MESHOD FOR REACTIVELY CRUSHING JATROPHA SEEDS
The present invention relates to a method for the
reactive grinding of Jatropha seeds which, starting
5 from specifically processed Jatropha seeds in the
presence of a light alcohol and a basic catalyst, makes
it possible to carry out the grinding and the reaction
for transesterification of the triglycerides present in
the Jatropha oil in a single step, simultaneously
10 producing an oil cake, glycerol and fatty acid esters.
Said esters are intended mainly for the production of
biodiesel. Furthermore, the method according to the
invention makes it possible to obtain a completely
detoxified oil cake. The oil cakes obtained by the
15 method for treating Jatropha (in particular Jatropha
curcas L.) seeds according to the invention retain a
nutritional value and can be directly used in animal
feed, without constituting a risk to the health of the
individuals who handle them.
20
It is known practice to prepare fatty acid esters from
seeds of oleaginous plants in two steps, namely a step
of extraction of oil in the presence of solvent and a
step of transesterification of this oil in the presence
25 of alcohol arid of catalyst, producing an ester phase
and a glycerol phase.
The-Jatropha genus comprises several species known for
the irritant properties of their seeds in humans and
30 animals. They are tropical plants grown in Latin
America, in Asia and in Africa and used mainly as
hedging. Their potential nutritional and technical
applications, in particular in controlling soil erosion
and preparing biodiesel, are at present limited owing
35 to their toxicity.
Jatropha seeds are rich in oil and in proteins, but
they are highly toxic and incompatible with human or
WO 2011 / 092430 2 PCT/FR2011 / 050155
animal consumption. The toxic and anti-nutritional
compounds of Jatropha include curcin (a lectin),
flavonoids, trypsin inhibitors, saponins, phytates and
phorbol esters. The lectin and the activity of the
5 trypsin inhibitors can be removed by heat treatment.
The high concentrations of phorbol esters, which are
thermally stable, remain the principal source of
toxicity of the oil extracted from the Jatropha seeds
and the oil cakes. Indeed, this family of compounds is
10 known for its harmful biological effects in humans and
animals, in particular in inflammation and promotion of
tumors. The phorbol esters do not induce a tumor by
themselves, but facilitate the growth of tumors after
exposure to doses considered to be noncarcinogenic of a
15 carcinogenic compound.
The phorbol ester content varies among the various
varieties of Jatropha, as shown by the study by Makkar
H.P.S et al. J. Agric. Food Chem. 45:8, 1997, 3152-
20 3157. The data presented in table 4 show that these
compounds were detected in most of the varieties
tested. There is one variety in :which the phorbol
esters are virtually absent, the Jatropha grown in
Mexico, while the other varieties are more or less rich
25 in these compounds (in particular the Jatropha
originating from Kenya or Nicaragua).
Jatropha oil cake can be used in animal feed only if
the "removal of toxic and anti-nutritional compounds can
30 be guaranteed. The toxic effects of Jatropha seeds on
animals appear to be linked to the dosage, as shown in
the publication by S.E.I. Adam, Toxicol. 2: 67-76,
1974 . The data presented in table 1 of the publication
Vet. Pathol. 16: 476-482, 1979 show that animals fed
35 with Jatropha seeds die after several days, probably
because of a cumulative effect of the toxic compounds.
The publication by W. Haas and M. Mittelbach, Ind.
Crops Prod. 12 (2000 ), 111-118 describes a method for
WO 2011/ 092430 3 PCT/F122011 / 050155
assaying phorbol esters in Jatropha oil and also
various treatments of the oil. It is shown that the
conventional oil degumming and deodorization treatments
have little influence on the concentration of these
5 compounds, whereas deacidification and bleaching make
it possible to reduce the phorbol ester content to 55%,
which remains insufficient.
Various Jatropha detoxification methods have been
10 tested.
Heating the Jatropha seeds at 160°C for 30 minutes does
not make it possible to remove the phorbol esters
(Aregheore E.M. et al., S. Pac. J. Nat. Sci. 21: 50-56,
15 200.3)
The injection of steam into the protein extracts
obtained from defatted oil cakes for 10 min at
approximately 92°C makes it possible to remove the
20 phorbol esters (Devappa R.K. and Swamylingappa B., J.
Sci. Food Agric. 88: 911-919, 2008 ) However, this
method consumes a great deal of energy and results in
isolating the proteins from the oil cake which is then
low in constituents of nutritional interest.
25
An extraction with 90% ethanol followed by treatment of
Jatropha oil cakes with NaHCO3 at 121°C for 20 minutes
made it possible to reduce the phorbol ester content by
close to 98% (Martinez-Herrera J. et al., Food Chem. 96
30 (2006 ), 80-89).
A basic treatment (aqueous solution of sodium hydroxide
or of lime at 2%), followed by heat treatment at 121°C
for 30 minutes, carried out on Jatropha oil cakes, have
35 made it possible to reduce their phorbol ester content
by 89%, but the detoxification is not complete
(Rakshit KD. et al., Food Chem. Toxicol. 46 (2008):
3621-3625).
WO 2011 / 092430 4 PCT/FR2011 / 050155
Other plants contain similar toxic compounds in their
seeds, in particular phorbol esters which are naturally
present in many plants of the family Euphorbiaceae and
the family Thymelaeaceae. By way of example, mention
5 may be made of Euphorbia lathyris (mole plant or
spurge) and Croton tiglium (purging croton) of the
family Euphorbiaceae, or else Bertholletia excelsa
(Brazil nut) , Prunus dulcis (almond tree) , Gossypium
hirsutum (cotton), Linum usitatissimum (flax), Ceiba
10 pentandra (kapok), Sapium indicum, S. Japonicum,
Euphorbia frankiana, E. cocrulescence, E. ticulli,
Croton spareiflorus, C. ciliatoglandulifer, Excoecaria
agallocha and Homalanthus mutans. The seed
detoxification method which is the subject of the
15 invention can be generalized to all these plants.
It is therefore desirable to have a method for treating
Jatropha seeds, and more generally any seed containing
toxic compounds such as phorbol esters and/or curcin or
20 other toxic proteins such as crotin (present in
particular in the seeds of Croton tiglium) and abrin
(in the seeds of Abrus precatorius), said method making
it possible to simply and inexpensively inactivate
these toxic compounds, which would then make possible,
25 firstly, risk-free handling by human beings and,
secondly, use of the oil cake, in particular Jatropha
oil cake, in animal feed. This is particularly
important for the economy of countries which are major
producers of Jatropha oil (India, Madagascar, Brazil),
30 since, while Jatropha oil has multiple industrial uses,
Jatropha oil cakes are not yet used on an industrial
scale, in particular because of the toxicity problems
mentioned above.
35 The present invention proposes to provide a method for
treating Jatropha seeds which limits the number of seed
treatment steps and the handling of the oil cake, with
a view to a continuous industrial application aimed at
producing fatty acid esters, and which makes it
WO 2011 / 092430 5 PCT/FR2011 / 050155
possible to destroy "at source" the toxin (curcin) and
the phorbol esters present in the Jatropha seeds, if
possible while maintaining a nutritive value for the
oil cake. The other advantage of the method compared
5 with the conventional methods lies in the small amounts
of water used, The operations for refining the crude
oil for example consume very large amounts of water.
This water saving is a major asset in the context of
the development of this technology in developing
10 countries and, to a lesser extent, in rich countries
since water is tending to become an increasingly
expensive commodity.
To this effect,.. the subject of the invention is a
15 method for treating seeds containing toxic components
such as curcin, abrin, crotin and/or phorbol esters, in
particular Jatropha seeds, said seeds preferably having
a degree of acidity of less than or equal to 3 mg
KOH/g, said method comprising the following steps:
20
25
30
a seed processing step;
a step of bringing the processed seeds into
contact with a light anhydrous alcohol and an
alkaline catalyst, under temperature and time
conditions sufficient to allow the
simultaneous extraction and
transesterification of the vegetable oil, and
producing a mixture comprising fatty acid
esters and glycerol, and an oil cake.
