Evidence for isomerizing hydroformylation of butadiene. A combined experimental and computational study

Article abstract of DOI:10.1016/j.molcata.2016.08.021

The (DIOP)rhodium-catalyzed hydroformylation of butadiene has been shown to give among the highest selectivities for formation of adipaldehyde, which is useful for the synthesis of nylon. Herein, isomerizing hydroformylation is shown to be a mechanism that is partially responsible for this selectivity and density functional theory studies are used to reveal the detailed pathway for the requisite alkene isomerization.

Full text of DOI:10.1016/j.molcata.2016.08.021

Journal of Molecular Catalysis A: Chemical 424 (2016) 145–152
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Evidence for isomerizing hydroformylation of butadiene. A combined
experimental and computational study
Tapan Majia,b, Camina H. Mendisa,b, Ward H. Thompsona,b,∗, Jon A. Tungea,b,∗
a Department of Chemistry, University of Kansas, Lawrence, KS 66045, United States
b Center for Environmentally Beneficial Catalysis, University of Kansas, Lawrence, KS 66047, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 11 June 2016
Received in revised form 27 July 2016
Accepted 21 August 2016
Available online 28 August 2016
The (DIOP)rhodium-catalyzed hydroformylation of butadiene has been shown to give among the highest
selectivities for formation of adipaldehyde, which is useful for the synthesis of nylon. Herein, isomerizing
hydroformylation is shown to be a mechanism that is partially responsible for this selectivity and density
functional theory studies are used to reveal the detailed pathway for the requisite alkene isomerization.
© 2016 Elsevier B.V. All rights reserved.
Keywords:
Hydroformylation
Carbonylation
Alkene
Isomerization
Rhodium
1. Introduction
acid is subsequently polymerized with hexamethylenediamine
(HMDA) to form nylon-6,6. HMDA is predominantly produced by
Transition-metal-catalyzed hydroformylation of olefins is one
of the most important homogeneous processes for industrial pro-
duction of aldehydes [1]. Hydroformylation, also known as the “oxo
process” was discovered by Otto Roelen in 1938 and subsequently
became a highly efficient and widely utilized industrial process for
the production of oxo chemicals globally [2]. The resulting alde-
hydes and aldehyde-derived compounds such as alcohols, amines,
carboxylic acids, and esters are also equally valuable for the synthe-
sis of bulk chemicals [3]. As such, hydroformylation of butadiene
has been suggested as an ideal process for the production of adi-
paldehyde.
Adipaldehyde is a potentially valuable intermediate for produc-
ing adipic acid and hexamethylene diamine (HMDA) which are key
monomers for nylon-6,6 synthesis [4,5]. Currently, nylon-6,6 syn-
thesis relies mainly on two processes [6–12]. First, adipic acid is
synthesized by the oxidation of KA oil, utilizing harsh oxidizing
agents such as concentrated sulfuric and nitric acids. These strong
acids produce large quantities of N2O gas as a side product, which
is a significant environmental concern [6–8]. The resulting adipic
Ni-catalyzed hydrocyanation, which requires extremely careful
handling of toxic HCN gas [9–12].
On the basis of the environmental concerns associated with
these processes, hydroformylation has emerged as a desirable
alternative for adipic acid and HMDA synthesis. Interestingly,
both monomers can be synthesized from a common adipaldehyde
intermediate by oxidation and reductive amination. Moreover,
reduction of adipaldehyde to hexane-1,6-diol would produce
a third high-valued monomer for the synthesis of polyesters
(Scheme 1) [1,2,13–15]. Therefore, the efficient synthesis of ali-
paldehyde via hydroformylation of butadiene has significant
potential to become industrially valuable. [16–23]
Several groups have reported the atom-economic synthesis of
adipaldehyde via the hydroformylation of 1,3-butadiene [16–23].
However, the reactions suffer from limited selectivity to the desired
adipaldehyde. For example, the Ohgomori group reported 31%
selectivity of adipaldehyde with a catalyst formed from rhodium(I)
and the commercially available DIOP ligand [13]. In 2011, the
Hofmann group reported that selectivities of up to 50% could
be achieved with rhodium(I) combined with bulky bisphosphite
ligands [17]. More recently, they have utilized isomerizing hydro-
formylation to achieve up to 73% yield of an adipaldehyde bis-acetal
derivative [24]. However, for industrial practicality the selectivity
and activity of the catalysts needs further improvement.
∗
Corresponding authors at: Department of Chemistry, University of Kansas,
Lawrence, KS 66045, United States.
