Note: Descriptions are shown in the official language in which they were submitted.
Ol~Z
This invention is concerned with the decaffeination of
vegetable materials.
There has long been a recognized demand for decaffeina-
ted vegetable materials, particularly beverages such as cof-
fee and tea. The customary prior art techniques for decaffei-
nation generally involve the use of organic solvents such as
trichlorethylene or chloroform, which solvents are contacted
either with the vegetable material or with an aqueous extract
thereof. When sufficient caffeine has been transferred to the
solvent, the resultant solution of caffeine is separated so
as to allow further processing of the decaffeinated material
or extract.
These organic solvent-based decaffeination techniques
have several disadvantages. Of particular concern to the ul-
timate consumer, the utilization of prior art decaffeinationsolvents often results in substantial loss, or denaturization,
of valuable flavour and aroma constituents of the eventual
beverage. Thus, decaffeination has frequently been responsible
for products lacking in high quality characteristics.
Further, because the prior art solvents themselves are
often detrimental, concern has been evidenced respecting con-
tacting them with vegetable materials from which comestibles
are to be produced. This concern has resulted in the develop-
ment of complex and stringent processing techniques in order
to insure complete solvent separation from the finished pro-
ducts.
- 2 -
~Q~019Z
In accordance wlth the present lnventlon, there ls pro-
vided a process for producing a decaffelnated vegetable mate-
rial comprising contacting a liquid, water-immiscible fatty
material with a caffeine-containing vegetable material, main-
taining said fatty material and vegetable material in contactfor a period of time sufficient to transfer caffeine from said
vegetable material into said fatty material, and separating
the resultant, caffeine-laden fatty material from the decaf-
feinated vegetable material.
By "fatty material" as the term is utilized herein, is
meant any of the animal or vegetable fats or oils or admix-
tures or fractions thereof which assume a liquid form within
the temperature range -- discussed hereinafter -- useful for
the removal of caffeine from a caffeine-containing composi-
tion. These fatty materials are customarily composed essen-
tially of esters of fatty acids -- usually glycerol esters --
and may be utilized in either their native form or in those
resultant from conventional treatments as are known in the
art. Moreover, the fatty materials used should desirably be
non-solvants for constituents of the vegetable material other
than caffeine.
Thus, for example, the present fatty materials may be in
either unsaturated or saturated fats or oils. Similarly, unre-
fined or conventionally refined oils as well as oils with or
without such normal additives as anti-oxidants and preserva-
tives are all useful within the scope of the present invention.
10~1019Z
It ls preferred, however, that the fatty materlal be essen-
tlally exempt of surfactants, elther naturally - present or
added. These materlals may stabilize emulsions which form
upon agitation of liquid compositions utilized in accordance
with the present invention and therefore increase the diffi-
culty of such processing steps as centrifugal separation as
may be required.
The fatty materials of the present invention include
commercial oils and fats and thus numerous examples are raa-
dily available. Of these fatty materials, however, thosewhlch are edible are highly preferred because their utiliza-
tion reduces the need for special care in separation from ve-
getable materials which will be further processed as comestib-
les.
These fatty materials are useful for effecting virtually
any desired degree of decaffeination of a vegetable material.
Thus, although essentially complete caffeine removal is ordi-
narily preferred, a lesser degree can also be provided upon
consumer demand. In either event, however, a corresponding
amount of caffeine will become available as a valuable, com-
mercial by-product.
The fatty materials are utilized to extract caffeine
from various vegetable materials, most commonly from coffee
or tea. In order, however, for decaffeination to proceed rea-
dily, it is desirable that there be water present. It is be-
lieved that this requirement is due to the desirability of
~(.~(~192
providing the caffeine in an initially water-solubilized or
partially water-solubilized form, so as to facilltate its
availability to the fatty material decaffeination solvent.
This belief as to a mechanism by which the present invention
may operate is not, however, intended to limit the scope of
the present invention but rather is set forth merely as an
attempt at explanation of what may be occurring incident
thereto.
