U.S. patent application number 12/969606 was filed with the patent office on 2011-05-26 for pure rebaudioside a.
This patent application is currently assigned to SGF Holdings, LLC. Invention is credited to Robert Gordon Chase, Gordon James Francis, Mel Clinton Jackson.
Application Number | 20110124587 12/969606 |
Document ID | / |
Family ID | 36181079 |
Filed Date | 2011-05-26 |
United States Patent
Application |
20110124587 |
Kind Code |
A1 |
Jackson; Mel Clinton ; et
al. |
May 26, 2011 |
Pure Rebaudioside A
Abstract
The invention provides a high throughput, high purity, high
yield system and method of isolating and purifying rebaudioside A
("Reb A"), with acceptable water solubility for all commercial
uses, from commercially available Stevia rebaudiana starting
material. The invention also provides a means of maximizing yields
of 99+% purity Reb A based on the attributes of a given batch of
Stevia starting material. The purity of the Reb A final product, up
to 99+% purity, can be selected based on a starting material
assay.
Inventors: |
Jackson; Mel Clinton;
(Honolulu, HI) ; Francis; Gordon James; (Issaquah,
WA) ; Chase; Robert Gordon; (Motueka, NZ) |
Assignee: |
SGF Holdings, LLC
Bellingham
WA
|
Family ID: |
36181079 |
Appl. No.: |
12/969606 |
Filed: |
December 16, 2010 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11252430 |
Oct 17, 2005 |
7923552 |
|
|
12969606 |
|
|
|
|
60620280 |
Oct 18, 2004 |
|
|
|
60678653 |
May 6, 2005 |
|
|
|
Current U.S.
Class: |
514/25 ; 426/549;
426/590; 426/614; 426/615; 426/629; 426/641; 426/658; 514/777;
536/18.1 |
Current CPC
Class: |
A23L 27/36 20160801;
C07H 15/252 20130101; A61P 3/04 20180101; C07H 1/00 20130101; A61P
3/10 20180101 |
Class at
Publication: |
514/25 ;
536/18.1; 426/658; 426/590; 426/549; 426/615; 426/629; 426/614;
426/641; 514/777 |
International
Class: |
A61K 31/70 20060101
A61K031/70; C07H 15/24 20060101 C07H015/24; A23G 3/00 20060101
A23G003/00; A23L 2/00 20060101 A23L002/00; A21D 13/00 20060101
A21D013/00; A23L 1/212 20060101 A23L001/212; A23L 1/36 20060101
A23L001/36; A23L 1/32 20060101 A23L001/32; A23L 1/31 20060101
A23L001/31; A61P 3/04 20060101 A61P003/04; A61P 3/10 20060101
A61P003/10; A61K 47/00 20060101 A61K047/00 |
Claims
1. Pure rebaudioside A.
2. The rebaudioside A of claim 1 used in beverages, broths, and
beverage preparations selected from the group comprising
carbonated, non-carbonated, frozen, semi-frozen, non-frozen,
ready-to-drink, concentrated, dairy, non-dairy, herbal, non-herbal,
caffeinated, non-caffeinated, alcoholic, non-alcoholic, flavored,
non-flavored, vegetable-based, fruit-based, root/tuber/corm-based,
nut-based, other plant-based, cola-based, chocolate-based,
meat-based, seafood-based, other animal-based, algae-based, calorie
enhanced, calorie-reduced, and calorie-free products, optionally
dispensed in open containers, cans, bottles or other packaging.
3. The rebaudioside A of claim 1 used in foods and food
preparations selected from the group comprising dairy-based,
cereal-based, baked, vegetable-based, fruit-based,
root/tuber/corm-based, nut-based, other plant-based, egg-based,
meat-based, seafood-based, other animal-based, algae-based,
processed, preserved, and synthesized products.