E-mail addresses: wthompson@ku.edu (W.H. Thompson), tunge@ku.edu
(J.A. Tunge).
http://dx.doi.org/10.1016/j.molcata.2016.08.021
1381-1169/© 2016 Elsevier B.V. All rights reserved.

146
T. Maji et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 145–152
Scheme 1. Versatile utility of adipaldehyde.
The hydroformylation of butadiene to produce adipaldehyde
exhibits significant selectivity for adipaldehyde, it was of interest
to study whether phosphine-ligated rhodium can engage in isomer-
izing hydroformylation.
Herein, we report the results of a combined experimental
and computational study of the hydroformylation-isomerization
of butadiene. It is shown that the (DIOP)Rh catalyst does indeed
facilitate conversion of pent-3-enal to adipaldehyde via isomeriz-
ing hydroformylation, albeit to a limited degree. Density functional
theory studies are then used to illuminate the mechanistic features
of the critical olefin isomerization.
requires two sequential hydroformylation reactions. Ideally, buta-
diene would react with syn-gas via 1,2-addition to produce
pent-4-enal (Scheme 2), which is known to undergo rapid hydro-
formylation to form adipaldehyde [16,17]. However, the poor
selectivity in hydroformylation of butadiene typically arises from
the predominance of 1,4-addition in the first hydroformylation
to produce pent-3-enal (Scheme 2). This product is preferen-
tially hydroformylated to produce undesired branched aldehydes
and also undergoes hydrogenation. Unfortunately, the formation
of pent-3-enal is usually strongly favored both kinetically and
thermodynamically. For example, dppe-ligated rhodium provides
pent-3-enal with 94% selectivity [16]. Since pent-3-enal can be
formed with high efficiency, we became curious whether a cata-
lyst, or catalysts, could effect the isomerization-hydroformylation
of pent-3-enal to form adipaldehyde (“our approach”, Scheme 2)
[25–27]. Such a process would require the development of a cata-
lyst that is highly active for isomerization and highly selective for
hydroformylation of the terminal olefin of pent-4-enal.
According to the previous report of the Ohgomori group [16],
there is no indication of this type of kinetically controlled in situ iso-
merization of pent-3-enal to pent-4-enal. Hofmann, however, has
recently found evidence for such a kinetic pathway with rohodium
phosphite-catalyzed bis-hydroformylation of butadiene [24], which
has been studied extensively (computationally and experimen-
tally). Since the related reaction with the simple DIOP ligand has not
been explored in detail from a mechanistic point of view, though it
2. Experimental
Hydroformylation reactions were conducted using a Parr Series
5000 Multiple Reactor System. 1,3-Butadiene was obtained as a
solution (20 wt.% in toluene) from Sigma-Aldrich. A typical hydro-
formylation of butadiene is performed in the following manner.
Rh(acac)(CO)2 (4.77 mg, 0.0184 mmol), DIOP (46 mg, 0.092 mmol)
and 5 g of butadiene solution (20 wt.% in toluene) were added
to an autoclave in the glove box. The autoclave was sealed and
removed from the glove box. The autoclave was flushed by syn-
gas (CO/H2 = 1:1) at 5–6 bar pressure. The procedure was repeated
three times at the same pressure to ensure the complete replace-
ment of the Ar by syngas, and then charged by the syngas to
20 bar. The solution was stirred by 1000 rpm while being heated
at 80 ◦C. The heating was continued for 3 h. After 3 h the auto-
clave was cooled to room temperature and the pressure was
Scheme 2. In situ isomerization-hydroformylation concept.

T. Maji et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 145–152
147
Table 1
Isomerizing hydroformylation of pent-3-enal.
Entry
CO/H2 (bar)
Temp. (◦C)
Conv. (%)b
%8 formed
1
2
3
4
5
6
7
6
10
20
37
20
20
20
80
80
80
80
70
85(5)
85(5)
85(5)
100
85(5)
100
7.75
6.92
6.59
8.52
4.07
2.74
2.78
90
110
100
aReaction conditions: pent-3-enal (19.5 mmol) in toluene, Rh(acac)(CO)2 (0.056 mmol), DIOP (0.11 mmol), syngas (CO:H2) = 1:1.
berror is due to slight overlap of GC peaks for 3 and n-pentanal.
