Tandem hydroformylation/isomerization/hydrogenation of bio-derived 1-arylbutadienes for the regioselective synthesis of branched aldehydes

Article abstract of DOI:10.1016/j.apcata.2020.117583

The rhodium-catalyzed hydroformylation of 1-arylbutadienes derived from lignocellulosic bio-resources has been carried out in toluene and green solvents. In the presence of DPPE and XANTPHOS ligands, a regioselective Markovnikov Rh-H insertion takes place resulting in branched aldehydes in high selectivity, which contrasts with previous results obtained from aliphatic conjugated dienes. Depending on the nature of the diphosphine ligand, conjugated enals or saturated aldehydes are obtained in good to excellent selectivity. The later have a potential interest for fragrance industry, as they are homologous to commercial fragrance ingredients.

Full text of DOI:10.1016/j.apcata.2020.117583

Journal Pre-proof
Tandem hydroformylation/isomerization/hydrogenation of bio-derived
1-arylbutadienes for the regioselective synthesis of branched aldehydes
Gabriel M. Vieira (Investigation), Artur V. Granato (Investigation),
Elena V. Gusevskaya<ce:contributor-role>Writing –review and
editing), Eduardo N. dos Santos (Writing - original
draft)<ce:contributor-role>Writing – review and editing)
(Supervision), Pierre H. Dixneuf<ce:contributor-role>Writing –
´
review and editing), Cedric
Fischmeister<ce:contributor-role>Writing – review and editing)
(Supervision), Christian Bruneau<ce:contributor-role>Writing –
review and editing) (Supervision)
PII:
S0926-860X(20)30176-9
DOI:
https://doi.org/10.1016/j.apcata.2020.117583
APCATA 117583
Reference:
To appear in:
Applied Catalysis A, General
Received Date:
Revised Date:
Accepted Date:
20 December 2019
15 April 2020
17 April 2020
Please cite this article as: Vieira GM, Granato AV, Gusevskaya EV, Santos ENd, Dixneuf PH,
Fischmeister C, Bruneau C, Tandem hydroformylation/isomerization/hydrogenation of
bio-derived 1-arylbutadienes for the regioselective synthesis of branched aldehydes, Applied
Catalysis A, General (2020), doi: https://doi.org/10.1016/j.apcata.2020.117583

This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.

Tandem hydroformylation/isomerization/hydrogenation of bio-derived 1-
arylbutadienes for the regioselective synthesis of branched aldehydes
Gabriel M. Vieiraa, Artur V. Granatoa, Elena V. Gusevskayaa, Eduardo N. dos
Santosa*nicolau@ufmg.br, Pierre H. Dixneufb, Cédric Fischmeisterb, Christian
Bruneaub* christian.bruneau@univ-rennes1.fr
a Departamento de Química – ICEx, Universidade Federal de Minas Gerais, Belo
Horizonte, Brazil
b Univ Rennes, CNRS, ISCR (Institut des Sciences Chimiques de Rennes) - UMR6226,
35000 Rennes, France.
;
Graphical abstrqact
Highlights
 Synthesis of 1-arylbutadienes from bio-renewable 2-propenylbenzenes
 Tandem hydroformylation/isomerization/hydrogenation of 1-arylbutadienes
 Branched aldehydes useful as fragrances in high selectivity
Abstract
The rhodium-catalyzed hydroformylation of 1-arylbutadienes derived from
lignocellulosic bio-resources has been carried out in toluene and green solvents.
In the presence of DPPE and XANTPHOS ligands, a regioselective Markovnikov
Rh-H insertion takes place resulting in branched aldehydes in high selectivity,
which contrasts with previous results obtained from aliphatic conjugated dienes.
Depending on the nature of the diphosphine ligand, conjugated enals or
saturated aldehydes are obtained in good to excellent selectivity. The later have
a potential interest for fragrance industry, as they are homologous to
commercial fragrance ingredients.
Keywords:
tandem catalysis, hydroformylation, fragrance
1. Introduction
1

