Bio-based crotonic acid from polyhydroxybutyrate: synthesis and photocatalyzed hydroacylation

Article abstract of DOI:10.1039/d1gc00421b

A novel thermolytic distillation process was developed to depolymerize polyhydroxybutyrate (PHB) for the selective production of crotonic acid. The conditions adopted (170 °C, 150 mbar) were applied to pure PHB and PHB-enriched bacteria containing 60 and 30% of PHB, giving a recovery of crotonic acid of 92, 78 and 58%, respectively. The high efficiency of the developed process poses the basis for a drop-in production of bio-based crotonic acid, whose versatility as a platform chemical has been investigated through a photochemical approach. The photocatalytic addition (promoted by tetrabutylammonium decatungstate - TBADT) of aliphatic and aromatic aldehydes to crotonic acid took place under solar-simulated light irradiation. TBADT triggered thein situformation of valuable acyl radicals from the corresponding aldehydes, thus inducing the desired hydroacylationviaradical conjugate addition. Notably, the functionalization took place in a satisfactory yield quite independently of the adopted sample of crotonic acid (whether commercial or bio-based).

Full text of DOI:10.1039/d1gc00421b

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Bio-based crotonic acid from polyhydroxybutyrate:
synthesis and photocatalyzed hydroacylation†
Cite this: Green Chem., 2021, 23,
3420
Adriano Parodi, a Alexandra Jorea, b Maurizio Fagnoni, b Davide Ravelli,
*
b
a
Chiara Samorì, *a Cristian Torri a and Paola Galletti
A novel thermolytic distillation process was developed to depolymerize polyhydroxybutyrate (PHB) for the
selective production of crotonic acid. The conditions adopted (170 °C, 150 mbar) were applied to pure
PHB and PHB-enriched bacteria containing 60 and 30% of PHB, giving a recovery of crotonic acid of 92,
78 and 58%, respectively. The high efficiency of the developed process poses the basis for a drop-in pro-
duction of bio-based crotonic acid, whose versatility as a platform chemical has been investigated
through a photochemical approach. The photocatalytic addition (promoted by tetrabutylammonium dec-
atungstate – TBADT) of aliphatic and aromatic aldehydes to crotonic acid took place under solar-simu-
lated light irradiation. TBADT triggered the in situ formation of valuable acyl radicals from the corres-
ponding aldehydes, thus inducing the desired hydroacylation via radical conjugate addition. Notably, the
functionalization took place in a satisfactory yield quite independently of the adopted sample of crotonic
acid (whether commercial or bio-based).
Received 3rd February 2021,
Accepted 1st April 2021
DOI: 10.1039/d1gc00421b
rsc.li/greenchem
through a drop-in strategy is the selective depolymerization of
specific biopolymers, such as polyhydroxybutyrate (PHB). PHB
Introduction
Crotonic acid (CA) and its esters find application in the belongs to the class of polyhydroxyalkanoates (PHAs), bio-
coating, paint, textile, binder, adhesive, flocculant, ceramics, based polyesters produced by a variety of aerobic bacteria able
and agrochemical industries.1 The main industrial use of CA is to ferment various carbon sources and accumulate PHA-gran-
as a building block in the synthesis of co-polymers with vinyl ules inside the cells for energy and carbon storage. Selective
acetate (trade name of Mowilith, Vinnapas, and Vinac). CA and depolymerization of PHB, isolated or as inclusions inside bac-
its cis-isomer (isocrotonic acid) are in fact quite prone to terial cells, has been explored applying both thermal (hydro-
polymerization in analogy to other α,β-unsaturated com- thermal treatment or pyrolysis) and chemical/enzymatic pro-
pounds (e.g. acrylic acid). The industrial synthesis of CA cesses demonstrating that, in principle, it is possible to
involves the oxidation of crotonaldehyde in a two-step process, produce different chemicals, including propene,3–5 CA and iso-
in which peroxocrotonic acid is initially formed as an inter- crotonic acid,4,6–11 methyl crotonate,12,13 methyl acrylate,12
mediate and then reacts with a further molecule of crotonalde- cyclic and linear oligomers,7,14 3-hydroxybutyric acid,4,8 methyl
hyde to yield CA. Crotonaldehyde, the direct precursor of CA, 3-hydroxybutanoate,15 and hydrocarbon oil.16 The cleavage
is industrially produced through the self-condensation of acet- and re-arrangement of PHB-bonds is a complementary strategy
aldehyde followed by dehydration of acetaldol, and acet- to the use of PHB as a polymer in the bio-based plastic
aldehyde in turn is directly derived from the ethylene platform market, and poses the basis for PHB-exploitation as a “plat-
(Wacker oxidation).2 An interesting alternative pathway for form chemical”. This strategy can be particularly useful to
synthesizing bio-based CA without the use of fossil resources manage “challenging PHAs”, such as: (i) PHA that does not
have suitable chemo-physical properties for the polymer
market (e.g. molecular weight below 0.5 MDa);17 (ii) PHA with
a fluctuating monomer ratio (and properties) due to time-
varying carbon source composition (e.g. waste);13 and (iii) low
amount-PHAs inside bacterial cells, thus difficult or non-con-
venient to be extracted.17,18 In these contexts, depolymeriza-
tion approaches dedicated to produce drop-in chemicals from
“challenging PHAs” appear economically and environmentally
