Selective ethenolysis and oestrogenicity of compounds from cashew nut shell liquid

Article abstract of DOI:10.1039/c4gc00111g

The ethenolysis of cardanol (2), a waste product from cashew kernel production, was carried out using a variety of metathesis catalysts. Surprisingly, the best activities and selectivities could be observed with ruthenium based 1st generation type catalysts converting cardanol (2) almost completely to the corresponding 1-octene (6) and 3-non-8-enylphenol (4), a potential detergent precursor. Detailed investigation of the reaction system showed that the high activity and selectivity were due to a combination of ethenolysis and internal self-metathesis of the unsaturated cardanol mixture, 2. Self-metathesis of cardanol (2) containing three double bonds led to the formation of 3-non-8-enylphenol (4) and 1,4-cyclohexadiene (7). The latter was crucial for a high selectivity and activity in the ethenolysis, not only of cardanol (2), but also of other substrates like methyl oleate (10) when using ruthenium based 1st generation catalysts. The endocrine disrupting properties of 3-nonylphenol and related compounds are compared. the Partner Organisations 2014.

Full text of DOI:10.1039/c4gc00111g

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Selective ethenolysis and oestrogenicity of
compounds from cashew nut shell liquid†
Cite this: Green Chem., 2014, 16,
2846
Jennifer Julis,a Stuart A. Bartlett,a Sabrina Baader,b Nicola Beresford,c
Edwin J. Routledge,c Catherine S. J. Cazina and David J. Cole-Hamilton*a
The ethenolysis of cardanol (2), a waste product from cashew kernel production, was carried out using a
variety of metathesis catalysts. Surprisingly, the best activities and selectivities could be observed with
ruthenium based 1st generation type catalysts converting cardanol (2) almost completely to the corres-
ponding 1-octene (6) and 3-non-8-enylphenol (4), a potential detergent precursor. Detailed investigation of
the reaction system showed that the high activity and selectivity were due to a combination of ethenolysis
and internal self-metathesis of the unsaturated cardanol mixture, 2. Self-metathesis of cardanol (2) con-
taining three double bonds led to the formation of 3-non-8-enylphenol (4) and 1,4-cyclohexadiene (7).
The latter was crucial for a high selectivity and activity in the ethenolysis, not only of cardanol (2), but also
of other substrates like methyl oleate (10) when using ruthenium based 1st generation catalysts. The
endocrine disrupting properties of 3-nonylphenol and related compounds are compared.
Received 21st January 2014,
Accepted 14th March 2014
DOI: 10.1039/c4gc00111g
www.rsc.org/greenchem
Introduction
In the last decade the up-grading of renewable resources has
received increasing attention. Not only the diminishing fossil
resources and the increasing anthropogenic CO2-emission, but
also the rising conscience of society regarding global warming
have intensified the efforts to establish new supply chains
based on sustainable resources. To avoid competition between
food and chemicals for land use, it is especially attractive to
use waste products, like agricultural waste, as sustainable
resources rather than renewable feedstocks which are also part
of the food supply chain.
In this context, cashew nut shell liquid (CNSL) is a very
interesting biodegradable and renewable natural resource,
which is obtained as a by-product during processing of cashew
nuts. Although 300 000 t per annum of CNSL are currently
available, only few applications exist and most of this raw
material is considered as a waste stream.1
Fig. 1 Components of CNSL.
method of the cashew nuts. Solvent extraction of the nuts gives
predominately anacardic acid whilst roasting of the nuts gives
mainly cardanol (2) due to the decarboxylation of anacardic
acid on heating. All three components have a fifteen carbon
linear chain in the meta-position to the phenolic group with a
varying degree of saturation, dependant on the origin of the
cashew nuts.2–4
CNSL is a phenolic mixture, which consists of three major
compounds, namely anacardic acid (1), cardanol (2) and
cardol (3) (Fig. 1). Their proportion depends on the processing
Anacardic acid (1) can be isolated from CNSL by precipi-
tation with calcium hydroxide. Separation and acidification of
the calcium anacardate gives the pure acid 1, which can be
transformed to cardanol (2) by heating to 200 °C.5 Further
purification by vacuum distillation gives pure cardanol (2)
without alteration of the side-chain. Its versatility as a renew-
able starting material arises from its structure. Due to the
phenol group and the unsaturated side-chain in the meta-posi-
aEaStCHEM, School of Chemistry, University of St. Andrews, St. Andrews, KY16 9ST,
UK. E-mail: djc@st-and.ac.uk; Fax: +44 (0)-1334-463808; Tel: +44 (0)-1334-463805
bFachbereich Chemie, Organische Chemie, TU Kaiserslautern, Erwin-Schrödinger-
Straße 54, 67663 Kaiserslautern, Germany
cInstitute for the Environment, Brunel University, Cleveland Road, Uxbridge,
Middlesex UB8 3PH, UK
†Electronic supplementary information (ESI) available: Origin of the products of
metathesis of cardanol, 1H and 13C NMR spectra and GC analysis of selected
reactions. See DOI: 10.1039/c4gc00111g
2846 | Green Chem., 2014, 16, 2846–2856
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
tion, it can be easily modified to valuable chemicals by intro- isolated from CNSL and heated to 200 °C for 3 h.3 After de-
ducing novel functionalities. carboxylation a brown viscous liquid remained, which was
Several applications of cardanol (2) or CNSL are known. For further purified by distillation at 230 °C under vacuum. 2
example, the synthesis of biscardanol derivatives as mono- could be isolated as a pale yellow, slightly viscous liquid,
mers, additives in surface coatings and resins6 and the syn- which was stored under nitrogen atmosphere. GC-MS and
thesis of sodium cardanol sulfate as detergents7 have all been 1H-NMR analysis showed almost pure cardanol (2), which con-
reported. 2 can also be functionalised to its corresponding tained 38% monoene 2b, 20% diene 2c, 38% triene 2d and 5%
ethers, which have been used as polymer additives8 or in nano- saturated cardanol 2a. Separation of these various unsaturated
fibers.9 However, the selective homogeneous catalysed trans- compounds via column chromatography proved difficult, so
formation of cardanol (2) to valuable intermediates is only the cardanol mixture 2 was used without further purification.
rarely described and just a few examples like the double bond
metathesis of cardanol (2) exist.6,9 Due to its unsaturated side-
Cardanol ethenolysis
chain, metathesis is an attractive tool for functionalising 2 to
intermediates of higher added-value. Vasapollo et al. reviewed
catalysts were tested in the ethenolysis of cardanol 2. During
the transformation of cardanol (2) to new fine chemicals as
well as new hybrid functional materials, such as cardanol por-
phyrins, cardanol phthalocyanines and cardanol fullerenes via
3-non-8-enyl phenol (4) in the case of mono-unsaturated carda-
In a first set of experiments various homogeneous metathesis
ethenolysis, a propagating methylidene species is formed,
which reacts with cardanol (2) to release a terminal alkene
alkene-metathesis.9 However, no applications of the newly
synthesised materials have been reported. Therefore, it is of
particular interest to convert cardanol (2) selectively to inter-
mediates which can be used directly as substitutes in the
value-added chain. We have recently reported the selective syn-
thesis of kairomone and other chemicals from 2.10 Amongst
other reactions, we reported that ethenolysis of cardanol (2), in
that case with a high monoene (2b) content, could lead to
3-non-8-enylphenol (4) and 1-octene (6), which are both poten-
tially important products. 1-Octene 6, is mainly used as a
comonomer for polyethylene with annual sales of 600 000
worth $ 1.2 billion11 and 3-nonylphenol has the potential for
replacement of 4-nonylphenol, which makes an excellent
detergent via ethoxylation, but has been banned in Europe
because of its endocrine disrupting properties.12,13 Using a
Grubbs–Hoveyda type catalyst for the ethenolysis of cardanol
(2), the selectivity towards the desired products was
moderate.10
nol 2b.15 Most alkene metathesis catalysts are unstable as
methylidene complexes and undergo rapid decomposition,
which affects the selectivity and productivity of the ethenolysis
reaction.14 Therefore, 1st generation type, 2nd generation type
and Grubbs–Hoveyda type catalysts were screened in order to
identify a suitable catalyst system for the selective ethenolysis
of cardanol (Fig. 2).