4. The rebaudioside A of claim 1 used in candies, confections,
desserts, and snacks selected from the group comprising
dairy-based, cereal-based, baked, vegetable-based, fruit-based,
root/tuber/corm-based, nut-based, gum-based, other plant-based,
egg-based, meat-based, seafood-based, other animal-based,
algae-based, processed, preserved, and synthesized products.
5. The rebaudioside A of claim 1 used in prescription and
over-the-counter pharmaceuticals, assays, diagnostic kits, and
therapies selected from the group comprising weight control,
nutritional supplement, vitamins, infant diet, diabetic diet,
athlete diet, geriatric diet, low carbohydrate diet, low fat diet,
low protein diet, high carbohydrate diet, high fat diet, high
protein diet, low calorie diet, non-caloric diet, oral hygiene
products, personal care products, professional dentistry products,
medical, veterinarian, and surgical products, and pharmaceutical
compounding fillers, syrups, capsules, gels, and coating
products.
6. The rebaudioside A of claim 1 used in consumer goods packaging
materials and containers selected from the group comprising plastic
film, thermoset and thermoplastic resin, gum, foil, paper, bottle,
box, ink, paint, adhesive, and packaging coating products.
7. The rebaudioside A of claim 1 used in goods selected from the
group comprising sweeteners, co-sweeteners, coated sweetener
sticks, frozen confection sticks, medicine spoons (human and
veterinary uses), dental instruments, pre-sweetened disposable
tableware and utensils, sachets, edible sachets, pot pourris,
edible pot pourris, hotch potches, edible hotch potches, artificial
flowers, edible artificial flowers, clothing, edible clothing,
massage oils, and edible massage oils.
Description
RELATED APPLICATIONS
[0001] This application is a divisional application of Ser. No.
11/252,430 entitled "High Yield Method of Producing Rebaudioside
A"; this divisional application claims the benefit of, and fully
incorporates by reference, that application Ser. No. 11/252,430
under 35 USC 120 and 121. The parent application, Ser. No.
11/242,430, claimed the benefit of two provisional patent
applications, Ser. No. 60/620,280, filed on Oct. 18, 2004, and Ser.
No. 60/678,653, filed on May 6, 2005, in the U.S. Patent and
Trademark Office for an invention entitled "High Yield Method of
Producing Rebaudioside A".
BACKGROUND OF THE INVENTION
[0002] 1. Technical Field
[0003] The invention is in the technical field of the purification
of rebaudioside A (also known herein as "Reb A" and "RA") from
crude extracts from the plant Stevia rebaudiana, L. Bertoni
("Stevia"). Reb A is a diterpene glycoside found, along with many
other glycosides, sterebins, and other compounds, in Stevia
extracts. Reb A is the sweetest tasting (roughly 250 to 450 times
the sweetness of sucrose at sweetener concentrations used in
comestibles) glycoside in Stevia and has enormous commercial
potential as a non-caloric sweetener. Generally speaking, the
Stevia glycosides, except rebaudioside A, have an undesirable
aftertaste (some have a bitter aftertaste), which generally limits
the use of such non-Reb A glycosides as sweeteners. Moreover,
sterebins (a group of diterpene compounds commonly known as "yellow
oil") in Stevia extracts have an extremely bitter taste, even in
minute concentrations. In short, the technical field, and the
technical problem, is to isolate, especially at industrial scale,
99+% purity Reb A from crude Stevia extract; such purity eliminates
the bitterness or off-taste that characterizes all currently
available Stevia sweeteners.
[0004] 2. Related Art
[0005] The six rebaudiosides (A to F), stevioside (the predominant
glycoside in extracts from wild type Stevia), dulcosides, and
sterebins are difficult to separate in solution. In particular, the
six rebaudiosides share the same diterpene skeleton and differ only
in glycoside moieties; all but Reb A has some degree of bitter
aftertaste. Scores of patents and journal articles describe efforts
to produce pure (i.e., 99+% purity) rebaudioside A, but to date no
method of producing pure Reb A in the lab, much less in commercial
quantities, has been reported. Many of the reported methods of
production require the use of ion exchange columns or gases, the
economics of which undermine scale-up to the commercial quantities
of sweetener required for the beverage, packaged food, baking,
candy, personal care product, pharmaceutical, and other
industries.