preferred to PHA recovery. When intracellular PHA is directly
used as feedstock to produce chemicals, no cell release is
aDepartment of Chemistry “Giacomo Ciamician”, University of Bologna,
Via S. Alberto 163, 48123 Ravenna, Italy. E-mail: chiara.samori3@unibo.it
bPhotoGreen Lab, Dipartimento di Chimica, Università di Pavia, viale Taramelli 12,
27100 Pavia, Italy. E-mail: davide.ravelli@unipv.it
†Electronic supplementary information (ESI) available: Detailed PHB production
through mixed microbial culture; description of thermal procedures (TP1–3);
NMR spectra of CAPHB, CA60 and CA360; characterization and NMR spectra of
products 6–12; yield determination via NMR spectra analysis. See DOI: 10.1039/
d1gc00421b
3420 | Green Chem., 2021, 23, 3420–3427
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
required, lowering the production costs of such chemicals and carboxylic acid),30 conjugated acids are rarely used as Michael
the carbon footprint due to a lower energy demand.12 Thermal acceptors in photocatalyzed syntheses.31,32 In particular, the
degradation of PHA and PHB (eventually assisted by Mg(OH)2, acylation via acyl radicals, a promising route for the preparation
MgO or CaO catalysts)9,19–21 occurs through β-elimination reac- of γ-keto acids, is virtually unexplored. As for the case of CA,
tions that randomly break the chain and give dehydrated trans- only a couple of examples have been reported for its acylation
alkenoic acids as major products (e.g. CA from hydroxybutyrate (to give VI), both exploiting the homolytic cleavage of the C–H
units).14 To the best of our knowledge, there are no reports bond of the formyl group of an aldehyde. The first involves
describing the production of alkenoic acids (specifically CA) heating in the presence of a peroxide (Scheme 1, path b),33
through thermolysis of PHA at temperatures below 200 °C, and while the other makes use of a photocatalyzed reaction (path c,
just three reports on the production of CA through thermolysis benzophenone as the photocatalyst).34
of PHB inclusions inside bacterial cells.4,5,11
The present paper aims to propose a novel and milder way
The chemical reactivity of CA explored so far includes poly- to get high-quality CA from PHB and PHB inclusions inside
merizations, esterifications, and additions to the double bond bacterial cells with different content of PHB (30 and 60%), and
to yield mono and di-substituted butanoic acids.1 Notably, the to explore CA versatility as substrate in valuable photocatalyzed
selective functionalization of this type of α,β-unsaturated acid is acylation reactions, offering a novel thermochemical-photo-
not an easy task because of the presence of acidic hydrogens in catalytic green route for synthesizing valuable γ-keto acids.
the COOH group and the γ-position. The nucleophilic alkylation To this purpose, the proposed approach makes use of a low-
of CA may lead to mixtures of products, as exemplified in temperature thermal treatment (below 200 °C), never reported
Scheme 1 (path a). In fact, along with β-alkylation (compound I) before but capable to selectively break the bonds of both pure
and O-alkylation (II), the deprotonation would lead to dienedio- PHB and PHB inclusions inside bacterial cells. This has been
lates causing a γ- (III) and an α-alkylation (IV). Moreover, simple combined to acylation reactions occurring under photocata-
protonation of the carbanion may induce a C–C double bond lyzed conditions35–38 by using the decatungstate anion,39–41
shift to give deconjugated derivative V.22–26 As a further draw- known to cleave the C(vO)–H bond in aldehydes and proved
back, at least two equivalents of the strong base/nucleophile are to be very robust in several acylation reactions.42–49
required, since one equivalent is simply devoted to the deproto-
nation of the COOH group. A milder and greener approach for
the functionalization of α,β-unsaturated acids could be conju-
gate addition of carbon-based radicals generated under photo-
Experimental
Chemicals
catalytic conditions via a hydrogen atom transfer (HAT) reaction
from suitable hydrogen donors.27–29 Although one of the
All reagents, solvents and chemicals used in this work were of
analytical grade and purchased from various commercial sup-
pliers. Commercial polyhydroxybutyrate (PHB) with a mean
molecular weight of 0.8 MDa and a polydispersity index of 5.9
first photocatalytic reactions ever-tested made use of an
α,β-unsaturated acid (the addition of isopropanol onto maleic
acid to give terebic acid, tetrahydro-2,2-dimethyl-5-oxo-3-furan-
(ref. 18) was bought from Biomer (Germany). Commercial cro-
tonic acid (hereafter named CAC) was purchased from Sigma-
Aldrich and had a purity of 98%. Aldehydes (5a–g) used in the
photocatalytic experiments were purified before use. HPLC-
grade acetonitrile was employed for the photochemical reac-
tions. Tetrabutylammonium decatungstate (nBu4N)4[W10O32
]
(TBADT) was synthesized as previously reported.47
PHB production by mixed microbial cultures
PHB-containing mixed microbial cultures (MMC) were culti-
vated in a 0.75 m3 prototype consisting of a sequencing batch
reactor (500 L) and an accumulation reactor (250 L; Fig. S1 and
detailed description in the ESI†). Two batches of freeze-dried
MMC samples were prepared and used: one with 30% of PHB
(hereafter named PHB-MMC-30) and the other with 60% of
PHB (hereafter named PHB-MMC-60).