Due to their varying stability and activity the performance
of the metathesis catalyst was investigated at various tempera-
tures. Beside the expected ethenolysis products, 3-non-8-enyl-
phenol (4) and 1-octene (6), which are formed by cross-
metathesis of ethene with 2 containing only 1 double bond,
1,4-cyclohexadiene (7), 3-dodecadienylphenol (5) and various
isomeric ethenolysis products were also observed in the reac-
tion mixture (Fig. 3).
In a recent study, Grubbs–Hoveyda catalysts containing
unsymmetrical N-heterocyclic carbenes where one N substitu-
ent is alkyl and the other aryl have been shown to be highly
selective catalysts for the ethenolysis of methyl oleate, the best
having di(2-propyl)phenyl and norbornyl substituents at load-
ings as low as 50 ppm, although conversions and reaction
rates are moderate (conversion 47%, selectivity to ethenolysis
products 98% at 500 ppm catalyst loading).14
We now report a detailed study of the selective ethenolysis
of cardanol (2). Unexpectedly, first generation catalysts give
very high selectivity to the desired products and very high con-
versions. We also report oestrogenicity studies on 3-nonyl
phenol which show that it 2 orders of magnitude less oestro-
genic than commercial 4-nonylphenol.
Results
As homogeneous metathesis catalysts are very sensitive
towards impurities, cardanol (2) was purified prior to use.
Fig. 2 Homogeneous metathesis catalysts tested for the ethenolysis of
Following a known literature procedure anacardic acid (1) was cardanol (2).
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2846–2856 | 2847

View Article Online
Paper
Green Chemistry
Fig. 3 Cross-metathesis of cardanol (2) with ethene.
The isomeric ethenolysis products are mainly formed via
isomerisation of the mono-unsaturated cardanol (2b), followed
by cross-metathesis with ethene. Ethenolysis at the unsatu-
rated C11-position of the diene (2c) and triene (2d) leads to the
side-product 3-dodeca-8,11-dienylphenol (5). However, due to
their volatility the corresponding side-products 1,4-pentadiene
and 1-pentene were not observed.
Reacting 2 with M1 at room temperature in the absence of
ethene gives almost exclusive conversion (43%) to 3-non-8-
enylphenol (4, 81% selectivity) and its isomers together with
1,4-cyclohexadiene (7), the remaining unreacted cardanol (2)
contains no 2d (absence of peaks with m/z 298, [M]+; entry 2,
Table 1). This reaction shows that 7 can be formed by direct
internal self-metathesis of the triene (2d) to 7 and 4. However,
in the presence of ethene, 7 might also be formed from the
ethenolysis of the triene (2d) at the unsaturated C8,9-position,
giving 3-non-8-enylphenol (4) and 1,4,7-octatriene, followed by
the internal self-metathesis of the 1,4,7-triene to ethene and 7
(see ESI, Fig. S1†).
Because of the high percentage of triene (2d) in the starting
material, the amount of 1,4-cyclohexadiene (7) is significant
and almost in the same range as 1-octene (6). The reactions
that are believed to occur during ethenolysis of cardanol (2)
are summarised in the ESI, Fig. S1.†
As some of the side-products such as the diene (7) and the
linear alkenes with an alkyl chain < C7 are volatile, it was not
possible reliably to determine their exact amounts via GC ana-
lysis. Therefore, we analysed the distribution of the corres-
ponding alkenylphenols.
Table 1 Conversion and selectivity in the ethenolysis of cardanol (2)
using a variety of ruthenium based catalystsa
Isomeric
products/%
Entry
Catalyst
T/°C
Conv./%
4/%
5/%
1
M1
20
20
40
20
20
40
20
40
70
20
40
70
20
40
70
40
70
90
20
40
70
20
40
70
95
43
96
84
90
91
57
61
84
26
83
87
67
83
87
61
58
92
82
93
90
88
95
89
94
81
93
87
91
90
19
34
68
43
54
49
46
53
59
9
22
75
50
16
22
23
24
22
2
0
3
8
5
4
19
4
5
4
2b
3
4c
5
6
7
8
Grubbs 1st
M2
5
5
14
12
12
25
20
15
24
19
22
12
14
12
12
4
67
54
20
32
27
37
30
28
20
79
65
13
38
80
72
70
66
72
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
M20
M31
Caz-1
M51
M52
6
7
10
7
a 2 (0.54 mmol), C2H4 (8 bar), CH2Cl2 (1.35 mL), catalyst (0.05 mol%),
analysis via GC using n-tetradecane as internal standard 6 h. b No
ethene. c Toluene (1.35 mL) as solvent.
(entries 1 and 3, 5 and 6, Table 1) and changing to the more
The results in Table 1 show that the conversion and the environmentally acceptable solvent, toluene from CH2Cl2 was
selectivity in the ethenolysis of cardanol (2) strongly depend only slightly detrimental (entry 4, Table 1). Furthermore, the
on the metathesis catalyst and the reaction temperature. Quite metathesis reaction of cardanol (2) with M1, but with no
unexpectedly the best results were obtained with the M1 and ethene present, showed that 2 can undergo self-metathesis to
the Grubbs 1st generation catalysts (entries 1–6, Table 1). Both form 1,4-cyclohexadiene (7) and the desired 3-non-8-enylphenol
catalytic systems exhibited an excellent performance in the (4) (entry 2, Table 1). Only the tri-unsaturated cardanol (2d) is
ethenolysis with yields and selectivities >90% towards the able to react to give 4 and 7 via self-metathesis, hence the
desired product, 4 (catalyst loading. 0.05 mol%). Only a minor conversion of the reaction (43%) is similar to the amount of
influence of the temperature on the activity was observed tri-unsaturated cardanol (2d) in the substrate mixture (38%;
2848 | Green Chem., 2014, 16, 2846–2856
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
the slight difference is within the experimental error of the GC the ethenolysis reaction of 2 with M1 did not improve the cata-
measurements). This result indicates that metathesis of 2 with lytic performance. In contrast, we observed no conversion of 2
M1 and ethene probably proceeds by a combination of self- indicating that, in our case, the addition of phenol inhibits the
metathesis (for 2d) and ethenolysis (for 2b and 2c).
cross-metathesis of ethene and cardanol (2).