slowly released into a gas outlet valve inside a fumehood to avoid
exposure to toxic CO. An aliquot was removed for immediate GC
analysis and the product concentrations were determined from
comparison to calibration curves. Pent-3-enal was produced by
similar treatment of 15 g of butadiene solution (20 wt.% in toluene)
with Rh(acac)(CO)2 (14.33 mg, 0.056 mmol) and diphenylphosphi-
noethane (dppe, 22.12 mg, 0.056 mmol) for 4 h and purified by
distillation at 150–160 ◦C. The colorless solution obtained by this
way was analyzed by GC and the concentration was determined
from the calibration curve. The solution can be stored under argon
at low temperature (–20 ◦C) for 2-3 days. A 5.5 g solution of pent-3-
enal in toluene (19.52 mmol) was then treated with Rh(acac)(CO)2
(14.33 mg, 0.056 mmol), DIOP (55.17 mg, 0.11 mmol) at various
temperatures and pressures outlined in Table 1. After 3 h, an aliquot
(150 mg) was removed from the reaction mixture and added to a
10 mL volumetric flask followed by 30 mg of decane and diluted to
10 mL with toluene. The sample was then analyzed by GC and the
product concentrations were calculated from comparison to cal-
ibration curves. Gas chromatographic analysis was performed on
Shimadzu QP2010 SE gas chromatograph equipped with a mass
selective detector (GC–MS) using helium as carrier gas. A SHRXI-
5MS (30 m, 0.25 mm ID, 0.25 um df) capillary column was used. The
He flow rate was kept at 0.92 mL/min. The column temperature was
initially held at 60 ◦C for 2 min, then ramped at 10 ◦C/min to 200 ◦C
and held at this temperature for 5 min. Retention times (min) of the
selected products are as follows: 1,3-butadiene (1.62), trans-pent-
2-enal (2.71), trans-pent-3-enal (2.52), pent-4-enal (2.43), pentanal
(2.55), 2-methylpentanedial (5.87), adipaldehyde (6.95).
(6) (Scheme 3). Hydroformylation of 1,3-butadiene can, in principal,
produce a number of products including the desired adipaldehyde
[17], but the catalytic (DIOP)Rh system only generated few of them
(viz. 3, 4, 6 in Scheme 3). This minimal number of byproducts
adds to our belief that proper mechanistic investigation followed
by rational DIOP/condition modification can increase the selectiv-
ity for adipaldehyde (8). Additionally, the factors providing 32%
selectivity of adipaldehyde by DIOP ligand are still unclear, and a
detailed mechanistic study of 1,3-butadiene hydroformylation by
the (DIOP)Rh catalyst is lacking.
3.2. Mechanistic investigation with DIOP ligand
The commonly accepted mechanism for 1,3-butadiene hydro-
formylation by the (DIOP)Rh catalyst was proposed by the
Ohgomori group [16], but needs further support from theoretical
and experimental investigations. The observed product distribu-
tion suggests that the first hydroformylation of 1,3-butadiene takes
place by the incorporation of active catalyst Rh-H via the 1,2-
and 1,4-addition pathways (Scheme 3). The 1,2-addition pathway
yields pent-4-enal (7) via intermediate 2, which is analogous to
the hydroformylation of non-conjugated/isolated terminal olefins.
The subsequent n-selective hydroformylation of the resulting pent-
4-enal (7) provides the desired adipaldehyde (8). Independent
hydroformylation of commercially available pent-4-enal (7) with
(DIOP)Rh catalyst by us, and (bisphosphite)Rh by others [17,18],
supports the linear hydroformylation preference to adipaldehyde
(Scheme 4). Alternatively, the 1,4-addition of Rh-H proceeds via a
stable ␩3-crotyl complex (1, Scheme 3) and results in pent-3-enal
(3), which upon a second hydroformylation produces dialdehyde
4 or 5 (while both are possible, interestingly only 4 is observed).
Additionally, pentanal (6) formation can be rationalized by direct
hydrogenation of 3 or 7 or via the double bond isomerization of
pent-3-enal (3) to the conjugated pent-2-enal followed by rapid
hydrogenation by the (DIOP)Rh catalyst [28,29].
Finally, we hypothesizedthat pent-3-enal (3) could also undergo
double bond isomerization to the non-conjugated pent-4-enal (7),
which is then rapidly hydroformylated to adipaldehyde (Scheme 2).
There is currently no reported evidence in the literature for this
in situ isomerization-hydroformylation of pent-3-enal (3) to adi-
paldehyde (8) with phosphine-ligated rhodium catalysts. This
prompted us to investigate the possibility of the isomerization-
hydroformylation of pent-3-enal to adipaldehyde, which we now
discuss.