The method according to the invention makes it possible
to react "in plants" the light alcohol with the oil
contained at the heart of the seed. In this method, the
35 alcohol plays both the role of solvent and the role of
reagent.
WO 2011 / 092430 6 PCT/ FR2011 / 050155
Characteristically, the seeds are processed by means of
a series of operations comprising a step of flattening
and a step of drying said seeds.
5 Preferably, said flattening step comprises triple
flattening on smooth rollers, in particular for the
hardest seeds such as Jatropha seeds.
According to the conditions used, the method according
10 to the invention may directly produce a detoxified oil
cake. In one embodiment variant, the oil cake is
subjected to an additional drying step, under
temperature and time conditions sufficient to
inactivate the curcin and to break down the phorbol
15 esters,
Advantageously, the oil cake thus treated loses its
harmful nature and can be handled without danger by
human beings so as to be used in animal feed.
20
In the context of the present invention, the term
"Jatropha seeds" is intended to mean seeds from
Jatropha plants, alone or as a mixture with seeds
originating from at least one other oleaginous,
25 oleaginous/protein-producing or protein-producing
plant, the seeds or the seed mixture producing an oil
containing at least 40% by weight of oleic acid. It
would not be a departure from the context of the
invention if the seeds used in the method according to
30 the invention were to wholly or partly originate from
genetically modified plants.
Oleaginous plants are cultivated specifically for their
oil-producing seeds or fruits rich in fats, from which
35 oil for food, energy or industrial use is extracted.
Protein-producing plants belong to the botanical group
of legumes, the seeds of which are rich in proteins.
Oleaginous/protein-producing plants are legumes, the
seeds of which also contain oil.
WO 2011 / 092430 - 7 PCT/FR2011 / 050155
According to the invention, the term "detoxified
Jatropha oil cake" is intended to mean a Jatropha oil
cake having both:
5
- a degree of curcin detoxification of at least
90% and preferably of at least 95%, with
respect to activity, when this degree is
measured by means of a quantitative test, or of
10 100% when this degree is measured by means of a
qualitative test;
15
and a degree of decomposition of the phorbol
esters of at least 95% and preferably of at
least 99%, with respect to activity, when this
degree is measured by means of a quantitative
test, or of 100% when this degree is measured
by means of a qualitative test.
20 Taking into account the results presented in the
publication by Makkar H.P.S. et al, Plant Foods for
Human Nutrition 1998 , vol. 52, No. 1, pp. 31-36, a
phorbol ester content of 0.11 mg/g corresponds to an
edible (non-toxic) oil cake. Specialists in animal feed
25 estimate in general that the oil cake is detoxified
when it has a phorbol ester content of less than or
equal to 0.3 mg/g and can be used in animal feed, in
particular in a mixture with other feed materials.
30 The detoxified oil cakes according to the invention
therefore have a phorbol ester content of at most
0.3 mg per g, preferably at most 0.11 mg per g of oil
cake treated.
35 The term "degree of curcin detoxification" is intended
to mean the percentage by weight of toxin inactivated
in the oil cake.
WO 2011 / 092430 8 POT/FR2011 / 050155
The term "degree of decomposition" of the phorbol
esters is intended to mean the percentage by weight of
phorbol esters broken down in the oil.
5 Other characteristics and advantages will emerge from
the detailed description of the method for treating
Jatropha seeds according to the invention that follows.
A subject of the invention is a method for treating
10 seeds containing toxic components such as curcin,
abrin, cretin and/or phorbol esters, in particular
Jatropha seeds, alone or as mixtures with seeds
originating from at least one other oleaginous,
oleaginous/protein-producing or protein-producing
15 plant, said seeds preferably having a degree of acidity
of less than or equal to 3 mg KOH/g, said method
comprising the following steps:
i) a step of processing the seeds without prior
20 hulling;
25
30
ii) a step of bringing the processed seeds into
contact with a light anhydrous alcohol and an
alkaline catalyst under temperature and time
conditions sufficient to allow the
simultaneous extraction and transesterification
of the vegetable oil, and
producing a mixture comprising fatty acid
esters and glycerol, and an oil cake,
characterized in that the seeds are processed by means
of a series of operations comprising a step of
flattening and a step of drying said seeds.
35 Another particularity of Jatropha seeds is linked to
their high toxicity, due in particular to the presence
of curcin and phorbol esters. After extraction of the
oil, the curcin becomes concentrated in the oil cakes
and the phorbol esters become concentrated in the oil
WO 2011/092430 9 PCT/FR2011/050155
and/or_ in the esters, malting their handling by human
beings problematic, or even dangerous.
The method according to the invention makes it possible
5 to simultaneously solve numerous problems associated
with the transesterification of Jatropha oil. This
method advantageously makes it possible to go directly
from the seed to the fatty acid esters, while avoiding
the grinding, refining and purification steps and the
10 production of by-products. The fatty acid esters
obtained by means of the method according to the
invention are particularly suitable for preparing
biodiesel, as mentioned above. The method according to
the invention makes it possible to obtain a fraction
15 rich in fatty acid esters which lacks toxicity and can
therefore be used free of risk, in particular for
producing biodiesel. Moreover, the method produces
detoxified oil cakes, which can be handled without
danger by human beings and can be used in animal feed
20 without risk of poisoning the animals.
Seed processing step
The first step of the method according to the invention
25 consists in processing the Jatropha seeds, used alone
or as a mixture with other seeds of oleaginous,
oleaginous/protein-producing or protein-producing
plants. This processing is carried out on whole seeds.
It comprises a first operation in which the seeds are
30 flattened, followed by an operation in which the
flattened seeds are dried.
The objective of the processing of the seed is to make
the oil as accessible as possible to the alcohol,
35 without, however, causing too much modification of its
mechanical strength. This prevents the formation of a
paste and of fines, which are respectively prejudicial
to the implementation of a continuous process and to
the final purification of the esters produced.
WO 2011/092430 10 PCT/FR2011/050155
Moreover, the processed seed should allow easy passage
of the reaction fluid (alcohol/basic catalyst mixture)
according to a simple percolation phenomenon.
5 According to one embodiment variant, fresh seeds are
flattened on a mechanical flattener with smooth or
fluted rollers.
The seeds thus flattened are dried, for example in a
10 ventilated oven which is thermoregulated or in a
continuous belt or rotary hot-air dryer. The drying
time and the temperature are chosen so as to obtain a
decrease in the moisture content of the seeds to values
of less than or equal to 2% by weight. Preferably, the
15 drying is carried out rapidly after flattening, in less
than one hour, preferably after 5 to 10 minutes, at a
temperature sufficient to reduce the moisture content
of the seeds to 2% by weight or less.
20 The residual moisture content of the seed is determined
by thermogravimetric analysis. The seed is ground
beforehand, and then the ground material obtained is
dried at 105°C in a thermobalance until stabilization
of the weight. The water content is expressed as
25 percentage of the crude material.
In one preferred embodiment variant, step i) of
processing the seeds also comprises a seed preheating
operation, carried out before the flattening operation.
30 This preheating operation gives the seed greater
plasticity and therefore more effective crushing during
flattening (gain in terms of contact surface, of
alcohol percolation rate and therefore of extracLive
capacity of the alcohol). It is preferably carried out
35 at a temperature of less than or equal to 100°C.
Extraction and transesterification step
WO 2011 / 092430 11 PCT/FR2011/050155
The seeds processed as described above are brought into
contact with a light anhydrous alcohol and an alkaline
catalyst under temperature and time conditions
sufficient to allow the extraction and the
5 transesterification of the vegetable oil, and producing
a mixture comprising fatty acid esters and glycerol,
and an oil cake.
The light alcohol used in step ii) is a lower aliphatic
10 alcohol, such as methanol, ethanol, isopropanol and
n-propanol, and is preferably methanol.