Hydroformylation (oxo synthesis) is an efficient atom economical reaction,
widely employed in industry, both in commodity and fine chemicals sector.[1]
Formally, hydroformylation corresponds to the addition of a formyl group and a
hydride to a C-C double bond (C=C), as represented in Scheme 1 for an alpha-
olefin. The actual mechanism involves the combination of the olefin, carbon
monoxide and hydrogen in the coordination sphere of a hydridocarbonyl
transition metal complex. Although discovered long ago by Otto Rölen[2], this
reaction still offers practical and scientific challenges because the selectivity
control is highly dependent on the reactants, the catalytic system, and the
reaction conditions.
Insert Scheme 1
State-of-art catalytic systems are predominately formed by a rhodium(I)
precursor and phosphorus(III) ligands, and the design of new ligands has
allowed to develop highly selective systems for specific applications.[3] For
alpha-olefins, the linear product is preferentially formed for steric reasons, but
high linear selectivity is only achieved by the employment of chelating
phosphorus ligands with large bite angles.[4] Obtaining branched aldehydes in
high selectivity when a methylidene group is in the alpha position of the C=C is a
far more complex task. We developed a catalytic system capable of producing the
branched aldehyde in selectivity higher than 65% from naturally occurring 2-
propenylbenzenes based on a rhodium(I) pre-catalyst and the commercially
available 1,3-bis(diphenylphosphino)propane (DPPP) as ancillary ligand.[5]
More recently, by the employment of specially designed ligands, this
regioselectivity was surpassed. [6-13] When the terminal C=C bond has an
electron-withdrawing group, a phenyl ring or a conjugated C=C in the alpha
position, the tendency to form branched products is higher. In this context, the
hydroformylation of conjugated dienes opens opportunities for the synthesis of
branched products, provided the selectivity can be controlled. The
hydroformylation of 1,3-butadiene, isoprene and piperylene is useful for the
synthesis of commodity products.[14-18] Also, 1,3-dienes moieties often occur in
natural products, and their hydroformylation leads to aldehydes of interest for
the fragrance industry.[19-21] However, hydroformylation of this kind of
substrates is particularly difficult to achieve. Special reactions conditions, such
as atypically high pressure, temperature or high ligand concentrations, are
required to get reasonable reaction rates. Although there are more possibilities
for the positioning of the formyl group, the primarily formed aldehydes in the
hydroformylation of 1,3-dienes are usually the ones represented in Scheme
2[22].
Insert Scheme 2
Subsequent C=C isomerization to form alpha,beta-unsaturated aldehydes
followed by their hydrogenation are often observed, which leads to the problems
of selectivity control. It is thus surprising that Nozaki et al. [23] were able to
develop a system that allowed a regio and enantioselective hydroformylation of
1-phenylbutadiene and other dienes, although under rather harsh pressure
conditions (100 bar of CO/H2), by the employment of (R)-2-
2

(diphenylphosphino)-1,1-binaphthalen-2-yl(S)-1,1-binaphthalene-2,2-diyl
phosphite ((R,S)-BINAPHOS) as a ligand. Years later, Landis et al. [22] employed
chiral diazaphospholane ligands and developed a very efficient protocol that
allows the regio and enantioselective hydroformylation of 1,3-dienes under
much milder conditions. Although very efficient, these systems involve non-
commercially available ligands whose synthesis is quite sophisticated. We
considered that it would be worth to develop catalytic systems for the selective
hydroformylation of conjugated dienes based on standard commercial ligands
that could be readily implemented by industry. Thus, in this paper we describe a
protocol to convert naturally occurring 2-propenylbenzenes[24], namely
estragole, eugenol and safrole, to the corresponding 1-arylbutadienes, which
were subsequently hydroformylated to produce in high selectivity novel
branched aldehydes that can be useful for Food & Flavor industry.
2. Experimental
The commercially acquired chemicals were used without an additional
treatment except indicated cases. Estragole (98%), eugenol (99%), palladium
acetate (99.9%), triphenylphosphine (99%), 1,2-bis(diphenylphosphino)ethane
(98%), 1,3-bis(diphenylphosphino)propane (97%) and 9,9-dimethyl-4,5-
bis(diphenylphosphino)xanthene (98%) were purchased from Acros Organics.
Safrole (97%), cesium carbonate (99%), tris(2,4-di-tert-butylphenyl) phosphite
(98%) were purchased from Merck and dichloro(2-isopropoxyphenylmethylene)
(tricyclohexylphosphine)ruthenium(II) (HGII) was supplied by Umicore.
Tribenzylphosphine and dicarbonyl(2,4-pentanedionato)rhodium(I) (99%) were
purchased from Alfa Aesar. Allyl chloride (98%) was purchased from Acros
Organics, distilled under argon and stored over 3 Å molecular sieves.
Dichloromethane, acetonitrile and toluene were dried using MBraun solvent
purification system.
Products were analyzed and characterized by GC and GC-MS using
Shimadzu GC 2014 and QP2010-PLUS instrument (70 eV), respectively, both
equipped with apolar column (Equity-5, 30 x 0.25x 0.25). NMR spectra were
recorded on a 400 MHz Avance III spectrometer from Bruker. Chemical shifts are
reported in ppm vs TMS using residual signals from deuterated solvents as
internal reference.
Preparation of the 1-arylbutadienes
1-Arylbutadienes were prepared from naturally occurring 2-propenylbenzenes,
namely, estragole (1a), eugenol (1b) and safrole (1c), using a previously
reported route (Scheme 3).[25]
Insert Scheme 3
Synthesis of 3-arylallyl chlorides (2a-c) (adapted from [25])
For the synthesis of compound 2a, a Schlenk tube was loaded under argon with
1.0 g (0.0067 mol) of estragole (1a), 0.084 g (0.00013 mol) of 2nd generation
Hoveyda-Grubbs (HGII), 3.1 g (0.040 mol) of allyl chloride and 20 mL of
3

