [6,7,9,10,18,19]. These biden-
such as NiL
4
(L = tri-O-p-tolyl phosphite) [4]. To facilitate the devel-
tate ligands can significantly enhance the catalytic activity. Moreover,
3
opment of the catalyst used in the DuPont process, the isomerization has
been widely used as a model reaction for catalyst evaluation [5–8] and
mechanistic investigation [9–12].
based on the crystal structure of Ni(
η
-1-Me-C
3
H
4
)(CN)(dppb), a reac-
tion mechanism of the isomerization has been proposed [9], which in-
volves five steps : (i) coordination of 2M3BN with nickel via its C–
–
C
Abbreviations: 2M3BN, 2-methyl-3-butenenitrile; 3PN, 3-pentenenitrile; 2PN, 2-pentenenitrile; 2M2BN, 2-methyl-2-butenenitrile; 4PN, 4-pentenenitrile; AdN,
adiponitrile; HCN, hydrogen cyanide; COD, 1, 5-cyclooctadiene; L, tri-O-p-tolyl phosphite; mL, tri-O-m-tolyl phosphite; oL, tri-O-o-tolyl phosphite; dppe, 1,2-bis
(
diphenylphosphino)ethane; dppp3, 1,3-bis(diphenylphosphino)propane; dppb, 1, 4-bis(diphenylphosphino)butane; dppp5, 1, 5-bis(diphenylphosphino)pentane;
′
dpph, 1,6-bis(diphenylphosphino)hexane; dppf, 1,1 -bis(diphenyphosphino)ferrocene; Xantphos, 9,9-dimethyl-4,5-bis(diphenylphosphino)xanthene; XantN, 4,6-Bis
(
diphenylphosphino)phenoxazine.
*
Corresponding authors.
E-mail addresses: wangtf@tsinghua.edu.cn (T. Wang), hanmh@tsinghua.edu.cn (M. Han).
https://doi.org/10.1016/j.mcat.2020.111259
Received 16 August 2020; Received in revised form 5 October 2020; Accepted 7 October 2020
Available online 21 October 2020
2
468-8231/© 2020 Elsevier B.V. All rights reserved.

K. Liu et al.
Molecular Catalysis 498 (2020) 111259
bond, (ii) C
–
CN bond breaking and formation of a
σ
-allyl species, (iii)
octane) were purchased from Aladdin. These nitriles were used as
received without purification. Gas chromatograph (GC, Purui Instru-
ment Company, Beijing) was used to analyze the product mixture and
the ReactIR15 technology (METTLER TOLEDO) was employed to
isomerization to the
π
-allyl species, (iv) -allyl formation, and (v) C
σ
–
CN
bond coupling to 3PN coordination. More recently, some mechanistic
details [8,11,12], including oxidative addition of 2M3BN, conforma-
tional rearrangement and reduction elimination forming 3PN, have been
discussed based on DFT calculations [20,21].
3
1
monitor the kinetic information of the isomerization process. P NMR
spectra were recorded using BRUKER 600 MHz nuclear magnetic reso-
nance spectrometer with cumene as solvent. And X-ray data was
collected on a Rigaku SuperNova AtlasS2 diffractometer equipped with
an Atlas CCD detector (see Supplementary material).
In the DuPont process, the conversion of 2M3BN to 2-methyl-2-bute-
nenitrile (2M2BN) and that of 3PN to 2-pentenenitrile (2PN) inevitably
occur as side reactions, as shown in Scheme 1, because the conjugated
2
M2BN and 2PN are more stable thermodynamically. Therefore, 2M2BN
The isomerization reactions were performed in a glove box as follows
and 2PN always exist in the DuPont process. These by-products reduce
the 3PN yield and they cannot be converted to AdN by subsequent
hydrocyanation. Using dppb [8,9] or dppf [6] as ligand can provide a
high selectivity to 3PN (>98 %). However, expensive organic nickel
4 2
(also called common conditions): 1 mmol NiL catalyst, 5 mmol ZnCl , 1
mmol ligand, and 10 mL cumene were added to a three-necked flask and
were thoroughly mixed. The reaction system was kept at 100 ℃ for 30
min, then 2M3BN (72.8 % 2M3BN or 90 % 2M3BN, see Supplementary
material) was added to the above system. Samplings (0.1 mL) were taken
out every hour, which were centrifuged and analyzed using GC. To test
2
complex such as Ni(COD) (COD : 1, 5-cyclooctadiene) must be used as
the nickel source, which limits their industrial applications. Therefore, it
is desired to improve the catalytic selectivity to 3PN in a cheaper way.