According to one embodiment variant, an organic Solvent
(cosolvent), which is miscible or immiscible with said
15 light alcohol, is also added to the reaction medium. As
cosolvent, mention may be made of: hexane, heptane,
benzene, bicyclohexyl, cyclohexane, decalin, decane,
hexane (Texsolve C) , spirit, petroleum ether, kerosene,
kerdane, diesel oil, paraffin oil, methylcyclohexane,
20 Texsolve S or 5-66, naphtha (Texsolve V) , skellite,
tetradecane, Texsolve (B, C, H, S, S-2, 5-66, S-LO, V),
supercritical C02, propane or butane which are
pressurized, natural solvents such as terpenes
(limonene, alpha- and beta-pinene, etc), ethers such as
25 dimethyl ether or diethyl ether, ketones such as
acetone, and mixtures of all these solvents.
The basic catalyst used in the method is chosen from
the- group: sodium hydroxide, alcoholic sodium
30 hydroxide, solid sodium hydroxide, potassium hydroxide,
alcoholic potassium hydroxide, solid potassium
hydroxide, sodium or potassium methoxide, sodium or
potassium ethoxide, sodium and potassium propoxide, and
sodium and potassium isopropoxide,
35
The reaction is carried out in a fixed bed reactor.
According to one embodiment, the fixed bed reactor is a
thermoregulated percolation column fitted with a
screen. A pump makes it possible to feed the column
WO 2011/092430 12 PCT/FR2011/050155
with alcohol-basic catalyst mixture. The alcohol and
the catalyst are therefore added simultaneously to the
reactor, which is maintained at a temperature ranging
from 30 to 75°C, preferably less than or equal to 50°C,
5 preferably less than 45°C, preferably approximately
equal to 40°C. The catalyst/alcohol/seeds weight ratio
is preferably included in the range 0.001 to 0.01/0.1
to 5/1, preferably in the range from 0.005 to 0.01/0.1
to 1/1, even more preferably in the range from 0.005 to
10 0.01/0.1 to 0.5/1.
In particular, a catalyst content of less than 0.001,
or even less than 0.005, does not make it possible to
obtain detoxified oil cakes, and, conversely, a content
15 of,greater than 0.01 leads to saponification and a poor
ester yield.
The feed is carried out at the top of the bed; the
reaction liquid then percolates through the bed and is
20 then recovered in a store located downstream, under the
bed. The liquid is sent back to the top of the bed, by
pumping, so as to again diffuse through the bed. The
duration of the alcohol/catalyst mixture recirculation
cycle is from 15 to 60 minutes, preferably from 20 to
25 40 minutes. At the end of the cycle, the liquid feed is
stopped. A part of the liquid still present in the
soaked seeds is then recovered by simple draining.
The-seeds are subsequently extracted and washed. For
30 this, the column is fed with anhydrous alcohol which
again diffuses by percolation without subsequent
recirculation of the alcohol. Preferably, the alcohol
extraction is carried out in 3 to 9 stages. The amount
of solvent is injected for a given period of time
35 (about 4 to 10 minutes), the liquid then being drained
for a period of from 10 to 20 minutes. The liquid
recovered can undergo a step of neutralization by
addition of acid, and then a step of evaporation of the
alcohol, so as to produce a mixture of phases
WO 2011/092430 13 PCT/FR2011/050155
consisting of a lighter phase rich in esters and a more
dense phase rich in glycerol. Neither of these phases
contains curcin.
5 The phase mixture is subjected to a decanting step
(consisting, for example, of static decanting in one or
more decanters in parallel or in series, centrifugal
decanting, a combination of static or centrifugal
decanting), making it possible to obtain an upper phase
10 composed predominantly of fatty esters of a fatty acid
(ester phase) and a lower phase composed predominantly
of glycerin and water (glycerin phase).
The ester phase is subsequently subjected to a sequence
15 of chemical reactions and/or separations/purifications
aimed at recovering the fatty esters, comprising, in a
known manner, a step of washing with water followed by
a step of drying under vacuum.
20 The resulting fatty acid ester is intended
particular for the preparation of biodiesel.
The other product resulting directly from the method
according to the invention is the Jatropha oil cake.
25
According to one embodiment variant, the reduced-fat
oil cake soaked with alcohol is dried, for example in a
ventilated oven, for 4 h, at a temperature of less than
or -equal to 200°C, preferably less than or equal to
30 150°C and even more preferentially less than or equal
to 120°C. The aim of this drying step is also to
destroy the curcin remaining in the oil cake. In
parallel, this drying step makes it possible to remove,
from the oil cake, the solvent (alcohol) used during
35 the extraction.
According to another embodiment variant, the method
according to the invention does not comprise a step of
drying the oil cake at high temperature (temperature
WO 2011/092430 14 PCT/FR2011/050155
above 120°C); according to the conditions used, the
curcin can be inactivated by virtue of the physical
and/or chemical treatments applied to the Jatropha
seeds during the processing and extraction/trans-
5 esterification steps described above, such that the
operation for drying the oil cake at high temperatures
becomes needless. In this case, the method comprises
only a step of drying the oil cake at temperatures of
less than 120°C, intended to remove the solvent
10 (alcohol) used during the extraction, in order to allow
said oil cake to be used in animal feed.
The quantitative test for determining the toxic nature
of the oil cakes and also of the liquid phases
15 recovered after the extraction/transesterification step
is the acute oral toxicity test.
The publication Makkar H.P.S. et al. J. Agric, Food
Chem. 45: 8, 1997, 3152-3157 describes a quantitative
20 curcin test (hemagglutination test) and also a method
for quantitatively assaying the phorbol esters
(successive extractions with dichloromethane, followed
by analysis by HPLC).
25 The method according to the invention can without
difficulty be implemented continuously on the
industrial scale, for example by means: of a
continuously operating, moving belt reactor-extractor
(of-De Smet extractor type); of a rotary filter or of a
30 centrifuge. Preferably, the reactive grinding is
carried out with methanol in a direction countercurrent
with respect to the oil cake, over several consecutive
stages. Preferably, the alcohol extraction is carried
out in 3 to 9 stages.
35
The reactive grinding method according to the invention
is particularly suitable for mixtures of seeds, such as
soybeans, castor beans, safflower seeds or rape seeds.
Advantageously, the oil cake of Jatropha, which cannot
WO 2011/092430 15 PCT/FR2011/050155
be used pure, but as a mixture with other protein
producers, is then directly mixed with other protein
sources.
5 A starting mixture consisting of Jatropha seeds (rich
in oil) and soybeans (rich in proteins) in a proportion
of 1:10 results, by means of the method according to
the invention, in a mixture of fatty acid methyl esters
containing from 15 to 40% by weight of oleic acid
10 methyl ester, particularly suitable for use as a
biofuel.
The method for reactive grinding of seeds according to
the invention has many advantages.
15
By virtue of the step of specific processing of the
seeds, it is possible to increase the contact surface
for better percolation of the alcohol-catalyst mixture
and therefore better extraction of the lipids and their
20 subsequent conversion to esters. No prior impregnation
of the processed seeds is necessary. The ester fraction
obtained from the mixture comprising fatty acid esters
and glycerol is particularly suitable for the
production of biodiesel.
25
Starting from whole seeds makes it possible:
firstly, to greatly limit the formation of fines,
making the subsequent filtration steps easier, and
30 limiting the toxic risk since the dry fines have a
tendency to dissipate/disperse in the ambient air;
and, secondly, to maintain a good mechanical
strength of the bed of flattened seeds (that will
35 form the oil cake), this being a very advantageous
property if it is desired to carry out the
reaction in a continuous mode.
WO 2011/092430 16 PCT/FR2011/050155
The oil cakes are obtained directly from the seeds,
according to the method of the invention. These oil
cakes are devoid of toxicity with respect to human
beings and can therefore be handled risk-free.
5 Moreover, these oil cakes keep their physical integrity
(cohesion, mechanical strength) and have an
advantageous nutritive value, which allows them to be
used in animal feed.
10 The invention and the advantages thereof will be
understood more clearly on reading the examples
hereinafter given purely by way of illustration.