3. Results and discussion
3.1. Benchmarked 1,3-butadiene hydroformylation
We started our experimental findings by establishing the bench-
mark hydroformylation reaction of 1,3-butadiene with (DIOP)Rh
catalyst as reported by Ohgomori [16]. A toluene solution of buta-
diene, Rh(acac)(CO)2, and DIOP ligand were exposed to syn-gas
(CO/H2 = 1:1) in a parallel Parr reactor at elevated tempera-
ture for several hours. The optimized conditions 18.51 mmol
of 1,3-butadiene in 5 mL toluene, Rh(acac)(CO)2 (0.1 mol%, DIOP
(0.5 mol%), 80 ◦C, syn-gas (20 bar), 3 h) resulted in the full con-
sumption of 1,3-butadiene with 32% selectivity to adipaldehyde,
together with some undesired side products. The major side prod-
ucts were pent-3-enal (3) and the hydrogenation product pentanal

148
T. Maji et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 145–152
Scheme 3. (DIOP)Rh-catalyzed 1,3-butadiene hydroformylation products.
3.3. Isomerization-hydroformylation of Pent-3-enal (3)
3.4. Computational results and discussion
To investigate the possibility that adipaldehyde arises from
isomerization-hydroformylation, we started with freshly prepared
toluene solutions of pent-3-enal (3) and subjected them to different
hydroformylation conditions.
In recent years density functional theory (DFT) calculations
have been used to examine the reaction mechanisms of vari-
ous transition-metal catalyzed isomerizations of alkenes yielding
results in good agreement with experimental studies [30–32]. Thus,
in order to better understand the experimental results presented
above, we undertook a DFT study of the mechanism of isomer-
ization of pent-3-enal (3) to pent-4-enal (7) and pent-2-enal. All
DFT calculations reported in this work were carried out using the
NWChem 6.5 quantum chemical software [33]. Structures were
fully optimized using the M06-L functional [34] with the 6–31g*
basis set on all atoms except Rh, for which the LANL2DZ effec-
tive core potential was used [35]. Harmonic frequency calculations
were carried out to identify the stationary points as minima or tran-
sition states and to estimate the zero-point vibrational energy. To
examine the isomerization of pent-3-enal to its isomers (Scheme 5)
catalyzed by (DIOP)Rh, calculations were carried out for (a) hydride
transfer from the (DIOP)Rh catalyst to pent-3-enal, and (b) ␤-
hydride elimination of the resulting intermediates to produce the
isomerization products, i.e., the desired pent-4-enal and the unde-
sired pent-2-enal.
In the (DIOP)Rh-catalyzed isomerization mechanism pathway
(Scheme 5), pent-3-enal first coordinates to the Rh center via the
alkene bond. Then, two possible isomerization pathways can occur
from this reactant complex that differ in the destination of the
hydride transferring from the catalyst to the bound pent-3-enal.
In path a the hydride on the catalyst transfers to the C3 carbon
of pent-3-enal resulting in the Inta intermediate. Then ␤-hydride
elimination from the C5 carbon chain of pent-3-enal will result in
the desired pent-4-enal. Alternatively, Inta can undergo ␤-hydride
elimination from the C3 carbon to regenerate pent-3-enal, the
reactant. In path b the hydride on the catalyst transfers to the C4 car-
bon of pent-3-enal to form the Intb intermediate. Then ␤-hydride
elimination from the C2 carbon of pent-3-enal will result in the
undesired pent-2-enal, while ␤-hydride elimination from the C4
carbon will reform pent-3-enal, the reactant.
As depicted in Table 1, a significant amount of adipaldehyde,
8 was formed by hydroformylation of pent-3-enal (3) under var-
ious conditions. The effect of the syn-gas (CO/H2 = 1:1) pressure
was explored by varying it from 6 to 37 bar while keeping the
temperature constant (Table 1, entries 1–4). These experiments
demonstrate that the syn-gas pressure has only a minor influ-
ence on the selectivity for adipaldehyde. In contrast, varying the
temperature (entries 5–7) decreases the aldipaldehyde selectivity
compared to the reaction at 80 ◦C. In all cases pentanal (6) was
the major byproduct that was observed. We note that it is diffi-
cult to quantitatively determine how much pentanal is produced at
incomplete conversion, since both pent-3-enal and pentanal have
similar retention times using our gas chromatographic method.
However, 1H NMR spectra of the crude reaction mixture of entry 3
(after 3 h) showed the presence of pentanal (6) and adipaldehyde
(8) in addition to starting pent-3-enal (3). Similar NMR spectro-
scopic analysis after 24 h showed complete consumption of 3 to
adipaldehyde and pentanal. Importantly, the derivation of sig-
nificant quantities of adipaldehyde from pent-3-enal shows that
isomerizing hydroformylation is kinetically competent to occur
during the hydroformylation of butadiene. Therefore, the forma-
tion of adipaldehyde (8) from hydroformylation of butadiene can
be rationalized by a “normal” 1,2-addition pathway combined with
a 1,4-addition hydroformylation-isomerization pathway involving
the thermodynamically unfavorable isomerization of pent-3-enal
(3) to pent-4-enal (7) followed by rapid linear selective hydro-
formylation to 8.