Caffeine-containing vegetable materials which may be de-
caffeinated in accordance with the present invention are mostsuitably provided in either aqueous liquid or solid form.
Where a solid form is employed, water should still be present,
although it may be bound within the solid. Accordingly, aqueous
solutions of vegetable material and vegetable materials having
a substantial moisture content are preferred compositions for
decaffeination in accordance with the present lnvention.
Aqueous extracts of tea or roast ground coffee are well-
known and may be produced by means conventional in the art.
Because these extracts are themselves eventually converted
into beverage products, however, they should ordinarily be
treated in a manner so as to minimize exposure to conditions
which might result in loss and/or degradation of valuable
flavour constituents. One particular class of constituents
of these brews -- the so-called volatiles or aromatics -- is
particularly sensitive in this regard.
Accordingly, they are preferably removed early during
9Z
processing and recombined with the more stable constituents
only at or near the end of the beverage production cycle.
This removal and preservation of the volatile or aroma-
tic constituents of vegetable materials may be accomplished
by means well-known in the art. Thus, for example, it is con-
ventional to subject aqueous extracts to stripping with steam,
through which technique an aqueous condensate containing the
aromatic and volatile constituents is readily obtained. Such
isolates are then preserved under conditions of low tempera-
ture until such time as they may be recombined with the pro-
cessed vegetable material constituents, for example, by ad-
mixture therewith immediately preparatory to drying or through
application to the dried material itself followed by a short,
secondary drying sequence. Accordingly, where aqueous extracts
are decaffeinated with the fatty materials, the aqueous solu-
tion of vegetable material is pr~ferably free of its customary
volatile constituents.
Another liquid vegetable material which may be decaffei-
nated in accordance with the present invention comprises an
aqueous extract of the vegetable material, which has been
formed specifically for the purpose of decaffeination and
will not constitute any portion of the eventual beverage pro-
duct. This embodiment of the present invention is most sui-
table for vegetable materials such as coffee which normally
require roasting or some other treatment to form many of their
desired beverage constituents.
~0~92
Exemplary of thls embodiment of the present inventlon i8
an aqueous extract of green coffee beans. Such beans may be
extracted with water so as to remove their caffeine content.
The resultant extract, however, contalns relatively few of
the normal coffee beverage constituents inasmuch as these
water-soluble constituents are largely produced only upon
subsequent roasting of the beans.
The formation of this extract is fairly simple. All that
is requlred is that the green beans be contacted with a weight
of water sufficient for dissolving their caffeine content, the
beans normally containing about 2 to 3 weight percent caffelne
depending on thelr origin. Ordinarily, dissolution is accom-
plished by counter-current flow of beans and water; however,
thls step can be effected simply by slurrying the beans in
water or any equivalent contact therebetween for a perlod of
time -- usually 10 to 60 minutes -- sufficient to allow the
desired degree of decaffeination.
Even where green beans are extracted with water, however,
some valuable beverage constituents may also be removed to the
aqueous phase. One technique by which the avoidance of any sub-
stantial loss of these desired constituents has been insured is
through closed, cyclic circulation of the aqueous extraction
medium. Pursuant to this technique, the aqueous medium rapidly
obtains its maximum concentration of the various water-soluble
constituents, including caffeine, of green beans. Upon subse-
quent selective removal of the caffeine from the medium, there
- 7 -
10!~01~2
is obtained a recyclable aqueous extraction medium which ra-
pidly approaches dynamic equilibrium with respect to those
water-soluble constituents of green beans which are not re-
moved by decaffeination of the medium.
With such a dynamic equilibrium in effect, the recycled
caffeine-free extraction medium will -- upon recontact with
green coffee beans -- remove essentially only caffeine there-
from. Thus, within a short time, a system may be obtained
whereby essentially only caffeine is removed from the green
beans.