dichloromethane. The mixture was refluxed for 15 h. After evaporation of the
solvent and the residual allyl chloride, the products were purified by column
chromatography on silica gel using heptane/2% ethyl acetate mixture as eluent.
An analogous protocol was employed to synthesize compounds 2b and 2c, but
for the former a mixture heptane/20% ethyl acetate was employed in the
purification. The isolated yields are summarized in Table 1, the side products, if
1
formed, were not characterized. The purified products were analyzed by H and
13C NMR, as well as GC-MS and the data are in Supporting Information S2.
Synthesis of 1-arylbutadienes (3a-c) (adapted from [25])
For the synthesis of compound 3a, a Schlenk tube was loaded under argon with
0.8 g (0.0041 mol) of compound 2a, 0.046 g (0.000205 mol) of palladium acetate,
0.107 g (0.00041 mol) of triphenylphosphine, 1.59 g (0.0049 mol) of cesium
carbonate and 25 mL of acetonitrile. This reaction mixture was stirred under
reflux for 17 h. The resulting mixture was centrifuged at 6000 rpm for 5 min and
the supernatant was collected. The residual solid was washed with cold acetone,
the mixture was centrifuged at 6000 rpm for 5 min and the supernatant was
collected. The supernatants were combined and the solvents were evaporated.
The products were purified by column chromatography on silica gel using
heptane/2% ethyl acetate mixture as eluent. An analogous protocol was
employed to synthesize compounds 3b and 3c, but for the former a mixture
heptane/10% ethyl acetate was employed in the purification. . These products
have already been described in the literature.[26, 27] The isolated yields are
summarized in Table 1, the side products, if formed, were not characterized. The
1
purified products were analyzed by H and 13C NMR, as well as GC-MS and the
data are in Supporting Information S3-S4.
Insert Table 1
Hydroformylation of 1-arylbutadienes (3a-c)
The 1-arylbutadiene (3a) and phosphorus ligand were added in a 1.5 mL glass
vial containing a magnetic stirrer. A solution of [Rh(acac)(CO)2] in toluene was
prepared in a Schlenk tube under argon atmosphere and the proper amount
(0.005 mol) of Rh per 1.0 mol of starting material of catalyst solution was
transferred to the glass vial containing the diene and closed with a cap
containing a rubber septum pierced with a small needle. The vial was set in a
stainless steel bomb, which was closed and purged with a CO/H2 (1:1) mixture.
The bomb was then pressurized with 40 bar of this gas mixture and kept under
o
magnetic stirring in an oil bath at the desired temperature (60-100 C) for 2 to
48 h. The bomb was cooled and carefully vented in a hood and then opened to
air. After solvent stripping, chloroform-d1 was added and the crude products
1
were analyzed by H NMR. The major products were separated by column
chromatography on silica gel using heptane/ 2% ethyl acetate mixture as eluent.
An analogous protocol was employed in the hydroformylation of compounds 3b
and 3c, but for the former a mixture heptane/20% ethyl acetate was employed in
the purification of the main products. The purified products were analyzed by 1H
and 13C NMR, as well as GC-MS and the data are in Supporting Information S4-S6.
4

Quantitative analysis of the hydroformylation products
1
The quantification of the products was made by H-NMR of the crude
reaction mixture. Depending on the catalytic system, up to 14 peaks were
observed in the aldehyde region of the spectra ( 9.40 -10.30 ppm), but only four
of them were predominant. These major signals were attributed to compounds
4a, 5a, 6a and 7a. To the best of our knowledge, compounds 4a and 5a have not
been reported before. They were isolated from the reaction solutions and fully
characterized (see spectroscopic data below). Aldehydes 6a[28] and 7a[29]
were previously described so that their aldehyde signals were attributed based
on the literature data. The remaining peaks in the aldehyde region in the NMR
spectra ( 9.40 - 10.30 ppm) of the crude product were added together and are
reported in the Tables as “Others” (other aldehydes). The quantification was
based on the intensities of the signals at 3.75 -3.90 ppm attributed to the methyl
hydrogens of the methoxy groups present in all the products (100% of the area =
3 H). The mass balance was checked by the comparison of the sum integration of
the aldehyde peaks (1H) and the methoxy peaks (3 H).
3. Results and Discussion
Hydroformylation of 1-arylbutadienes
Estragole is the most available compound of the series and for this reason, the
diene 3a derived from estragole was chosen as the substrate for
hydroformylation
in
the
initial
studies.
(Acetylacetonato)bis(carbonyl)rhodium(I), [Rh(acac)(CO)2], a widely used Rh(I)
pre-catalyst in hydroformylation was used in combination with various
phosphorus(III) ancillary ligands. Under the hydroformylation conditions, the
precatalyst [Rh(acac)(CO)2] is regarded to readily form in situ the 16-electron
active catalyst [RhH(L)m(CO)n], in which m+n = 3 and L is a phosphorus atom of
the ancillary ligand.[30] We employed representative phosphorus(III) ligands of
the class of monophosphines (PPh3 and PBn3), monophosphite (TBDP) and
chelating diphosphines with different bite angles (DPPE, DPPP, XANTPHOS),
whose structures are depicted in Chart 1.
Insert Chart 1
The major products were 4a, 5a, 6a and 7a (Scheme 4). They differ from
the generic ones in Scheme 2 because some of them are formed by consecutive
reactions after hydroformylation, such as isomerization and hydrogenation.
While linear aldehydes 6a and 7a can be formed directly by the
hydroformylation of the terminal C=C in 3a, the formation of 4a should involve a
further double bond migration that can occur either in the catalytic cycle or
through a fast consecutive isomerization process of the -unsaturated
branched aldehyde, which, nevertheless, was not observed by the analytical
techniques employed. The hydrogenation of 4a gives the saturated branched
aldehyde 5a.
5