Metathesis of methyl protected cardanol 8. The effect of the
The N-heterocyclic carbene (NHC) bearing 2nd generation
type16–19 catalysts M2, M20, M31 and Caz-1 (entries 7–18, phenolic group in cardanol (2) was further tested by etherifica-
Table 1) were generally less active and selective in the etheno- tion of the phenolic –OH with methyl iodide (Fig. 4).24 The
lysis of cardanol (2) compared to the Grubbs 1st generation methyl cardanol (8) was tested in the ethenolysis with M1
catalyst and M1, especially at 20 °C where the conversion and under standard reaction conditions. In comparison to the
the selectivity were much lower in comparison to the 1st gene- unprotected cardanol (2) the conversion (62%) and selectivity
ration type catalysts. With increasing temperature the activity towards the desired 3-non-8-enylphenolmethylether (9) (84%)
and selectivity improved with M2 (entries 8 and 9, Table 1) and were both lower than when using cardanol (2) itself as sub-
Caz-1 (entries 17 and 18, Table 1) catalysts giving good per- strate, but higher than when using any of the other metathesis
formances at 70 °C and 90 °C respectively.
catalysts shown in Fig. 2. The result shows that, although the
This increased activity with higher reaction temperature phenolic structure of 2 may have some beneficial effect on the
could be related to the activation energy of the catalyst since it ethenolysis reaction, there must also be some other reason for
is known that the latent Caz-1 must isomerise from cis to trans the excellent results obtained when using M1 or Grubbs 1st
before it becomes active.20,21 For the metathesis reaction the generation catalysts.
catalyst must provide a free coordination site, which is gener-
Ethenolysis of mono-unsaturated cardanol (2b). Further
ated via dissociation of a ligand. In contrast to the tricyclo- information as to the important influences involved in the
hexylphosphine ligands, which readily dissociate at room ethenolysis reactions came from a study of mono-unsaturated
temperature, phosphite ligands are much less labile and need cardanol (2b), which was originally initiated to avoid the for-
higher temperature to leave the metal center.20,21 Nevertheless, mation of 1,4-cyclohexadiene (7) and to maximise the pro-
even at elevated temperature the activities of the 2nd gene- duction of 1-octene (6).
ration type catalysts are lower than those of the 1st generation
Recently, Perdriau et al. reported the selective transfer
and M1 catalysts. The reduced activity in the ethenolysis is not hydrogenation of a cardanol mixture 2 to mono-unsaturated
only visible in the lower conversion, but is also seen in the compound 2b with RuCl3·xH2O in 2-propanol.25 The transfer-
lower selectivities to product 4.
hydrogenation gives access to almost pure 2b, although some
The Grubbs–Hoveyda type catalysts show even less selec- double bond isomerisation occurs during the transfer hydro-
tivity in the ethenolysis of cardanol (2). Only in the case of M51 genation reaction. The mono-unsaturated cardanol 2b was
is significant selectivity towards the desired 3-non-8-enyl- tested in the ethenolysis with M1, Caz-1 and M51 under the
phenol (4) observed at 20 °C (entry 19, Table 1). Furthermore, standard reaction conditions (Table 2).
the activity and selectivity of M51 and M52 show only minor
The reaction was somewhat dependent upon the batch of
dependencies on the reaction temperature (entries 19–24, starting material, but surprisingly, we only observed zero or
Table 1). Both catalytic systems are active at 20 °C and 40 °C very low conversion of the mono-unsaturated cardanol with M1
respectively, but catalyse mainly the self-metathesis reactions (entry 1, Table 2 and footnote b, Table 2). The catalytic system
of 2. In comparison to the 2nd generation type catalysts, the Caz-1 (entry 2, Table 1) and M51 (entry 3, Table 2) showed
boomerang-type ligand in M51 and M52 can easily dissociate to some conversion in the ethenolysis, but in comparison to the
generate a free coordination site. It has been proposed that the unsaturated cardanol mixture 2 the activity and selectivity were
ligand remains close to the metal centre and re-coordinates also much lower. This observation indicates that the di- and
after the catalytic cycle to stabilise the complex,15 but this tri-unsaturated cardanol (2c and 2d) are highly beneficial for
recoordination has been disputed.22 In any case, they are much
more active in the conversion of 2 than the 2nd generation type
catalysts. In general, NHC-based ruthenium catalysts are known
to be more active and stable than the first-generation catalyst
but are significantly less selective in ethenolysis, as they tend to
promote self-metathesis.14
Forman et al. showed that the performance of certain
alkene metathesis reactions by 1st generation catalysts could
be enhanced by the addition of phenols.23 In the presence of
phenol only small quantities of undesired by-products were
detected and the activity of the catalytic system was signifi-
cantly increased. We reasoned, therefore, that the phenol
present in cardanol (2) might be responsible for the excellent
activity and selectivity provided by 1st generation and M1 cata-
Fig. 4 Synthesis of methyl cardanol (8) and ethenolysis to 3-nonenyl-
lysts in the ethenolysis reaction. However, adding phenol to phenylmethyl ether (9).
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2846–2856 | 2849

View Article Online
Paper
Green Chemistry
Table 2 Ethenolysis with mono-unsaturated cardanol (2b)a
Table 4 Ethenolysis of mono-unsaturated cardanol 2b with different
catalysts with or without 1,4-cyclohexadiene (7)a
Entry Catalyst T/°C Conversion/% 4/% Isomeric products/%
Entry Catalyst
Conversion/% 4/% Isomeric products/%
b
1
2
3
M1
0
70
70
0
28
72
0
13
9
60
87
93
1b
2b
3c
4c
Caz-1
Caz-1 + 1,4-CHD 78
M51
M51 + 1,4-CHD
28
13
11
7
87
90
93
98
Caz-1
M51
72
64
a Conditions as in Table 1. b Using a different batch of 2b, the
conversion was 12% and 4 made up 40% of the alkyl phenols.
2
a Conditions as in Table 1. b 70 °C. c 40 °C.
the ethenolysis of cardanol. Especially in the case of M1 their
presence seems to be essential.
mixture 2 is very important for the stabilisation of M1 and its
activity in the metathesis reaction of cardanol (2) with ethene.
Caz-1 and M51 were also tested in the ethenolysis of mono-
unsaturated cardanol (2b) with and without 1,4-cyclohexadiene
(7) as additive, see Table 4.
In the case of Caz-1 the addition of 7 leads to an increase in
conversion of 2b up to 80%. However, mainly isomeric pro-
ducts of cardanol (2) were formed and only 10.5% of the
desired 3-non-8-enylphenol (4) was detected. With M51 no
enhancement of the catalytic performance was observed.
These results indicate that the effect of the diene (7) strongly
depends on the ethenolysis catalyst employed.
Due to its positive effect in the ethenolysis of mono-unsatu-
rated cardanol 2b with M1, we also analysed the influence of
1,4-cyclohexadiene (7) on the ethenolysis of 2b protected via
etherification with methyl iodide (see Table 5).
In contrast to the results of Tables 3 and 4, the addition of
1,4-cyclohexadiene (7) to the reaction mixture only led to a
minor improvements of the catalytic activity when using
methyl protected monounsaturated cardanol 8b as the
substrate.
Ethenolysis of oleate 10 and linoleate esters 12. The highly
beneficial combination of M1 + 1,4-cyclohexadiene (7) was also
tested in the ethenolysis of oleate and linolenate based sub-
strates (Fig. 5).