Thermal treatment of PHB and PHB-MMC samples
Scheme 1 (a) Functionalization of crotonic acid (CA) by nucleophilic
alkylation under basic conditions (path a) and by addition of an acyl
radical generated from the corresponding aldehyde under thermal
(path b) and photocatalytic (path c) conditions. (b) Thermochemical-
photocatalytic route for the eco-sustainable synthesis of γ-keto acids via
bio-based CA.
Three different thermal procedures (TP1–3, detailed in the
ESI† and summarized in Table 1) with various configurations
(Fig. S2 and S3†) have been applied to PHB (sample of 5 g),
PHB-MMC-60 (sample of 8.34 g) and PHB-MMC-30 (sample of
16.67 g), to give a distilled fraction enriched in CA.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3420–3427 | 3421

View Article Online
Paper
Green Chemistry
Table 1 Thermal procedures (TP1–3) applied in this work
GC-MS analysis
GC-MS analysis of chemicals produced by thermolysis of PHB
or PHB-enriched bacteria was performed using an Agilent
7820A gas chromatograph connected to an Agilent 5977E
quadrupole mass spectrometer. The injection port tempera-
ture was 280 °C. Analytes were separated on a DB-FFAP polar
column (30 m length, 0.25 mm i.d., 0.25 μm film thickness),
with a helium flow of 1 mL min−1. Mass spectra were recorded
under electron ionization (70 eV) at a frequency of 1 scan per s
within the 29–450 m/z range. The temperature of the column
was set to 50 °C (5 min) and increased to 250 °C (10 °C min−1).
A calibration curve was recorded with CAC in a 50–500 ppm
range; the yield of CA (YieldCA-distillates) and by-products (isocro-
tonic acid (1), 3-butenoic acid (2), crotonamide (3), and dimer
of PHB (4)) in each distilled sample was determined by using
Thermal
procedure
Pressure
(mbar)
Time
(min)
Step
T (°C)
TP1
TP2
Thermolysis
Distillation
Thermolysis
290
170
240
(or 290)
290
170
170
150
150
150
30
60
30
(or 15)
30
60
60
Thermolysis
Distillation
Thermolytic distillation
150
150
150
TP3
The yield of the distilled fractions produced through each
TP (on PHB or biomass weight basis in the case of pure PHB
or bacterial biomass, respectively) was determined as follows:
Yielddistillatesðw=wPHB%Þ
1
this calibration curve and confirmed by H NMR (see below).
¼ distillatesðgÞ=PHBðgÞ ꢀ 100
The identification of 1–3 was achieved by comparing their
mass spectra with the NIST spectra database (https://chem-
data.nist.gov/). The identification of the dimer of PHB (4) was
achieved by comparison with mass fragmentation reported in
the literature.50
Yielddistillatesðw=wbiomass%Þ
¼ distillatesðgÞ=biomassðgÞ ꢀ 100:
The yield of CA on a distilled fraction weight basis
(YieldCA-distillates, w/wdistillates%) was calculated by GC-MS ana-
lysis as described below.
GC-FID analysis
GC-FID analysis to monitor the progress of the photocatalytic
reactions was performed on an Agilent 7820A instrument. The
injection was performed at 250 °C in split mode. The initial
oven temperature of 80 °C was maintained for 2 min,
increased by 10 °C min−1 to 250 °C and maintained for 5 min.
An Agilent HP5 capillary column (30 m length, 0.32 mm i.d.,
0.25 µm film thickness) was used, with nitrogen as the carrier
Consequently, the yield of CA on PHB or biomass weight
basis was determined as follows:
YieldCAðw=wPHB%Þ ¼ yielddistillatesðw=wPHB%Þ
ꢀ yieldCA‐distillatesðw=wdistillates%Þ
YieldCAðw=wbiomass%Þ ¼ yielddistillatesðw=wbiomass%Þ
ꢀ yieldCA‐distillatesðw=wdistillates%Þ:
gas at a constant flow rate of 6.0 mL min−1
.