Despite of the importance of this topic, there are limited researches
concerning the formation of 2M2BN and 2PN. In general, commercially
available 2M3BN contains ~30 % other nitriles (see Supplementary
material). It is unclear how the presence of 2M2BN and 2PN changes the
catalytic activity and selectivity. In addition, it is difficult to remove
4
the effect of 2M2BN and 2PN on the activity of the NiL catalyst, a
certain amount of 2M2BN or 2PN was added to 90 % 2M3BN to obtain
2M3BN with different contents of 2M2BN or 2PN. Herein, to get the
mixed nitriles containing 20 % 2M2BN, 0.65 mL 2M2BN is added to 3.23
mL 90 % 2M3BN. Similarly, 0.80 mL 2PN is added to 3.26 mL 90 %
2M3BN to provide the mixed nitriles containing 20 % 2PN. (see Sup-
plementary material)
2
M2BN and 2PN from 2M3BN by traditional distillation because they
have very close boiling points (2M2BN: 115~120℃, 2M3BN: 124℃,
PN: 127℃). Hence, it is urgently required to figure out the negative
2
2.2. Computational methods
effects of undesired 2M2BN and 2PN and develop an efficient strategy to
increase the selectivity to 3PN in the isomerization of 2M3BN.
All calculations were carried out using the Gaussian 09 program
[22]. The geometry optimizations and calculations of the zero-point
vibrational energies (ZPVEs) were performed in the gas phase using
the B3LYP [23,24] functional with the SDD [25] basis set for Ni and
6ꢀ 31 G(d) [26] basis set for C, H, N, O and P. The optimized structures
were verified by the calculated vibrational frequencies. No imaginary
frequencies were found for the optimized structures that correspond to
local minima on the potential energy surface. On the basis of the
gas-phase optimized geometries, the solvation effect was incorporated
with the SMD solvent model [27]. The solution-phase Gibbs free energy
was calculated by adding the correction of the Gibbs free energy in the
gas phase to the electronic energy in the solvent. All energies discussed
in this work were Gibbs free energies. The buried volume values %Vbur
of the nitriles and intermediates were calculated using SambVca 2.0 [28,
In this work, the isomerization of 2M3BN to 3PN using NiL
catalyst was studied as the model reaction. It was found that 2M2BN and
PN had a significant inhibition effect on the activity of the NiL cata-
4
as the
2
4
lyst. DFT calculations were carried out to understand this inhibition
effect. A catalytic mechanism was proposed, in which the inhibition by
2
M2BN and 2PN was attributed to the formation of (2M2BN)NiL
2PN)NiL that made the NiL catalyst off-cycle of the reaction. Based on
the structural differences of phosphorus ligands, suitable ligands were
used to modify the NiL catalyst to increase its selectivity to 3PN. Due to
the steric effect, the modification of the NiL catalyst by dppp5 effec-
2
and
(
2
4
4
4
tively suppressed the inhibition effect of 2M2BN and 2PN in the 2M3BN
isomerization reaction, increasing the 2M3BN conversion from 75.4 %
to 93.5 % and the 3PN selectivity from 90 % to 98 %.
2
9]. The crystal structures were drawn by the Ortep3 program [30], and
2
. Materials and methods
.1. Experimental methods
All the catalyst (NiL ) synthesis experiments were performed with
the CYLview [31] was employed to represent the intermediates.
2
3. Results and discussion
4
3.1. Inhibition effect of 2M2BN and 2PN
standard Schlenk techniques and all the isomerization reactions were
performed in glove box techniques under dry nitrogen with the con-
centration of oxygen and water below 0.1 ppm, and the entire operation
process does not involve water and oxygen. Nitiriles (2M3BN, 3PN, 2PN,
and 2M2BN) were purchased from J&K, ligands were purchased from
To clarify the effect of 2M2BN and 2PN on the activity of the NiL4
catalyst, the isomerization of 2M3BN was used as the model reaction.
The 2M3BN of 90 % purity is known as the purest, although it contains
7% 2M2BN and 3% 2PN. For each experiment, after the isomerization
reaction proceeded for 4 h, the nitriles were carefully distilled from the
mixture under reduced pressure (10 kPa) and 60 ℃, then a second
aliquot of 2M3BN was added to the system. As shown in Fig. 1, the
conversion of 2M3BN was 74.0 % at 1 h when using 90.0 % 2M3BN as
reactant (Fig. 1(a)), while it decreased to 35.5 % when 72.8 % 2M3BN
was used (Fig. 1(b)), and it took 3 h that the conversion reached to 74.5
3 2 2
Laajoo, and the other chemicals (Ni(NO ) , ZnCl , cumene, toluene, and
%
. The concentrations of 2M2BN and 2PN were only the factor that
varied, while other factors remained unchanged. These results showed
that 2M2BN and 2PN had an inhibition effect on the NiL catalyst in the
M3BN isomerization reaction. In addition, the catalytic activity further
dropped as the aliquot of 2M3BN ((a): 90 % 2M3BN, (b): 72.8 %
4
2
2
M3BN) was added to the reaction system.
Next, a varied amount of 2M2BN or 2PN was added to the 90 %
Scheme 1. The eight nitriles revolved in the reaction system considering cis (C-
and trans (T-) isomers.
2M3BN to further study which one or both have the effect on the cata-
lytic activity of the NiL catalyst. As shown in Fig. 1(c), the reaction rate
)
4
2

K. Liu et al.
Molecular Catalysis 498 (2020) 111259
[
a]
Fig. 1. The catalytic results of NiL
4
using different 2M3BN as reactants.
