Note: Descriptions are shown in the official language in which they were submitted.
CA 0221~ 1997-09-11
WO 96131S81 ~ IIZ42
Fractionation o~ triglyceride oils
The present invention is concerned with a process for
fractionating triglyceride oils, particularly lauric oils.
The fractionation (fractional crystallization) of
5 triglyceride oils is described by Gunstone, Harwood and
Padley in The Lipid ~n~hook~ 1986 edition, pages 213-215.
Generally triglyceride oils are mixtures of various
triglycerides having different melting points. The present
invention refers to a specific lauric oil: palmkernel oil
10 me composition of triglyceride oils may be modified e.g
by fractionation_yielding a fraction having a different
melting point or solubility.
One fractionation method is the so-called dry fractionation
process which comprises cooling the oil until a ~olid phase
15 crystallises and separating the crystallised phase from the
liquid phase. The liquid phase is denoted as olein
fraction, while the solid phase is denoted as stearin
fraction.
The separation of the phases is usually carried out by
20 filtration, optionally applying some kind of pressure.
CA 0221~ 1997-09-11
WO 96131581 PCT/EP96/01242
The major problem encountered with phase separation in the
dry fractionation process is the inclusion of a lot of
liquid olein fraction in the separated stearin fraction.
The olein fraction is thereby included in the inter- and
5 intracrystal spaces of the crystal mass of the stearin
fraction. Therefore the separation of the solid from the
liquid fraction is only partial.
The solids content of the stearin fraction is denoted as
10 the separation efficiency. In dry fractionation it seldom
surpasses 50 to 60 wt.~. This is detrimental to the quality
of the stearin as well as to the yield of the olein.
For the related solvent fractionation process, where the
fat to be fractionated is crystallised from e.g. a hexane
15 or acetone solution, separation efficiencies may be up to
9S~ .
Dry fractionation, however, is a process which is more
economical and more environmentally friendly than solvent
fractionation. An increase of separation efficiency for dry
20 fractionation is therefore much desired, particularly for
the commercially important palmkernel oil.
It is known to interfere with the crystallization by adding
to a crystallising oil a substance which will generally be
25 indicated as a crystallization modifying substance. The
presence of small quantities of such a substance in the
cooling oil may accelerate, retard or inhibit
crystallization. In certain situations the above substances
are more precisely indicated as crystal habit modifiers.
30 Known crystallization modifiers are e.g. fatty acid esters
of sucrose, described in US 3,059,010, US 3,059,010,
JP 05/125389 and JP 06/181686, fatty acid esters of glucose
and derivatives, described in US 3,059,011. These
crystallization modi~iers are e~fective in speeding up the
35 crystallization rate.
- CA 022l~ l997-09-ll
WO 96~31S81 PCI~/EP96/01242
Other crystallization modifiers, e.g. as described in US
3,158,490 when added to kitchen oils have the effect that
solid fat crystallization is prevented or at least
retarded. Other types of crystallization modifiers,
5 particularly referred to as crystal habit modifiers, are
widely used as an ingredient for mineral fu~l oils in which
waxes are prone to crystallize at low temperatures. US
3,536,461 teaches the addition of a crystal habit modifier
to fuel oil with the effect that the cloud point (or pour
10 point) temperature is lowered far enough to prevent crystal
precipitation. Or, alternatively, the solids are induced to
crystallize in a different habit so that the crystals when
formed can pass fuel filters without clogging them.
Other crystal habit modifiers are actually able to change
15 the habit of the crystallized triglyceride fat crystals in
a way such that after crystallization the crystals, the
stearin phase, can be more effectively separated from the
liquid phase, the olein phase. Publications describing such
crystal habit modifiers are e.g. GB 1 015 354, US
20 2,610,915, co-p~nA; ng PCT application WO 95/04122,
US 3,059,008, US 3,059,009 and US 3,059,010.
Separation efficiency also depends on the mode of
crystallisation, either stagnant or stirred. Often good
25 results are obt~; n~A with stagnant crystallisation rather
than with stirred crystallisation. From the point of view
of process economy, however, stirred crystallization is
preferred.
P~lmk~rnel oil crystallizes in a needle-like morphology.