The recovery of CA (%) achievable from each sample was
calculated as follows, on the assumption that 1 g of PHB will
give 1 g of CA (YieldCA-theoretical, w/wPHB%), independent of the
fact that pure PHB or PHB inclusions have been treated:
NMR analysis
1H and 13C NMR spectra were recorded on a 300 or 200 MHz
and a 75 MHz spectrometer, respectively. All NMR spectra were
acquired using CDCl3 as the solvent. Attributions were made
1
1
based on H and 13C NMR. Data for H NMR are reported as
follows: chemical shifts reported in ppm and referred to
residual chloroform (CHCl3), multiplicity (s = singlet, bs =
broad singlet, d = doublet, t = triplet, q = quadruplet, m = mul-
tiplet), integration and coupling constant (J, Hz). Data for 13C
NMR are reported as chemical shifts (ppm) and referred to
residual chloroform (CHCl3). The reported NMR yields are
RecoveryCAð%Þ
¼ yieldCAðw=w%Þ=yieldCA‐theoreticalðw=wPHB%Þ:
General procedure for CA acylation
A degassed solution (by argon bubbling for 10 minutes) of CA
(0.65 mmol, 0.13 M, 56 mg), the chosen aldehyde 5a–g
(0.75 mmol, 0.15 M) and TBADT (2 mol%, 43 mg) in 5 mL of
acetonitrile was exposed for 24 h to simulated sunlight in a
closed Pyrex vessel, using a Solarbox (Solarbox 1500e; CO.FO.
ME.GRA., Italy, equipped with a 1.5 kW Xenon lamp; light
intensity: 500 W m−2). The progress of the reaction was moni-
tored by GC-FID as described below and, upon completion, the
crude mixture was poured into a round-bottom flask and the
solvent was removed via rotary evaporation. Then, the reaction
product was isolated by column chromatography using SiO2 as
1
based on 200 MHz H NMR spectra and have been calculated
by adding to the crude reaction mixture a known amount of
dibromomethane (CH2Br2) used as an external standard (see
the ESI† for details).
Results and discussion
Thermal treatment of PHB and PHB-MMC samples
the stationary phase and mixtures of cyclohexane/ethyl acetate The first thermal treatment here applied to PHB and PHB-con-
as eluants. Column chromatography was performed on an taining biomass (TP1, thermolysis–distillation) consisted of
Isolera Spektra One (Biotage, Sweden), using Sepachrom thermolysis at 290 °C (the maximum degradation temperature
Purezza Daily-Open Load cartridges (Sepachrom Srl, Italy).
of PHB),6,7,9,19 followed by a distillation at 170 °C, both under
3422 | Green Chem., 2021, 23, 3420–3427
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
Table 2 Yield of the distilled fraction (on PHB or PHB-MMC input basis) and its composition in GC-MS detectable compounds, yield of CA (on PHB
or PHB-MMC input basis) and recovery of CA obtained through thermal procedures TP1–3 applied to PHB and PHB-MMC samples
Product yield
Yielddistillates (w/wdistillates%)
(w/wPHB% or
a
YieldCA
b
(w/wPHB% or RecoveryCA
w/wbiomass%) (%)
Entry TP
TP steps
Substrate
w/wbiomass%) CA
1
2
3
4
1
2
3
4
5
6
7
8
TP1 Thermolysis
PHB
PHB
99
92
85
96
3.5 0.6
2.6 0.2
—
—
9.3 84
84
88
26
51
52
78
61
77
92
58
78
TP1 Thermolysis–distillation
TP1 Thermolysis–distillation
TP1 Thermolysis–distillation
—
88
8
PHB-MMC-30 12
PHB-MMC-60 39
64 10.6 3.8 0.7 5.2
79
86
94
87
94
98
92
96
8.5 3.5 0.4 1.0 31
TP2 Thermolysisc–thermolysis–distillation PHB-MMC-30 18
TP2 Thermolysisc–thermolysis–distillation PHB-MMC-60 50
TP2 Thermolysisd–thermolysis–distillation PHB-MMC-30 21
TP2 Thermolysisd–thermolysis–distillation PHB-MMC-60 49
8.4 1.8 0.3
1.4 0.6
8.7 1.8 0.6
—
—
—
—
15
47
18
46
—
4.1 0.9
—
—
0.5
—
9
10
11
TP3 Thermolytic distillation
TP3 Thermolytic distillation
TP3 Thermolytic distillation
PHB
94
0.1
0.1
0.5
—
—
—
0.2 92
PHB-MMC-30 19
PHB-MMC-60 49
—
—
17
47
a Calculated as: YieldCA = Yielddistillates × YieldCA-distillates. b Calculated on the assumption that 1 g of PHB would give 1 g of CA, independent of the
fact that pure PHB or PHB inclusions have been treated, as: RecoveryCA = YieldCA/YieldCA-theoretical. c First thermolysis step performed at 290 °C.
d First thermolysis step performed at 240 °C.
reduced pressure (150 mbar). TP1 was initially tested on pure TP1 applied on PHB-MMC-30 was 8%, corresponding to a
PHB (entries 1 and 2, Table 2), verifying the effect of each step recovery of CA of 26%, whereas the yield of CA (w/wPHB%)
of the thermal procedure (thermolysis alone, entry 1, or ther- applied on PHB-MMC-60 was 31%, corresponding to a recovery
molysis followed by distillation, entry 2) on the yield of the dis- of CA of 51%.