(a) 90.0 % 2M3BN, (b) 72.8 % 2M3BN, (c) 2M2BN and (d) 2PN were added to the 90 %
2
M3BN, and (e) and (f) dppp5 was added to reaction system.
[
a]
For (a), (b), (c) and (d), NiL
4
: L : ZnCl
2
: 2M3BN = 1 : 1 : 5 : 30, T = 100 ℃, and 10 mL cumene; for (e), NiL : dppp5 : 2M3BN = 1 : 1 : 30, T = 100 ℃, and 10 mL
4
cumene; for (f), NiL
4
: dppp5 : 2M3BN = 1 : 1 : 30, T = 110 ℃, and 10 mL cumene.
decreased significantly as the amount of 2M2BN was increased from 7%
to 50 %. A rapid decrease in catalytic activity was also observed with the
addition of 2PN (Fig. 1(d)). Apparently, both 2M2BN and 2PN are cat-
alytic inhibitors. The formations of 2M2BN and 2PN not only decreased
calculated. Using the energy of (2M3BN)NiL
calculated relative energies of X-NiL are depicted in Fig. 2. The lowest
energy of 2PN binding to the NiL catalyst, which is -11.0 kcal/mol for
(C2PN)NiL and -13.1 kcal/mol for (T2PN)NiL , indicates that both
C2PN and T2PN are preferentially bound to the NiL catalyst to form the
most stable intermediates (C2PN)NiL and (T2PN)NiL . However, these
intermediates are not reactive because no conversion of 2PN is detected
when adding 2PN to the NiL catalytic system at the same reaction
2
as the reference, the
2
4
2
2
the yield of 3PN, but also decreased the activity of the NiL
4
catalyst in
4
the isomerization of 2M3BN. Moreover, 2PN showed a stronger inhibi-
tion effect than 2M2BN, as revealed by the experimental results with the
addition of 20 % 2PN or 20 % 2M2BN.
2
2
4
conditions (see Supplementary material). Thus the decrease in catalytic
activity in Fig. 1(d) could be attributed to the increased formation of the
3
.2. Explanation of the inhibition effect
stable (2PN)NiL
2M2BN could also bind to NiL
(2M2BN)NiL intermediate, which also decreased the catalytic activity
(Fig. 1(c)). Similarly, no conversion of 2M2BN was detected when
2M2BN was added to the NiL catalytic system (see Supplementary ma-
terial). Because (2M2BN)NiL had a higher energy than (2PN)NiL , 2PN
2
intermediate. In addition, the most stable by-product
To further understand the inhibition effect, the relative energies of
4
and produce the relatively stable
the eight nitriles and their X-NiL
2
complex have been determined by DFT
2
calculations, as shown in Fig. 2. The reactant 2M3BN is selected as the
reference, and the relative energies of 2PN and 2M2BN are negative
4
–
–
N moiety. Thus, the
–
because of the conjugation of C
–
C bond with C
2
2
formation of 2PN and 2M2BN is more favorable thermodynamically,
especially for branched 2M2BN with the lowest energy. In contrast,
showed a stronger inhibition effect on the catalytic activity than 2M2BN,
which was consistent with the experimental results in Fig. 1.
2
M3BN and 3PN are non-conjugated isomers with a higher energy and
Using the data shown in Fig. 1 (c) and (b), the quantitative rela-
tionship between the conversion of 2M3BN and the content of by-
products (2M2BN and 2PN) has been determined through multivariate
regression (see Supplementary material), as shown in Formula (1). Based
show higher reactivity.
Next, the energies of all the detected nitriles combined with the NiL
4
2
catalyst, forming corresponding intermediates X-NiL (X: nitriles), were
2
Fig. 2. The relative energies of nitriles and X-NiL (X: nitriles).
(
No C3PN was detected in the isomerization reaction experiment.)
3

K. Liu et al.
Molecular Catalysis 498 (2020) 111259
on this formula, the calculated conversion of commercially available
is replaced by 2M2BN or 2PN, resulting in a catalytic inhibition; (3) at
the end of the isomerization reaction, 2M2BN or 2PN replaces 3PN
forming dead-end intermediates. Compared to the previous results on
2
M3BN (72.8 %) is 32.5 %, which is very close to the experimentally
measured conversion (35.5 %). In addition, the quantitative relationship
also provides useful insights for understanding the inhibiting effects. As
2
the deactivation caused by the formation of Ni(CN) via the reaction of
2
2
M2BN and 2PN present in the isomerization of 2M3BN, 2M2BN and
NiL and HCN [3], other new ways accounting for the lower catalytic
4
PN decrease the reaction rate constant (it is 1.4961 in Formula (1)
activity in the isomerization of 2M3BN are found in the current study. To
improve the catalytic activity in the isomerization of 2M3BN, it is
desired to increase the selectivity to 3PN to prevent the increase of
2M2BN and 2PN and suppress the combinations of 2M2BN and 2PN in
when no 2M2BN and 2PN exist in reaction system) during the reaction,
which means that the presence of 2M2BN and 2PN makes some catalyst
lose their catalytic ability. Here, what needs to be explained is that
Formula 1 is only used to quantitatively predict the catalytic activity of
the reaction system rather than the reaction kinetic equation.