30 The resulting crystal agglomerates (see Figure lA) easily
~ include olein. Fractionation by stirred crystallization is
not possible because the hydrodynamic shear would crush the
needles and produces crystal slurries which can not be
separated in a stearin and an olein phase. Quiescent
35 crystallization (the stagnant mode) produces large, flake-
like crystallites which can be separated, provided a high
CA 0221~ 1997-09-11
WO 96/31581 PCTlEPg6/01242
pressure regime o~ 25-30 bar is applied. There~ore
palmkernel oil ~ractionation is only possible in a stagnant
mode which is a very labour intensive process. An e~fective
crystallization modifying substance which enables stirred
5 mode ~ractionation is badly needed.
STATEMENT OF lN v~NllON
10 It has been found that palmkernel oil can be crystallized
in a stirred mode so that the stearin phase and the olein
phase can be separated. By crystallization large and non-
porous spherulites are ~ormed which are easily separated
and - also when working in stagnant mode - the separation
15 efficiency is considerably increased. Accordingly the
invention relates to a process ~or separating solid fatty
material crystallised from palmkernel oil, which comprises
the steps:
a. heating the oil until no longer a substantial
20 amount o~ solid triglyceride is present in the oil,
b. cooling and crystallising the triglyceride oil
resulting in a solid stearin phase besides a liquid olein
phase and
c. recovering the stearin phase by separating it
25 from the olein phase,
where be~ore crystallization starts a crystallization
modifying substance is added to said triglyceride oil or to
a solution of said triglyceride oil in an inert solvent,
characterized in that the crystallization proceeds in a
30 stirred mode.
DESCRIPTION OF T~E FIGURE
Figure lA shows agglomerates o~ needle-like crystals o~
35 palmkernel oil obtained by quiescent crystallization
without additive.
CA 022l~ l997-09-ll
WO 96/31581 PCT/EP96~0I242
Figure lB shows spherulite crystals of palmkernel oil
obtained by stirred crystallization in the presence of
sucrose polylaurate.
DETAILS OF ~1~ lN V~ llON
The palmkernel oil to be ~ractionated is m; ~ with the
crystallization modifying substance (the additive) be~ore
10 crystallization starts, preferably before the oil is heated
or dissolved in the solvent so that all solid triglyceride
fat and preferably also the modifying substance is
lique~ied.
Then the oil or solution is cooled to the chosen
15 crystallisation temperature. A-suitable temperature range
is 15-35~C. To each temperature belongs a specific
composition of the olein and stearin phases.
Crystallisation proceeds at the chosen temperature until
the crystallised oil stabilises to a constant solid phase
20 content. The crystallisation time increases when more solid
phase is desired and the temperature is lowered. Usual
times are in the range of 4-16 hours. During
crystallisation the oil is stirred, e.g. with a gate
stirrer.
- 25
The additive is a sugar polyester o~ fatty acids. Preferred
polyesters are chosen from the group consisting of sucrose
polylaurate (100wt.~ laurate), sucrose polyester (75wt.
laurate, 25wt.~ palmitate), sucrose polyester (50wt.
30 laurate, 25wt.~ palmitate, 25wt.~ stearate), sucrose
polyester (25wt.~ laurate, 35wt.~ palmitate, 40wt.
stearate), sucrose polyester (75wt.~ caprate, 25wt.
palmitate), sucrose polyester (75wt.~ myristate, 25wt.~
palmitate) and sucrose polypalmitate (100 wt.~ palmitate),
35 where the ratios are denoted as weight percentages.
Sucrose polyester in the context of this description
=~
CA 0221~ 1997-09-11
WO g6131581 PCTIEP96/01242
denotes a sucrose ester of which on average at least four,
preferably ~ive to six o~ the eight hydroxyl groups have
been esterified with one or more types of fatty acids The
ester can be obt~;n~ by well-known usual processes such as
5 esterification of sucrose with ~atty acids or a mixture of
fatty acids or of reactive fatty acid derivatives. Sucrose
with more than four free hydroxyl groups has an
insu~icient oil solubility.
10 Sucrose polylaurate as mentioned in this specification is a
highly esterified (50-100~) sucrose ester with a lauric-
content of 95 wt.~. It is a readily available commercial
product (e.g. ~yoto Sugar Ester L195, ex Mll~u~ISHI).
15 The stearin and olein phases may be separated by filtration
but for an effective separation o~ the solid from the
liquid phase the higher pressures of a membrane filter
press are used. Suitable pressures are 3-50 bar, exerted
for about 20-200 minutes. However, the invention allows a
20 low or moderate pressure. As a rule with a pressure of 6-12
bar it takes about 30-60 minutes to get a proper separation
of the stearin phase from the olein phase.