tilled fraction (w/wPHB%) and CA (w/wdistillates%). The yields of
Since these results were lower than those achievable from
the distilled fraction were almost quantitative in both cases, in pure PHB (entry 2, Table 2), another thermal procedure named
line with those reported by other authors,9 but the final distil- TP2 (thermolysis–thermolysis–distillation) was set up and
lation step increased the yield of CA up to 96% (entry 2). The tested. TP2 consisted of a first thermolysis at 240 or 290 °C,
dimer of PHB (4) was the by-product whose amount decreased followed by a second thermolysis at 290 °C, and a final distilla-
to the largest extent after the distillation, both for its physical tion at 170 °C (each step under reduced pressure, 150 mbar).
separation under reduced pressure and for its further thermo- The main difference between configurations TP1 and TP2
lysis that could eventually occur at 170 °C under the same con- relied on preliminary thermolysis (at 240 or 290 °C) dedicated
ditions. CA yield slightly increased after the distillation to the pre-formation of PHB-oligomers, to physically separate
(entries 1 and 2, Table 2), whereas the amount of the other two this fraction from bacterial biomass residue that could have a
alkenoic acids (isocrotonic acid 1 and 3-butenoic acid 2) was role in promoting side-reactions and by-product formation.
substantially the same before and after such step. The yield of This preliminary thermolysis performed at 290 °C, followed by
CA on the PHB input basis (w/wPHB%) achieved through TP1 the two steps already adopted in TP1, gave a large improve-
was 88%. TP1 was also tested on bacterial biomass containing ment in terms of CA yield (entries 5 and 6, Table 2). Although
30 and 60% of PHB (entries 3 and 4, Table 2). It appeared the yield of the distilled fraction was slightly higher than that
evident that the presence of the “non-PHB cell mass” (NPCM) obtained with TP1 (18 vs. 12% in the case of PHB-MMC-30,
had a detrimental effect on the production of CA. The yield of and 50 vs. 39% in the case of PHB-MMC-60), the yield of CA in
CA on the distilled fraction basis (w/wdistillates%) was signifi- this fraction increased from 64 to 86% in the case of
cantly lower than that obtained when the same procedure was PHB-MMC-30, and from 79 to 94% in the case of
applied to pure PHB, whereas all GC-MS detectable by-pro- PHB-MMC-60, optimizing the purity of the distillate. The yield
ducts drastically increased. The amount of non-GC-MS detect- of CA (w/wPHB%) achieved through TP2 with the first thermoly-
able by-products became relevant, especially in the case of sis at 290 °C applied on both PHB-MMC-30 and PHB-MMC-60
PHB-MMC-30. The yield of CA (w/wPHB%) achieved through was 15 and 47%, respectively (CA recovery of 52 and 78%,
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3420–3427 | 3423

View Article Online
Paper
Green Chemistry
respectively). Similar results were obtained when the first ther-
molysis was applied at a lower temperature (240 °C, entries 7
and 8, Table 2). It is noteworthy that this finding suggested
the possibility that a full thermal decomposition of PHB could
occur even operating at a lower temperature than that at which
PHB usually decomposes. From this evidence, a last thermal
procedure (TP3, here named “thermolytic distillation”) has
been set up. TP3 consisted of a distillation at 170 °C under
reduced pressure (150 mbar), largely below the maximum
decomposition temperature of PHB (250–310 °C).6,7,9,19
Differently from TP1 and TP2, distillation was the only process
applied to PHB or PHB-MMC (entries 9–11, Table 2), thus the
only one potentially able to break PHB bonds. The yield of CA
(92 w/wPHB%, entry 9) here achieved by applying TP3 to PHB
was significantly higher than the results obtained through
other catalytic/non-catalytic thermal procedures reported in
the literature (68–83%).4,7,17 The yield of CA (w/wPHB%) from
PHB-MMC-30 with TP3 (entry 10) was analogous to that
obtained with TP2 in three steps (compare entries 5, 7 and 10),
but with a composition of the distilled fraction much more
enriched in CA than in other by-products. The same held for
PHB-MMC-60 treated with TP3 (entry 11), with a final CA yield
(w/wPHB%) of 47%, and a distilled fraction more selectively
enriched in this compound. To the best of our knowledge, the
only three thermal procedures already applied in the literature
to convert PHB inclusions (10 or 66% PHB content) into
CA 4,5,11 gave CA recoveries of 20–26 and 60%, respectively,
mainly due to a large presence of dimers of PHB. By applying
TP3 to MMC with 30 or 60% of PHB, the CA recovery was 58
and 78%, respectively. The thermolysis conditions adopted in
TP3 seem to be the key aspect underpinning the improved per-
formance of the procedure here developed with respect to lit-
erature reports, usually performed at the decomposition temp-
erature of PHB (280–290 °C). The temperature of 170 °C
applied under reduced pressure (150 mbar) permits slow
decomposition of PHB and selective recovery of the monomers
through distillation. Since oligomers have higher boiling
points than monomers, they cannot be distilled under these
conditions and remain in the flask, so that their complete
monomerization can thus be reached. Moreover, it is reported
that heat favours isomerization,51 and thus it is reasonable to
suppose that at 170 °C (TP3) isomerization side-reactions are
inhibited, favouring the formation of the more stable
monomer (CA, the trans-isomer). To the best of our knowledge,
the “thermolytic distillation” at 170 °C adopted in TP3 is
peculiar of the present study and represents a novel one-step
procedure for degrading selectively PHB and PHB-inclusions
into CA, favouring its recovery.