The results in Table 6 show that 7 also has a positive effect
in the ethenolysis of methyl oleate (10) (entries 1 and 2,
Table 6), increasing the conversion of 10 to the ethenolysis
product methyl 9-decenoate by more than 25%. However, M1
is inactive for the metathesis of methyl linolenate (12) in the
presence or absence of the diene 7 (entries 5 and 6, Table 6).
To confirm that it was not the phenol in cardanol (2) that
was allowing the excellent results obtained in ethenolysis reac-
A major difference between the mono-unsaturated com-
pound 2b and the natural cardanol mixture (2) is the for-
mation of 1,4-cyclohexadiene (7) during the metathesis
reaction of 2d in the latter. To analyse the role of 1,4-cyclohexa-
diene (7) in the ethenolysis reaction, we added 0.1 equivalent
of 1,4-cyclohexadiene (7) to 2b and repeated the metathesis
reaction under the standard reaction conditions (Table 3).
The addition of 1,4-cyclohexadiene (7) had a major impact
on the ethenolysis of mono-unsaturated cardanol (2b)
(compare entries 1 and 2 in Table 3 and footnotes b and c in
Table 3). In the presence of 1,4-cyclohexadiene 2b (64%)
underwent metathesis and 3-non-8-enylphenol (4) was formed
(55% selectivity). The other products were mainly different
chain length metathesis products, which arose from double
bond positional isomers of the monoene (2b) that were
formed during the transfer hydrogenation reaction.25 Since
without diene 7 we observed no or very little reaction, these
results indicate that 7 has a positive effect on the M1 catalyst
during the ethenolysis reaction. We also tested other dienes as
additives in the metathesis of mono-unsaturated cardanol 2b
(entries 3–5, Table 3). With all three additives we could see an
improved activity of the M1 catalyst in the ethenolysis of 2b.
The effect of 1,5-cyclooctadiene on the catalytic activity was
less pronounced, giving a lower conversion (31%, entry 4,
Table 1). This is possibly because 1,5-cyclooctadiene can co-
ordinate through both double bonds to the metal centre and
hence may block coordination of the double bond in 2b.
However, it still gives better catalysis than is obtained in its
absence. These results suggest that the formation of 1,4-cyclo-
hexadiene (7) in the ethenolysis of the natural cardanol
Table 3 Ethenolysis of mono-unsaturated cardanol (2b) with additivesa
Table 5 Ethenolysis of methyl-protected mono-unsaturated cardanol
8b with different catalysts and 7 as additivea
Isomeric
Entry Additive
Conversion/% 4/% products/%
Isomeric
products/%
1
2
3
4
5
Noneb
0
0
55
46
47
46
0
46
55
53
54
1,4-Cyclohexadiene (7)c 64
Entry
Catalyst
Conversion/%
4/%
1,4-Hexadiene
1,5-Cyclooctadiene
1,7-Octadiene
63
31
69
1b
2b
3c
4c
5d
6d
M1
53
56
71
80
77
86
50
53
6
8
2
50
37
94
92
98
97
M1 + 1,4-CHD
Caz-1
Caz-1 + 1,4-CHD
M51
M51 + 1,4-CHD
a Conditions as in Table 1; additive (0.1 equiv.). b Using a different
batch of 2b, the conversion was 12% and 4 made up 40% of the alkyl
phenols. c Using the batch of 2b used for the reaction described in
footnote b, the conversion was 49% and 4 made up 38% of the alkyl
phenols.
3
a Conditions as in Table 1. b rt. c 70 °C. d 40 °C.
2850 | Green Chem., 2014, 16, 2846–2856
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Fig. 5 Oleate (10) and linoleate ester (11).
Table 6 Ethenolysis of different olefinic substrates (10–13, Fig. 5)
a
with M1
Entry
Substrate
Conversion/%
Selectivity/%
1
2
3
4
5
6
7
8
10
47
76
7
13
—
—
36
5
89
94
62
57
—
—
75
—
10 + 1,4-CHD
11
11 + 1,4-CHD
12
12 + 1,4-CHD
Fig. 6 (a) Catalysts used for the ethenolysis of methylated anacardic
acid (14); (b) synthesis and ethenolysis of 14. DMS = dimethylsulfate.
Table 7 Ethenolysis of dimethylated anacardic acid 14a
13
13 + 1,4-CHD
Entry
Metathesis catalyst
T/°C
Conversion/%
Selectivity/%
a Conditions as in Table 1.
1
2
3
4
HG1
M31
M41
M51
M74
HG1
M51
M31
M41
M74
HG1
HG1
HG1
25
25
25
25
25
60
60
60
60
60
25
25
25
93
17
30
96
90
74
88
56
44
93
98
92
78
93
13
18
36
37
57
16
11
19
7
tions using M1, we synthesised the resorcinol esters of oleic
(11) and linolenic (13) acids (Fig. 5). Introducing the phenol
moiety into the oleic ester dramatically reduced the conversion
and the selectivity towards the desired dec-9-enoic acid ester
(entry 3, Table 6), a result that was hardly improved by adding 7
(entry 4, Table 6). In the case of linolenate, there was some
small improvement as a result of using the resorcinol ester 13
(entry 7, Table 6), but this was nullified by adding the diene (7)
(entry 8, Table 6). Overall, the results of these ethenolysis reac-
tions with oleate and linoleate esters 10–13 confirm that the
phenol moiety is at best neutral or inhibiting to the reactions,
but that 1,4-cyclohexadiene (7) can provide a positive effect.
Ethenolysis of anacardic acid (1) and its derivatives. Anacar-
dic acid (1) can also be obtained directly from CSNL, but using
fewer steps than are required for the isolation of cardanol (2).
It can potentially give the desired 3-non-8-enylphenol (4) by
ethenolysis followed by decarboxylation. Preliminary studies
on the ethenolysis of anacardic acid itself using M1 catalyst
(conditions reported in entry 1, Table 1) showed no evidence
for reaction. We therefore methylated anacardic acid at both
the phenolic and acidic positions. The ester 14 could then be
purified by distillation without decarboxylation of the carboxyl
group. Because of the lower activity of M1 in the ethenolysis of
methyl cardanol (8, see above), we directly investigated a range
of Grubbs–Hoveyda type catalysts shown in Fig. 1 and 6 which
showed high efficiency in our previously investigated isomeris-
ing ethenolysis of methyl oleate26 and of alkenyl benzenes.27
The results of these studies, collected in Table 7, show that
of all the catalysts tested only the phosphine containing cata-
lyst, HG1, shows high activity and selectivity at 25 °C and a
5
6b
7b
8b
9b
10b
11c,d
12c,e
13c, f
92
93
93
a Conditions: 14 (0.25 mmol), cat. (1 mol%), CH2Cl2 (1 mL), 16 h;
yields were determined using n-dodecane as internal standard. b THF
(1 mL). c Reaction time 6 h. d Cat. (0.5 mol%). e Cat. (0.1 mol%). f Cat.
(0.05 mol%).
catalyst loading of 1 mol% over 16 h (entry 1, Table 7). Amongst
the other catalysts (entries 2–5, Table 7), only M51 and M74 show
good activity (entries 4 and 5, Table 7), but their selectivity
towards the desired product 15 bearing an 8-nonenyl substitu-
ent is low. There is little or no improvement for any of the cata-
lysts at higher temperature (entries 6–10, Table 7) but HG1 is
adversely affected. At 25 °C HG1 performs well even at lower
catalysts loadings (entries 11–13, Table 7) with a slight drop in
activity but with the high selectivity being retained.
