This disclosure relates to the dewaxing of citrus oils
and to a method of achieving this.
It is desirable to dewax natural oils in order to avoid
the formation of precipitate in finished citrus-based products, and there have been
described a number of methods for the dewaxing of vegetable oils. These include
filtration through one of the variety of porous membranes available on the market,
typically porous ceramic or polymeric membranes, and such methods have produced
One test that is carried out in industry to check the efficiency
of dewaxing is the cooling test, in which a sample of the dewaxed material is allowed
to stand at 4°C for a prolonged period. There is no standard period, but 48
hours is a useful time. In the case of citrus oils, there is often a good result,
but not always. In some cases, there is a visibly turbid solution or some precipitate
at the end of the 48 hours, indicating that a substantial amount of wax remains.
In addition, samples may pass the 48 hour test immediately after production, but
when tested a few weeks later they may fail. This generally means that additional
processing is needed, which adds to the expense of the processing. In addition,
the fact that a good result cannot be guaranteed with citrus oils introduces uncertainties
into the whole process.
It has now been found that this problem can be substantially
or even completely overcome by a particular method. There is therefore provided
a method of dewaxing a citrus oil by subjecting the oil to a filtration process,
in which the oil is caused to flow parallel to a porous membrane during which process
it is maintained at a temperature of less than 11°C, said membrane having an
average pore size of from 0.05-5 micrometres.
There is additionally provided a substantially wax-free
citrus oil, prepared by a process as hereinabove described.
By "citrus oil" is meant any oil derived from any of the
citrus family of fruits, such as orange, lemon, lime, tangerine, grapefruit and
The filtration process involves causing the oil to flow
(without limitation, such as by pumping) parallel to a porous membrane surface,
as opposed to the more usual so-called "dead end" filtration method, in which the
flow is perpendicular to the filtration surface.
By "about 11°C" is meant that this temperature does
not represent a sharp cut-off between what works and what does not. Citrus oils
are complex mixtures of materials, the natures and proportions of which often differ,
depending on source. Thus, although there may be oils that will give good results
above 11°C, the incidence of failure increases sharply as 11°C is approached,
and the majority of citrus oils will fail above 11°C. The choice of 11°C
therefore represents the results of practical observation and a slight temperature
variation above 11°C is, for the purposes of this disclosure, considered to
lie within its scope.
Suitable membranes that are able to withstand the physical
and chemical rigors of the application are well known in the art and are readily
available. Some typical types include membranes of metal, ceramic, graphite or polymeric
materials, which may be self-supporting or deposited on a support. Examples of the
latter type include metals, ceramics and polymers on ceramic supports, zirconia
on a graphite support, titanium dioxide on a stainless support or polymer on a support
of the same or a different polymer. Well-known commercial products include the FICL
filter of Doulton USA and the Membralox™ membranes of Pall Corp.,
All such membranes and any associated supporting apparatus
are generally available in modular form for easy installation and replacement.
Pumps for use with such a filter are well known to the
art, and a suitable pump can readily be selected from any of the commercially-available
pumps known to the art.
The pressure causing the oil to flow through the filter
may be provided by the pump. This creates a pressure on the feed or "retentate"
side of the membrane, and if this is greater than the pressure on the product or
"permeate" side, the oil will flow through the membrane. Appropriate pressures to
give optimal flux may be determined in each case by simple experimentation.
The average pore size may be between 0.05 and 5 micrometres.
Particular examples have average pore sizes between 0.2 and 3 micrometres, and between
0.2 and 1.4 micrometres. The provision of a suitable degree of porosity in any selected
membrane material is well within the skill of the art.
The oil is caused to flow parallel to the membrane at a
relatively high linear velocity. The skilled person can determine a suitable linear
velocity by simple experimentation, but as a general (but by no means rigidly binding)
rule, a typical linear velocity is one of from 1 to 7 M/sec. Thus, the linear velocity
for polymer membranes is typically from 1 to 2 M/sec, and ceramic membranes are
typically operated at 4 to 7 M/sec.
