A mild entry to isoindolinones from furfural as renewable resource

Article abstract of DOI:10.1039/c3nj41050a

A convenient transformation of furfural and derivatives into arene compounds is described. This process provides a straightforward access to isoindolinones from a renewable resource via the sequence of a tandem Ugi-Diels-Alder process followed by a ring-opening, dehydrating aromatisation. Five bonds are created and water is the sole by-product during this two-step sequence.

Full text of DOI:10.1039/c3nj41050a

NJC
View Article Online
View Journal | View Issue
PAPER
A mild entry to isoindolinones from furfural as
renewable resource†
Gilles Caillot,ab Shridhar Hegdec and Emmanuel Gras*ab
Cite this: NewJ.Chem., 2013,
37, 1195
A convenient transformation of furfural and derivatives into arene compounds is described. This process
provides a straightforward access to isoindolinones from a renewable resource via the sequence of a
tandem Ugi–Diels–Alder process followed by a ring-opening, dehydrating aromatisation. Five bonds are
created and water is the sole by-product during this two-step sequence.
Received (in Montpellier, France)
19th November 2012,
Accepted 23rd January 2013
DOI: 10.1039/c3nj41050a
www.rsc.org/njc
produced from petroleum), starting from a renewable starting
material like furfural, would represent an overall process
Introduction
Furfural 1 (from the Latin furfur, bran) is a compound directly
derived from C5 sugars that are extracted from otherwise
unexploited plant wastes such as those of wheat bran, corncobs
or sugarcane.1–3 Current large-scale uses of furfural mostly
involve simple processes like hydrogenation, leading to
widely-used compounds such as furan itself, tetrahydrofuran
(THF) and 2-methyltetrahydrofuran (MeTHF).4 Although well-
established, these processes are still under investigation in
order to increase their efficiency and enhance their E-factors.5
This clearly illustrates the significant degree of optimisation
required to establish an environmentally sustainable industrial
process. Considering that petroleum resources are dwindling,
and that the use of petroleum in energy production might be
prioritised over petrochemicals manufacturing, we have engaged
in a program aimed at enabling building-block production from
renewable resources such as furfural. It is noteworthy that
although a wide array of furan scaffolds (at different levels of
hydrogenation) may be accessed from furfural, processes giving
rise to other scaffolds (such as arenes) starting from furfural are
scarce. This is surprising, since furan derivatives are known to
react with activated double bonds in [4+2] cycloadditions,
leading to bicyclic structures possessing six-membered carbo-
cyclic sub-units. These oxabicyclic adducts can in theory lead to
benzenic products through aromatisation with loss of a single
molecule of water. Such access to aromatic scaffolds (typically
fulfilling a number of green chemistry principles (use of renewable
feedstocks, carbon-atom economy, waste prevention). Indeed,
pericyclic reactions are highly atom-efficient, and proceed with
excellent levels of regio- and stereo-control. The regioselectivity
combined with a simple ring opening–dehydration domino
sequence shall therefore give a sustainable access to structurally
defined polyfunctionalised arenes (Scheme 1). One would expect
the reaction to proceed smoothly if it was carried out in an
intramolecular fashion.6 For that purpose, the dienophile could
be tethered to the pendant arm of the furan arising from the initial
aldehyde function of the furfural. Such a tether and further
molecular complexity could be achieved by taking advantage of
multi-component reactions (MCR) such as the well described
Ugi reaction.7–10 This would then represent a very step-efficient
conversion of the initial five-membered ring to the six-membered
ring. The overall process would complement nicely the tandem
processes involving a Ugi reaction, yielding heterocycles.11 Also, it
would be an efficient process in terms of taking advantage of every
a CNRS, LCC (Laboratoire de Chimie de Coordination), 205 route de Narbonne,
BP 44099, F-31077 Toulouse Cedex 4, France
b
´
Universite de Toulouse, UPS, INPT, F-31077 Toulouse Cedex 4, France.
E-mail: emmanuel.gras@lcc-toulouse.fr
c Pennakem, 3324 Chelsea Avenue Memphis, TN 38108, USA
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3nj41050a
Scheme 1 Arenes from furfural, an alternative access to petrochemicals from
‘‘renewable wastes.’’