4
2M3BN with the NiL catalyst.
1
ꢀ X
3.3. Suppressing the inhibition effect
ln
= 1.4961*eꢀ 0.04353*x2M2BN ꢀ 0.05964*x2PN
(1)
1
For catalytic isomerization of 2M3BN, the research on diphosphines
indicates that the space a ligand occupies around the Ni center is an
extremely important parameter for modulating the catalytic perfor-
mance and the optimal steric congestion improves the catalyst activity
X means the conversion of 2M3BN at 1 h, x2M2BN and x2PN represent the
percentage content of 2M2BN and 2PN in the isomerization process,
respectively.
Furthermore, under the same reaction conditions as used in Fig. 1(c)
[
8]. With this in mind, a reasonable solution is explored to suppress the
and (d), 0.65 mL 2M2BN was mixed with NiL
, then 3.23 mL 2M3BN was added so that the reaction system contained
0 % 2M2BN. The experimental results showed that the activity of NiL
4
and stirred for 1 h at 100
inhibition effect.
ꢀ
Several parameters have been proposed to quantify the steric
congestion, such as the Tolman cone angle (θ) [15], solid angle (Ω) [32]
and percent buried volume (%Vbur) [28,29]. To better describe the steric
environment conferred on Ni center by ligands, the %Vbur was employed
to calculate the percent of volume occupied by ligands around the Ni
center and was used to assist the selection of suitable ligands with the
appropriate %Vbur value. The nitriles are regarded as ligands, and the %
2
4
sharply decreased, and the conversion of 2M3BN was only 18.6 % at 4 h
while it was 83 % when 2M3BN including 20 % 2M2BN was used as
reactant (Fig. 1(c)). For 2PN, 0.80 mL 2PN was mixed with NiL
4
and
stirred for 1 h at 100 ℃, then 3.26 mL 2M3BN was added to have 20 %
2
2
PN in the reaction system for comparison, and the conversion of
M3BN was 10.8 % at 4 h while it was 77.6 % using 2M3BN including 20
V
bur of the intermediates X-NiL
2
in Fig. 2 are calculated. As shown in
have bigger %Vbur value. This
%
2PN as reactant as shown in Fig. 1(d). The decrease of reaction rate
indicates that 2M2BN or 2PN can bind directly with the NiL catalyst
and form (2M2BN)NiL and (2PN)NiL , respectively, when they are
mixed with the catalyst. In addition, the inhibition on the catalytic ac-
tivity is more noticeable as the NiL catalyst is reused employing 72.8 %
M3BN as reactant. In the second isomerization, 90 % conversion of 90
Table 1, (2M2BN)NiL and (2PN)NiL
2
2
4
prompts us that rationally designing or regulating the steric congestion
of the catalytic intermediate provides desirable method to inhibit the
2
2
combination of these by-products with the NiL
bulky ligands are used to prevent the internal coordination of 2M2BN or
PN.
On the basis of the %Vbur results, several ligands are selected,
4
catalyst. Therefore, the
4
2
2
%
2M3BN was obtained at 4 h (Fig. 1(a)) while the conversion was only
4
5 % at 4 h using 72.8 % 2M3BN as reactant (Fig. 1(b)). The reason for a
including mL, oL, dppe, dppp3, dppb, dppp5, dpph, dppf, Xantphos and
XantN, and their structures are depicted in Table 2. In order to give a
greater loss of activity showed in Fig. 1(b) was that more NiL catalyst
4
was combined with 2M2BN or 2PN in the reaction system. The lower
conversion of 2M3BN in the above situations is that the active nickel is
taken out of the productive catalytic cycle as dead-end intermediates
Table 1
a
The %VBur of X-NiL
2
(X: nitriles).
T2M2BN
83.3
(
2 2
(2M2BN)NiL and (2PN)NiL ). Here, three types of the formation of the
X
2M3BN
79.7
C2M2BN
83.7
T2PN
82.8
C2PN
82.3
T3PN
83.3
4PN
81.3
dead-end intermediates are proposed to understand the inhibiting effect
in mechanistic insights, as shown in Scheme 2: (1) 2M2BN or 2PN
%VBur
a
4
directly combines with NiL causing a lower catalytic activity; (2) at the
Bondi radii scaled by 1.17, 4.0 Å for the sphere radius and the mesh spacing
beginning of the isomerization reaction, 2M3BN bound with the catalyst
value s =0.1 Å.
2 2
Scheme 2. The ways of forming (2M2BN)NiL and (2PN)NiL intermediates.
4

K. Liu et al.
Molecular Catalysis 498 (2020) 111259
Table 2
The structure of ligands used and the %VBur value of (2M3BN)Ni(Ligand) intermediate.