The calculated zero-point energy-corrected electronic energy
profiles for the two pathways are shown in Fig. 1 and the corre-
sponding structures of the stationary points are presented in Fig. 2.
All reported electronic energies are relative to the reactant complex
R (pent-3-enal coordinated to the (DIOP)Rh catalyst). We note that
the reactant complex is a saturated 18 valence electron (VE) species
with a trigonal bipyramidal structure. The minimum energy struc-
ture has the anionic hydride in the axial position and pent-3-enal is
coordinated to the Rh center in the equatorial plane. The structure
Scheme 4. Hydroformylation of pent-4-enal.

T. Maji et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 145–152
149
Scheme 5. (DIOP)Rh-catalyzed isomerization mechanism.
of the most stable conformer is shown in Fig. 2 (and the coordinates
are given in the Supporting information).
a 16 VE complex with a distorted square planar geometry (shown
in Fig. 2) and is stabilized relative to the bound pent-3-enal by
10.3 kcal/mol.
The green curve (R → TS1a → Inta → TS2a → pent-4-enal) in
Fig. 1 represents the energy profile for isomerization of pent-3-
As noted above, two ␤-hydride elimination steps are possible
from Inta. Elimination at the C3 carbon represents a return over the
TS1a barrier to the pent-3-enal reactant. While this route is unpro-
ductive, it is likely important in the reaction kinetics as it ties up
both a (DIOP)Rh catalyst and a substrate molecule. Due to the lower
energy of Inta, the barrier to reform pent-3-enal is 22.0 kcal/mol. If
the hydride transfer instead occurs from the C5 carbon, the desired
pent-4-enal is formed through the transition state TS2a. This bar-
enal to the desired pent-4-enal. The first transition state, TS1a
,
corresponds to hydride transfer from Rh to the C3 carbon of pent-3-
enal and has a barrier of ꢀE‡ = 11.7 kcal/mol. This hydride transfer
involves the rotation of the coordinated pent-3-enal out of the
equatorial plane. In the TS1a transition state structure (Fig. 2), the
transferring hydride sits equidistant between the Rh and C3 car-
bon, with Rh H and H C3 distances of 1.67 Å. The resulting Inta is
TS2a
19.8
20.0
19.0
TS1a
15.0
TS2b
11.7
H
5
P
10.0
Rh
11.6
3
P
4
CO
TS1b
5.0
1
2
CHO
0.0
R
0.5
pent-4-enal
0.0
5
Inta
-5.0
-10.0
-15.0
H
H
5
4
P
P
-10.3
4
P
P
3
Rh
1
Rh
CHO
1
-13.1
3
CO
2
CHO
2
pentC-2O-enal
-12.6
pent-3-enal
Intb
Reaction Coordinate
Fig. 1. Calculated zero point-corrected electronic energy profile for isomerization of pent-3-enal to pent-4-enal in green and isomerization of pent-3-enal to pent-2-enal in
red. All energies are reported in kcal/mol with respect to the stable conformation of pent-3-enal coordinated to the Rh center (R). (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article.)

150
T. Maji et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 145–152
Fig. 2. Stationary point structures for paths a and b from the DFT calculations. First row: pent-3-enal (R), pent-4-enal, and pent-2-enal. Second row: TS1a, Inta, and TS2a. Third
row: TS1b, Intb, and TS2b. Geometry parameters are shown in Table S1.
rier is higher than that to regenerate pent-3-enal, with a barrier of
30.1 kcal/mol relative to Inta. As with TS1a the structure of TS2a has
the hydride directly between the Rh and C5 atoms with effectively
the same Rh H and H C5 distances in both cases of 1.67 Å.
The red curve (R → TS1b → Intb → TS2b → pent-2-enal) in Fig. 1
represents the energy profile for isomerization of pent-3-enal to
undesired pent-2-enal. As described above, TS1b, corresponds to
hydride transfer from Rh to the C4 carbon of pent-3-enal, which
requires the rotation of the coordinated pent-3-enal out of the
equatorial plane, and has a barrier of ꢀE‡ = 11.6 kcal/mol, In TS1b
(Fig. 2) the transferring hydride sits equidistant between Rh and
the C4 carbon with Rh H and H C4 distances of 1.67 Å. Similar to
the path a intermediate Inta, Intb is also a 16 VE species with a dis-
torted square planar geometry and is stabilized by 12.6 kcal/mol
relative to R.