Test for oestrogenicity
Due to their excellent properties, ethoxylated alkylphenols
(APE) are widely used in various applications, for example as
emulsifiers, detergents or surfactants in household products.
Nevertheless, they are being replaced by ethoxylated alcohols
because of environmental concerns. One of the most impor-
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2846–2856 | 2851

View Article Online
Paper
Green Chemistry
Fig. 7 One isomer of 4-nonylphenol showing how it can adopt a struc-
ture related to that of oestradiol.
tant APEs has been banned in Europe because its precursor,
4-nonylphenol is an endocrine disrupter.12,13 For example, it
has been shown to induce testes-ova, and intersex condition,
in the post-hatching stages of development of male Japanese
Medaka fish.28 The form of 4-nonylphenol used has a variety
of differently branched C9 chains in the 4 position of the
phenol and its endocrine disrupting properties have been
attributed to its ability to mimic the structure of oestradiol
(Fig. 7).
We reasoned that 3-nonylphenol might be less endocrine
disrupting than 4-nonylphenol since it has a linear C9 chain in
the 3-position and should not so readily mimic oestradiol.
Some of us have shown13 that the oestrogenicity of alkyl
phenols increases in the order 2 < 3 < 4 alkyl substitution on
the ring and primary ∼ secondary < tertiary. Linear 4-nonyl
phenol has been studied before13 and is known to show lower
oestrogenicity than the commercial 4-nonylphenol with mixed
alkyl chains, but 3-nonyl phenol has not been examined.
In order to test our hypothesis that 3-nonylphenol might
show less oestrogenicity than either the linear or the branched
4-nonylphenol, we have carried out a yeast oestrogen screen
(YES) assay which is specific for oestradiol mimics. For com-
parison, we also tested oestradiol, 4-nonylphenol with mixed
C9 chains (4-NP), 4-nonylphenol with a linear chain (4-n-NP),
3-nonylphenol (3-NP) prepared in this study by hydrogenation
of 3-non-8-enylphenol, cardanol and crude cashew nut shell
liquid. We appreciate that oestrogenicity is only one environ-
mental hazard and further testing of any compounds that
show promise for reducing oestrogenicity would be required.
This aspect is discussed in more detail in the concluding
section of the paper.
Fig. 8 Oestrogenic response curves for a variety of substrates obtained
using a YES assay.
for the difference between the pure and contaminated samples
is unknown, but may suggest that impurities containing satu-
rated rings in the sample may have contributed to the loss of
oestrogenicity, and/or that the ring hydrogenated products are
less oestrogenic.
Cardanol produced a very weak oestrogenic response at the
highest concentrations tested whereas the cashew nut shell
liquid was not oestrogenic in the YES assay when tested over
the same concentration range (Fig. 9).
Experimental
All reagents were purchased from Sigma-Aldrich and used as
Fig. 8 shows the oestrogenic response of 3-NP and 4-n-NP received unless otherwise stated. 17 β-Oestradiol (≥98% pure)
tested over a concentration range of 1 × 10−3 mol dm−3 to was purchased from Sigma Chemical Company Ltd (Dorset,
5 × 10−7 mol dm−3. Consistent with previous reports,13 moving UK) and ethanol (>99.7%) was purchased from Hayman
the linear alky group from the 4- to the 3-position resulted in a Speciality Products (Essex, UK). All other solvents were pur-
10-fold reduction in oestrogenic activity to produce a full dose chased from Sigma-Aldrich and were distilled under N2 using
response curve with a potency approximately 1.5 × 106 and the appropriate drying reagent.29 CNSL was extracted from
300-fold less than 17 β-oestradiol and 4-NP (mixed isomers), shells collected from Naliendele in Mtwara, Tanzania. Anacar-
respectively. The initial sample of 3-NP, which had been dic acid was obtained from the oil by a literature method3 and
obtained using Pd/C as the hydrogenation catalyst contained cardanol (2) from the anacardic acid (1) as previously pub-
small amounts of ring hydrogenated product as a result of lished.10 The cardanol (2) was vacuum dried before it was sub-
over hydrogenation. A second sample was prepared using jected to ethenolysis reactions.
[RhCl(PPh3)3] as the hydrogenation catalyst. This sample
which did not contain any ring hydrogenation products,
Instrumentation
exhibited a 2-fold increase in oestrogenicity, which was still
150 times lower than that of 4-NP mixed isomers. The reason
All weighing manipulations of air- and moisture-sensitive
chemicals were carried out in the glovebox of model type
2852 | Green Chem., 2014, 16, 2846–2856
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
degrees of saturation. The composition of saturated, mono-,
di- and tri-unsaturated cardanol was calculated from inte-
gration of the olefinic proton signals the protons adjacent to
the double binds and the aromatic protons.
Monounsaturated cardanol (2d).25 RuCl3·xH2O (17 mg,
1.1 mmol) was dissolved in 2-propanol (5 mL) and cardanol
(0.5 g, 1.66 mmol) was added. The reaction was refluxed for
18 h under an N2 atmosphere. The resulting brown solution
was cooled to room temperature and the solvent removed to
give a viscous brown oil. The oil was dissolved in CH2Cl2
(20 ml) and filtered over a 5 cm3 plug of silica to give a yellow
oil. Yield: 0.43 g, 1.4 mmol (86%). 1H NMR (CDCl3): δ 0.9
(t, 3H, CH3); 1.3 (m, 16 H, CH2); 2.0 (m, 4H, CH2–CvC); 2.6
(m, 2H, Ar–CH2–); 5.4 (m, 2H, HCvCH); 6.7–7.2 (m, 4H, Ar)
ppm. MS (m/z): 302, 304 (saturated cardanol).
Methyl protected monounsaturated cardanol (8b).24 Mono-
unsaturated cardanol (1 g, 3.3 mmol) and potassium carbon-
ate (0.9 g, 6.6 mmol) were suspended in dry acetone (15 mL).
Methyl iodide (0.4 mL, 6.6 mmol) was added dropwise and the
mixture allowed to reflux for 6 h. The reaction was then
allowed to cool to room temperature and the solvent was
removed under reduced pressure. The residue was dissolved in
ethyl acetate (50 ml). The organic layer was washed with water
(3 × 20 ml), dried over MgSO4, filtered and evaporated to give a
yellow oil. The oil was then purified over a silica column using
hexane–EtOAc (5 : 1). Yield: 55%. 1H NMR (CDCl3): δ 0.9 (t, 3H,
CH3); 1.3 (m, 16 H, CH2); 2.0 (m, 4H, CH2–CvC); 2.6 (m, 2H,
Ar–CH2–); 3.8 (s, 3H, OCH3); 5.4 (m, 2H, HCvCH); 6.7–7.2
(m, 4H, Ar) ppm. MS (m/z): 316.
Methyl linolenate (12). Oleic acid (0.92 g, 3.5 mmol) and
polyethyleneglycol-750 (1.24 g) were dissolved in CH2Cl2
(5 mL). KOH (0.37 g) was added and, after stirring for 1 h, MeI
(0.5 g, 0.22 mL, 3.4 mmol). After stirring for 5 h, during which
time a white precipitate formed, water was added followed by
NaCl to break the emulsion. The organic phase was collected
combined with CH2Cl2 washings of the aqueous phase
(2 × 5 mL), dried over anhydrous MgSO4 and evaporated to
dryness. The product was separated on a silica column using
hexane–EtOAc (4 : 1). GCMS analysis if the product showed it
to be contaminated with up to 50% methyl linoleate and
traces of methyl oleate. NMR integration also suggests that sig-
nificant amounts of methyl linoleate are present.