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 1195--1200 1195

View Article Online
Paper
NJC
chemical function of furfural (aldehyde involved in MCR to anchor of the requirement for a second electron-withdrawing group
the dienophile, diene involved in the Diels–Alder cycloaddition, (compared to previous reports12–15) enhances the versatility of
oxygen released in the final water elimination). We disclose this reaction. In order to shorten the reaction time, we explored
here our recent results toward this approach.
the use of microwave irradiation (Table 1, entry 2). Although the
reaction time was indeed shortened, we observed no significant
enhancement of the reaction efficiency. Having established that
one single electron-withdrawing group is sufficient for the
Results and discussion
Based on previous reports using maleic and fumaric derivatives,12–15 reaction to proceed smoothly, we explored the effect of further
we first carried out the Ugi reaction on furfural with tert-butyl substitution of the double bond, as it is well known to strongly
isocyanide, 4-methoxybenzylamine and differently substituted affect the reversibility of the Diels–Alder reaction of furans.19
The introduction of a methyl group appeared rather detrimental,
acrylic acids in methanol at 50 1C, as protic solvents are known to
favor the Ugi reaction.16 We also showed that substituted furan allowing the isolation of the oxabicycle in only 26% yield,
carboxaldehydes can be successfully engaged in this reaction, as together with 47% of the uncyclised product 6 (Table 1, entry
illustrated below (Scheme 2). Our first attempt (Table 1, entry 1) was 3; see Fig. 1). Interestingly, the cyclised product was obtained as
performed with acrylic acid. We noticed that a slight heating of the a single diastereoisomer, corresponding to an exo approach. The
reaction medium was required for complete consumption of the isolation of this thermodynamic adduct fits well with the known
starting furfural, and we observed after one day a single reaction reversibility of Diels–Alder reactions with furans,20 and the
product exhibiting the bicyclic core of the expected product of the epimerisable nature of the carbon a to the amide, since no
stereocontrol is expected in the course of the Ugi reaction.
In contrast, in the presence of a phenyl group, no uncyclised
tandem process in a satisfying yield of 54% for the four-component,
five-bond-forming process. This result contrasts with the previous
reports from McCluskey and co-workers, who showed that compound was obtained and a better isolated yield of the
the Ugi reaction of furfural occurs within 30 minutes at room polycyclic structure was achieved (Table 1, entry 4). Again, only
temperature.17,18
the exo diastereoisomer was isolated. Finally, the best yields
Yet, in our case, the ethylene moiety proved more reactive were obtained for reactions involving bromine-substituted
than the previously reported acetylene analogue, since the acrylic acids. Interestingly, introducing the bromine atom at
Diels–Alder reaction proceeded spontaneously under the reac- position 2 of the acrylic partner promoted a scrambling of the
diastereoselectivity of the reaction, a mixture of both diastereo-
tion conditions, therefore avoiding the need for solvent
exchange and more forcing conditions. No uncyclised product isomers being obtained in 80% overall yield. Contrastingly,
could be isolated (the only intermediate observed by TLC using the bromoacrylic acid of E geometry allowed the isolation
analysis is thought to be the imine, as it appears to be hydro- of the cyclised adduct in 90% yield as a single diastereoisomer.
lysed back to furfural on preparative scale). No intermolecular This enhanced efficiency has been nicely correlated with a
Diels–Alder reaction was observed under these reaction condi- higher exergonic reaction profile.21 Having established the
tions, clearly indicating the order of reactions during the usefulness of the tandem process using unactivated acrylic
cascade. From a sustainable development perspective the lower derivatives, we also explored the effect of the substitution
pattern of the furan ring on the reaction profile. Introduction
temperature (compared to the 200 1C in a sealed tube reported
by McCluskey et al.) is a significant improvement; and the lack of substituents at the 5-position of the furan did not hamper the
Scheme 2 Tandem Ugi–Diels–Alder involving differently substituted furyl aldehydes and various acrylic acids.