Both, the recirculating extraction medium and the aqueous
extract discussed more completely above all essentially aqueous
solutions which contain both caffeine and various water-soluble
vegetable material constituents. Accordingly, the present tech-
nique of treating liquid vegetable materials with fatty mate-
rial to remove caffeine therefrom may be applied to each in
much the same manner. Thus, a liquid vegetable material is ad-
mixed with a suitable volume of a liquid water-immiscible fat-
ty material, maintained in admixture therewith until caffeine
has migrated into the fatty material, and then separated with
a corresponding decrease in its caffeine content. These steps
may be accomplished quite simply, because the fact that both
phases are liquid permits easy and thorough admixture under
agitation, while the immiscibility of the two phases -- aqueous
and fatty -- allows substantially complete separation by many
known techniques, including decantation.
l(!~iO~
Of ma~or lmportance for the efflclency of decaffelna-
tlon are the caffeine solubillty characterlstlcs of the fatty
materlal. These propertles are dependent upon the particular
materlal selected and the temperature durlng admlxture wlth
the liquid vegetable material. Thls effect may be discussed
ln terms of the distribution coefficient for caffelne between
equal volumes of the fatty and aqueous phases durlng admlx-
ture and at equilibrium. More particularly, the afflnity of
the fatty material for caffeine is defined by the relation-
shlp:
Distrlbution coefficient = caffeine concentration in fatty phasecaffeine concentration in aqueous phase
for any glven temperature. Clearly then, higher dlstribution co-
efficients evidence a superior ability to effect decaffeination.
In Table I which follows, exemplary data for various fatty
materials at different temperatures are provided. The data re-
flect the equilibrium achieved by single admixtures of volumes
of fatty material and aqueous caffeine solutions. It should be
understood that the fatty materials actually utilized are only
representative commercial products. Thus, depending on the par-
ticular history of a given material, some variation in the dis-
tribution coefficient would be expected. With this data and the
additional description provided herein, however, the characte-
ristics and optimum conditions of use for other fatty materials
within the scope of the present invention may easily be deter-
mined.
_ g _
0192
TABLE I
SOLUBILITY CHARACTERISTICS OF FATTY MATERIALS
Fatty Material Temperature Caffeine Distribution Coeffi-
cient for Equal Volumes of
_ Aqueous and Fattv Phases
Safflower Oil 20 C .064
10 C .067
Soy Bean Oil 20 C .064
10 C .05g
Corn Oil 20 C .064
15 C .064
10 C .071
5 C .085
Peanut Oil 10 C .067
Coffee Oil 23 C .140
Triolein (oleic acid
ester of glycerol) 23 C .085
Olive Oil 23 C .076
Lard 65 C .197
The time of contact between fatty and vegetable phases
is relatively unimportant. Only a few minutes are required
for approaching the equilibrium degree. The optimum tempera-
ture for decaffeination with any particular fatty material
within the scope of the present invention hcwever should be
determined prior to utilization. Exemplary data are reflected
in Table I, but additional determinations may be made by well-
known techniques and thus this aspect of the present invention
may be ascertained by simple experimentation after selection
of the particular fatty material to be utilized.
-- 10 --
1~019Z
In determining the optlmum temperature, the partlcular
materlals lnvolved place llmlts on the degree of deslrable
varlation. Thus, the freezlng polnt of the aqueous vegetable
material and the solidlfication point of the fatty material
define the lower limit for useful temperatures. At the other
end of the range, the degradation to flavour whlch may result
upon exposure of flavour and aroma constltuents to hlgher tem-
perature should also be avolded. Normally, however, decaffel-
natlon can be effected within the range of from 0 to 50 C,
wlth from 10 to 30 C being preferred for aqueous extracts.
Where these constituents are essentlally absent, stlll hlgher
temperatures, up to the lnstability point for the particular
vegetable material remain useful.
Once a particular fatty material and temperature for the
decaffeination step have been selected, there remains the con-
sideration of the desired degree of decaffelnatlon. This is
usually controlled, at least partly through the ratlo of vege-
table to fatty materials. Ordinarily, a ratlo of fatty mate-
rial to aqueous vegetable materlal of about 20 : 1 will achleve
only partial decaffeination in a single contacting sequence.