[0006] Most methods for extraction and purification of diterpene
glycosides from Stevia use complicated processing of crude extracts
and require high concentrations of rebaudioside A in starting
materials, yet have low yields and low purities (<95%) of
rebaudioside A. Some methods use alkanols and/or hydrous alkanols
as solvents. For instance, in paragraphs 49 to 53 of U.S. Published
Patent Application No. 2003/0138538 A1 of Kitazume, et al., a
method is described whereby plants or dried leaves from Stevia
rebaudiana are processed using methanol, hydrous methanol, or
hydrous ethanol (but not using a reflux process) and column
purification to produce an extract containing a minimum of 40%
rebaudioside A by weight and with a rebaudioside A to stevioside
ratio is 1.5:1. In Kitazume's method, if the starting plant
material is not 40% rebaudioside A by weight and with a
rebaudioside A to stevioside ratio of 1.5:1 ("Kitazume Threshold
Purity"), column purification or recrystallization must be used to
produce rebaudioside A of sufficient purity needed to support the
object of the Kitazume patent application (the substitution on
rebaudioside A of .beta.-1,4-galactosyl using a
.beta.-1,4-galactosyl transferase enzyme). The preferred starting
material in Kitazume is a Stevia extract with at least 70% Reb A
content. In the Kitazume method, when recrystallization is used to
produce rebaudioside A of Kitazume Threshold Purity for the
transferase reaction, the Stevia extract is dissolved to saturation
in a hydrophilic organic solvent, such as methanol or ethanol, with
no water in the solvent, and the solution concentrated or cooled,
then filtered to collect the precipitate, a higher purity
rebaudioside A; process parameters, e.g., heating and cooling
rates, filtration specifications, etc., of the purification of Reb
A are not disclosed. Kitazume also does not disclose the purity of
the Reb A produced by recrystallization, only that
recrystallization can be used when the starting material is below
Kitazume Threshold Purity. Importantly, in Kitazume, all Reb A
material is produced using cation-exchange resin and anion-exchange
resin steps (paragraph 50) following any crystallization or
recrystallization steps. A Reb A purity higher than 91.3% (Example
3) is not disclosed, nor is the solubility of Kitazume's Reb A
disclosed. In fact, Kitazume does not disclose whether the 91.3%
purity Reb A used in the Examples was produced using
recrystallization or only by column purification. Reb A with 91.3%
purity may be acceptable for the transferase reaction, but it is
not acceptable for commercial uses in foods and beverages.
[0007] U.S. Pat. No. 5,962,678 to Payzant et al., describes a
seven-step extraction and purification of Reb A from Stevia plant
material, including processing through two ion exchange columns and
precipitation out of a methanol solution. For purification of Reb
A, after removing mixed sweet glycosides from the second ion
exchange column with methanol, the collected eluent is dried. Upon
refluxing the dried solids in a methanol solution and then cooling
the solution, stevioside precipitates and is captured by
filtration. The filtrate is then concentrated (presumably by
evaporating the solvent) and cooled and rebaudioside A
precipitates. The method in Example 5 of Payzant is described as
yielding rebaudioside A with purity up to 98.6%. However, using 80%
purity Reb A Stevia extract as starting material, repeated efforts
by the inventors of the present invention to replicate Payzant's
purification of Reb A failed; typical results were a Reb A yield of
less than 5% by weight and amorphous crystals with a noticeable,
bitter aftertaste. Attempts by a third party to replicate Payzant
also failed (unpublished data). In attempts to replicate Example 5,
precipitation of Reb A did not begin until more than 40 hours after
heat was removed from the methanol solution. The Payzant
purification method in Example 5 apparently requires Reb A starting
material obtained by following the steps in Payzant claim 1 (the
seven-step process using methanol) rather than by procuring
commercially available Stevia extracts. The Payzant method,
therefore, is neither generally applicable to Stevia extracts nor
scaleable to commercial production of Reb A.