The solids content of the crystal slurry before separation
and o~ the stearin phase obtained after separation is
25 measured according to the known pulse NMR method (ref.
Fette, Seifen, Anstrichmittel 1978, 80, nr. 5, pp. 180-
186).
The e~fect of the invention is believed to be caused by
30 alteration of the crystal structure or crystal habit of the
stearin under the influence o~ the additive. These might
interfere in different ways with the growth of the various
crystal faces.
At microscopic inspection (see Figure lB) the effect of the
35 additive is that the crystals and crystal aggregates ~ormed
in the oil are conspicuously different from the crystals
CA 0221~ 1997-09-11
WO g6131581 P~ nl242
obt~; ne~ without the crystallization modifying substance
(Figure lB). Instead of brittle needle-like crystals, large
and non-porous spherulites are formed Since a stearin
~raction with such crystals retains less of the olein
5 fraction, even at low or moderate filtration pressure, the
altered crystallization results in a considerable increase
of the separation e~ficiency and facilitates stirred
crystallization.
10 Although the invention is useful for solvent fractionation
or detergent fractionation, the process is carried out
preferably as a dry fractionation process.
The sucrose polyester is suitably applied in an amount of
15 0.005 - 2 wt.~ on the total amount of oil. A useful amount
is about 1 wt.~.
The invention is illustrated by the following examples:
r-- _les 1-2
A sample was prepared cont~;n;ng 1000 g of palmkernel oil
(neutralised, bleached, deodorised) and 10 g (1~) of
25 sucrose polylaurate. The sample was heated and stirred at
65~C until completely liquefied (no ~olid fat content) and
then slowly cooled. Crystallization proceeded in a stagnant
(0 rpm) mode at the chosen temperature of 23~C until a
constant solid phase content was reached. The sample was
30 filtered and pressed at 12 bar for 30 minutes. After
filtration and pressing, the solid phase content (SE
= separation efficiency) of the cake was measured by NMR.
For comparison the fractionation was repeated with the only
difference that no sucrose polylaurate was added.
CA 0221~ ss7-os-11
W096/31581 PCTtEP96tO1242
In example 2, palmke~n~l oil was fractionated following the
same procedure as described above, but in a stirred mode (5
rpm).
Examples 3-8
The dry fractionation process of example 2 is repeated with
various sucrose polye~ters. Each experiment has been
carried out with and without crystallization modifying
lO substance. Table I indicates the SE and the relative
improvements caused by the modifier.
The used crystal habit modifiers are: #3 sucrose polyester
(75~ laurate, 25~ palmitate), #4 sucrose polyester (50~
laurate, 25~ palmitate, 25~ stearate), #5 sucrose polyester
15 (25~ laurate, 35~ palmitate, 40~ stearate), #6 sucrose
polyester (75~ caprate, 25~ palmitate), #7 sucrose
polyester (75~ myristate, 25~ palmitate) and #8 sucrose
polypalmitate, where the ratios are denoted as weight
percentages.
The results of Table I show that the addition of various
sucrose polyesters allow the stirred fractionation of
palmkernel oil with a surprisingly high separation
efficiency.
CA 022l~ l997-09-ll
WO 96~31581 P< ~ 0~242
. .
TABLE I
# C~ rpm c SE
sucrose polylaurate 0 0.0 46
+59
1.0 73
2 sucrose polylaurate 5 0.0 n.d.
1.0 71
3 sucrose polyester 50.0 n.d.
(75~ laurate,
25~ palmitate) 1.0 69 ++
4 sucrose polyester 50.0 n.d.
(50~ laurate,
25~ palmitate 1.0 73 ++
25~ stearate)
sucrose polyester 50.0 n.d.
(25~ laurate,
35~ palmitate 1.0 73 ++
40~ stearate)
6 sucrose polyester 50.0 n.d.
(75~ caprate,
25~ palmitate) 1.0 60 ++
7 sucrose polyester 50.0 n.d.
(75~ myristate
25~ palmitate) 1.0 72 ++
8 sucrose 50.0 n.d.
polypalmitate 1.0 48
c in wt~ conc. o~ crystallization modi~ying substance
rpm stirrer rotation speed
SE in wt~ separation e~îiciency
~ in ~ enhancement SE relative to blank experiment
n.d. not done: unable to separate solid/liquid phase.