Thus, for example, at that ratlo, dlstrlbutlon coefficients of
.035 and .085 yield about 40 % and 65 % decaffeinatlon respec-
tlvely. Increases in the ratlo of fatty to vegetable materlal
wlll, of course, lncrease thls degree of deoaffeination just
as lower ratios decrease it.
Additionally, however, the means through which contactlng
-- 11 --
~Ol~Z
wlth fatty materlal ls accompllshed affects the degree of de-
caffelnatlon. Decaffelnatlon can occur as prevlously descrlbed,
through a one-step decaffelnatlon sequence includlng contact-
lng a partlcular welght of fatty materlal wlth a partlcular
welght of aqueous extract, malntalnlng such materlals ln con-
tact for a perlod sufflcient to approach or reach the caffeine
dlstrlbutlon equlllbrlum, and then separatlng the extract. The
efflclency of decaffelnation may however be increased by ln-
creasing the number of steps. Accordlngly, in a preferred em-
bodiment of the present lnventlon, decaffelnatlon i8 performedln a multl-step sequence, whereln the vegetable materlal is
contaoted wlth successlve allquots of the fatty materlal untll
the deslred degree of decaffelnation has been reached.
Thls preferred embodiment may most slmply be followed by
successlvely contactlng the caffelne-contalning composltlon
wlth allquots of a partlcular fatty materlal, malntalnlng the
fatty and aqueous phases ln contact for a perlod of tlme suf-
flclent to effect a substantial transfer of caffeine lnto the
fat~y materlal (such a transfer ordlnarily belng in an amount
greater than about 70 % of the equlllbrium dlstributlon there-
ln), separating the aqueous and fatty phases and then repeat-
lng thls sequence of steps wlth additlonal al$quots of caffelne-
free fatty materlal.
Stlll another form of this preferred embodiment comprlses
counter-current decaffeination. Therein, an initially caffelne-
free fatty materlal is passed through consecutive volumes of
:103~)~9~
vegetable material which are arrayed in reverse order by
caffelne content. Thus, the fresh fatty material flrst con-
tacts the most decaffeinated vegetable material volumes, and
then those of higher caffeine content. During this process
embodiment, when the first vegetable material volume reaches
the desired degree of decaffeination, it is simply by-passed
while a new volume -- having the highest caffeine content --
is connected down-stream, to maintain a constant number of vo-
lumes on stream and a proper order of contact with the fatty
material.
It is additionally noted that aqueous solutions or ex-
tracts of vegetable material may contain, for example, from
2 to 60 % soluble solids by total weight. Ordinarily, however,
it is preferred that the solutions to be decaffeinated have a
soluble solids concentration of from 10 to 50 %, preparatory
to contacting with the fatty material caffeine solvent. Such
concentrations are preferred for the purpose of minimizing the
subsequent volumes of liquids which may be utilized in the de-
caffeination sequence, as well as for the purpose of reducing
the quantity of water which must eventually be removed from a
brew or extract which is to be dried to solid form.
Concentration may be obtained through techniques well
known to the prior art. Again, because the aqueous extract
will eventually provide the dried beverage product, it is de-
sirable that such concentration be obtained under conditionswhich will minimize the possibility of adverse effect on fla-
0192
vour. Accordingly, lt ls preferred that technlques such aslow pressure evaporatlon or freeze concentratlon, which
avold exposlng the brew to higher temperature for any sub-
stantial period of tlme, be utilized.
The present invention also includes utilization of a
fatty material for direct decaffeination of solid vegetable
materials. Exemplary of such solids are green coffee beans
which may be provlded in ground, crushed and, most desirably,
whole form. Roast coffee beans may also be utilized; however
volatiles should first be removed to avoid undue loss of va-
luable beverage constituents. Consequently, the decaffeination
of whole green beans is most preferred and the following dis-
cussion of this embodiment is particularly directed thereto,
although other vegetable materials may be treated in similar
manner.