Fig. 9 YES assay for cardanol and CNSL compared with oestradiol.
FF100 Recirc 13649 series, where the port was evacuated for
30 minutes and flooded with nitrogen gas for 3 cycles. All reac-
tions which used air sensitive chemicals were carried out
under nitrogen atmosphere using standard Schlenk line and
catheter techniques.
GC-MS analyses were carried out using a Hewlett-Packard
6890 series gas chromatograph instrument equipped with a
flame ionization detector for quantitative analysis and a
Hewlett-Packard 5973 series mass selective detector fitted with
hp1 film for mass spectral identification of products. Helium
was used as the carrier gas with initial flow of 1 mL min−1
.
1
The H NMR and 13C NMR spectra were recorded on a Bruker
AM 400 NMR spectrometer at 400 and 100 MHz or a Bruker
AM 300 spectrometer at 300 and 75 MHz, respectively. Samples
were dissolved in deuterated solvents which were referenced
internally relative to tetramethylsilane (TMS) at δ = 0 ppm.
Chemical shifts, δ, are reported in ppm relative to TMS. All 13
C
NMR spectra were proton-decoupled.
3-Hydroxyphenyl oleate (11). Oleic acid (1.6 g, 4.21 mmol)
and resorcinol (1.6 g, 14.82 mmol) were dissolved in THF
(10 mL). The reaction mixture was cooled to 0 °C and slowly
N,N-dicyclohexylcarbodiimide (991 mg, 4.81 mmol) and DMAP
(22 mg, 0.18 mmol) were added. The reaction mixture turned
turbulent and a white precipitate was formed. The suspension
was allowed to warm and was stirred for 85 h at room tempera-
ture. Ethyl acetate (25 mL) was added to the mixture and the
precipitate was collected by filtration. The filtrate was evapor-
ated and the remaining residue was purified via column
chromatography with hexane–EtOAc (4 : 1) as eluent. Yield:
62%. 1H NMR(CDCl3): δ 0.91 (t, 3J = 6.9 Hz, 3 H, CH3),
1.27–1.45 (m, 20 H, CH2), 1.75–1.79 (m, 2 H, CH2), 2.01–2. 07
(m, 4 H, CH2), 2.57 (t, 3J = 7.8 Hz, CH2), 5.36–5.42 (m, 2 H,
Analysis of cardanol (2). Cardanol was analysed via GC and
1
NMR. H NMR (300 MHz, CDCl3): δ 0.93–1.00 (m, 1.9 H, CH3/
CH2), 1.33–1.45 (m, 12.4 H, CH2), 2.05–2.14 (m, 2.1 H, CH2),
2.57–2.62 (m, 3.1 H, CH2), 2.83–2.92 (m, 1.9 H, CH2), 5.03–5.16
(m, 0.8 H, CH), 5.33–5.56 (m, 4.0 H, CH), 5.82–5.95 (m, 0.4 H,
CH), 6.70–7.22 (m, 4H, Ar-H) ppm. 13C (75 MHz, CDCl3): 14.3,
14.6 (CH3), 32.1, 23.3, 26.0, 26.1, 27.6, 27.7, 29.4, 29.7, 29.8,
29.9, 30.1, 30.1, 30.2, 31.7, 32.0, 32.2, 36.3 (CH2). 127.3, 128.0,
128.4, 128.6 129.7, 129.8 (CH), 130.4, 130.6, 130.8, 137.3,
145.3, 155.8 (Ar-C) ppm. The integration of the 1H NMR-
signals did not result in even numbers, as cardanol is a
mixture of compounds of different saturation. Furthermore,
more than 21 C-signals can be observed due to the different
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2846–2856 | 2853

View Article Online
Paper
Green Chemistry
CH), 6.23 (b, 1 H, OH), 6.57–6.70 (m, 3 H, Ar-H), 7.21 (t, 3J = 9.0 cooled to room temperature and quenched with ammonium
Hz, 1 H, Ar-H) ppm. ESMS (m/z): 373 [M − H]−.
chloride (10 mL). Distilled water (30 mL) was added to the
The product was contaminated with a product of slightly reaction mixture, which was then extracted with ethyl acetate
lower GC retention time (see ESI†). (3 × 30 mL). The organic layer was washed with distilled water
3-Hydroxyphenyl linolenate (13) was similarly prepared (1 × 10 mL), dried over anhydrous sodium sulfate, and concen-
from linolenic acid (1.2 g, 4.43 mmol), resorcinol (1.6 g, trated in vacuo. The crude product was further purified by
14.82 mmol), N,N-dicyclohexylcarbodiimide (998 mg, Kugelrohr distillation (270 °C, 1 × 10−3 mbar) to yield 4.31 g
1
1
4.84 mmol) and DMAP (18.3 mg, 0.15 mmol). Yield: 58%. H (11.6 mmol, 80%) of a bright yellow oil. H NMR (CDCl3): δ =
3
NMR (CDCl3): δ 1.01 (t, J = 7.5 Hz, 3 H, CH3), 1.29–1.46 (m, 0.9 (m, 1.77 H, CH3); 1.3 (m, 12.07 H, CH2); 1.6 (m, 2.23 H,
8 H, CH2), 1.73–1.82 (m, 2 H, CH2), 2.06–2.16 (m, 4 H, CH2), CH2); 2.0 (m, 3.02 H, CH2–CvC); 2.5 (m, 1.89 H, CvC–CH2–
2.58 (t, 3J = 7.5 Hz, 2 H, CH2), 2.81–2.86 (m, 4 H, CH2), CvC); 2.8 (m, 2 H, Ar–CH2–); 3.8 (s, 3 H, OCH3); 3.9 (s, 3 H,
5.31–5.48 (m, 6 H, CH), 6.03 (b, 1 H, OH), 7.56 (b, 1 H, COOCH3); 5.0 (m, 0.72 H, CvCH2); 5.4 (m, 2H, HCvCH); 5.8
3
3
Ar-H), 6.66 (t, J = 8.4 Hz, 2 H, Ar-H), 7.21 (t, J = 8.4 Hz, 1 H, (m, 0.72 H, HCvCH2); 6.7–7.2 (m, 4H, Ar) ppm. MS (m/z): 374.
Ar-H) ppm.
Typical catalytic experiments
The product was contaminated with a product of slightly
lower GC retention time (see ESI†).
The catalyst (M1; 2 mg, 2.2 µmol) was weighed in the glove box
3-Nonylphenol. 3-non-8-enylphenol (1 g, 4.6 mmol), in and made up to a standard solution in CH2Cl2 (2 mL), in a
degassed toluene (10 ml), was added to a solution of Schlenk tube under dinitrogen atmosphere. When required,
[RhCl(PPh3)3]30 (30 mg, 46 µmol) dissolved in degassed the additive was added to the standard solution in the correct
toluene (5 ml) and transferred to a Fischer–Porter bottle which molar amounts (1,4-cyclohexadiene; 27 µl, 55 µmol). The sub-
was charged with H2 (6 bar) and left to stir overnight at 60 °C. strate (0.18 ml, 0.55 mmol) was degassed and added to a pre-
The solution was filtered over a plug of silica (5 cm3) using viously dried and inert Fischer–Porter bottle with magnetic
CH2Cl2 (100 mL). The solvent was removed to give a colourless stirrer bar under dinitrogen atmosphere, in CH2Cl2 (1.3 mL).