c
1196 New J. Chem., 2013, 37, 1195--1200
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
NJC
Paper
Table 1 Tandem Ugi–Diels–Alder reaction of furyl aldehydes with tert-butyl
decomposition may occur and that the by-products are elimi-
nated during the work-up/purification procedure.
isocyanide, 4-methoxybenzylamine and various acrylic acids
Having shown that the tandem process using unactivated
acrylate can be efficiently performed on furfural and furfuryl-
amine, we explored the transformation of the oxabridged
bicycle into an aromatic ring. Various strategies have been
reported for the conversion of oxabicyclic adducts to the
corresponding aromatic moiety. Treatment with concentrated
sulfuric acid has been shown to perform this ring opening-
dehydration process albeit with variable efficiencies.22,23 One
recent report has shown that a Lewis acid can promote
Product and
Entry Aldehyde 1 Acrylate 4 Reaction time (h) isolated yield (%)
1
2
3
4
5
6
7
8
1a
1a
1a
1a
1a
1a
1b
1c
1d
1e
1f
4a
4a
4b
4c
4d
4e
4a
4a
4a
4a
4a
25
4a
24
31
5a 54
5a 59
5b 26 (47b)
5c 69
4.5
5d 80c
5e 90
5f 44
24
24
24
24
24
24
5g 56
5h 54
5i 54
9
10
11
a deselenisation–aromatisation sequence from selenated
oxabridge bicycles.24 This approach did not meet our aim of
developing an environmentally friendly process. An alternative
approach may be envisaged, in which an electron-withdrawing
group is attached to the oxabridge bicycle by performing the
ring opening-dehydration process under basic conditions in
the presence of LiI.25 With the acidic treatment appearing as the
most mass-efficient one, we explored the use of various acids and
found that one equivalent of hydrated PTSA promoted the desired
transformation within a few minutes in toluene under reflux
(Scheme 4).26 As expected, the substitution of the oxabicycle
strongly affected not only the reaction time but also its outcome.
The unsubstituted oxabicycle arising from the Ugi–Diels–Alder
tandem required two hours to reach full conversion and although
the crude reaction mixture appeared relatively clean, the isolated
yield remained moderate though reproducible (Table 3, entry 1).
A shorter reaction time was necessary to fully convert the
phenyl-substituted oxabicycle (Table 3, entry 2). Although the
yields are still low, this method is an interesting approach
5j 68
a
The reaction was conducted under microwave irradiation (CEM
b
c
Explorer). Uncyclised product. A 1 : 1 mixture of the two diastereo-
isomers was obtained.
Fig. 1 Uncyclised product from the reaction involving crotonic acid.
overall efficiency of the process: the desired cyclised adducts to the 1,2,3-trisubstituted biphenyl moieties in a metal-free
were systematically obtained as single diastereoisomers, with process beginning from a renewable resource. Higher yields
moderate yields similar to those obtained with furfural and and shorter reaction times were achieved from oxabicycles
acrylic acid (Table 1, entries 7, 8, 10, and 11). Disubstitution of arising from diversely substituted furans.
the furan at positions 4 and 5 was also tolerated in this process.
Thus, introduction of a methyl on the bridgehead cut the
In addition to the studies described above, we sought to reaction time to 30 minutes, resulting in an average isolated
develop similar tandem processes involving furfurylamine, yield of 51% (Table 3, entry 3). A 62% yield was achieved in the
another derivative of furfural. Furfurylamine is routinely produced case of the more inductive electron-donating ethyl substitution
from furfural under heterogeneous catalytic conditions, and is at the bridgehead (Table 3, entry 5), with the reaction time
another interesting renewable building-block. It may be engaged decreased further. Disubstitution of the bicyclic moiety allowed
similarly in a tandem Ugi–Diels–Alder reaction to give cyclised another increase in yield (Table 3, entry 4). Introduction of a
adducts exhibiting an alternative substitution pattern (compounds 4-bromo-substituted phenyl ring on the bridgehead also
7a, 7d and 7e) as shown in Scheme 3. Under the same reaction allowed the access to a biphenyl adduct, now 1,3,4,40 substi-
conditions using furfurylamine, 4-methoxybenzaldehyde, tert-butyl tuted (Table 3, entry 6). Interestingly, introducing a bromine on
isocyanide and acrylic acid, we obtained the expected bicyclic the bridgehead led to the formation of the corresponding
adduct (Table 2, entry 1) in a yield similar to that achieved phenol (Table 3, entry 7). Finally, among the oxabicyclic
previously with furfural (see Table 1, entry 1). Increasing the adducts obtained from furfurylamine only 7a underwent the
substitution of the acrylic reagent proved to be detrimental in this ring opening-dehydrative aromatisation process, although
case, as reactions involving crotonic and cinnamic acid gave good the isolated yield was 95%. The other substrates obtained from
yields of the Ugi adduct, but no Diels–Alder cycloaddition was the Ugi–Diels–Alder tandem reaction were shown to be either
observed (Table 2, entries 2 and 3).