As described in Scheme 5, Intb can also undergo two ␤-hydride
elimination steps. Elimination at the C4 carbon leads to the forma-
tion of pent-3-enal (R) via TS1b with a barrier of 24.2 kcal/mol, a
step that is important in the overall reaction kinetics. If the hydride
transfer occurs from the C2 carbon, the undesired isomer pent-2-
enal is formed through the transition state TS2b with a barrier of
31.6 kcal/mol. This barrier is higher than the barrier to form Intb
from R, and thus ␤-hydride elimination from the Intb is the rate-
culations were carried out with toluene as the solvent. [36] The
reaction energy profile is shown in Fig. S1 of the Supporting infor-
mation. The results indicate that toluene solvent has only a minor
effect on the reaction energetics. It does not change the relative
ordering of the barriers with respect to formation of pent-4-enal
and pent-2-enal. The key difference with respect to the gas phase
results shown in Fig. 1 is that in the weakly polar toluene solvent
the barriers are slightly larger, by ∼1.2–4.6 kcal/mol. These small
changes upon incorporating solvation indicate that gas phase cal-
culations represent the relevant processes well.
It is interesting to compare the isomerization pathway of pent-
3-enal to pent-4-enal (Fig. 1, green curve) with that for pent-3-enal
isomerization to pent-2-enal (Fig. 1, red curve). The rate-limiting
step in both cases is the second transition state (TS2a, and TS2b
,
respectively) from the intermediate to the pentenal product. The
barriers for these two pathways are similar, 30.1 kcal/mol to
form pent-4-enal compared to 31.6 kcal/mol to form pent-2-enal.
However, both are larger than that for converting the respec-
tive intermediate back to pent-3-enal, 22 kcal/mol for Inta and
24.2 kcal/mol for Intb. Thus, the calculations suggest that the rate-
limiting barrier for pent-4-enal formation is less than that for
pent-2-enal. However, the kinetics should be strongly influenced
by the formation and decomposition of the intermediates Inta and
Intb, which involve substantially lower barriers.
limiting step. Further, as with all the other transition states (TS1a
,
TS2a, and TS1b), in the TS2b structure the transferring hydride has
identical Rh—H and H C2 distances of 1.67 Å.
To model the solvent effects on the isomerization of pent-3-enal
reaction, solvation model based on density (SMD) single-point cal-
The experimental results indicate that the isomerization of
pent-3-enal in the presence of the (DIOP)Rh catalyst occurs to form
both pent-4-enal and pent-2-enal, with the latter the major prod-
uct. An additional consideration in examining these results is that

T. Maji et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 145–152
[4] G. Bellussi, C. Perego, Industrial catalytic aspects of the synthesis of
151
pent-2-enal is the most stable isomer and is thus the thermody-
namically favored isomer. This is enhanced upon binding to the
(DIOP)Rh catalyst, which further stabilizes the pent-2-enal iso-
mer compared to the (DIOP)Rh-bound pent-3-enal and pent-4-enal
forms. Moreover, pent-2-enal bound to the (DIOP)Rh catalyst is
stabilized by 13.1 and 13.6 kcal/mol relative to the catalyst-bound
pent-3-enal and pent-4-enal, respectively (Fig. 1). These relative
stabilities are modified compared to the isolated pentenals, where
pent-2-enal is 5.0 and 9.3 kcal/mol more stable than pent-3-enal
and pent-4-enal, respectively. The (DIOP)Rh-bound pent-2-enal
appears to be further favored over the other two isomers due to
favorable interactions of the aldehyde group with a ligand phenyl
ring.
monomers for nylon production, Cattech 4 (2000) 4–16.
[5] R. Franke, D. Selent, A. Bo¨rner, Applied hydroformylation, Chem. Rev. 112
(2012) 5675–5732.
[6] S. Van de Vyver, Y. Román-Leshkov, Emerging catalytic processes for the
production of adipic acid, Catal. Sci. Technol. 3 (2013) 1465–1479.
[7] Teles, B. Roessler, R. Pinkos, T. Genger, T. Preiss, Method for Producing
Cyclododecanone. WO2005030689A2, May 26, (2005).
[8] B.D. Herzog, R.A. Smiley, Ullmann’s Encyclopedia of Industrial Chemistry,
Wiley-VCH Weinheim, Germany, 2000.
[9] P.W.N.M. Van Leeuwen, J. Frauendiener, R. Penrose Kluwer, Homogenous
Catalysis Understanding the Art, Academic Publishers, Dordrecht, 2004, pp.
231.
[10] M.E. Tauchert, D.C. 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.
[11] P. Hofmann, M.E. Tauchert, Molecular Catalysis, Wiley-VCH Weinheim,
Germany, 2014, pp. 161.