[0008] In addition to not being replicable, Payzant teaches away
from using hydrous solvents. Payzant uses anhydrous methanol as a
reflux solvent to purify stevioside. In Example 2, Payzant states,
"Since the sweet glycosides were absorbed on to the XAD-7 resin
from water, there is water contained within or on the surface of
this resin and this water is removed with the methanol. The only
reason for going to dryness is to remove this water. If the minor
amount of water wasn't there or if it were removed by the some
other means, then the methanol could be removed by evaporation to
some level and then the sweet glycosides would crystallize from the
solution." In Example 3, Payzant states, "By being anhydrous or as
free from water as possible the crystalline yield is maximized and
the crystallization time is minimized. For example, 0.5% water in
the methanol in this initial crystallization will significantly
reduce the yield. The Stevioside can be further purified, if
desired, by using the purification technique set forth in Example 5
below." In Example 5, Payzant discloses a purification in which a
Stevia extract containing 90.2% Reb A (in Example 3, however, the
final product had 7.9% Reb A content) is refluxed in a solvent of
86.67% methanol and 13.33% water by mass; this procedure is said to
yield 98.6% Reb A; heating rate, cooling rate, stirring, and
temperatures are not disclosed. Payzant notes that the extract used
in Example 5 is "not from Example 4", but elsewhere discloses
achieving only 7.9% purity Reb A. Payzant states in Example 5 that
including water in the solvent is "a convenience in the
laboratory", but "on a commercial scale anhydrous solvent might be
more convenient". In Examples 3 and 5, Payzant expressly teaches
away from using hydrous alkanol solvents for purification of Stevia
glycosides. Payzant does not disclose how the 90.2% purity Reb A
used in Example 5 was obtained, i.e., the method of improving Reb A
purity from 7.9% to 90.2% is not enabled, and therefore Example 5
is not enabled. Payzant also does not disclose the method used to
determine the Reb A purities claimed. Under the Reb A reference
standards currently available from Chromadex (www.chromadex.com) or
Wako (www.wako-chem.cojp), Payzant's claimed purities are thought
to be overstated. As noted above, Payzant's Reb A purification
method failed when existing commercial quality Stevia extracts were
used as starting material. The Payzant method of purifying Reb A
requires first precipitating out stevioside in a methanol reflux,
and Payzant should be so limited as a reference. Putting aside the
omission of how 90.2% Reb A purity was obtained, the lack of a
reference standard, and the instructions to use anhydrous solvents,
the process complexity, low throughput, and low yield in the
Payzant method mean that it cannot be used as an industrial
process. Moreover, the use of a methanol reflux in isolating Reb A,
with concomitant risk of methylated contaminants in the final
product, may present a barrier to obtaining "generally recognized
as safe" status with government regulators of food products.
[0009] The existing art of producing higher purity rebaudioside A
uses elaborate sequences of ion exchange columns and other
low-throughput procedures. The existing methods used by Kitazume,
Payzant, and others do not produce 99+% purity Reb A, whether by
crystallization, recrystallization, column purification, or
combinations thereof. There are heretofore unsolved problems that
have caused unacceptably high levels of contamination in Reb A
final products. The existing art in the industrial production of
Stevia glycosides, including rebaudioside A, typically starts with
Stevia plant material and produces a Stevia extract that has 40% to
90% purity Reb A mixed with various Stevia glycosides and "yellow
oil"; such extracts are commonly called "Stevia extract powder" or
"Stevia extract", and are herein called "Stevia starting material".