Ligand
L
mL
oL
dppe
dppp3
dppb
Structure
%
%
V
V
Bur of partial substitution
Bur of total substitution
7
9.7
86.2
84.2
87.6
86.5
84.0
81.4
85.1
84.4
85.6
86.2
79.7
Ligand
dppp5
dpph
dppf
Xantphos
XantN
Structure
%
%
V
V
Bur of partial substitution
Bur of total substitution
8
6.2
85.6
89.3
–
85.2
–
87.2
–
87.6
87.5
larger value of %Vbur, the formation of the new catalytic intermediates
via partial or total substitution reactions is designed as these ligands are
conversion of 2M3BN : 74.5%–92% at 3 h, and the selectivity to 3PN :
90%–98%). However, adding dppe, dppf, Xantphos or XantN deceases
the 2M3BN conversion from 82 % to below 40 % at 4 h of reaction (see
Supplementary material). Diphosphines (dppf, Xantphos or XantN) with
rigid backbone failed to meet the requirements for improving the cata-
added to the NiL
4
system. Here, partial substitution means only one L in
(
2M3BN)NiL is replaced by the added ligands while total substitution
2
refers that both L are replaced. In order to clearly illustrate these two
situations, the corresponding structures with dppp5 ligands are shown
in Fig. 3. The %Vbur values of (2M3BN)Ni(Ligand) are used to evaluate
the steric effect of the ligands. After the substitution reaction, the %Vbur
of the catalytic intermediates increases from 79.7 % to 84.0–89.3 %.
Under the circumstance of partial substitution, (2M3BN)NiL(mL) and
2
lyst performance. In addition, Ni(COD) and ligands were used to
simulate the total substitution experimentally, and the catalytic results
were area C in Fig. 3, (2M3BN)Ni(dppp3), (2M3BN)Ni(dppb) and
(2M3BN)Ni(dppp5) with the %Vbur of 84.8 %, 86.2 % and 87.5 %,
respectively, provides 99 % selectivity to 3PN, but only (2M3BN)Ni
(dppb) offeres a high conversion (93.4 %) of 2M3BN. However, As
(
2M3BN)NiL(oL) provide bigger %Vbur (86.2 % and 87.6 %) while
2M3BN)NiL(diphosphine) gives close %Vbur values (dppe : 84.0 %,
(
2
mentioned earlier, expensive Ni(COD) limits its large-scale use. As for
dppp3 : 85.1 %, dppb : 85.6 %, dppp5 : 86.2 %, and dpph : 85.6 %), as
listed in Table 2. For total substitution, the increased %Vbur of (2M3BN)
Ni(Ligand) are also observed.
(2M3BN)Ni(dppe) and (2M3BN)Ni(dpph), a smaller (81.4 %) or larger
(89.3 %) %Vbur value would provide the negative consequences (see
4
Supplementary material). Therefore, a method for combining NiL with
To test the above strategy, the selected ligands are evaluated in the
dppp5 ligand to effectively improve the catalyst performance was
discovered.
4
NiL catalyzed isomerization reaction, and the results are shown in
Fig. 4. Although the addition of L increased the selectivity to 90 % and
thus hindered the formation of 2M2BN and 2PN, the excess ligand
4
Moreover, using dppp5 and NiL as the catalyst, the selectivity to
3PN could be maintained above 98 % when commercially available 72.8
% 2M3BN was added in 3 successive aliquots, and the catalytic activity
moderately decreased from 90 % to 71 %, as shown in Fig. 1(e). The
optimal experimental conditions (see Supplementary material) were
4
inhibited the dissociation of the L in NiL , resulting in a decrease in
catalytic activity (area A in Fig. 4). Using extra mono-ligands (mL or oL)
increased the selectivity to 3PN from 85 % to 92 %. Furthermore, the
3
PN selectivity was 90 % when the Ni(mL)
4
was employed as catalyst.
4
determined as follows: reaction temperature of 110 ℃ and NiL : dppp5 :
Recalling that the %Vbur increased from L (79.7 %) to oL (87.6 %), as
listed in Table 2, the improvement in 3PN selectivity could be attributed
to the increased steric effect. Further, diphosphines were added to the
2M3BN = 1 : 1 : 30. The 2M3BN conversion could be further increased to
93.4 % at 3 h at the optimal conditions and the selectivity to 3PN
maintained at 98 % in Fig. 1(f).
NiL
M3BN conversion and 3PN selectivity are shown in area B in Fig. 4. The
NiL catalyst with dppp5 ligand give a 92 % conversion of 2M3BN and >
8 % selectivity to 3PN, which is the best result in this work. Compared
with the experimental results using only the NiL catalyst (Fig. 1(b)),
adding dppp5 greatly improves the catalytic performance (the
4
catalytic system to improve the catalytic activity. The results of the
2
All our attempts to isolate (dppp5)NiL have failed, to gain pre-
2
liminary insights in the isomerization of 2M3BN, the mixture of NiL
4
and
◦
4
dppp5 in a 1 : 1 ratio was heated for 30 min at 100 C and then analyzed
9
by NMR. To better visualize the nickel species, NiL
Ni(COD)
and dppp5 were also monitored by NMR. The 31P NMR
spectrum were shown in Fig. 5. Chemical shifts of free P atoms in L and
4
, L and the mixture of
4
2
Fig. 3. The structure of the (dppp5)NiL (left) and the Ni(dppp5) (right).