[12] A.P. Göthlich, M. Tensfeldt, H. Rothfuss, M.E. Tauchert, D. Haap, F. Rominger, P.
Hofmann, Novel chelating phosphonite ligands: syntheses, structures, and
nickel-catalyzed hydrocyanation of olefins, Organometallics 27 (2008)
2189–2200.
[13] K.-S. Chen, T.L. Yu, Y.-S. Chen, T.-L. Lin, W.-J. Liu, Soft-and hard-segment phase
segregation of polyester-based polyurethane, J. Polym. Res. 8 (2001) 99–109.
[14] G. Teng, F.N. Jones, End-grafting of oligoesters based on terephthalic acid and
linear diols for high solids coatings, J. Appl Polym. Sci. 60 (1996) 1609–1618.
[15] T. Kiyotsukuri, N. Takemoto, N. Tsutsumi, W. Sakai, M. Nagata, Regular
network polyesters from benzenepolycarboxylic acids and glycol, Polymer 36
(1995) 5045–5049.
After isomerization, pent-4-enal can undergo a hydroformyla-
tion to form adipaldehyde; this is expected to be rapid but must
compete with reformation of Inta via TS2a as shown in Fig. 1. In
contrast, the hydroformylation of pent-2-enal is expected to be
slower. Fig. 1 also indicates that return to the intermediate (Intb
via TS2b) should be less likely for pent-2-enal due to the substan-
tially larger barrier of 32.1 kcal/mol (compared to the same process
for pent-4-enal). On the other hand, pent-2-enal can also undergo
a hydrogenation reaction to form pentanal side product.
[16] Y. Ohgomori, N. Suzuki, N. Sumitani, Formation of 1 6-hexanedial via
hydroformylation of 1, 3-butadiene, J. Mol. Catal. A: Chem. 133 (1998)
289–291.
[17] S.E. Smith, T. Rosendahl, P. Hofmann, Toward the rhodium-catalyzed
bis-hydroformylation of 1, 3-Butadiene to adipic aldehyde, Organometallics
30 (2011) 3643–3651.
[18] J. Mormul, M. Mulzer, T. Rosendahl, F. Rominger, M. Limbach, P. Hofmann,
Synthesis of adipic aldehyde by n-selective hydroformylation of 4-pentenal,
Organometallics 34 (2015) 4102–4108.
[19] J. Briggs, D. Packett, D. Bryant, A. Phillips, D. Schreck, K. Olson, E. Tjaden, A.
Guram, T. Eisenschmid, E. Bragham, Process for Producing Hydroxyaldehydes,
WO199740001A1 Oct. 30, (1997).
[20] C. Botteghi, M. Branca, A. Saba, Optically active aldehydes via
hydroformylation of 1, 3-dienes with chiral diphosphinerhodium complexes,
J. Organomet. Chem. 184 (1980) C17–C19.
[21] B. Fell, P. Hermanns, H. Bahrmann, J. Prakt. Chem. Ztg. 340 (1998) 459–467.
[22] H. Bahrmann, B. Fell, The hydroformylation of conjugated dienes VI tertiary
aryl-and arylalkyl-phosphines and secondary alkyl-and arylphosphines as
ligands in the rhodium catalyzed hydroformylation reaction of conjugated
dienes to dialdehydes, J. Mol. Catal. 8 (1980) 329–337.
4. Conclusions
Isomerization of pent-3-enal to pent-4-enal, leading to adi-
paldehyde formation through subsequent hydroformylation, has
been demonstrated using the (DIOP)Rh catalyst. The thermo-
dynamically favored isomer, pent-2-enal, is also formed and
represents the major isomerization product; pent-2-enal is iden-
tified through the products of its subsequent and hydrogenation.
While the yield of pent-4-enal is low, the results support the
notion that its hydroformylation is rapid and thus an isomeriza-
tion catalyst may be used to direct the branching of butadiene
hydroformylation toward adipaldehyde formation.
A DFT study of the reaction pathways has been carried out to
better understand the experimental results. The energy profiles
for pent-3-enal isomerization to both pent-2-enal and pent-4-enal
have been determined. They indicate that the barriers to form the
two isomers are similar. The lowest barriers are for forming inter-
mediates, that can then lead to pent-2-enal or pent-4-enal, by
hydride transfer from the catalyst to pent-3-enal and its reverse
reaction. Thus, the overall reaction kinetics are expected to depend
not only on the barriers to isomerization alone, but the equilibria
between (DIOP)Rh-bound pent-3-enal and these intermediates.
[23] D.L. Packett, Hydroformylation Process for Producing 1,6-Hexandials, U.S.