Fig. 1 CA samples obtained by applying the TP3 procedure to PHB
(CAPHB), PHB-MMC-60 (CA60) and PHB-MMC-30 (CA30).
procedure promoted by tetrabutylammonium decatungstate
(TBADT). The procedure was initially set up on commercial
crotonic acid (CAC), focusing on its functionalization with a
small library of aldehydes under solar-simulated light
irradiation.
As shown in Fig.
2
and according to literature
precedents,45,46 the process is triggered by excited TBADT that
abstracts a hydrogen atom from the formyl C(sp2)–H bond in
Photocatalyzed hydroacylation of CA
After selecting TP3 as the most promising approach to Fig. 2 Photocatalytic hydroacylation of CAC with aldehydes 5a–g.
Reaction conditions: An Ar-bubbled acetonitrile solution (5 mL) contain-
ing CAC (0.13 M, 0.65 mmol), aldehydes 5a–g (0.15 M, 0.75 mmol, 1.15
produce bio-based CA, the procedure was applied to PHB,
PHB-MMC-60 and PHB-MMC-30 to obtain three samples of
CA, hereafter named CAPHB (almost colourless), CA60 (light
brown) and CA30 (dark brown), respectively (Fig. 1). These
equiv.) and TBADT (2 mol%) was irradiated with a Solarbox (1.5 kW Xe
lamp; light intensity: 500 W m−2) for 24 h. The reported yields are
referred to the isolated products after column chromatography (for
samples were subjected to a photocatalyzed hydroacylation further details see the Experimental section and ESI†).
3424 | Green Chem., 2021, 23, 3420–3427
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
aldehydes 5a–g to afford the corresponding acyl radical. This
in turn adds onto the β-position of CAC to give a radical adduct
that leads to the formation of the final product via a back
hydrogen-atom transfer (b-HAT) step. The optimized con-
ditions to reach complete conversion of the starting materials
are as follows: an acetonitrile solution of CAC (0.13 M) and a
slight excess of the chosen aldehyde (5a–g, 0.15 M; 1.15 equiv.)
was irradiated in the presence of a catalytic amount of TBADT
(2 mol%) for 24 h. To our delight, the process worked well
both with aliphatic and aromatic aldehydes to give the corres-
ponding adducts in yields ranging from good to excellent.
Thus, hydrocinnamaldehyde 5a and heptaldehyde 5b gave
adducts 6 and 7 in 71 and 58% isolated yield, respectively.
Control experiments related to the preparation of 6 demon-
strated that the presence of both light and TBADT is manda-
tory for the occurrence of the process. A decrease of the yield
to 36% (according to NMR analysis) was observed in the syn-
thesis of 6 by adding TEMPO (2,2,6,6-tetramethylpiperidine
N-oxyl, 1.15 equiv.) to the reaction mixture, thus confirming
the radical nature of this photocatalytic hydroacylation (see
Table S1†).
Fig. 3 Glass vessels used for the preparation of γ-keto acid 6 before (a)
and after (b) the irradiation in the SolarBox.
Moving to aromatic aldehydes, the reaction worked well on
parent benzaldehyde 5c to give 8 (70% yield), while the pres-
ence of a substituent in the para-position was tolerated, inde-
pendent of its electronic character, as demonstrated by para-
anisaldehyde 5d and para-chlorobenzaldehyde 5e (products 9
Scheme 2 Large scale preparation (3.25 mmol) of 6 starting from CAC
or CA60
.
A difference of only 4% in terms of yield from the best to
and 10). Further modification of the substitution pattern on the worst sample was observed, albeit the colour of the irra-
the aromatic ring, as in the case of meta-chlorobenzaldehyde diated mixture was markedly different (see Fig. 3a and Fig. S4–
5f and 3,4,5-trimethoxybenzaldehyde 5g, smoothly gave the S6†). However, in each case, the final irradiated solutions
expected adducts 11 and 12 in up to 89% isolated yield.
showed the blue colour characteristic of the TBADT-H•
Next, selected hydroacylations of different CA samples species,39,40 although less apparent in the case of CA30
(CAPHB, CA60 and CA30) obtained via the thermolytic distilla- (Fig. 3b). On the other hand, the colour of the starting solution
tion approach optimized in the present work (TP3) were tested, seems to have a role in the preparation of 8, wherein, in con-
also to check whether the coloured impurities present in the trast to the case of CAPHB and CA60, the functionalization of
samples may affect the reaction course (see Fig. 1). Thus, the CA30 afforded the desired product in a markedly lower yield
hydroacylations of CAPHB, CA60 and CA30 to give 6 (from hydro- (50% by NMR).
cinnamaldehyde) and 8 (from benzaldehyde) were tested as
Finally, we repeated the preparation of 6 on a larger scale
model reactions. As shown in Table 3, the substitution of CAC (3.25 mmol, 25 mL solution) by reacting CAC and CA60 with
with the hereby prepared CAPHB, CA60 and CA30 did not aldehyde 5a. To our delight, a very high isolated yield of 6 has
hamper the observed reactivity, with a trend following the been obtained in both cases (88 and 85% from CAC and CA60,
order: CAPHB ∼ CA60 > CA30. Indeed, a limited yield drop respectively), delivering the desired product in >600 mg upon
(around 10%) has been observed in the formation of 6, with irradiation for 48 h (Scheme 2).
the samples prepared via TP3 showing
a very similar
performance.