Use of solid vegetable materials is a particularly pre-
ferred embodiment of the present invention inasmuch as it ob-
viates certain disadvantages of dealing with aqueous solutions
of vegetable material. Thus, the separation of the fatty mate-
rial from the caffeine-containing composition is facilitated
by the fact that, while the fatty material remains liquid, the
vegetable material is in solid form. Separation of the beans
and fatty material may even conveniently be effected by simple
drainage of the beans and, where centrifugal force is utilized
to facilitate this separation, fairly simple machinery is ade-
quate.
- 14 -
~0~0192
Further, in the separatlon of green coffee beans from
a liquid fatty material, the degree of separation may be es-
sentially 100 %. Whereas even the most sophisticated of immis-
cible liquid separations may result in some retainment of one
liquid in the other, no such problem is encountered here. Once
most of the fatty material has been separated from the beans,
a secondary physlcal purification step, such as dlrection of
a burst of steam through the beans, permlts essentially 100 %
separation of retained fatty material. Moreover, even this
additional step may be rendered unnecessary. Because of the
post-decaffeination roasting and extractive processing of the
beans, where a substantially flavourless fatty material is uti-
lized, little if any effect upon the flavour of the eventual
beverage product will result even if separation is incomplete.
In order to be in a form readily susceptible to caffeine
removal, green coffee beans should contain some moisture. The
beans ordinarily naturally contain from 8 to 10 % moisture,
although higher amounts, of at least about 20 ~ by total weight,
are preferred. The upper limit of moisture content, however, is
more variable. Decaffeination of a caffeine-containing composi-
tion comprising green beans is desirably performed in the ab-
sence of free liquid water, so as to avoid the separation pro-
blems incident to an admixture of liquids and the possible loss
of valuable vegetable material constituents solubilized in that
water. Accordingly, in this embodiment of the present invention,
it is preferred that the green beans contain between about 20
- 15 -
,
0192
and 60 %, most preferably between about 40 and 60 %, water
by total weight.
The incorporatlon of water into green beans is easily
accomplished. Thus, for example, the beans may simply be im-
mersed in water and there maintained for several hours, untilthey have absorbed the desired amount of moisture. Thereafter,
they may easily be separated from the excess water, for exam-
ple, by centrifugation. Through the use of heat and/or pres-
sure, this incorporation or "swelling" of the beans may be
accelerated. Thus, for example, where beans are immersed in
water at a temperature of about 80 to 90 C, they achieve the
desired moisture content much more quickly. Also, upon being
subjected to steam at about 2 atmospheres, even less tlme --
normally about 1 to 30 minutes -- is sufficient to reach the
desired moisture content.
Once the beans have been swollen to an appropriate mois-
ture content, they are placed in a bath or stream of fatty ma-
terial until the desired degree of decaffeination has been
achieved. Here, the time of admixture becomes significant :
transfer of caffeine from the beans is much slower than from
dilute solution. Accordingly, it is desirable that periods of
at least 30 minutes, with longer periods for more complete de-
grees of decaffeination, be permitted for this step.
In processes leading to high degrees of decaffeination,
it is also important to ensure that the moisture content of
the swollen beans does not significantly decrease during treat-
; - 16 -
lQ~:)19Z
ment. Contact between swollen beans and fatty material can
result in lower moisture contents due to loss of water from
the beans to the fatty material. Where this decrease brings
the beans to a moisture content below the preferred 40 to
60 % range, there occurs a corresponding decrease in the ef-
ficlency of decaffeination.
It is therefore preferred that the fatty material uti-
lized for decaffeination contain a small amount of water.