Insert Scheme 4
The first attempts for the hydroformylation of 3a were made employing
the standard PPh3 as ancillary ligand at 100 oC, 40 bar - CO: H2 = 1, P/Rh = 10, 24
o
h, but the substrate conversion was very low. Under harsher conditions (100 C,
80 bar - CO: H2 = 1, P/Rh = 20, 48 h) it was possible to obtain a higher conversion
(92%), but the selectivity for individual aldehydes was very poor and other
ligands were tested.
The results for the ligand screening in the hydroformylation of 3a are
presented in Table 2. Employing the bulkier and more basic PBn3 (as compared
to PPh3) as ancillary ligand, the substrate was completely converted after 24 h to
give the saturated aldehyde 5a as a major product with 56% selectivity (Table 2,
entry 1). TBDP is a bulky phosphite known to be an efficient promoter in the
rhodium-catalyzed hydroformylation of olefins.[31] Employing this compound
as ancillary ligand under the same conditions also allowed a complete substrate
conversion within 24 h; however, the reaction was poorly selective providing a
complex mixture of other aldehydes (Table 2, entry 2). As the data previously
reported in the literature suggested that chelating ligands could give better
results in the hydroformylation of conjugated dienes as compared to
monodentated ligands[15] , a set of chelating diphosphines with different bite
angles was tested in the hydroformylation of 3a. The ligands employed and the
corresponding bite angles were: DPPE (85o), DPPP (91o), and XANTPHOS
(108o).[32] In the presence of DPPE, the ligand with the lowest bite angle in the
series, not only a full conversion was obtained after 24 h, but also a high
selectivity (91%) for the saturated aldehyde 5a (Table 2, entry 3). Thus, this
system is very efficient for the tandem isomerization/hydroformylation/C=C
hydrogenation, as observed previously by Behr et al.[18] employing isoprene as
substrate. Nevertheless, 1-arylbutadiene gives the branched aldehyde, opposed
to isoprene, which yields the terminal aldehyde. This is remarkable finding from
the practical point of view, considering that such kind of branched aldehydes are
difficult to obtain in high selectivity by the direct hydroformylation of alpha-
olefins.[6]
Insert Table 2
DPPP is a ligand with similar electronic features as DPPE, but with a
larger bite angle. Although both systems presented the same trends, the
selectivity of hydroformylation was lower in the system with DPPP (Table 2, cf.
entries 4 and 3). [15] XANTPHOS is a bidentated ligand, well known to promote
the rhodium-catalyzed hydroformylation of alpha-olefins with
a high
regioselectivity for linear aldehydes.[4] Thus, a higher selectivity for aldehydes
6a and 7a would be expected at the hydroformylation of 3a in the presence of
XANTPHOS, but it was not actually observed (Table 2, entry 5). The branched
aldehydes 4a and 5a were the major reaction products, with the combined
selectivity being similar to that in the system promoted by DPPP (Table 2, cf.
entries 4 and 5). It is noteworthy that there is a lower tendency to hydrogenate
the -unsaturated aldehyde 4a by the XANTPHOS-promoted system, since
even after 24 h, 37% of this aldehyde is present among the products.
6

Considering that DPPE as ancillary ligand (L) gave the best results, further
studies concerning the reaction conditions was carried out with this ligand and
they are presented in Table 3. In entries 1-3, the effect of L/Rh molar ratio can be
observed. The systems with L/Rh= 2.5 (entry 1) and 5.0 (entry 2) showed nearly
the same performance, but at a higher ratio (L/Rh= 10, entry 3) the contribution
of linear aldehydes 6a and 7a in the product distribution increased. The effect of
the reaction temperature can be evaluated by comparing the results presented in
o
entries 2, 4, and 5 (Table 3). Although at 80 C the conversion was completed
after 24 h, a significant amount of the branched aldehyde remained in the non-
hydrogenated form (4a). This effect was even more evident in the reaction at 60
oC, in which aldehyde 4a was a predominant product, thus clearly showing that
higher temperature favored the hydrogenation step.
Insert Table 3
Although toluene is presently considered an acceptable solvent, there are
some environmental issues assigned to this compound[33]. In a further study,
we decided to test other solvents, with higher sustainability rankings than
toluene, as a reaction medium for the hydroformylation of 3a. p-Cymene has
been successfully employed by us in hydroformylation[34] and metathesis.[35]
For the system with p-cymene, the catalytic results were similar to those
obtained in toluene (Table 3, cf. entries 2 and 6). Anisole has a much better
sustainability rank than toluene and even than p-cymene,[33] and we have
recently found it was an excellent solvent for hydroformylation[34]. Although
the selectivity for aldehyde 5a in the reaction in anisole was slightly lower than
in toluene, it was still reasonably high (85% vs. 91%, cf. entries 2 and 7 in Table
3). Thus, the more sustainable solvents p-cymene and anisole can be
recommended as possible alternatives to toluene for the tandem
hydroformylation/hydrogenation of conjugated diene 3a aiming at obtaining
saturated aldehyde 5a with high selectivity.
In order to determine the reaction evolvement, a kinetic follow-up was
performed and the results are presented in Figure 1. Under these reaction
conditions, the substrate 3a is completely converted into products within 4 h.
The nearly straight line for the substrate consumption up to 100% conversion
showed an independence of the reaction rate on the substrate concentration,
thus suggesting saturation kinetics. After 2 h, the predominant product was the
unsaturated aldehyde 4a, along with minor amounts of aldehydes 6a and 5a.
Then, the concentration of 4a started decreasing due to its conversion into 5a,
whereas the concentration of 6a continued to increase until 4h, and then
decreased. The slight decrease of 6a until 24 h is probably due to the C=C double
bond isomerization to form other unsaturated aldehydes. Only after 24 h the
formation of product 7a increased, likely due to the hydrogenation of ,-
unsaturated linear aldehydes rather than 6a directly. The variation in 5a
percentage from 24 to 48 h (less than 4%) is within the experimental
reproducibility, although a decrease due to formation of minor amounts of non-
aldehyde derivatives cannot be ruled out. The -unsaturated aldehyde 4a’
isomer of 4a was not observed, but it would be the logic precursor of 4a, firstly
formed by the Markovnikov-type hydroformylation of the terminal double bond,
followed by fast isomerization. Nevertheless, the direct formation of 4a as a
7