Conclusions
Table 3 Synthesis of products
6 and 8 by adopting different CA
This work illustrates an improved green route for the efficient
synthesis of bio-based crotonic acid from PHB and extends the
knowledge on its potential use as a platform chemical through
photocatalyzed transformations. The novel thermolytic distilla-
tion process here developed was characterized by milder con-
ditions (170 °C without catalysts) than those reported so far in
the field of PHB-depolymerization processes, but enough to
give better results in terms of PHB depolymerization and CA
purity. The recovery of CA from pure PHB and bacteria contain-
ing 30 and 60% of PHB was 92, 58, and 78%, respectively, con-
samples as starting materialsa
Yield (%)
Product
CAC
CAPHB
CA60
CA30
6
8
74
79
66
75
65
73
62
50
a NMR yields calculated by adding to the crude reaction mixture a
known amount of dibromomethane (CH2Br2) used as the external stan-
dard (for further details see the Experimental section and ESI†).
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3420–3427 | 3425

View Article Online
Paper
Green Chemistry
firming that this procedure can be exploited in a chemical re- 13 J. Spekreijse, J. Le Nôtre, J. P. M. Sanders and E. L. Scott,
cycling perspective independent of the actual PHB “form”
(extracted polymer or bacterial inclusions).
J. Appl. Polym. Sci., 2015, 132, 1–8.
14 H. Nishida, H. Ariffin, Y. Shirai and M. Hassan,
Biopolymers, 2010, 369–386.
The bio-based crotonic acid so obtained was easily deriva-
tized under mild conditions (room temperature with the help 15 X. Song, F. Liu, H. Wang, C. Wang, S. Yu and S. Liu, Polym.
of a small amount of an inorganic photocatalyst), avoiding the Degrad. Stab., 2018, 147, 215–221.
use of aggressive bases and nucleophiles, to give valuable 16 S. Kang and J. Yu, RSC Adv., 2014, 4, 14320–14327.
γ-keto acids. Despite its high purity, the CA derived from PHB- 17 C. Samorì, F. Abbondanzi, P. Galletti, L. Giorgini,
enriched bacteria containing 30 and 60% of PHB shows a
brown coloration. This, however, did not hamper the use of
solar light for the hydroacylation step.
L. Mazzocchetti, C. Torri and E. Tagliavini, Bioresour.
Technol., 2015, 189, 195–202.
18 C. Samorì, M. Basaglia, S. Casella, L. Favaro, P. Galletti,
L. Giorgini, D. Marchi, L. Mazzocchetti, C. Torri and
E. Tagliavini, Green Chem., 2015, 17, 1047–1056.
19 H. Ariffin, H. Nishida, M. A. Hassan and Y. Shirai,
Biotechnol. J., 2010, 5, 484–492.
20 J. C. A. Flanagan, J. Myung, C. S. Criddle and
R. M. Waymouth, ChemistrySelect, 2016, 1, 2327–2331.
21 F. D. Kopinke, M. Remmler and K. Mackenzie, Polym.
Degrad. Stab., 1996, 52, 25–38.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
22 B. Plunian, J. Mortier, M. Vaultier and L. Toupet, J. Org.
Chem., 1996, 61, 5206–5207.
23 M. J. Aurell, S. Gil, R. Mestres, M. Parra and L. Parra,
Tetrahedron, 1998, 54, 4357–4366.
24 M. J. Aurell, R. Mestres and E. Muñoz, Tetrahedron Lett.,
1998, 39, 6351–6354.
The study was conducted within the project “B-PLAS DEMO”
funded by EIT-CLIMATE KIC. We thank Regione Emilia-
Romagna (Bando “Alte competenze per la ricerca e il trasferi-
mento tecnologico” – POR FSE 2014/2020, Obiettivo tematico
10) for PhD funding.
25 B. Plunian, Chem. Commun., 1998, 81–82.
26 M. J. Aurell, L. R. Domingo, R. Mestres, E. Muñoz and
R. J. Zaragozá, Tetrahedron, 1999, 55, 815–830.
Notes and references
1 J. Blumenstein, J. Albert, R. P. Schulz and C. Kohlpaintner, 27 S. Protti, M. Fagnoni and D. Ravelli, ChemCatChem, 2015,
in Ullmann’s Encyclopedia of Industrial Chemistry, Wiley- 7, 1516–1523.
VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2015, 28 L. Capaldo and D. Ravelli, Eur. J. Org. Chem., 2017, 2056–
pp. 1–20. 2071.