Reactive grinding o£ Jatropha sees
15
Table 1: characterization of the Jatropha seed tested
Characteristics Jatropha seed
(Nov 2009)
Moisture content, % 7.5
Fat, % DM 35.0
Acidity of the fat, mg KOH/g 1.8
Fatty acid distribution (relative o)
Palmitic (C16:0) 12.R
Palmitoleic (C16:1) 0.7
Stearic (C18:0) 6.4
Oleic (C18:1) 42.2
Linoleic (C18:2) 37.2
Linolenic (C18:3) 0.2
Arachidic (C20:0) 0.2
Eicosenoic (C20:1) 0.3
Phorbol ester content, mg/g 3.6
20 In terms of the oil content and the fatty acid
distribution, the Jatropha seed is in accordance with
the literature (Biodiesel & Jatropha Cultivation,
S. Lele, 2006). Its acidity of less than 2 mg KOH/g
WO 2011/092430 17 PCT/FR2011/050155
allows it to be used in the method according to the
invention.
Finally, by virtue of its phorbol ester content, very
5 much higher than 0.3 mg/g, the Jatropha seed belongs to
the toxic varieties.
Cosolvent°-free reactive grinding test
10 Reactive grinding of Jatropha seeds with methanol
extraction carried out in 3 stages (method carried out
in a fixed bed reactor)
500 g of fresh unhulled Jatropha seeds were processed
15 on a Henry flattener with smooth rollers having a fixed
gap of 0.05 mm. The flattened seeds are in the form of
petals 0.2 mm thick and 0.2 mm in diameter
approximately. The flattened seeds were dried at 60°C
for 16 h. Their final water content is 1.3% by weight.
20
In a thermoregulated fixed-bed percolation column,
these flattened and dried seeds were brought into
contact with a mixture of sodium hydroxide and
methanol, containing 0.5% by weight of sodium hydroxide
25 relative to the seed and having an alcohol/seed weight
ratio of 1.15. The extraction and transesterification
reactions are carried out at a temperature of 50°C for
30 minutes. The bed is drained for 15 minutes.
Extraction and washing of the seeds are then carried
30 out with methanol in three stages and in a
countercurrent direction.
The liquid phase obtained is subjected to decanting in
order to recover a lighter phase rich in esters and a
35 more dense phase rich in glycerol. The ester yield is
77.2%.
The oil cake obtained is subjected to drying in a
ventilated oven at 120°C for 4 h. It is noted that the
Wa 2011/092430 18 PCT/FR2011/050155
reduced-fat oil cake is relatively well depleted, with
a residual fat content of 5.4% (determined in
accordance with standard NF ISO 659).
5 Tests carried out in order to assay the toxic compounds
in the oil cake show that the oil cake is detoxified.
Optimization of the reaction in a stirred-bed reactor
10 In order to test the reactivity of the Jatropha seed,
tests are carried out in a closed stirred-bed reactor
in which the reaction is carried out on a ground seed.
In greater detail, the stirred-bed reaction is carried
out under the following conditions:
1. Drying of the whole seed at 100°C for 16 h.
20
2. Grinding of the seed at ambient temperature in
the solution in methanolic sodium hydroxide for
5 minutes.
3. Maintaining of the stirring in a reactor heated
at 50°C for 30 minutes.
25 4. Filtration on a Buchner filter (simulation of a
rotary filter) followed by washing with
anhydrous methanol.
Table 2: Mass balance for the fractionation of the
30 Jatropha seed in a stirred-bed reactor
TEST LA10-01 TEST LA1O-02
Catalyst content (vs
seed), % by mass
0.5 1.5
Methanol/ seed mass ratio 1.5 1.5
Test balance yield (%) yield (%)
Dry extract yield (1), % 34.0 110.0
Methyl ester yield, % zero 75.1
Loss in terms of esters**
WO 2011/ 092430 19 PCT/FR2011/050155
(calculated value) 100 24.
(Crude glycerin yield, % - 453El
loss in terms of esters = [theoretical mass of esters] - [mass
of esters produced] - [potential mass of esters in the reduced-fat
oil cake]
(1): The dry extract yield is the ratio of the dry extract
5 obtained after evaporation of the miscella to the sum of the
theoretical ester and of the theoretical glycerin
- In the first test (LA10-01), the amount of dry
extract of the miscella obtained represents only 34% of
10 the theoretical amount expected. Furthermore, this dry
extract, firstly, is not a two-phase extract (absence
of glycerin) and, secondly, has a neutral pH.
Consequently, under these conditions, the
extractability and the reactivity of the lipids are not
15 optimal.
- In the second test (LA10-02), carried out with 3
times more catalyst used, the dry extract content of
the miscella is at a maximum with probably other
20 extracted products (yield = 110%). The ester yield is
75%, while the glycerol yield (>> 400%) attests to the
formation of soaps. Therefore it clearly appears that
the amount of catalyst is still too high. The optimum
in terms of catalyst is therefore indeed intermediate
25 between 0.5 and 1.5%, and preferably between 0.5 and 1%
by mass, relative to the mass of seed used.
- From the qualitative point of view, the esters
produced have reasonable glyceride contents (table 3).
30
Table 3: Analytical balance of Jatropha esters
Method TEST LA10-02
Acid number (mg KOH/g) EN14104 no*
Monoglyceride content (%) EN14104 0.8
Diglyceride content (%) EN14104 0.7
Triglyceride content (%) EN14104 nd**
WO 2011/092430 20
analysis not carried out
** not detected
PCT/F122011/050155
Implementation of the reaction in a fixed-bed reactor
5 after double flattening
In parallel to the tests in a stirred-bed reactor,
tests were carried out on a fixed-bed reactor. In
summary, the fixed-bed reaction is carried out under
10 the following conditions:
1. Flaking of the fresh Jatropha seed on a flutedroller
flattener. The rollers are first apart (5.0 mm)
in order to allow a first crushing of the seed. The
15 crushed seed is then passed through the flattener
again, with rollers as tightly together as possible
(0.1 mm).
2. The flakes are then dried for 16 h at 100°C.
3. The flakes are introduced into the percolation
20 column.
4. The methanolic sodium hydroxide solution is
then recirculated over the bed for 30 minutes at 50°C.
5. The miscella is then withdrawn and the flake
bed is then washed with 5 successive washes using
25 methanol at 50°C (5 minutes per wash).
Initially, the processing of the seed was carried out
only on a fluted-roller flattener.
30 Table 4: Mass balance for the fractionation of the
Jatropha seed on a fixed bed
TEST
09-E43
TEST
10-E01
TEST
10-E02
1st flattening (pre- Yes Yes Yes
crushing) on a flattener
with separated fluted
rollers
2nd flattening with Yes Yes Yes
WO 2011/092430 21 PCT/FR2011/050155
fluted rollers tightly
together
Flake thickness 0.5-0.7 mm 0.5-0.7 mm 0.5-0.7 mm
Catalyst content (vs 0.3 0.6 1.5
flake), %
Methanol/seed mass ratio 2 2 2
Test balance Yield (%) Amount (g)
Yield (%)
Seed mass used, g 350 350 350
Dry extract yield (1), % 36.0 49.0 64.0
Methyl ester yield, % No phase 29.1 No phase
separation separation
Losses in terms of nc* 56.4 33.0
methyl esters in the oil
cake, %
Other losses in terms of nc* 14.5 66.9
esters in terms of
methyl esters**
(calculated value), %
Crude glycerin yield, % 395 249 699
* not carried out
** loss in terms of esters = [theoretical mass of
esters] - [mass of esters produced] - [potential mass
of esters in the reduced-fat oil cake]
5 (1) : The dry extract yield is the ratio of the dry
extract obtained after evaporation of the miscella to
the sum of the theoretical ester and of the theoretical
glycerin.
10 - The 3 tests carried out reveal low contents of
extracted matter in the miscellas (36, 49 and 64%),
clearly indicating that the processing of the flake
(morphology, thickness) is not optimal.