unreactive (5d and 7d bearing a bromine atom a to the CQO
Contrastingly, when bromoacrylic acids were involved the bond) or underwent degradation (7e). Although rather disap-
bridged bicyclic adducts were obtained in good yields but with pointing, these results provide evidence for the reaction mecha-
low diastereoselectivities (Table 2, entries 4 and 5). It should be nism of the process. Acidic activation of the oxabicycle likely
pointed out that in every case only the isolated product was promotes the ring opening by elimination. Among the two
detected. We believe that in the less efficient reactions hydrogen atoms available for this elimination, the one a to
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 1195--1200 1197

View Article Online
Paper
NJC
Scheme 3 Tandem Ugi–Diels–Alder involving furfurylamine and various acrylic derivatives.
Table
2
Tandem Ugi–Diels–Alder reaction of furfurylamine with tert-butyl
Table 3 Ring opening-dehydrative aromatisation yielding isoindolinones
Entry Reactant Reaction time (min) Product and isolated yield (%)
isocyanide, 4-methoxybenzaldehyde and various acrylic acids
Entry Acrylate 4 Reaction time (h) Product and isolated yield (%)
1
2
3
4
5
6
7
8
5a
5c
5f
5h
5g
5i
120
60
30
40
10
10
30
100
9a, 39
9b, 21
9c, 51
9d, 75
9e, 62
9f, 31
9g, 54
10, 95
1
2
3
4
5
4a
4b
4c
4d
4e
24
24
16
24
24
7a, 57
8b, 70
8c, 81
7d, 87a
7e, 70b
5j
7a
a
b
A 1 : 1 mixture of diastereoisomers was obtained. A 2 : 1 mixture of
diastereoisomers was obtained.
bond). The lack of reactivity of the compounds in which this
hydrogen is replaced by a bromine atom accounts for this first
step. From this point, the alcohol obtained undergoes an acid-
catalysed elimination to yield the desired arene as shown in
Scheme 5.
the CQO bond is the one involved (being more acidic and
giving access to a tetrasubstituted doubly conjugated double
This second elimination appears to be favored when R4
promotes the formation of a carbocation. This effect might
not be observed from 5i, as the bromine atom on the phenyl
substituent deactivates it, such that the process is only as
efficient as in the unsubstituted cases. In contrast, when R3 is
a phenyl group and R4 a hydrogen atom, its position trans to the
leaving oxygen disfavours the elimination and makes the
Scheme 4 Aromatisation of Ugi–Diels–Alder adducts.
Scheme 5 Mechanism of the ring opening-aromatisation sequence.
c
1198 New J. Chem., 2013, 37, 1195--1200
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
NJC
Paper
6.20 (d, 1H, J 5.9); 5.06 (dd, 1H, J 4.5 J 1.6); 4.62 (d, 1H, J 15.2);
4.25 (s, 1H); 3.83 (d, 1H, J 15.2); 3.72 (s, 3H); 2.49 (m, 1H);
1.87 (ddd J 11.5 J 4.5 J 3.8); 1.49 (dd, 1H, J 11.5 J 8.9); 1.26
(s, 9H). 13C NMR (d6-DMSO, 100 MHz, ppm): 174.1; 167.0;
159.0; 137.6; 132.6; 129.3; 128.8; 114.3; 91.1; 78.7; 62.6; 55.6;
51.2; 46.4; 44.7; 28.7; 28.4. IR (ATR, cmꢀ1): 3314; 3083; 2966;
1686; 1663; 1553; 1517; 1453; 1249; 1224; 1189; 1030; 929; 791;
714; 709. MS (DCI, NH3, m/z): 371 (MH+), 388 (MNH4+); HRMS
(ESI) expected for C21H27N2O4 (MH+): 371.1971; found:
371.1966.