primary product resulting from 1,3-H migration in the coordination sphere of
the metal within the catalytic cycle cannot be ruled out.[16]
Insert Figure 1
The substrate scope for the tandem hydrofomylation/hydrogenation
process was extended to other 1-arylbutadienes derived from the natural
products, eugenol (3b) and safrole (3c), being the results presented in Table 4.
Insert Table 4
The products, 4a-c and 5a-c obtained from these substrates are new
compounds and the corresponding spectroscopic data are presented in the
experimental section. As the diene double bonds of 3a-3c are conjugated to the
aromatic ring, an influence of the ring substituents on the reaction output could
be expected, being the more influent substituent at the para position.
Nevertheless, the effect of the substituent of the aromatic ring was not so
marked after 24 h for the systems promoted by DPPE (Table 4, entries 1-3). The
high regioselectivity for the branched product (4 + 5) was about the same for all
three substrates. The difference in reactivity was more evident for the systems
promoted by XANTPHOS. The hydroformylation rate was significantly lower for
substrate 3b as compared to the other two as only 80% conversion was reached
after 24 h, whereas the product distributions were comparable (Table 4, entries
4, 5 and 6). Again, it was obvious that the XANTPHOS-promoted system was less
efficient to promote the hydrogenation of 4 into 5, as 37-45% of 4 remained
after 24 h. At 60 oC (Table 4, entries 7 and 8) unsaturated aldehyde 4a accounted
for 76% of the products at 62% conversion. Interestingly enough, if the reaction
was prolonged to 48 h (entry 8) the conversion rose to 89%, but no important
change on product distribution occurred, which indicated that consecutive
hydrogenation processes are not significant under these conditions. The
unsaturated aldehyde 4a was thus obtained in 66% yield. It is noteworthy that
the catalytic reaction can be driven selectively towards the unsaturated aldehyde
4a or the saturated aldehyde 5a, simply by changing the reaction temperature
and the auxiliary phosphorous ligand, DPPE or XANTPHOS, respectively.
Regarding to catalyst recovery and reuse, some of us published recently a
methodology for the hydroformylation of 2-propenylbenzenes in which the
catalyst is supported in an aqueous phase by means of the water-soluble
sulfonated version of DPPE, inter alia.[36] The catalyst is recovered by
decantation of the aqueous phase after hydroformylation. We believe that this
system could be adapted for the 1-arylbutadienes 3a-c, and this should be a
matter for further studies.
The results are fully consistent with the reaction paths shown in Scheme
5, based on Nozaki’s proposal[23], which is supported by spectroscopic
evidences. Although there would be other possibilities, the dominant interaction
between 1-arylbutadienes 3 and the active rhodium species (LnRhH) occurs
through the terminal C-C double bond. In Path A, the hydrogen is transferred to
carbon in position 3 at the chain, leading to an Rh-alkyl species A, which, after
carbonylation and hydrogenolysis, yields aldehyde 6. Alternatively, path B, in
which the hydrogen is transferred to carbon in position 4 at the chain, leads to
8