- In the presence of 0.3% of catalyst, the pH of the
15 miscella obtained is neutral and, moreover, no ester
formation is observed.
wo 2011/092430 22 PCT/FR2011/050155
- When the amount of catalyst is doubled, there is
indeed ester formation (yield 290), but the oil cake
remains very rich in lipids and the high glycerin yield
indicates a hyper production of soaps.
5 - When the amount of catalyst is tripled, the miscella
is very basic (pH > 12) and the medium is again singlephase,
the esters being saponified (explosion of the
glycerin yield >> 6000).
- From the qualitative point of view, the esters
10 produced during test 10-E01 have reasonable glyceride
contents (table 5).
Table 5: Analytical balance of Jatropha esters
Method TEST 10-E01
Acid number (mg KOH/g) EN14104 nc
Monoglyceride content (%) ARKEMA 0.8
Diglyceride content (%) ARKEMA 0.4
Triglyceride content (%) ARKEMA nd
15 * analysis not carried out
** not detected
Thus, it is envisioned to optimize the morphology of
the flake by adding a step of flattening on smooth
20 rollers in order to obtain a more extractable oil cake.
Optimization of the preparation of the Jatropha flake
after triple flattening
25 The flake preparation mode is improved in order to
reduce the loss of fat in the oil cake. The flaking is
carried out under the following conditions:
1. Flaking of the fresh Jatropha seed on a fluted-
30 roller flattener. The rollers are first apart in order
to allow a first crushing of the seed (0.5 mm) The
crushed seed is then passed through the flattener again
with the rollers as tightly together as possible
(0.1 mm). This flake is then flattened on smooth
WO 2011/092430 23 PCT/FR2011/050155
rollers with a spacing of 0.05 mm. The thickness of the
flake obtained is approximately 0.2 to 0.3 mm.
2. The flakes are then dried for 16 h at 100°C.
3. The flakes are then introduced into the
5 percolation column.
4. The methanolic sodium hydroxide solution is
then sent back over the bed for 30 minutes at 50°C.
5. The miscella is then withdrawn and the flake
bed is then washed with five successive washes using
10 fresh methanol at 50°C (5 minutes per wash).
WO 2011/092430 24 PCT/FR2011/050155
Table 6: Mass balance for the fractionation of the Jatropha seed on a fixed bed after triple
flattening
TEST 10-E08 l0-Ell l0-El3 10-E12 10-E14 10-E06
1st flattening (pre-crushing) on a
flattener with fluted rollers apart
Yes Yes Yes Yes Yes Yes
2nd flattening with fluted rollers tightly
together
Yes Yes Yes Yes Yes Yes
3rd flattening with smooth rollers tightly
together
Yes Yes Yes Yes Yes Yes'
Drying of the flake at 100°C for 16 h Yes Yes Yes Yes Yes Yes
Flake thickness (mm) 0.2-0.3 0.2-0.3 0.2-0.3 0.2-0.3 0.2-0.3 0.2-0.3
Catalyst content (vs flake), % 0.3 0.6 0 . 7 0.8 0.9 1.0
Methanol/seed mass ratio 2 2 2 2 2 2
Dry extract yield (1), % 31 77 91 98.9 99.3 97
Ester/glycerin phase separation no yes yes yes yes no
Methyl ester yield, % nc* 44.4 67.3 71.0 72.5 nc*
Crude glycerin yield, % 338 405 327 361 366 1067
Losses in terms of methyl esters in the
oil cake, %
53.9 9.7 16.7 7.5 4.2 6.3
Other losses in terms of esters in terms 46.1 45.9 16.0 21.5 23.3 93.7
PTO 2011/092430 25 PCT/FR2011/050155
of methyl esters** (calculated value), %
(1) : The dry extract yield is the ratio of the dry extract obtained after evaporation of the
miscella to the sum of the theoretical ester and of the theoretical glycerin.
* could not be carried out
5 ** loss in terms of esters = [theoretical mass of esters] - [mass of esters produced]
[potential mass of esters in the reduced-fat oil cake]
Table 7: Analytical balance of Jatropha esters
Method TEST
10-E08
TEST
l0-Ell
TEST
10-E13
TEST
10-E12
TEST
10-E14
TEST
10-E06
Acid number (mg KOH/g) EN14104 nd 0.33 0.17 0.15 0.14 nd
Monoglyceride content (%) ARKEMA nd 0.77 0.57 1.11 0.68 nd
Diglyceride content (%) ARKEMA nd 0.35 <0.1 <0.1 <0.1 nd
Triglyceride content (%) ARKEMA rid <0.1 <0.1 <0.1 <0.1 nd
WO 2011/092430 - 26 PCT/FR2011/050155
Comments:
- the triple flattening provides a boost in terms of
lipid extractability since, in the presence of at least
0.80 of catalyst, the dry extract yields are found to
5 be greater than 96%;
- the maximum ester yield observed is 71%, even though
the extractability is very high (980). Furthermore, the
high glycerin yield clearly reflects a still
substantial lipid saponification;
10 - on the other hand, from the qualitative point of
view, the ester of test 10-E12 is not very acidic and
is very low in monoglycerides and corresponds, on the
basis of these criteria, to a biodiesel quality.
Generally, the final acidity of the esters decreases
15 with the amount of basic catalyst used;
under the conditions of test 10-E12, the miscella
before evaporation is clear, but still strongly basic.
Thus, it is presumed that, after evaporation of the
methanol, the high concentration of catalyst leads to
20 parasitic saponification of the esters. For this
reason, the miscella will be neutralized before
evaporation of the methanol in the next test.
Glycerin neutralization test
25
Procedure:
1. Flaking of the fresh Jatropha seed on a flutedroller
flattener. The rollers are first apart in order
to -allow a first crushing of the seed (0.5 mm). The
30 crushed seed is then passed through the flattener again
with the rollers as tightly together as possible
(0.1 mm). This flake is then flattened on smooth
rollers with a spacing of 0.05 mm. The flake thickness
obtained is approximately 0.20 to 0.30 mm.
35 2. The flakes are then dried for 16 h at 100°C.
3. The flakes are introduced into the percolation
column.
4. The methanolic sodium hydroxide solution is then
sent over the bed again for 30 minutes at 50°C.
WO 2011/092430 27 PCT/FR2011/050155
5. The miscella is then withdrawn and the flake bed
is then washed with 5h successive washes using fresh
methanol at 50°C (5 minutes per wash).
6. The miscellas are combined and conveyed to the
5 distillation (90°C, 100 mbar).
7. Once the methanol has been evaporated off, the
glycerin and the ester are separated by decanting.
8. The ester is washed to neutrality and then dried
under vacuum (90°C, 20 mbar).
10 9. The crude glycerin is treated with an aqueous
solution of sulfuric acid in which the acid represents
5% of the mass of crude glycerin and the water
represents 100% of the mass of glycerin. The mixture is
kept stirring at, 90°C for 30 min. The mixture is then
15 separated by decanting. The fatty phase (fatty acids)
is washed to neutrality and dried under vacuum (90°C,
100 mbar).
20
Table 8: Effect of the reaction temperature
TEST 10-E20
1st flattening (pre-crushing) on a flattener
with fluted rollers apart
Yes
2nd flattening with fluted rollers tightly
together
Yes
3rd flattening with smooth rollers tightly
together
Yes
Drying at 100°C for 16 h Yes
Flake thickness 0.2-0.3 mm
Catalyst content (vs flake), % 0.8
Reaction and extraction temperature, °C 50
Methanol/seed mass ratio 2
Dry extract yield (1), % 100
Ester/glycerin phase separation Yes
Methyl ester yield, % 67.8
Crude glycerin yield before neutralization, 421
Yield of fatty acids resulting from 25.5
WO 2011/092430 28 PCT/FR2011/050155
neutralization of the crude glycerin, °
Crude glycerin yield after neutralization, 143
a
Losses in terms of methyl esters in the oil 6.7
cake, %
Other losses in terms of esters in terms of 0.0
methyl esters** (calculated value), %
(1): The dry extract yield is the ratio of the. dry
extract obtained after evaporation of the miscella to
the sum of the theoretical ester and of the theoretical
glycerin.