5

K. Liu et al.
Molecular Catalysis 498 (2020) 111259
a]
[
Fig. 4. The catalytic results using different ligands
.
[
a]
area A is the result of the isomerization reaction of using
NiL combined with monodentate phosphorus ligands (L, mL,
and oL) as catalyst, area B is the result of the isomerization
reaction of using NiL and bidentate phosphine ligands (dppp3,
dppb, dppp5, and dpph) as a catalyst, and area C is the result of
using Ni(COD) and the bidentate phosphines (dppp3, dppb,
4
4
2
dppp5, and dpph) as catalyst in the catalyzed isomerization
reaction.
4
Fig. 6. The catalytic kinetics using NiL and dppp5 as catalyst.
Therefore, a first-order reaction means that the reaction is not catalyzed
by Ni(dppp5) (in Fig. 3). According to the NMR and in-situ IR results, the
improved result in Fig. 1(f) could be attributed to the formation of
(
dppp5)NiL
.4. The crystal structure
The crystal of the (dppp5)Ni(CN) species suitable for single-crystal X-
2
.
3
3
1
Fig. 5. P NMR spectra.
ray diffraction studies were obtained from the reacted system by vapor
diffusion using the hexane and cumene system. As shown in Fig. 7, the
structure of the (dppp5)Ni(CN) species depicts that the nickel atom is in
a planar configuration and is attached to four groups: two CN groups and
two P atoms in different dppp5 ligands. The parameters of the crystal
can be found in Supplementary material.
dppp5 were observed at 128.15 ppm and -13.50 ppm, respectively. The
signal at 130.40 ppm was assigned to the NiL complex. A new signal at
4.52 ppm indicated that only one phosphorus atom in dppp5 was
4
2
linked to nickel while the sharp signal appeared at 31.78 ppm when both
phosphorus atoms in dppp5 were coordinated to the nickel atom (see Ni
Although the combination of 2M2BN and 2PN with the nickel cata-
(
dppp5) in Fig. 3). Based on the peak area (see Supplementary material),
the nickel atom linked to one phosphorus atom in dppp5 was combined
with two L to form (dppp5)NiL
To study the changes of the NiL
4
lyst is suppressed with the coexistence of NiL and ligand dppp5, the
catalytic activity still decreases gradually under optimal conditions. The
conversion of 2M3BN is 65 %, 55.4 % and 43 % for the first, second and
third times at 1 h, respectively, as shown in Fig. 1(e). When (dppp5)Ni
2
.
4
catalyst after adding dppp5 in the
catalytic process, the kinetics of the 2M3BN isomerization was studied
by ReactIR15 (more details in Supplementary material). The results
showed that the reaction was first order with respect to 2M3BN, as
(
CN) is used as the catalyst to further investigate this phenomenon, no
activity is found in the isomerization reaction. (dppp5)Ni(CN) accounts
for the decrease of the catalytic activity in Fig. 1(e).
ꢀ
1
shown in Fig. 6. The calculated reaction rate constant (k) was 1.232 h
ꢀ
1
for 2M3BN and 1.211 h for 3PN, indicating that the selectivity to 3PN
is 98.3 %. As previously reported that when Ni(dppp5) was used as
catalyst, the isomerization reaction followed a zero-order kinetics [8].
4
. Conclusions
In summary, the current study reports the inhibition effect of 2PN
6

K. Liu et al.
Molecular Catalysis 498 (2020) 111259
Fig. 7. The crystal structures of (dppp5)Ni(CN) species.
and 2M2BN on the catalytic activity in the isomerization of 2M3BN and
provides an effective method to weaken the inhibition effect. The
quantitative relationship shows that 2PN and 2M2BN in the feedstock
are detrimental to the catalytic activity, and their inhibition effect is
elucidated by experimental studies and DFT calculations. The genera-
complexes in ionic liquid, J. Mol. Catal. A Chem. 246 (2006) 242–247, https://doi.
org/10.1016/j.molcata.2005.11.006.
6] A. Acosta-Ramírez, M. Mu n˜ oz-Hern ´a ndez, W.D. Jones, J.J. García, Isomerization of
[
[
2
-methyl-3-butenenitrile with (bis-diphenylphosphinoferrocene)nickel
compounds: catalytic and structural studies, J. Organomet. Chem. 691 (2006)
895–3901, https://doi.org/10.1016/j.jorganchem.2006.05.042.
3
7] M.E. Tauchert, D.C.M. Warth, S.M. Braun, I. Gruber, A. Ziesak, F. Rominger,
P. Hofmann, Highly efficient nickel-catalyzed 2-Methyl-3-butenenitrile
isomerization: applications and mechanistic studies employing the TTP ligand
family, Organometallics 30 (2011) 2790–2809, https://doi.org/10.1021/
om200164f.