the formation of a much more stable 3-allyl intermediate B, which seems to be
the resting state for rhodium species and would explaining the observed
saturation kinetics. The 3-allyl intermediate B does not undergo directly
carbonylation/hydrogenolysis,[15, 16, 19, 23] but it is in equilibrium with 1-
allyl intermediate C, which can suffer carbonylation/hydrogenolysis. The
primarily formed products can be further transformed by C=C isomerization
(migration) and hydrogenation. Under hydroformylation conditions catalyzed by
rhodium, hydrogenation is only a major path when the C=C double bond is
conjugated with the aldehydic C=O, and thus the efficient hydrogenation
occurring for the formation of 5 in high yield should be preceded by the
formation of 4. Likewise, the formation of 7 should also be preceded by the
isomerization of 6.
Insert Scheme 5
Although, the 3-allyl (B) – 1-allyl (C) equilibrium is widely favored
towards the former, it can be shifted towards C by the enhancement of the steric
encumbrance on rhodium, and this is the key point to explain the ligand effect.
Monodentated phosphorus ligands can promote the 3-allyl–1-allyl shift
by increase in their concentration or their binding ability. Increasing the
concentration of certain ligands such as PPh3 may suffice in some cases,[16] but
apparently the presence the conjugated phenyl ring in B makes it more stable, so
that other ligands are needed. PBn3, a more basic and bulkier ligand than PPh3,
acted better as a promoter in the hydroformylation of 3a, probably, due to the
shift of the 3– 1 equilibrium towards C. Employing TBDP as ancillary ligand,
which is also bulky and binds strongly rhodium species, the same promoting
effect on the transformation of 3-intermediate B into 1-intermediate C was
observed. However, this ligand yields a more acidic rhodium intermediate, which
is prone to promote the C-C double bond isomerization and less prone to
perform its hydrogenation. As a result, a complex mixture of isomeric aldehydes
with low individual selectivity for each compound was formed when TBDP was
employed as ligand.
Bidentated phosphines, such as DPPE, DPPP and XANTPHOS form more
stable Rh species containing two coordinated phosphorus(III) due to the chelate
effect. This forces the 3– 1 equilibrium to 1 species (intermediate C, Scheme 5)
and for that reason, diphosphines are more efficient for the hydroformylation of
this kind of dienes than the corresponding monodentated ligands. DPPE is more
efficient than DPPP because it forms more stable 5-membered ring chelates than
the 6-membered ring of DPPP.[37] .XANTPHOS has a wider bite angle and is
conformationally more rigid than DPPE, resulting in a more encumbered
environment around the Rh atom and favoring the formation of linear products
in hydroformylation.[32] Thus, it is remarkable that the regioselectivity for the
branched products (4 + 5) is comparable for both DPPE and XANTPHOS, since
the latter is expected to favor Path A (less encumbered)[4] This finding
reinforces the hypothesis that the regioselectivity in the hydroformylation of 1-
phenylbutadienes is biased by the formation of stable 3-allyl B, so that Path B
becomes a dominant reaction path, even in the presence of ligands know for
directing hydroformilation for linear products.
9

Although -unsaturated aldehyde 4’ (Scheme 5) has not been detected
in our work by the analytic technique used, its intermediate formation cannot be
ruled out. On the other hand, the aldehydes similar to 4’ were reported as major
reaction products at the hydroformylation of related dienes with other catalytic
systems and under different reaction conditions (in particular, lower
o
temperatures in the range of 30-40 C).[22, 23] We suggest that in our systems,
either the fast isomerization of -unsaturated aldehyde 4’ into -unsaturated
aldehyde 4 biased by the favorable conjugation of the C=C and C=O bonds is
occurring, or the double bond migration starting from intermediate C to give
directly aldehyde 4 is operating, as previously suggested.[16] Either way, under
appropriate conditions, the -unsaturated aldehyde 4 can undergo
hydrogenation in tandem to give branched aldehyde 5 in excellent yields. DPPE
is the most efficient ancillary for this purpose. With XANTPHOS, the tandem
hydrogenation is less efficient, probably, because the coordination of the
trisubstituted C-C double bond in 4 to the rhodium species is more difficult due
to the greater steric encumbrance caused by this ligand. On the other hand,
under appropriate temperature conditions, the use of the XANTPHOS-promoted
system offers the possibility to obtain selectively another useful product, the
-unsaturated aldehyde 4a.
4. Conclusions
Bio-sourced 2-propenylbenzene derivatives have been transformed into
4-aryl-2-methylbutanals and 4-aryl-2-methylbutenals via a sequence of catalytic
transformations involving ruthenium-catalyzed cross metathesis, palladium-
catalyzed dehydrochlorination and rhodium-catalyzed hydroformylation. The
hydroformylation
of
the
synthesized
1,3-dienes
employing
rhodium/monophosphorus ligands was slow and the selectivity was poor. Only
by the employment of chelating diphosphines as ancillary ligands at relatively
high concentration were the activity and selectivity increased. The selectivity
was also highly dependent on the ligands. At 100oC, consecutive C-C double bond
isomerization and hydrogenation took place, resulting in a useful tandem
process. Employing XANTPHOS as ancillary ligand, the C-C double bond
hydrogenation was not efficient, allowing the synthesis of the intermediate enals
4a-c in good yields. Employing DPPE as ancillary ligand at 100 oC, a clean tandem
sequence took place: although this system was efficient to reduce the C-C double
bond of the enals, the C-O double bond remained untouched, leading to saturated
branched aldehydes 5a-c in very good selectivity. These branched aldehydes are
difficult to obtain through other routes and are potentially interesting for food
and flavor industry as 5a and 5c are homologous to the commercial fragrances
canthoxal and helional, respectively. Noteworthy is that these transformations
could be carried out in green solvents such as anisole.
CRediT author statement
Gabriel M. Vieira (investigation), Artur V. Granato(investigation), Elena V.
Gusevskaya (writing –Review & Editing), Eduardo N. dos Santos (writing-
Original Draft, writing –Review & Editing, supervision), Pierre H. Dixneuf
10