From 0.9 to 1.2 %, most desirably about 1.0 %, of water by
weight of fatty material is ordinarily utilized in this pre-
ferred embodiment of this invention. This amount acts to
maintain in equilibrium the amount of water present in the
beans and the fatty material, so that during contacting there
is no net migration of water from the beans to the fatty phase,
nor vice-versa. On the other hand, it prevents excessive remo-
val of water from the beans, and on the other it minimizes the
total amount of water present during decaffeination, thereby
avoiding undue loss of non-caffeine, water-soluble bean con-
stituents.
Again, the use of multi-step as opposed to single-step
contact of fatty material with the vegetable material may be
practiced in the manner previously described. Thus co-current
and, more preferably, counter-current extraction with fatty
material are desirable embodiments of the present invention.
One aspect by which the decaffeination of swollen green
beans differs substantially from decaffeination of aqueous
- 17 -
lQ~3019Z
caffelne-contalnlng solutlons lies ln the effect of tempera-
ture upon the efficlency of decaffelnatlon. As previously no-
ted, where decaffeination involves the contacting of fatty ma-
terial with an aqueous caffeine-containing solution, differen-
ces in temperature during such contact~ng do not have a largeeffect upon the efficlency of decaffelnatlon. In treating so-
lid beans, however, increases in the prevaillng temperature
for decaffelnation may dramatically increase the rate of caf-
feine removal.
Accordingly, to achieve maximum efficiency of caffeine
removal, decaffeination of beans is preferably carried out at
as high a temperature as is practicable. Degradation of fatty
materials customarily occurs at or around a temperature of
about 150 C, and therefore this temperature usually repre-
sents a practical maximum limit for decaffeination. Also, pro-
longed contact times at very high temperatures may produce some
degradation of flavour constituents. Accordingly, it is prefer-
red that the decaffeination of green beans be performed within
the temperature range of from about 50 to 120 C.
A further aspect of the present invention involves rege-
nerating caffeine-containing fatty material so as to permit
reuse thereof in the decaffeination process. This is most ef-
ficiently achieved by contacting the separated caffeine-con-
talning fatty material with water, permitting the transfer of
the caffeine into aqueous phase, and then separating the fatty
material to permit its recycle for further decaffeination. In
- 18 -
9Z
large measure, the regeneration sequence reverses the steps
already described above with respect to decaffeinatlon. In
addition, however, it readily permits isolation of the caf-
feine from the regenerate aqueous phase.
For regeneration of the fatty material, the efficiency
of caffelne removal to aqueous phase is again governed by the
same parameters of temperature, the caffeine distribution co-
efficient of the particular fatty material, and the weight
ratio of fatty material to water as discussed above with res-
pect to the separation of caffeine from an aqueous solution
of vegetable material.
Because flavour degradation is not a serious problem du-
ring regeneration, as constituents of the final product are
not present, the temperature during regeneration may be raised
to improve the efficiency of caffeine transfer to the aqueous
phase. Thus where, with increasing temperature, the solubility
of caffeine in water increases more rapidly than in the fatty
material, it is advantageous to effect regeneration at a tem-
perature up to 100 C (and even higher where pressure is uti-
lized to avoid evaporation). If on the other hand, lower tem-
peratures favour this transfer then they should be employed.
Also, because it is here desired to transfer caffeine
from the fatty material to an aqueous phase, the relatively
greater solubility of caffeine in water permits the utiliza-
tion of low fatty to aqueous phase ratios, even for substan-
-- 19 --
Z
tially complete transfer. Additionally, where lt is desired
further to minlmize the amounts of water utlllzed in regene-
ration of the fatty material, a multi-step regeneratlon se-
quence comprising co-current or counter-current extraction of
fatty material with water may be utilized in the same manner
as has already been described hereinabove so as to facilitate
the efficient regeneration of the fatty material.
Incident to regeneration of the fatty material through
removal of caffeine with water, it has been discovered that
the separated fatty material can contain -- even when separa-
tion of these immisclble liquids is accomplished through such
normally efficient means as centrifugal separation -- about
1 % by weight of water. This water is desirably removed from
the fatty material before recontacting with vegetable mate-
rial so as to avoid dilution of liquid vegetable materials.Removal of the entrained water in the fatty material can be
accomplished by such means as a flash distillation under va-
cuum conditions or equivalent known techniques.