5 * could not be carried out
** loss in terms of esters = [theoretical mass of
,esters] - [mass of esters produced] - [potential mass
of esters in the reduced-fat oil cake]
10
Comments:
- The sulfuric acid treatment of the crude glycerin
clearly makes it possible to extract glycerin and 25%
of free fatty acids (ex-soaps) and to bring the overall
15 glycerin yield back to a more :conventional level
(1430). These fatty acids may be recycled into the
process, in particular by esterification in the
presence of an acid catalyst (sulfuric acid) and of
methanol;
20 - under the conditions of acid treatment of the
glycerin, it is noted that the esters are not
hydrolyzed (cf. table 9: analysis of the esters
resulting from test 10-E20 FFA), but that they are very
acidic (AN = 19.4, i.e. approximately 10% of free fatty
25 acids) and relatively loaded with residual glycerides;
- as regards the methyl ester phase resulting from test
10 E20, recovered after removal of the glycerin
coproduced, it still remains relatively loaded with
glycerides.
30
WO 2011/092430 29 PCT/FR2011/050155
Table 9: Analytical balance of the esters
Methyl ester Method TEST 10-E20 TEST 10-E20
ester "Free fatty acids"
Acid number EN14104 0.16 19.4
(mg KOH/g)
Monoglyceride EN14104 0.95 2.41
content (%)
Diglyceride EN14104 <0.1 <0.1
content (%)
Triglyceride EN14104 <0.1 <0.1
content (%)
Reactive grinding test in the presence of ethanol
A reactive grinding method was carried out in the
presence of ethanol under the conditions presented in
table 10.
10 Table 10: Conditions and mass balance for the method of
reactive grinding in the presence of ethanol
TEST 10-E26
1st flattening (pre-crushing) on a
with fluted rollers apart
flattener Yes
2nd flattening with fluted rollers
together
tightly Yes
3rdflattening with smooth rollers
together
tightly Yes
Drying of the flake at 100°C, 16 h Yes
Flake thickness 0.2-0.3 mm
Catalyst content (vs flake), % 0.8
Reaction and extraction temperature, °C 50
Ethanol/seed mass ratio 2
Dry extract yield (1), % 94.1
Ester/glycerin phase separation No phase
separation
Ethyl ester yield, ° nc*
WO 2011 /09243 0 30 PCT/FR2011/ 050155
Losses in terms of ethyl esters in the oil 12.1
cake, %
Other losses in terms of esters in terms of 100
ethyl esters ** ( calculated value), °
(1) : The dry extract yield is the ratio of the dry
extract obtained after evaporation of the miscella to
the sum of the theoretical ester and of the theoretical
glycerin.
5 * could not be carried out
** loss in terms of esters = [theoretical mass of
esters] - [mass of esters produced] - [potential mass
of esters in the reduced-fat oil cake]
10
Comments:
- under the conditions of the test, the medium is too
saponifying since the reaction medium is in a
nonextractable pasty form (soaps);
15 - with regard to the potential of residual ester in the
oil cake, ethanol appears to extract less fat than
methanol (12.1% vs 7.5% in test 10-E12);
- despite these observations, the "ethanolic" oil cake
was analyzed in order to determine its phorbol ester
20 content (table 17).
Reactive grinding tests in the presence of a cosolvent
Test with a methanol/hexane (28/72) (m/m) mixture
25 In the context of these tests, given the high
volatility of hexane, the reaction temperature was
lowered to 40°C.
WO 2011 / 092430 31 PCT/FR2011 / 050155
Table 11: Influence of the presence of a methanol/hexane (28/72) (m/m) cosolvent
TEST 10-E21 10-E19 10-E18(2)
1st flattening (pre-crushing) on flattener with fluted rollers
apart
Yes Yes Yes
2nd flattening with fluted rollers tightly together Yes Yes Yes
3rd flattening with smooth rollers tightly together Yes Yes Yes
Drying of the flake at 100°C, 16 h Yes Yes Yes
Flake thickness 0.2-0.3 mm 0.2-0.3 mm 0.2-0.3 mm
Catalyst content (vs flake), % 0.5 0.7 0.9
Reaction and extraction temperature, °C - 40 40 40
Solvent (hexane+methanol)/seed mass ratio 2 2 2
Dry extract yield (1), % 100.1 103.3 101.6
Ester/glycerin phase separation Yes Yes Yes
Methyl ester yield, % 101.4 98.2 107.4*
Crude glycerin yield, % 87.8 154 43.9*
Losses in terms of methyl esters in the oil cake, % 3.3 3.5 3.0
Other losses in terms of esters in terms of methyl esters**
(calculated value), %
-4 .7 -1.7 -10.4
(1) : The dry extract yield is the ratio of the dry extract obtained after evaporation of the
miscella to the sum of the theoretical ester and of the theoretical glycerin.
WO 2011/092430 32 PCT/ 2011/050155
(2) Relative to test 10-E18, the ester yield is higher than the theoretical yield, and the glycerin
yield is abnormally low. As it happens, it is found that the ester phase has an emulsified
appearance and therefore seems to retain nondecantable glycerin.
* could not be carried out
5 ** loss in terms of esters = [theoretical mass of esters] - [mass of esters produced] - [potential
mass of esters in the reduced-fat oil cake]
Comments:
- In the presence of the methanol-hexane mixture, the oil cakes are correctly depleted.
10 - While the ester yield in test 10-E19 appears to be high, it is found that said esters are loaded
with glycerides (table 12), indicating that the reaction medium (depending on the methanol/hexane
ratio) is not transesterifying enough. This result is, moreover, confirmed regardless of the
catalyst content. It is also noted that, at very high catalyst content (0.9%), the medium even
becomes saponifying.
15
Table 12: Analytical balance of esters
Method 10-E21 10-E19 10-E18
Acid number (mg KOH/g) EN14104 0.43 0.17 0.18
Monoglyceride content (%) EN14104 2.29 1.4 2.2
Diglyceride content (%) EN14104 8.53 2.8 3.9
Triglyceride content (%) EN14104 <0.1 <0.1 <0.1
5
WO 2011/092430 33 PCT/ 011/050155
Perspectives:
A test with a higher methanol content was therefore carried out:
Reactive grinding tests in the presence of a hexane/alcohol mixture enriched in methanol:
Table 13. Influence of methanol content
TEST 10-E24 10-E-25
1st flattening (pre-crushing) on flattener with fluted rollers apart Yes Yes
2nd flattening with fluted rollers tightly together Yes Yes
3rd flattening with smooth rollers tightly together Yes Yes
Drying of the flake at 100°C, 16 h Yes Yes
Flake thickness 0.2-0.3 mm 0.2-0.3 mm
Catalyst content (vs flake), % 0.7 0.7
Reaction and extraction temperature, °C 40 40
Methanol/hexane mass ratio 90/10 50/50
Solvent (methanol/hexane)/seed mass ratio 2 2
Dry extract yield (1), % 102.2 104.3
Ester/glycerin phase separation Yes Yes
WO 2011 /0 92430 34 PCT/ 2011/050155
Methyl ester yield, % 74.4 88.3
Crude glycerin yield, % 379 263
Losses in terms of methyl esters in the oil cake, % 10.9 2.8
Other losses in terms of esters in terms of methyl esters** (calculated
value), %
14.7 8.9
(1) : The dry extract yield is the ratio of the dry extract obtained after evaporation of the
miscella to the sum of the theoretical ester and of the theoretical glycerin.
* could not be carried out
** loss in terms of esters = [theoretical mass of esters] - [mass of esters produced] - [potential
5 mass of esters in the reduced-fat oil cake]
wO 2011/092430 35 PCT/FR2011/050155
Comments:
- in the proportions 90/10, the addition of a large
amount of methanol significantly alters the lipid
extractability (cf. oil cake ester potential) and
5 therefore the methyl ester yield. Under these
conditions, the medium is however more transesterifying
than in the presence of the methanol/hexane mixture =
28/72 (table 14, cf. % glycerides);
- when the hexane is increased (50/50 mixture), the oil
10 cake is, finally, correctly depleted (2.8% of losses of
methyl esters in the oil cake) . In fact, the overall
methyl ester yield is much improved (88.3%), clearly
indicating that the transesterifying activity is
regained under these conditions;
15 - from the qualitative point of view (table 14), the
methyl esters produced are not very acidic even though
their monoglyceride content is still high (probably due
to a retro-conversion of the methyl esters to
glycerides).