General procedure for aromatisation
To a 0.1 M solution of cycloadduct in toluene in a two-necked
round-bottomed flask equipped with a reflux condenser was
added PTSAꢁH2O (1 eq.). The mixture was heated to reflux until
Fig. 2 Direct phenol formation from 5j.
process less efficient (as it must go through a secondary completion of the reaction (indicated by disappearance of the
carbocation). Finally when R4 is a bromine, the ring opening starting material on TLC), then cooled back to RT and poured
leads directly to the phenol via its tautomeric keto form in a 3 : 2 mixture of dichloromethane and saturated aqueous
(see Fig. 2).
sodium bicarbonate. The organic layer was separated, washed
again with saturated aqueous sodium bicarbonate and brine,
dried over anhydrous sodium sulfate and evaporated to dryness.
When required, the product was purified by flash chromatography
using ethyl acetate–cyclohexane mixtures. Spectroscopic data for 9c:
1H NMR (CDCl3, 400 MHz, ppm): 7.67 (d, 1H, J 0.9); 7.48 (d, 1H,
J 7.8); 7.37 (dd, 1H, J 7.8 J 0.9); 7.27 (m, 2H); 6.87 (m, 2H); 5.40 (sbr,
1H); 5.06 (d, 1H, J 14.6); 4.69 (s, 1H); 4.37 (d, 1H, J 14.6); 3.80 (s, 3H);
2.46 (s, 3H); 1.19 (s, 9H). 13C NMR (CDCl3, 100 MHz, ppm): 170.1;
166.8; 159.4; 139.2; 139.0; 133.5; 130.7; 130.1; 129.5; 128.7; 124.2;
114.4; 65.2; 55.4; 51.4; 45.8; 28.4; 21.4. IR (ATR, cmꢀ1): 3304; 2957;
2929; 1694; 1663; 1543; 1515; 1444; 1307; 1244; 1207; 1188; 1029;
849; 832; 818; 768; 741. MS (DCI, NH3, m/z): 367 (MH+). HRMS
(DCI, CH4) expected for C22H27N2O3 (MH+): 367.2022; found:
367.2024
Conclusions
We have shown that a two-step approach to arenes can be
achieved via a tandem Ugi–Diels–Alder/acid-catalysed aromatisa-
tion sequence from furfural and derivatives. This represents an
interesting entry to isoindolinone, a moiety found in various
biologically active compounds. Moreover, differently substituted
biphenyl moieties have been reached without using any transi-
tion metal catalysis under mild conditions and from a renewable
resource. This approach is therefore a cost-effective access to
polysubstituted arenes (including biphenyls) and represents an
interesting starting point for further investigations required to
make the process not only economically – but also environmen-
tally – friendly; this is the goal of our current investigations since
only the combination of these two factors can make a process
truly sustainable.
Further analytical data are provided in the ESI.†
Acknowledgements
The authors thank the C.N.R.S. and the University Paul Sabatier
for their support. Minakem holding and the Region Midi-
Representative experimental details
General procedure for the tandem Ugi–Diels–Alder reactions
´ ´
Pyrenees are gratefully acknowledged for a PhD grant (G.C.)
and generous sponsorship. Prof. Peter Faller is thanked for his
Methanol (6 mL) was heated at 50 1C in a two-necked round- continuous support, as is Prof. Donald Craig for his accurate
bottomed flask equipped with a condenser, then furfural suggestions.