1
oil. Yield: 0.96 g, 4.4 mmol (96%). H NMR (CDCl3): δ 0.91 (t, The appropriate amount of catalyst was syringed from the stan-
3 H, CH3); 1.3 (m, 12 H, CH2); 1.6 (m, 2H, CH2CH2Ar); 2.6 (m, dard solution (0.26 mL) and added to the Fischer–Porter bottle
2H, CH2Ar); 4.7 (s, 1H, OH); 6.7–7.2 (m, 4H, Ar-H) ppm. to give the correct substrate concentration (typically 0.35 mol
13C NMR (CDCl3): δ 14.1 (CH3); 22.7 (CH2); 29.2 (CH2); 29.3 dm−3, 1 : 2000 [catalyst : substrate]). The glass bottle was sealed,
(CH2); 29.6 (CH2); 29.8 (CH2); 31.2 (CH2); 31.9 (CH2); 36.0 flushed with ethene five times and pressurised to 8 bar. The
(Ar-CH2); 113.1 (Ar); 115.6 (Ar); 120.7 (Ar); 129.9 (Ar); 144.8 solution was allowed to stir for 6 h at the predetermined temp-
(qAr); 156.1 (qAr) ppm. MS (m/z): 220.
erature in the sealed vessel. The reaction was quenched with
Hydrogenation of 3-non-8-enylphenol (4) with Pd/C. 3-Non- ethylvinyl ether (0.05 ml) and analysed by GC and GCMS.
8-enylphenol (4) (800 mg, 3.6 mmol) and Pd/C (5 wt%, 39 mg)
Recombinant yeast oestrogen screen
were placed in an autoclave and dichloromethane (5 mL) was
added. The autoclave was pressurized with hydrogen to 10 bar The recombinant hER yeast strain was developed by Glaxo
and the suspension was stirred at 40 °C for 6 h. The reaction Wellcome and details of the yeast oestrogen screen have been
was then allowed to cool to room temperature and the auto- described previously.32
clave was slowly depressurized. The solvent was removed
In brief, yeast cells were transfected with the human oestro-
under reduced pressure and the residue was purified via gen receptor gene together with expression plasmids; the oes-
column chromatography with hexane, ethylacetate and diethyl- trogen response element and the lac-Z gene encoding the
ether as eluent (7 : 1 : 1). Yield 78%. 1H NMR (300 MHz, enzyme β-galactosidase. The yeast cells were incubated in a
3
CDCl3): δ 0.91 (t, J = 6.9 Hz, 3 H, CH3), 1.19–1.40 (m, 12 H, medium containing the test chemical and the chromogenic
3
CH2), 1.54–1.67 (m, 2 H, CH2), 5.57 (t, J = 8.1 Hz, 2 H, CH2), substrate, chlorophenol red-β-D-galactosidase (CPRG). Active
4.94 (b, 1 H, OH), 6.64–6.71 (m, 2 H, Ar-H), 6.78 (bd, 3J = ligands induced β-gal expression. The β-galactosidase secreted
8.4 Hz, 1 H, Ar-H), 7.16 (t, 3J = 8.4 Hz, 1 H, Ar-H) ppm. 13C into the medium causes the yellow CPGR to change into a red
(75 MHz, CDCl3): 14.6 (CH3), 23.2 (CH2), 29.8 (CH2), 30.0 product, and this is measurable by absorbance.
(CH2), 30.1 (CH2), 30.2 (CH2), 31.8 (CH2), 32.4 (CH2), 36.3
Assay procedure
(CH2), 112.9 (Ar-CH), 115.9 (Ar-CH), 121.5 (Ar-CH), 129.8
(Ar-CH), 145.4 (Ar-C), 155.8 (Ar-C) ppm.
Some completely hydrogenated 3-non-8-enylphenol to assay procedure was followed.32 Chemicals were serially
3-nonylcyclohexanol can be observed in the 1H NMR spectrum.
diluted in ethanol and 10 µL volumes were transferred to
The medium components were prepared and the standard
Anacardic acid methyl ester (14).31 To a stirred solution of 96-well flat-bottom plates where the ethanol was allowed to
anacardic acid (1) (5.00 g, 14.6 mmol) in acetone (30 mL) was evaporate to dryness. Then, 200 µL medium containing CPRG
added potassium carbonate (8.07 g, 58.4 mmol). Dimethylsul- and yeast (final cell number of 5 × 105 cells mL−1) was added
fate (3.72 g, 29.2 mmol) was added in portions over about to each well. Included with every assay was a negative control,
10 min at room temperature. After the addition was complete, ethanol, and a positive control, 17 β-oestradiol (stock solution
the solution was heated to reflux for 4 h. The solution was of 17 β-oestradiol (2 × 10−7 mol dm−3) serially diluted in
2854 | Green Chem., 2014, 16, 2846–2856
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
ethanol to achieve final concentrations of 1 × 10−8 mol dm−3 of 3-NP33 and following the outcomes of internationally agreed
to 4.88 × 10−12 mol dm−3 in the wells).
and validated OECD test methods developed for the identifi-
The plates were incubated at 32 °C for 3 days, after which cation of endocrine disrupters, including oestrogenicity, (anti)-
absorbance readings were taken at 540 and 620 nm (the androgenicity and thyroid disruption. Finally, this work is an
second absorbance being a measure of cell density and hence example of how green chemistry can be used to design endo-
yeast growth). The absorbance values were corrected for cell crine disruption out of the next generation of chemicals as
density using the following equation:
proposed in the Tiered Protocol for Endocrine Disruption
(TiPED) approach.34
Corrected value ¼ chemical540 nm ꢀ ðchemical620 nm
ethanol blank620 nmÞ:
ꢀ
Acknowledgements
All chemicals were tested in duplicate and each YES was
carried out at least three times.
We thank the EPSRC for Postdoctoral and other funding
(J. J. and S. A. B.) and Deutsche Bundesstiftung Umwelt
(Fellowship to S. B.). We also thank Umicore AG for the
donation of chemicals. C. S. J. C. thanks the Royal Society for a
University Research Fellowship.
Conclusion
M1 and Grubbs’ 1st generation catalysts are very active and
selective for the ethenolysis of cardanol (2) to 3-non-8-enyl-
phenol (4), an intermediate for potential surfactants. The
unexpectedly high catalytic performance of M1 is due to the
side-product 1,4-cyclohexadiene (7), which is formed during
the ethenolysis from the tri-unsaturated component of carda-
nol (2d). 1,4-Cyclohexadiene (7) stabilises the catalytically
active species and prevents inhibition by the phenolic group.
Absence of 7 leads to a complete deactivation of M1, as the
results of the ethenolysis of mono-unsaturated cardanol
showed. The positive impact of 7 on the catalytic performance
was also observed with other substrates such as methyl oleate.
It also enables the use of less expensive homogeneous meta-
thesis catalysts in the ethenolysis reaction, one of the most
challenging reactions in metathesis.