20
Table 14: Analytical balance of the esters
Method 10-E24 10-E25
Acid number (mg KOH/g) EN14104 0.17 0.20
Monoglyceride content (%) EN14104 1.26 1.36
Diglyceride content (%) EN14104 0.2 0.3
Triglyceride content (%) EN14104 <0.1 <0.1
Evaluation of the detoxifying effect of the method by
25 measuring the change in phorbol ester content
The preparation of the samples and also the assaying of
the phorbol esters were carried out according to the
method of Makkar (Makkar HPS, Becker K, Sporer F, Wink
30 M (1997) Studies on nutritive potential and toxic
constituents of different provenances of Jatropha
curcas. J Agric Food Chem 45:3152-3157).
3.1 Sample preparation
Wa 2011 / 092430 36 PCT/FR2011 / 050155
The liquid samples are diluted in methanol and then
injected. For the solid samples, the phorbol esters are
first of all extracted with a pestle and mortar in the
presence of methanol. The alcoholic extracts obtained
5 are then analyzed by high performance liquid
chromatography.
3.2 Operating conditions:
Chromatographic conditions:
10 - Detector: diode array (peak integration at 280 nm).
- Column: C18 reverse phase (LiChrospher 100, 5 mm),
250 x 4 mm + precolumn.
- Oven: 22°C (amb T).
- Eluents:
15 B Acidified water (1.75 ml H3PO4 (85%) in 1 litre of
demineralised water).
A = acetonitrile
3.3 Results
20 Reactive grinding with methanol on triple flattened
flake with retreatment of the glycerine (test 10E20)
25
Table 15: Phorbol ester distribution in a fractionation
with retreatment of the glycerine/test 10E20
Mass
g
PE
content
mg/g
PE
mass
mg
PE
distribution
(in % PEs of
the seed)
Seed used 346.5 3.5 1230 -
Dried flake 346.5 2.6 925 75.2
Methyl ester 82.2 4.7 386 31.4
Ester resulting from
the retreated glycerin
(FFA)
30.9 7.6 235 19.1
Retreated glycerin 17.3 2.7 47 3.8
Oil cake 216. 1 0.3 65 5.3
Total losses - - 286 40.4J
WO 2011/092430 - 37 RCT/FR2011/050155
Comments:
- the losses of phorbol esters (PEs) after flaking and
drying are approximately 250. The PEs appear to be
relatively sensitive to increased temperature. Overall,
5 the method results in PE losses of 40%;
- approximately 1/3 of the PEs of the seed are found in
the methyl esters and 5% in the dried oil cake;
- in view of the PE contents in the esters directly
resulting from the method or after acid retreatment of
10 the glycerin, the PEs appear to have a relative
affinity for lipophilic compounds (methyl ester phase);
- the residual content in the oil cake is 0.3 mg/g,
i.e. very close to the values of the non-toxic Mexican
varieties (0.1 mg/g). There is therefore clearly a
15 positive and detoxifying effect of the method according
to the invention.
Reactive grinding with cosolvent on triple-flattened
flakes
20
Table 16: Phorbol ester distribution in a method with
cosolvent/test lOE25
Mass
g
PE
content
mg/g
PE
mass
mg
PE
distribution
(in % of PEs
of the seed)
Seed used 346.5 3.5 1230 -
Methyl ester 107.1 5.5 589 47.9
Glycerin 31.8 2.7 86 7.0
Oil cake 207.6 0.3 62 6.7
Losses - - 493 38.4
25
Comments:
- in the presence of cosolvent (hexane), the method
according to the invention produces an oil cake with a
low PE content (0.3 mg/g), whereas the methyl esters
WO 2011/092430 38 PCT/FR2011/050155
capture 50% of the PEs of the seed. This result appears
to confirm the relatively liposoluble nature of PEs;
- here again, overall, the method results in PE losses
of 40%.
5
Moreover, with regard to the low PE contents in the oil
cakes resulting from the method, it may be concluded
that said method is detoxifying in nature. It can also
be readily considered that this content will be further
10 improved at the industrial level, comprising drying of
the flake and of the oil cake in a toaster.
Transformation of the Jatropha seed by means of a
conventional method (pressing + semi-refining of the
15 oil + methanolysis; comparative example)
The Jatropha seed is ground by pressing in order to
obtain a crude press oil and an oil cake.
20 For this, the seed undergoes the following steps:
Grinding:
The seed is crushed on a fluted-roller
flattener.
The flakes are then conveyed to the heated Taby
25 press with no die.
The crude press oil obtained is then filtered
through an 11 um cellulose filter.
Before being esterified, the crude oil undergoes semirefining
which comprises the following steps:
30 - Mucilage removal
- Neutralization
Mucilage removal:
the oil is heated to 65°C;
35 - when the temperature of 65°C has been reached,
a mixture composed of 1.5%0 of phosphoric acid
and 6% of water (% by mass relative to the mass
of dry oil) is added;
WO 2011/092430 39 PCT/FR2011/050155
- the mixture is then kept stirring for 10 min.
The temperature is then increased to 75°C and
maintained for 30 min. The mixture is then
centrifuged for 5 min at 4500 rpm.
5
Neutralization:
The dephosphorized oil is neutralized with an
aqueous solution of sodium hydroxide composed
of 6% water (relative to the mass of oil) and
10 of sodium hydroxide required to neutralize all
the free fatty acids with an excess of 5%. The
sodium hydroxide solution is added to the
dephosphorized oil heated to 75°C, and the
mixture is maintained for 10 minutes. The
15 temperature is then increased to 90°C for 30
minutes. The mixture is then centrifuged for 5
min at 4500 rpm in order to remove the soapy
heavy phase. The oil is then washed to
neutrality with demineralized water by
20 successive additions of 20% water with stirring
for 5 min and centrifugation for 5 min at
4500 rpm. The oil is then dried under vacuum at
90°C (20 mbar).
25 The semi-refined oil is then transesterified with
methanol in the presence of a basic catalyst.
Transesterificat on and purification of esters
30 - in a first step, the semi-refined oil is brought into
contact with anhydrous methanol in an oil/methanol mass
ratio of 5/1;
- the mixture is then brought, with stirring, to the
reflux temperature of methanol (65-70°C);
35 - sodium methoxide (methanolic solution of catalyst at
25%) is then gradually added (3 additions) in an
oil/methanol/catalyst mass ratio of 5/1.02/0.03;
- the mixture is then refluxed with stirring for 2
hours;
WO 2011/092430 40 PCT/FR2O11/050155
- after having left the glycerin (heavy phase) to
decant for 1 hour, the latter is removed by drawing off
from the reactor;
- the ester phase is then washed to neutrality with
5 demineralized water (each wash is carried out with
stirring for 15 min at 90°C);
- finally, the esters are dried under vacuum at 90°C
(20 mbar).