(2 mmol), p-methoxybenzylamine (2 mmol), acrylic acid
(2 mmol) and tert-butylisocyanide (2.4 mmol) were added
successively. The mixture was stirred at 50 1C until completion
Notes and references
of the reaction (monitored by TLC and GC), then cooled back to
RT and diluted with dichloromethane (60 mL). The solution
was washed with saturated aqueous sodium bicarbonate
(30 mL), saturated aqueous ammonium chloride (30 mL) and
brine (30 mL); the organic layer was dried over anhydrous
sodium sulfate and evaporated to dryness. The residue was
purified by flash chromatography with ethyl acetate–cyclo-
hexane as system of eluents to yield the pure polycyclic adducts.
Spectroscopic data for 5a: 1H NMR (d6-DMSO, 300 MHz, ppm):
8.16 (s, 1H); 7.10 (m, 2H); 6.88 (m, 2H); 6.44 (dd, 1H, J 5.9 J 1.6);
1 K. Gairola and I. Smirnova, Bioresour. Technol., 2012, 123,
592–598.
2 L. Mao, L. Zhang, N. Gao and A. Li, Bioresour. Technol., 2012,
123, 324–331.
3 S. Dutta, S. De, B. Saha and M. I. Alam, Catal. Sci. Technol.,
2012, 2, 2025.
4 P. Gallezot, Chem. Soc. Rev., 2012, 41, 1538.
5 J. G. Stevens, R. A. Bourne, M. V. Twigg and M. Poliakoff,
Angew. Chem., Int. Ed., 2010, 49, 8856–8859.
6 M. Juhl and D. Tanner, Chem. Soc. Rev., 2009, 38, 2983.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 1195--1200 1199

View Article Online
Paper
NJC
7 L. El Kaim and L. Grimaud, Tetrahedron, 2009, 65, 2153–2171. 17 C. P. Gordon, N. Byrne and A. McCluskey, Green Chem.,
¨
8 A. Domling, Chem. Rev., 2006, 106, 17–89.
2010, 12, 1000.
9 A. Domling, W. Wang and K. Wang, Chem. Rev., 2012, 112, 18 C. P. Gordon, K. A. Young, M. J. Robertson, T. A. Hill and
3083–3135. A. McCluskey, Tetrahedron, 2011, 67, 554–561.
10 G. Masson, L. Neuville, C. Bughin, A. Fayol and J. Zhu, 19 R. C. Boutelle and B. H. Northrop, J. Org. Chem., 2011, 76,
¨
Multicomponent Syntheses of Macrocycles, Springer, Berlin,
7994–8002.
Heidelberg, 2010, vol. 25.
20 D. Craig, Chem. Soc. Rev., 1987, 16, 187.
11 A. Y. Shaw, Z. Xu and C. Hulme, Tetrahedron Lett., 2012, 53, 21 S. N. Pieniazek and K. N. Houk, Angew. Chem., Int. Ed., 2006,
1998–2000.
45, 1442–1445.
12 K. Paulvannan, Tetrahedron Lett., 1999, 40, 1851–1854.
22 M. S. Newman and V. Lee, J. Org. Chem., 1977, 42, 1478–1479.
13 D. Lee, J. K. Sello and S. L. Schreiber, Org. Lett., 2000, 2, 709–712. 23 R. F. Guignard and S. Z. Zard, Chem. Commun., 2011, 47,
14 M. Oikawa, M. Ikoma and M. Sasaki, Tetrahedron Lett., 2005,
46, 415–418.
12185.
24 X. Huang and J. Xu, J. Org. Chem., 2009, 74, 8859–8861.
15 M. Ikoma, M. Oikawa and M. Sasaki, Eur. J. Org. Chem., 25 S. H. Chan, C. Y. Yick and H. N. C. Wong, Tetrahedron, 2002,
2009, 72–84. 58, 9413–9422.
16 N. Cheron, R. Ramozzi, L. E. Kaım, L. Grimaud and 26 P. S. Sarang, A. A. Yadav, P. S. Patil, U. M. Krishna, G. K. Trivedi
´
¨
P. Fleurat-Lessard, J. Org. Chem., 2012, 77, 1361–1366.
and M. M. Salunkhe, Synthesis, 2007, 1091–1095.
c
1200 New J. Chem., 2013, 37, 1195--1200
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013