2 2
tion of the dead-end intermediates (2PN)NiL and (2M2BN)NiL makes
nickel atom out of the productive catalytic cycle in the isomerization
process, which decreases the catalytic activity. Therefore, it is desired to
suppress the inhibition effect of 2PN and 2M2BN by rational catalyst
design. This is realized by increasing the steric effect of the catalytic
[
[
8] K. Liu, H. Xin, M. Han, Elucidation of key factors in nickel-diphosphines catalyzed
isomerization of 2-methyl-3-butenenitrile, J. Catal. 377 (2019) 13–19, https://doi.
org/10.1016/j.jcat.2019.07.016.
9] A. Chaumonnot, F. Lamy, S. Sabo-Etienne, B. Donnadieu, B. Chaudret, J.-
C. Barthelat, J.-C. Galland, Catalytic Isomerization of Cyanoolefins Involved in the
4
intermediates with the addition of dppp5 to the NiL catalytic system to
prevent 2PN and 2M2BN from binding to the catalyst. This simple
approach operation provides improved conversion of 2M3BN (75.4 %–
Adiponitrile Process. C-CN Bond Cleavage and Structure of the Nickel
π
-Allyl
3
Cyanide Complex Ni(
η
-1-Me-C H )(CN)(dppb), Organometallics 23 (2004)
3 4
9
3.4 %) and excellent selectivity to 3PN (90 %–98 %). Undoubtedly,
3
363–3365, https://doi.org/10.1021/om049706i.
using the NiL
4
combined with dppp5 is a better choice for the DuPont
[
[
10] J. Wilting, C. Müller, A.C. Hewat, D.D. Ellis, D.M. Tooke, A.L. Spek, D. Vogt,
Nickel-catalyzed isomerization of 2-methyl-3-butenenitrile, Organometallics 24
(2005) 13–15, https://doi.org/10.1021/om049206p.
process. Finally, based on the steric effect, a method for combining NiL
4
with other phosphine ligands to effectively improve the catalyst per-
formance is discovered. This method is suitable for any ligand, and
greatly accelerates the efficiency of ligand screening. In order to further
improve the conversion of 2M3BN in the isomerization process, more
ligands will be screened in our next research.
11] T. Li, W.D. Jones, DFT calculations of the isomerization of 2-Methyl-3-butenenitrile
by [Ni(bisphosphine)] in relation to the DuPont adiponitrile process,
Organometallics 30 (2011) 547–555, https://doi.org/10.1021/om100907y.
[12] K. Liu, K.-K. Liu, M.-J. Cheng, M.-H. Han, Theoretical and experimental study of
the nickel-catalyzed isomerization of 2-Methyl-3-butenenitrile and the effect of a
Lewis acid, J. Organomet. Chem. 822 (2016) 29–38, https://doi.org/10.1016/j.
jorganchem.2016.08.016.
Declaration of Competing Interest
The authors report no declarations of interest.
Acknowledgements
[13] C.A. Tolman, Steric effects of phosphorus ligands in organometallic chemistry and
homogeneous catalysis, Chem. Rev. 77 (1977) 313–348, https://doi.org/10.1021/
cr60307a002.
[
[
[
14] C.A. Tolman, Electron donor-acceptor properties of phosphorus ligands.
Substituent additivity, J. Am. Chem. Soc. 92 (1970) 2953–2956, https://doi.org/
1
0.1021/ja00713a006.
15] C.A. Tolman, Phosphorus ligand exchange equilibriums on zerovalent nickel.
Dominant role for steric effects, J. Am. Chem. Soc. 92 (1970) 2956–2965, https://
doi.org/10.1021/ja00713a007.
16] C.A. Tolman, W.C. Seidel, J.D. Druliner, P.J. Domaille, Catalytic hydrocyanation of
olefins by nickel(0) phosphite complexes-effects of Lewis acids, Organometallics 3
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
(
1984) 33–38, https://doi.org/10.1021/om00079a008.
Appendix A. Supplementary data
[17] B.D. Swartz, N.M. Reinartz, W.W. Brennessel, J.J. García, W.D. Jones, Solvent
effects and activation parameters in the competitive cleavage of C-CN and C-H
bonds in 2-Methyl-3-Butenenitrile using [(dippe)NiH]
2
, J. Am. Chem. Soc. 130
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.mcat.2020.111259.
(
2008) 8548–8554, https://doi.org/10.1021/ja8000216.
[18] M.J. Baker, P.G. Pringle, Chiral aryl diphosphites: a new class of ligands for
hydrocyanation catalysis, J. Chem. Soc. Chem. Commun. 0 (1991) 1292–1293,
https://doi.org/10.1039/C39910001292.