2 R. A. Fernandes, A. K. Jha and P. Kumar, Catal. Sci. 29 L. Capaldo, L. L. Quadri and D. Ravelli, Green Chem., 2020,
Technol., 2020, 10, 7448–7470. 22, 3376–3396.
3 C. Torri, T. D. O. Weme, C. Samorì, A. Kiwan and 30 G. O. Schenck, G. Koltzenburg and H. Grossmann, Angew.
D. W. F. Brilman, Environ. Sci. Technol., 2017, 51, 12683–12691. Chem., 1957, 69, 177–178.
4 Y. Li and T. J. Strathmann, Green Chem., 2019, 21, 5586– 31 D. Dondi, S. Protti, A. Albini, S. M. Carpio and M. Fagnoni,
5597. Green Chem., 2009, 11, 1653.
5 J. M. Clark, H. M. Pilath, A. Mittal, W. E. Michener, 32 K. Zhu, T. Ohtani, C. B. Tripathi, D. Uraguchi and T. Ooi,
D. J. Robichaud and D. K. Johnson, J. Phys. Chem. A, 2016,
120, 332–345.
Chem. Lett., 2019, 48, 715–717.
33 R. L. Huang, J. Chem. Soc., 1956, 1749.
6 M. R. Z. Mamat, H. Ariffin, M. A. Hassan and 34 H. Cerfontain and P. C. M. Van Noort, Synthesis, 1980, 490–
M. A. K. M. Zahari, J. Cleaner Prod., 2014, 83, 463–472. 492.
7 H. Morikawa and R. H. Marchessault, Can. J. Chem., 1981, 35 C. Raviola, S. Protti, D. Ravelli and M. Fagnoni, Green
59, 2306–2313. Chem., 2019, 21, 748–764.
8 X. Yang, K. Odelius and M. Hakkarainen, ACS Sustainable 36 A. Banerjee, Z. Lei and M.-Y. Ngai, Synthesis, 2019, 303–
Chem. Eng., 2014, 2, 2198–2203. 333.
9 H. Ariffin, H. Nishida, Y. Shirai and M. A. Hassan, Polym. 37 G. N. Papadopoulos, E. Voutyritsa, N. Kaplaneris and
Degrad. Stab., 2010, 95, 1375–1381. C. G. Kokotos, Chem. – Eur. J., 2018, 24, 1726–1731.
10 C. A. Mullen, A. A. Boateng, D. Schweitzer, K. Sparks and 38 For a related photoinitiated protocol, see: I. K. Sideri,
K. D. Snell, J. Anal. Appl. Pyrolysis, 2014, 107, 40–45.
11 C. Samorì, A. Kiwan, C. Torri, R. Conti, P. Galletti and
E. Voutyritsa and C. G. Kokotos, ChemSusChem, 2019, 12,
4194–4201.
E. Tagliavini, ACS Sustainable Chem. Eng., 2019, 7, 10266– 39 D. Ravelli, S. Protti and M. Fagnoni, Acc. Chem. Res., 2016,
10273. 49, 2232–2242.
12 C. Fernández-Dacosta, J. A. Posada and A. Ramirez, 40 D. Ravelli, M. Fagnoni, T. Fukuyama, T. Nishikawa and
J. Cleaner Prod., 2016, 137, 942–952.
I. Ryu, ACS Catal., 2018, 8, 701–713.
3426 | Green Chem., 2021, 23, 3420–3427
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
41 For
K. van der Wal, D. Ravelli, M. Nuño, M. Fagnoni,
D. Guthrie, Y. Sun and T. Noël, Science, 2020, 369, 47 S. Protti, D. Ravelli, M. Fagnoni and A. Albini, Chem.
92–96. Commun., 2009, 7351–7353.
42 Y. Kuang, H. Cao, H. Tang, J. Chew, W. Chen, X. Shi and 48 F. Bonassi, D. Ravelli, S. Protti and M. Fagnoni, Adv. Synth.
J. Wu, Chem. Sci., 2020, 11, 8912–8918. Catal., 2015, 357, 3687–3695.
43 P. Fan, C. Zhang, Y. Lan, Z. Lin, L. Zhang and C. Wang, 49 L. Capaldo, M. Fagnoni and D. Ravelli, Chem. – Eur. J.,
Chem. Commun., 2019, 55, 12691–12694. 2017, 23, 6527–6530.
44 P. Fan, Y. Lan, C. Zhang and C. Wang, J. Am. Chem. Soc., 50 F. Abbondanzi, G. Biscaro, G. Carvalho, L. Favaro,
a
recent example, see: G. Laudadio, Y. Deng, 46 D. Ravelli, M. Zema, M. Mella, M. Fagnoni and A. Albini,
Org. Biomol. Chem., 2010, 8, 4158–4164.
2020, 142, 2180–2186.
45 S. Esposti, D. Dondi, M. Fagnoni and A. Albini, Angew.
Chem., Int. Ed., 2007, 46, 2531–2534.
P. Lemos, M. Paglione, C. Samorì and C. Torri, New
Biotechnol., 2017, 39, 29–35.
51 M. B. Hocking, Can. J. Chem., 1972, 50, 1224–1232.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 3420–3427 | 3427