As previously described, certain preferred embodiments
of this invention relating to decaffeination of solid vege-
table materials rely upon utilizing a fatty material contai-
ning a small amount of water. In the practice of these embo-
diments, the dilution factor becomes negligible. Therefore,
the regenerated fatty material containing entrained water may
be used directly for further decaffeination. Alternatively,
its aqueous content may be adjusted -- for example, by adding
- 20 -
lO9(~19Z
water or by partlal strlpplng -- as requlred to brlng lt to
an optlmum water content.
Even where the presence of water ln the fatty material
is desired, however it ls preferred to remove the entrained
water and then add the desired amount of water to dry fatty
material. This sequence of steps ensures an optlmum water
content and avolds the difficulties and/or interruptlons re-
quired for monitoring the entrained water content of regene-
rated fatty materlal and then adjusting lt, as desired.
The followlng examples are lllustrative of the present
inventlon. Unless otherwise noted, the percentages are on a
weight basis.
EXAMPLE 1
An aqueous extract of roast ground coffee beans is
stripped with steam to remove volatiles. 10 kilograms of the
stripped extract, at a soluble solids concentration of 19 %
and a temperature of 22 C, are then added to 179 kilograms
of corn oil at 60 C. The resultant admixture is agitated
for 30 minutes and then passed through a centrifugal separa-
tor at a rate of 3.16 kilograms per minute. The brew, whichis separated from the oil in the centrifuge, has a 51 ~ de-
gree of decaffeination.
By repeating this treatment of the brew with additional
corn oll, the degree of decaffeination is successively in-
creased until essentially complete caffelne removal is effec-
ted.
- 21 -
105~01~2
EXAMPLE 2
A tea extract having a soluble solids concentration of
27.6 % and a temperature of 22 C is mixed with corn oil in
a volume ratio of 1 : 20, respectively, and maintained under
agitation for 10 minutes. The admixture is then subjected to
centrifugal separation to yield a tea extract exhibiting 63 %
decaffeination. Again retreatment affords a means for achiev-
ing any greater desired degree of decaffeination.
ExAMæLE 3
Green coffee beans are decaffeinated utilizing a recir-
culating aqueous medium which has reached an equilibrium so-
luble solids concentration of 29 % and is at a temperature of
22 C. Decaffeination is performed by passing the aqueous me-
dium in counter-current through a column of green beans, with
essentially caffeine-free beans being removed from the column
at one end and natural green beans added at the other. Within
the circulating aqueous system, and at a point removed from
the column, aqueous medium is diverted through a centrifugal
extractor to which coffee oil at 50 C, in a ratio of 21 : 1
to the medium, is added for removal of caffeine from the aqueous
extract. Heat exchangers in the system are used to maintain the
temperatures of these two liquids essentially as indicated.
With a single pass of the medium and oil through the cen-
trifuge, over one-third of the caffeine in the medium is re-
moved. A second pass of the medium and a further equal aliquot
- 22 -
10~
of oil increases the decaffeinatlon of the medlum to over 60 %.
After roastlng, grlndlng and extracting separate samples of
decaffelnated beans obtalned after the slngle and second pas-
ses, lt ls found that both sample aqueous extracts are essen-
tlally caffelne-free.
EXAMPLE 4
Green coffee beans are subjected to steam at 110 C untll
they reach a molsture content of 45 ~ by total weight, and 10
kg aliquots are placed lnto separate extraction chambers. Each
of the allquots is decaffeinated for four hours at 95 C wlth
corn oil whlch ls passed through the chambers at a rate of
1.1 kg/mlnute.
In one Trial, "A", oil is passed through only one chamber.
Thereafter it is regenerated by extraction wlth water to remove
lts caffeine content, the water is removed, and the oil is then
recirculated to maintain a continuous flow of caffeine-free corn
oll to swollen beans in the chamber.