(writing –Review & Editing), Cédric Fischmeister(writing –Review & Editing,
supervision), Christian Bruneau(writing –Review & Editing, supervision)
Declaration of interests
The authors declare that they have no known competing financial interests or
personal relationships that could have appeared to influence the work reported
in this paper.
Acknowledgements
The authors thank CAPES-COFECUB (project PHC 884-17-France; 883/2017-Brazil),
INCT-Catalysis, and CNPq for financial support. Umicore is thanked for supplying
Hoveyda-Grubbs catalyst M72 (C627).
References
[1] A. Borner, R. Franke, Hydroformylation: Fundamentals, Processes, and
Applications in Organic Synthesis, Wiley-VCH, Weinheim, 2016.
[2] O. Rölen, Chem. Abstr., 38 (1944) 3631.
[3] E.V. Gusevskaya, J. Jimenez-Pinto, A. Borner, ChemCatChem, 6 (2014) 382-
411.
[4] P.W.N.M. van Leeuwen, P.C.J. Kamer, Catal. Sci. Technol., 8 (2018) 26-113.
[5] A.C. da Silva, K.C.B. de Oliveira, E.V. Gusevskaya, E.N. dos Santos, J. Mol. Catal.
A Chem., 179 (2002) 133-141.
[6] G.M. Noonan, J.A. Fuentes, C.J. Cobley, M.L. Clarke, Angew. Chem. Int. Ed., 51
(2012) 2477-2480.
[7] P. Dingwall, J.A. Fuentes, L. Crawford, A.M.Z. Slawin, M. Buhl, M.L. Clarke, J.
Am. Chem. Soc., 139 (2017) 15921-15932.
[8] L. Iu, J.A. Fuentes, M.E. Janka, K.J. Fontenot, M.L. Clarke, Angew. Chem. Int. Ed.,
58 (2019) 2120-2124.
[9] V.F. Slagt, P.C.J. Kamer, P. W.N.M. van Leeuwen, J.N.H. Reek, J. Am. Chem. Soc.,
126 (2004) 1526-1536.
[10] A. Phanopoulos, K. Nozaki, ACS Catal., 8 (2018) 5799-5809.
[11] T. Besset, D.W. Norman, J.N.H. Reek, Adv. Synth. Catal., 355 (2013) 348-352.
[12] P. Dydio, R.J. Detz, B. de Bruin, J.N.H. Reek, J. Am. Chem. Soc., 136 (2014)
8418-8429.
[13] X.W. Wang, S.S. Nurttila, W.I. Dzik, R. Becker, J. Rodgers, J.N.H. Reek, Chemi.
Eur. J., 23 (2017) 14769-14777.
[14] H. Bahrmann, B. Fell, J. Mol. Catal., 8 (1980) 329-337.
[15] P.W.N.M. Van Leeuwen, C.F. Roobeek, J. Mol. Catal., 31 (1985) 345-353.
[16] H.J.V. Barros, C.C. Guimaraes, E.N. dos Santos, E.V. Gusevskaya,
Organometallics, 26 (2007) 2211-2218.
[17] P. Neubert, S. Fuchs, A. Behr, Green Chem., 17 (2015) 4045-4052.
[18] A. Behr, S. Reyer, N. Tenhumberg, Dalton Trans., 40 (2011) 11742-11747.
[19] C.M. Foca, H.J.V. Barros, E.N. dos Santos, E.V. Gusevskaya, J.C. Bayon, New J.
Chem., 27 (2003) 533-539.
[20] H.J.V. Barros, C.C. Guimaraes, E.N. dos Santos, E.V. Gusevskaya, Catal.
Commun., 8 (2007) 747-750.
11

[21] H.J.V. Barros, J.G. da Silva, C.C. Guimaraes, E.N. dos Santos, E.V. Gusevskaya,
Organometallics, 27 (2008) 4523-4531.
[22] A.L. Watkins, C.R. Landis, Org. Lett., 13 (2011) 164-167.
[23] T. Horiuchi, T. Ohta, E. Shirakawa, K. Nozaki, H. Takaya, Tetrahedron, 53
(1997) 7795-7804.
[24] M. Hassam, A. Taher, G.E. Arnott, I.R. Green, W.A.L. van Otterlo, Chem. Rev.,
115 (2015) 5462-5569.
[25] H. Bilel, N. Hamdi, F. Zagrouba, C. Fischmeister, C. Bruneau, Catal. Sci.
Technol., 4 (2014) 2064-2071.
[26] M.H. Al-Huniti, M.A. Perez, M.K. Garr, M.P. Croatt, Org. Lett., 20 (2018) 7375-
7379.
[27] T.Q. Wang, Y.Y. Hu, S.L. Zhang, Org. Biomol. Chem., 8 (2010) 2312-2315.
[28] Y. Jeong, D.Y. Kim, Y. Choi, J.S. Ryu, Org. Biomol. Chem., 9 (2011) 374-378.
[29] B. Meka, S.R. Ravada, M.K.K. Muthyala, P.N. Kurre, T. Golakoti, Bioor. Chem.,
80 (2018) 408-421.
[30] R. Franke, D. Selent, A. Borner, Chem. Rev., 112 (2012) 5675-5732.
[31] H. Tricas, O. Diebolt, P.W.N.M. van Leeuwen, J. Catal., 298 (2013) 198-205.
[32] M.N. Birkholz, Z. Freixa, P. W.N.M. van Leeuwen, Chem. Soc. Rev., 38 (2009)
1099-1118.
[33] C.M. Alder, J.D. Hayler, R.K. Henderson, A.M. Redman, L. Shukla, L.E. Shuster,
H.F. Sneddon, Green Chem., 18 (2016) 3879-3890.
[34] F.G. Delolo, E.N. dos Santos, E.V. Gusevskaya, Green Chem., 21 (2019) 1091-
1098.
[35] A.V. Granato, A.G. Santos, E.N. dos Santos, ChemSusChem, 10 (2017) 1832-
1837.
[36] P.J. Baricelli, M. Rodriguez, L.G. Melean, M.M. Alonso, M. Borusiak, M.
Rosales, B. Gonzalez, K.C.B. de Oliveira, E.V. Gusevskaya, E.N. dos Santos, Appl.
Catal. A Gen., 490 (2015) 163-169.
[37] P. Dierkes, P.W.N.M. van Leeuwen, J. Chem. Soc. Dalton Trans., (1999) 1519-
1529.
12