10 Table 17: Balance for the conventional grinding of the
datropha seed
TEST 10-E27
Pre-crushing on z flattener with fluted
rollers apart
Yes
Flattening with fluted rollers tightly
together
No
Flattening with smooth rollers tightly
together
No
Drying 100°C 16 h No
Taby press with no die Yes
Press oil/oil cake proportion (ms) 21/79
Press yield, % 60.2
Semi-refining yield 97%
Transesterification yield 98%
Table 18: Analytical balance of the Jatropha oils and
15 esters obtained by means of the conventional method
Method Crude
press
oil
Semirefined
oil
Methyl
ester
Acid number (mg KOH/g) EN14104 2.0 0.20 0.1
Phosphorus content NFT60-227 >25 <5 <5
Monoglyceride content (%) ARKEMA - - 1.31
Diglyceride content (%) ARKEMA - - 0.75
Triglyceride content (%) ARKEMA - - 0.0
WO 2011/092430 41 PCT/FR2011/0501555
Table 19: Phorbol ester distribution in a fractionation
with the conventional method 10-E27
Mass
g
PE
content
mg/g
PE
mass
mg
PE
distribution
(in % of PEs
of the seed)
Seed used 100 3.5 350 -
Oil cake 73. 0 2.4 175.2 50.1
Crude press oil 19.5 11.5 214.5 61.3
Semi-refined oil 19.0 7.4 140.6 40.2
Methyl ester 18.6 3 .1 57.7 16.5
Crude glycerin 2.3 0.8 1.8 0.5
Total losses - - 115.3 32.9
5 Comments:
- a large part (61.3%) of the phorbol esters is found
in the press oil, but said esters degrade over the
course of the semi-refining and transesterification
steps;
10 - the press oil cake has a very high PE content
compared with the oil cake resulting from the method
according to the invention, approximately 12 times
higher. It is noted in passing that the balance with
regard to the PEs in the oil cake and in the press oil
15 is slightly in excess (+10%)g
- the semi-refining (neutralization + mucilage removal
+ drying) leads to a loss of 1/3 of the PEs of the
crude oil;
- the methanolysis leads to a loss of 60% of the PEs of
20 the semi-refined oil;
- the esters obtained are identical in terms of PEs to
the esters resulting from the method according to the
invention;
- finally, approximately 33% of the phorbol esters were
25 degraded in the conventional method, compared with 40%
in the method according to the invention.
Reactive grinding with ethanol
WO 2011/092430 42
5
PC'T/FR2011/050155
Table 20: Phorbol ester distribution in the products
resulting from a reactive grinding method carried out
in the presence of ethanol (E10-E26)
Product Mass
g
PE
content
mg/g
PE
mass
mg
PE
distribution
(in % of PEs
of the seed)
Seed used 346.5 3.5 1230 -
Dried flake (1) 346.5 2.6 925 75.2
Paste 126.9 NC* - -
Oil cake (2) 233.9 1.0 234 19.0
(1) T = 100°C, 16 hours
(2) T = 100°C, 16 hours
* NC = analysis not carried out
10 Comments:
- the oil cake resulting from a method with ethanol
proves to be 3 times more concentrated in terms of PEs
than a methanolic oil cake, although its residual fat
content is only 1.7 times higher. Consequently, the PEs
15 appear to be less soluble in ethanol than in methanol;
- as previously observed, and inexplicably, more
extensive drying of the oil cake leads to an increase
in the PE content in the oil cake.
20 In fact, the reactive grinding method according to the
invention, in particular with methanol, and preferably
in the presence of cosolvent and/or of a flake prepared
by triple flattening of the seed, makes it possible to
go directly from Jatropha seeds to fatty acid esters
25 with a yield greater than 70%, or even greater than
80%, and simultaneously to obtain a detoxified oil cake
containing a maximum of 0.03 mg/g of phorbol esters,
said content being compatible with use of the oil cake
in animal feed.
30
WO 2011/092430 43

CLAIMS
PCT/FR2011/050155
1. A method for treating seeds containing toxic
components such as curcin, abrin, cretin and/or
5 phorbol esters, in particular Jatropha seeds, said
seeds preferably having a degree of acidity of
less than or equal to 3 mg KOH/g, said method
comprising the following steps:
i) a seed processing step;
10 ii) a step of bringing the processed seeds into
contact with a light anhydrous alcohol and an
alkaline catalyst under temperature and time
conditions sufficient to allow the extraction
and the transesterification of the vegetable
15 oil, and producing a mixture comprising fatty
acid esters and glycerol, and an oil cake,
characterized in that step i) comprises seed
flattening and drying operations.
20 2. The method as claimed in claim 1, in which step
i) also comprises preheating the seeds at a
temperature of less than or equal to 100°C, the
preheating operation being carried out before the
flattening.
25
3. The method as claimed in either of claims 1 or 2,
in which the flattened-seed drying operation of
step i) is carried out rapidly after flattening,
in less than one hour, preferably after 5 to 10
30 minutes, at a temperature sufficient to reduce the
moisture content of the seeds to 2% by weight or
less.
4. The method as claimed in any one of claims 1 to 3,
35 in which step ii) comprises a first reaction
carried out at a temperature ranging from 30 to
75°C, preferably less than or equal to 50°C,
preferably less than 45°C, preferably
approximately equal to 40°C, for 15 to 60 minutes,
WO 2011 / 092430 44 PCT/FR2011 / 050155
preferably 20 to 40 minutes, followed by
extraction with alcohol carried out in 3 to 9
stages and in a countercurrent direction.
5 5. The method as claimed in any one of claims 1 to 4,
in which the flattening operation is carried out
by means of a mechanical roller flattener, and
preferably comprises triple flattening on smooth
rollers.
10
6. The method as claimed in any one of claims 1 to 5,
in which steps i) and ii) are carried out
continuously.
15 7. The method as claimed in any one of claims 1 to 6,
in which the light alcohol is methanol.
8. The method as claimed in any one of claims 1 to 7,
in which the alkaline catalyst is sodium
20 hydroxide.
9. The method as claimed in any one of claims 1 to 8,
in which the catalyst/alcohol/seeds weight ratio
is included in the range 0.001 to 0.01/0.1 to 5/1,
25 preferably in the range from 0.005 to 0.01/0.1 to
1/1, even more preferably in the range from 0.005
to 0.01/0.1 to 0.5/1.
10. The method as claimed in any one of claims 1 to 9,
30 in which, in step ii), a cosolvent is also added,
selected from the group: hexane, heptane, benzene,
bicyclohexyl, cyclohexane, decalin, decane,
spirit, petroleum ether, kerosene, kerdane, diesel
oil, paraffin oil, methylcyclohexane, naphtha
35 (Texsolve V) , skellite, tetradecane, Texsolve (B,
C, H, S, S-2, S-66, S-LO, V), supercritical CO2,
propane or butane which are pressurized, natural
solvents such as terpenes (limonene, alpha- and
beta-pinene) , ethers such as dimethyl ether or
WO 2011/092430 45 PCT/FR2011/050155
diethyl ether, ketones such as acetone, and
mixtures of all these solvents.
11. The method as claimed in any one of claims 1 to
5 10, in which the mixture comprising fatty acid
esters and glycerol is subjected to a decanting
step which makes it possible to obtain an upper
phase composed predominantly of fatty esters of a
fatty acid and a lower phase composed
10 predominantly of glycerin and of water.
12. The method as claimed in claim 11, in which said
upper phase is subjected to a succession of
chemical reactions and/or of
15 separations/purifications producing biodiesel.
13. The method as claimed in any one of claims 1 to
12, in which the Jatropha seeds are mixed with
soybeans in a ratio of 1 to 10.
20
14. A mixture of fatty acid methyl esters which can be
obtained by means of the method as claimed in
claim 13, comprising from 15 to 40% by weight of
oleic acid methyl esters.
25
15. The use of the mixture as claimed in claim 14, as
a biofuel.
16. - The method as claimed in any one of claims 1 to
30 12, in which the oil cake obtained is subjected to
a drying step under temperature and time
conditions sufficient to inactivate the curcin.
17. The method as claimed in claim 16, in which the
35 drying of the oil cake is carried out for 4 h at a
temperature of less than or equal to 200°C,
preferably less than or equal to 150°C and even
more preferentially less than or equal to 120°C.
WO 2011/092430 46 PCT/FR2011/050155
18. A detoxified' jatropha oil cake which can be
obtained by means of the method as claimed in one
of claims 1 to 12, 16 and 17, which has:
- a degree of curcin detoxification of at least
5 90% and preferably of at least 95%, with
respect to activity, when this degree is
measured by means of a quantitative test;
- a phorbol ester content of less than or equal
to 0.3 mg/g.
10
19. The use of the oil cake as claimed in claim 18, in
animal feed.