In a second Trial, "B", four chambers are connected in se-
ries so that the corn oil flows through each. Regeneration and
recirculation of the oil is accomplished only after it has pas-
sed through all four chambers. One chamber - the first contac-
ted by oil - is removed each hour and a new chamber is added
at the downstream end. In this manner, and after a start-up pe-
riod of 6 hours a system is achieved wherein the four chambers
contaln beans of varying caffeine content by virtue of the fact
- 23 -
)l9Z
that they have been subjected to different duratlons of on-
stream oil decaffeination.
The caffeine contents of beans from Trial "A" and from
a chamber which has passed through all four stages of Trial
"B" after start-up were analyzed. Despite the fact that each
of these beans had been decaffeinated under essentially the
same physical conditions, their respective degrees of decaf-
feination are markedly different. Thus while the beans of
Trial "A" exhibit 52 % decaffeination, those of Trial "B" ex-
hibit 92 %. It is therefore evident that multi-step extraction
substantially increases the efficiency of decaffeination.
EXAMPLE 5
An aqueous extract of roast ground coffee beans is strip-
ped with steam to remove volatiles. 200 grams of the stripped
brew having a soluble solids concentration of 18.4 % are then
admixed with 2 kg of safflower oil and agitated for 30 minutes
at 20 C. This admixture was then centrifuged to break the
emulsion and the brew separated by decantation. The separated
brew exhibits a 56 % degree of decaffeination.
EXAMPLE 6
The process of Example 5 is repeated with the change
that 2 kg of soy bean oil are substituted for the safflower
oil. After separation, the brew exhibits a 56 % degree of de-
caffeination.
- 24 -
()192
EXAMPLE 7
The process of Example 5 is repeated substltuting 2 kg
of peanut oil for the safflower oil. Additionally, the brew
and oil are maintained at 10 C - instead of 20 C - through-
out the process. The separated brew exhibits a 56 % degree ofdecaffeination.
EXAMPLE 8
Green coffee beans are decaffeinated with coffee oil in
a four-chamber countercurrent extraction zone ln the manner
described for Trial "B" of Example 4. Each chamber, or cell,
contains 6.8 kg of beans by dry weight. The oil is maintained
at 105 C and extraction is performed over a total extraction
period of 6 hours (1.5 hours for each cycle). A total oil to
bean weight ratio of 120 : 1 is utilized.
After each pass of the recirculating oil through all
four chambers of the extraction zone, the oil regenerated by
aqueous extraction for caffeine and then is stripped to remove
its aqueous content. A predetermined measure of water is ad-
mixed with the oil preparatory to its recirculation.
, ~ ~
Utilizing the foregoing procedure, five different Trials
are made. These Trials differ essentially in the addition of
varying amounts of water to stripped, caffeine-free fatty ma-
terial. Data for steady-state conditions of operation of each
Trial are as follows:
- 25 -
lO9015~Z
Tri~l Water Content Water Content Peroent Non- TyFe of
No of Bean Charge maintained in Decaffein- Caffeine Coffee
Oilation Solids Loss Beans
1 54 % 0.3787 % 2.5 % "Milds" (1.33%
2 55 % 1.0097 % 1.1 %
3 55 % 1.0097 % 1.1 ~ ll
¦ 4 ¦ 5 ~ ¦ 0.75¦ 69 ~ ¦ 2.4 ~ ¦~Rnb ~ta
57 % 1.2097 % 5.2 % "
This data indicates that proper moisture contents permit
optimum decaffeination with minimal loss of non-caffeine bean
constituents. The efficiency of decaffeination decreases where
the concentration of water present in the fatty material during
decaffeination decreases. It is also shown that, although the
efficiency of decaffeination remains high at higher aqueous con-
tents for the fatty material, the high level of water present in
the system results in an increase loss of non-caffeine solubles
from the beans.
- 26 -