Notes and references
1 S. H. Azam-Ali and E. C. Judge, Small scale cashew nut pro-
cessing, Food and Agriculture Organisation, Rome, 2004.
2 J. H. P. Tyman, Chem. Soc. Rev., 1979, 8, 499.
3 R. Paramashivappa, P. P. Kumar, P. J. Vithayathil and
A. S. Rao, J. Agric. Food Chem., 2001, 49, 2548.
4 P. P. Kumar, R. Paramashivappa, P. J. Vithayathil, P. V. S. Rao
and A. S. Rao, J. Agric. Food Chem., 2002, 50, 4705.
5 L. P. L. Logrado, C. O. Santos, L. A. S. Romeiro, A. M. Costa,
J. R. O. Ferreira, B. C. Cavalcanti, O. Manoel de Moraes,
L. V. Costa-Lotufo, C. Pessoa and M. L. dos Santos,
Eur. J. Med. Chem., 2010, 45, 3480.
One application of 3-non-8-enylphenol (4) is for it to be
hydrogenated to 3-nonylphenol for use as a possible replace-
ment for 4-nonylphenol which has been banned in many
countries on the basis of its endocrine disrupting properties. A
YES assay shows that 3-nonylphenol prepared by ethenolysis of
cardanol followed by hydrogenation using [RhCl(PPh3)3] is at
least 150 times less potent in oestrogenicity than the banned
6 Y.-C. Guo, G. Mele, F. Martina, E. Margapoti, G. Vasapollo
and W.-J. Xiao, J. Organomet. Chem., 2006, 691, 5383.
7 P. Peungjitton, P. Sangvanich, S. Pornpakakul, A. Petsom
and S. Roengsumran, J. Surfactants Deterg., 2009, 12, 85.
8 R. K. Paul and C. K. S. Pillai, Synth. Met., 2000, 114, 27.
9 G. Vasapollo, G. Mele and R. Del Sole, Molecules, 2011, 16,
6871.
substance and some 10−6 times as oestrogenic as 17 β-oestra- 10 J. A. Mmongoyo, Q. A. Mgani, S. J. M. Mdachi,
diol. Cardanol and cashew nut shell liquid are even less potent
P. J. Pogorzelec and D. J. Cole-Hamilton, Eur. J. Lipid Sci.
showing very little oestrogenicity in the YES assay. We recog-
Technol., 2012, 114, 1183.
nise that the >150-fold reduction in oestrogenic activity 11 http://www.icis.com/Articles/2007/08/06/9050162/chemical-
between 4-NP and 3-NP, although desirable, is not enough to profile-1-octene.html.
secure the long-term use of 3-NP as a replacement for use in 12 F. L. P. Gabriel, E. J. Routledge, A. Heidlberger, D. Rentsch,
alkylphenol ethoxylate surfactants where non-estrogenic
alternatives (including alcohol ethoxylates) already exist.
K. Guenther, W. Giger, J. P. Sumpter and H.-P. E. Kohler,
Environ. Sci. Technol., 2008, 42, 6399.
However, in countries lacking restrictions on the use and 13 E. J. Routledge and J. P. Sumpter, J. Biol. Chem., 1997, 272,
release of 4-NP into the environment, or in industrial appli- 3280.
cations where alternatives are not readily available, the 14 R. M. Thomas, B. K. Keitz, T. M. Champagne and
reduction in oestrogenicity of 3-NP over 4-NP may be seen a R. H. Grubbs, J. Am. Chem. Soc., 2011, 133, 7490.
useful stop-gap measure in the transition away from 4-NP use. 15 D. Astruc, New J. Chem., 2005, 29, 42.
The significance of the loss of oestrogenicity of 3-NP in terms 16 L. Ackermann, A. Furstner, T. Weskamp, F. J. Kohl and
of its safety in use can only be determined following robust
W. A. Herrmann, Tetrahedron Lett., 1999, 40, 4787.
investigations on its fate and persistence in the environment, 17 J. K. Huang, E. D. Stevens, S. P. Nolan and J. L. Petersen,
further knowledge of the wider biological/toxicological hazard
J. Am. Chem. Soc., 1999, 121, 2674.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2846–2856 | 2855

View Article Online
Paper
Green Chemistry
18 M. Scholl, S. Ding, C. W. Lee and R. H. Grubbs, Org. Lett., 27 S. Baader, D. M. Ohlmann and L. J. Goossen, Chem.–Eur. J,
1999, 1, 953. 2013, 19, 9807.
19 M. Scholl, T. M. Trnka, J. P. Morgan and R. H. Grubbs, 28 M. A. Gray and C. D. Metcalfe, Environ. Toxicol. Chem.,
Tetrahedron Lett., 1999, 40, 2247. 1997, 16, 1082.
20 X. Bantreil, A. Poater, C. A. Urbina-Blanco, Y. D. Bidal, 29 C. L. L. C. Wilfred and L. F. Armarego, Purification of
L. Falivene, R. A. M. Randall, L. Cavallo, A. M. Z. Slawin
and C. S. J. Cazin, Organometallics, 2012, 31, 7415.
21 X. Bantreil, T. E. Schmid, R. A. M. Randall, A. M. Z. Slawin
and C. S. J. Cazin, Chem. Commun., 2010, 46, 7115.
22 T. Vorfalt, K. J. Wannowius, V. Thiel and H. Plenio,
Chem.–Eur. J, 2010, 16, 12312.
Laboratory Chemicals, Elsevier, North Holland, 2003.
30 J. A. Osborn, F. H. Jardine, J. F. Young and G. Wilkinson,
J. Chem. Soc. A, 1966, 1711.
31 R. Paramashivappa, P. P. Kumar, P. V. S. Rao and A. S. Rao,
J. Agric. Food Chem., 2002, 50, 7709.
32 E. J. Routledge and J. P. Sumpter, Environ. Toxicol. Chem.,
1996, 15, 241.
23 G. S. Forman, A. E. McConnell, R. P. Tooze, W. J. van
Rensburg, W. H. Meyer, M. M. Kirk, C. L. Dwyer and 33 R. R. Tice, C. P. Austin, R. J. Kavlock and J. R. Bucher,
D. W. Serfontein, Organometallics, 2005, 24, 4528. Environ. Health Perspect., 2013, 121, 756.
24 R. Holzwarth, R. Bartsch, Z. Cherkaoui and G. Solladie, 34 T. T. Schug, R. Abagyan, B. Blumberg, T. J. Collins,
Chem.–Eur. J, 2004, 10, 3931.
25 S. Perdriau, S. Harder, H. J. Heeres and J. G. de Vries,
ChemSusChem, 2012, 5, 2427.
26 D. M. Ohlmann, N. Tschauder, J. P. Stockis, P. Goossen,
M. Dierker and L. J. Goossen, J. Am. Chem. Soc., 2012, 134,
13716.
D. Crews, P. L. DeFur, S. M. Dickerson, T. M. Edwards,
A. C. Gore, L. J. Guillette, T. Hayes, J. J. Heindel,
A. Moores, H. B. Patisaul, T. L. Tal, K. A. Thayer,
L. N. Vandenberg, J. C. Warner, C. S. Watson, F. S. vom
Saal, R. T. Zoeller, K. P. O’Brien and J. P. Myers, Green
Chem., 2013, 15, 181.
2856 | Green Chem., 2014, 16, 2846–2856
This journal is © The Royal Society of Chemistry 2014