Patent 5,312,996, May 17, (1994).
[24] J. Mormul, J. Breitenfeld, O. Trapp, R. Paciello, T. Schaub, P. Hofmann, Synthesis
of adipic acid, 1, 6-hexanediamine, and 1 6-hexanediol via double-n-selective
hydroformylation of 1, 3-butadiene, ACS Catal. 6 (2016) 2802–2810.
[25] J. Ternel, J.L. Couturier, J.L. Dubois, J.F. Carpentier, Rhodium-catalyzed tandem
isomerization/hydroformylation of the bio-sourced 10-undecenenitrile:
selective and productive catalysts for production of polyamide-12 precursor,
Adv. Synth. Catal. 355 (2013) 3191–3204.
[26] A. Behr, D. Obst, A. Westfechtel, Isomerizing hydroformylation of fatty acid
esters: formation of ␻-aldehydes, Eur. J. Lipid Sci Technol. 107 (2005)
213–219.
Acknowledgement
[27] Y. Yuki, K. Takahashi, Y. Tanaka, K. Nozaki, Tandem
isomerization/hydroformylation/hydrogenation of internal alkenes to
n-alcohols using Rh/Ru dual-or ternary-catalyst systems, J. Am. Chem. Soc.
135 (2013) 17393–17400.
We thank the NSF-EPA Networks for Sustainable Materials
Design and Synthesis (NSMDS) under NSF Grant CHE-1339661 for
funding.
[28] C.J. Scheuermann née, C. Taylor, Jaekel, Enantioselective hydrogenation of
enones with a hydroformylation catalyst, Adv. Synth. Catal. 350 (2008)
2708–2714.
[29] T. Ohshima, H. Tadaoka, K. Hori, N. Sayo, K. Mashima, Highly enantio-and
S-trans CC bond selective catalytic hydrogenation of cyclic enones:
alternative synthesis of (−)-Menthol, Chem. Eur. J. 14 (2008) 2060–2066.
[30] M.A.n. Carvajal, S. Kozuch, S. Shaik, Factors controlling the selective
hydroformylation of internal alkenes to linear aldehydes 1. The isomerization
step, Organometallics 28 (2009) 3656–3665.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molcata.2016.08.
021.
[31] J. Tao, F. Sun, T. Fang, Mechanism of alkene isomerization by bifunctional
ruthenium catalyst: a theoretical study, J. Organomet. Chem. 698 (2012) 1–6.
[32] S. Biswas, Z. Huang, Y. Choliy, D.Y. Wang, M. Brookhart, K. Krogh-Jespersen,
A.S. Goldman, Olefin isomerization by iridium pincer catalysts. experimental
evidence for an ␩3-Allyl pathway and an unconventional mechanism
predicted by DFT calculations, J. Am. Chem. Soc. 134 (2012) 13276–13295.
[33] M. Valiev, E.J. Bylaska, N. Govind, K. Kowalski, T.P. Straatsma, H.J. Van Dam, D.
Wang, J. Nieplocha, E. Apra, T.L. Windus, N.W. Chem, A comprehensive and
scalable open-source solution for large scale molecular simulations, Comp.
Phys. Comm. 181 (2010) 1477–1489.
References
[1] H.-J. Arpe, Industrielle Organische Chemie, Bedeutende Vor- und
Zwischenprodukte, Wiley-VCH Weinheim, 2007, pp. 138.
[2] H.-J. Arpe, K. Weissermel, Industrial Organic Chemistry, Wiley-VCH
Weinheim, 2010.
[3] K. Weissermel, H.-J. Arpe, Industrial Organic Chemistry, 4th ed., Wiley-VCH
Weinheim, Germany, 2003.

152
T. Maji et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 145–152
[34] Y. Zhao, D.G. Truhlar, The M06 suite of density functionals for main group
thermochemistry, thermochemical kinetics, noncovalent interactions, excited
states, and transition elements: two new functionals and systematic testing
of four M06-class functionals and 12 other functionals, Theor. Chem. Acc. 120
(2008) 215–241.
calculations illuminate the role of long-range noncovalent interactions,
Organometallics 33 (2014) 4183–4191.
[36] A.V. Marenich, C.J. Cramer, D.G. Truhlar, Universal solvation model based on
solute electron density and on a continuum model of the solvent defined by
the bulk dielectric constant and atomic surface tensions, J. Phys. Chem. B 113
(2009) 6378–6396.
[35] M. Kumar, R.V. Chaudhari, B. Subramaniam, T.A. Jackson, Ligand effects on the
regioselectivity of rhodium-catalyzed hydroformylation: density functional