The method hereinabove described allows the recovery of
a dewaxed citrus oil that not only passes the 48 hour test hereinabove mentioned
immediately after production, but also will pass the same test if left for days,
or even months.
The method is now further described with reference to the
following non-limiting examples that describe particular embodiments.
Filtration of ten fold orange oil
The oil is ten-fold orange oil, produced by distillation
from single fold Valencia oil. Folded orange oil is the concentrated product obtained
from the removal of bulk terpenes from peel oil via distillation, which concentrates
the desired flavor compounds.
The membrane element used was tubular and contained a ceramic
(&agr;-alumina) membrane coated onto an &agr;-alumina support. The pore diameter
was 0.2 µm. The element had an inside diameter of 7 mm, outside diameter of
10 mm, and length of 25 cm. This membrane element fit into a housing to form a filtration
module, which in turn was fitted into a Pall T1-70 benchtop crossflow microfiltration
unit (Pall, East Hills, NY, USA). The membrane used was a Membralox® membrane
(ex Pall, East Hills, NY, USA).
The filtration unit comprised a one-gallon jacketed feed/retentate
tank, a circulation pump with a variable frequency drive, the membrane module with
inlet and outlet ports, module inlet and outlet pressure gauges, and a module outlet
temperature gauge. The permeate line was equipped with a permeate valve, which can
shut off the flow. The apparatus was also equipped with a backpulse device, which
periodically directs a pulse of permeate through the membrane opposite the normal
direction of flow, to minimize membrane fouling.
2500 g of ten fold orange oil was added to the feed tank,
then the oil was chilled to -10°C by circulating propylene glycol/water at
-16°C through the tank jacket. The permeate valve was closed, the tank was
pressurized with air, and the circulation pump was started. The air pressure was
adjusted to obtain a module inlet pressure of 0.90 bar and an outlet pressure of
0.70 bar. The permeate valve was then opened to begin collection of wax-free orange
oil. The module outlet temperature was maintained between 5°C and 11°C,
and the filtered oil collected was subjected to a 48-hour chilled wax test, which
This test was performed as follows. A sample of filtered
oil was incubated in a refrigerator at 4°C for 48 hours. The sample was then
visually inspected for the presence of precipitate. The oil was considered to have
passed the test when it remained clear, with no solids present.
During the filtration, the average flux of oil through
the membrane was 8 liters/M2-hour and the resulting yield was 84.5% (wt/wt)
filtered oil per oil feed charged to the system.
Filtration of ten fold orange oil depending on the temperature
The filtration was performed utilising the apparatus as
described in example 1, with some modifications in the method as detailed hereinafter.
In this case, to determine the temperature at which the filtered oil would no longer
pass the 48-hour wax test due to incomplete removal of the wax, the temperature
was allowed to rise above 11°C as the added heat from the pump increased the
temperature of the oil.
2190 g of ten-fold orange oil was charged to the feed tank
and chilled to 0°C using -13°C propylene glycol/water. The inlet pressure
to the membrane module was 0.85 bar and the membrane module outlet pressure was
The module outlet temperature was monitored and collection
of the filtered oil product started when the module outlet temperature became 8.7°C
and was stopped when it reached 12°C. During filtration, an average flux of
9 liters/M2-hour was observed along with a yield of 87% (wt/wt) filtered
oil per oil feed charged to the system.
Filtered oil samples were collected throughout the course
of the filtration at varying module outlet temperatures, and the samples were subjected
to the 48-hour chilled wax test. The results obtained are set out in the following
Module outlet temperature
48 h wax test
As can be seen from the table, all of the filtered samples
collected below 11°C passed the test, in that they were all clear (no precipitate).
The remaining samples fail because precipitate was observed.
It will be understood that the embodiments hereinabove
described are specific embodiments, and are in no way to be construed as being in
any way limiting on the scope of the disclosure. The skilled person will readily
be able to conceive of modifications, including the combination of embodiments or
elements thereof, that lie within the scope of the disclosure.