References
[
19] J.I. Van der Vlugt, A.C. Hewat, S. Neto, R. Sablong, A.M. Mills, M. Lutz, A.L. Spek,
C. Müller, D. Vogt, Sterically Demanding Diphosphonite Ligands-Synthesis and
Application in Nickel-Catalyzed Isomerization of 2-Methyl-3-Butenenitrile, Adv.
Synth. Catal. 346 (2004) 993–1003, https://doi.org/10.1002/adsc.200303240.
[
[
[
1] L. Bini, C. Müller, D. Vogt, Mechanistic studies on hydrocyanation reactions,
ChemCatChem 2 (2010) 590–608, https://doi.org/10.1002/cctc.201000034.
2] W.C. Seidel, C.A. Tolman, Homogeneous nickel-catalyzed olefin hydrocyanation,
Ann. N. Y. Acad. Sci. 415 (1983), 201-201.
3] C.A. Tolman, R.J. McKinney, W.C. Seidel, J.D. Druliner, W.R. Stevens,
Homogeneous nickel-catalyzed olefin hydrocyanation, Adv. Catal. 33 (1985) 1–46,
https://doi.org/10.1016/S0360-0564(08)60257-6.
[20] P. Hohenberg, W. Kohn, Inhomogenous electron gas, Phys. Rev. 136 (1964)
864–871, https://doi.org/10.1103/PhysRev.136.B864.
[21] W. Kohn, L.J. Sham, Self-consistent equations including exchange and correlation
effects, Phys. Rev. A 140 (1965) A1133–A1138, https://doi.org/10.1103/
PhysRev.140.A1133.
[
4] C.A. Tolman, W.C. Seidel, L.W. Gosser, Formation of three-coordinate nickel(0)
[22] Gaussian 09, E. Revision, M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.
A. Robb, J.R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson,
H. Nakatsuji, M. Caricato, X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng,
J.L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida,
4
complexes by phosphorus ligand dissociation from NiL , J. Am. Chem. Soc. 96
(
1974) 53–60, https://doi.org/10.1021/ja00808a009.
[
5] V. Lecocq, C.C. Santini, Y. Chauvin, J.-M. Basset, J.-C. Galland, Reactivity of 2-
methyl-3-butenenitrile in the presence of catalytic amounts of various cobalt
7

K. Liu et al.
T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J.A. Montgomery Jr., J.
Molecular Catalysis 498 (2020) 111259
[27] A.V. Marenich, C.J. Cramer, D.G. Truhlar, Performance of SM6, SM8, and SMD on
the SAMPL1 test set for the prediction of small-molecule solvation free energies,
J. Phys. Chem. B 113 (2009) 4538–4543, https://doi.org/10.1021/jp809094y.
[28] L. Falivene, R. Credendino, A. Poater, A. Petta, L. Serra, R. Oliva, V. Scarano,
L. Cavallo, SambVca 2. A web tool for analyzing catalytic pockets with topographic
steric maps, Organometallics 35 (2016) 2286–2293, https://doi.org/10.1021/acs.
organomet.6b00371.
E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.
N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J.
C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J.M. Millam, M. Klene, J.
E. Knox, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.
E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, R.
L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P. Salvador, J.J. Dannenberg,
S. Dapprich, A.D. Daniels, O. Farkas, J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.
J. Fox, Gaussian, Inc., Wallingford CT, 2015.
[29] L. Falivene, Z. Cao, A. Petta, L. Serra, A. Poater, R. Oliva, V. Scarano, L. Cavallo,
Towards the online computer-aided design of catalytic pockets, Nat. Chem. 11
(2019) 872–879.
[
23] A.D. Becke, Density-functional exchange-energy approximation with correct
asymptotic behavior, Phys. Rev. A 38 (1988) 3098–3100, https://doi.org/
[30] L.J. Farrugia, ORTEP-3 for Windows - a version of ORTEP-III with a Graphical User
Interface (GUI), J. Appl. Crystal. 30 (1997), https://doi.org/10.1107/
S0021889897003117, 565-565.
1
0.1103/PhysRevA.38.3098.
[
[
24] A.D. Becke, Density-functional thermochemistry. III. The role of exact exchange,
J. Chem. Phys. 98 (1993) 5648–5652, https://doi.org/10.1063/1.464913.
25] M. Dolg, U. Wedig, H. Stoll, H. Preuss, Energy-adjusted ab initio pseudopotentials
for the first row transition elements, J. Chem. Phys. 86 (1987) 866–872, https://
doi.org/10.1063/1.452288.
[31] C.Y. Legault, CYLview, Universit ´e de Sherbrooke, 2009. http://www.cylview.org.
[32] I.A. Guzei, M. Wendt, An improved method for the computation of ligand steric
effects based on solid angles, Dalton Trans. 33 (2006) 3991–3999, https://doi.org/
10.1039/B605102B.
[
26] J.-P. Blaudeau, M.P. McGrath, L.A. Curtiss, L. Radom, Extension of Gaussian-2 (G2)
theory to molecules containing third-row atoms K and Ca, J. Chem. Phys. 107
(
1997) 5016–5021, https://doi.org/10.1063/1.474865.
8