Table 1 - Synthesis of 3-arylallyl chlorides (2a-c) and 1-arylbutadienes (3a-c)
Starting material
Conversion (%)
Isolated Yield (%)
Synthesis of 2a-ca
1a
1b
1c
95
81 (2a)
71 (2b)
62 (2c)
85
90
Synthesis of 3a-cb
2a
2b
2c
100
100
100
62 (3a)
47 (3b)
62 (3c)
a Conditions: 1a-c - 1.0 g, allyl chloride (6 equiv.), HGII (2 mol%), solvent - dichloromethane (20
mL), 40 ºC, 15 h.
b Conditions: 2a-c - 0.5-1.0 g, Pd(OAc)2 (5 mol%), PPh3 (10 mol%), Cs2CO3 (1.2 equiv.), solvent -
acetonitrile, 90 ºC, 17 h.
Table 2 – Hydroformylation of 3a in the presence of different ligandsa
Conversion b
Product distribution (%)
Entry
Ligand
4a
0
0
0
2
5a
56
21
91
73
41
6a
1
0
3
7
7a
6
3
0
2
Othersc
(%)
100
100
100
100
100
1d
2d
3
4
5
PBn3
TBDP
DPPE
37
76
6
16
12
DPPP
XANTPHOS
37
10
0
a Conditions: 3a - 1.0 mol.L-1, [Rh(acac)(CO)2] - 0.5 mol%; ligand - 2.5 mol%, solvent - toluene,
100 ºC, 40 bar (CO:H2 = 1), 24 h. b Determined by 1HNMR, zero value means not observed by the
technique. c Other products, mainly other aldehydes. d Ligand - 5.0 mol%.
13

Table 3 – Hydroformylation of 3a using DPPE as ancillary under different reaction
conditionsa
Product distribution b (%)
Ligand
(mol %)
1.25
2.5
T
Conversion b
(%)
Entry
(oC)
100
100
100
80
4a
0
0
5a
87
91
83
73
35
88
85
6a
3
3
5
7
3
4
6
7a
0
0
6
0
0
3
3
Othersc
1
2
3
4
5
6d
7e
100
100
100
100
91
100
100
9
6
6
7
8
5
6
5.0
2.5
2.5
2.5
0
13
54
0
60
100
100
2.5
0
a Conditions: 3a - 1.0 mol.L-1; [Rh(acac)(CO)2] - 0.5 mol%; solvent - toluene, 40 bar (CO:H2 = 1),
24 h. bDetermined by 1HNMR, zero value means not observed by the technique. c Other products,
mainly other aldehydes. d Solvent - p-cymene. e Solvent - anisole.
Table 4 – Hydroformylation of 1-arylbutadienes a
Conversion b
Product distribution b (%)
4(a-c) 5(a-c) 6(a-c) 7(a-c)
Entry
Ligand
Substrate
Othersc
%
100
99
100
100
80
100
62
89
1
2
3
4
5
DPPE
DPPE
DPPE
3a
3b
3c
3a
3b
3c
3a
3a
0
4
0
37
45
45
76
74
91
86
92
41
42
46
3
3
7
5
10
13
9
12
13
0
3
3
0
0
0
0
0
6
0
0
12
0
0
XANTPHOS
XANTPHOS
XANTPHOS
XANTPHOS
XANTPHOS
6
7d
8d,e
9
11
2
a Conditions: substrate - 1.0 mol.L-1, [Rh(acac)(CO)2] - 0.5 mol%, ligand - 2.5 mol%, solvent -
toluene, 40 bar (CO:H2 = 1), 24 h. bDetermined by 1HNMR, zero value means not observed by the
technique. c Other products, mainly, other aldehydes. d 60oC. e 48h.
14

Scheme 1
Scheme 2
Scheme 3
Scheme 4
Scheme 5
15

Chart 1
Figure 1: Kinetic follow-up for the hydroformylation of 3a promoted by Rh/DPPE.
Conditions: [3a]0 = 1.0 mol.L-1; [Rh(acac)(CO)2] , 0.5 mol%; DPPE, 2.5 mol%; toluene;
100ºC; 40 bar (CO:H2 = 1); 24 h. b determined by 1HNMR. Ot: Other products.
16