Stevia starting material is abbreviated "SSM" in the Drawings. The
lack of an industrial method of producing 99+% purity rebaudioside
A has meant that Stevia sweeteners are today typically used (e.g.,
in pickling) where other flavourants mask the bitter or astringent
aftertastes arising from residual contaminants in the Stevia
sweeteners. Even 1% contamination by other Stevia compounds
produces a noticeable, bitter aftertaste in Reb A sweeteners, which
led Dobberstein, in U.S. Pat. No. 4,612,942, to limit use of Stevia
glycosides in orally consumable compositions to a level (the
"sweetness level threshold") that modified the taste of the
compositions, but could not be perceived as sweet. The term "orally
consumable composition", as used herein, means substances that are
contacted with the mouth of a human or animal, including substances
which are taken into and subsequently ejected from the mouth and
substances which are drunk, eaten, swallowed or otherwise ingested.
There is a need for non-caloric materials that can sweeten, modify,
or enhance the flavor of orally consumable compositions such as
foodstuffs, smoking compositions, chewing compositions, oral
hygiene compositions, and medicinal compositions to improve or vary
the sensory perceptions thereof.
[0010] In view of the enormous market for a non-caloric, natural
sweetener for use in foodstuffs, beverages, medicines, tobacco
products, candies, etc., a method of obtaining 99+% purity
rebaudioside A has long been sought. 99+% purity Reb A is useful as
a reagent, as an ingredient in foods and beverages, as a standalone
sweetener, and as a co-sweetener. There is an unmet demand for (1)
a method of producing 100% pure, water soluble, rebaudioside A in
the lab and (2) a method of producing industrial quantities of 99+%
purity, water soluble, rebaudioside A suitable for use as a
reagent, as an ingredient in foods and beverages, as a standalone
sweetener, and as a co-sweetener. It will be shown that, counter to
Payzant's teaching, having a significant water content in the
reflux solvent is essential to producing 99+% purity Reb A and to
avoiding the contamination reported by all other researchers.
SUMMARY OF THE INVENTION
[0011] The primary technical problem to be solved, and the primary
object of the invention, is to provide a high throughput, high
purity, high yield system and method of isolating and purifying Reb
A, with acceptable water solubility for all commercial uses, using
commercially available Stevia starting material. A second technical
problem to be solved by the invention is to maximize yields of 99+%
purity Reb A based on the attributes of a given batch of Stevia
starting material. Solutions to the technical problems must be
usable in laboratory-scale volumes as well as in industrial
production volumes. Industrial production involves processing
hundreds or thousands of kilograms of Stevia starting material per
batch. Quantitative proof of the inventive step of the invention
disclosed herein is that the Reb A final product of the invention
has a higher purity than the current primary analytical standard
for the highest available purity rebaudioside A (i.e., Chromadex
reference standard, Catalog No. ASB-00018226, www.chromadex.com);
the Chromadex standard has a certified purity of 98.7%. Reb A of
purity greater than the cited Chromadex reference standard is a
signature that the invention disclosed herein was likely employed,
since no other disclosed method obtains such purities. The higher
purity of the Reb A produced by the invention arises from use of
the "starting material assay" and the "selected EtOH solvent", as
described below. The unmodified word "ethanol" means ethanol
denatured with 1% ethyl acetate. The phrase "absolute ethanol"
means 100%, non-denatured, anhydrous ethanol. "EtOH" means a
solvent consisting of ethanol and water; the percentages of ethanol
and water, respectively, are denoted as "xxEt/xxW". For instance, a
solvent consisting of 92% ethanol and 8% water would be denoted as
"92Et/08W" solvent; a mixture of solids in 92% ethanol and 8% water
would be denoted as a "92Et/08W mixture". "Reb A" and "RA" both
mean rebaudioside A. The term "RAxx" means a material with a
rebaudioside A content of approximately xx percent. For instance,
RA80 means a material with a rebaudioside A content of
approximately 80%. The term "RAxx.yy" means a material with a
rebaudioside A content of xx.yy percent. For instance, RA81.58
means a material with a Rebaudioside A content of 81.58%.
[0012] In a laboratory embodiment of the invention, Stevia starting
material is mixed with a solvent comprising, by volume, between 4%
to 15% water and the remaining percentage ethanol, refluxed, and
then filtered after a stirred wash. "Water" means potable water,
and preferably filtered water compliant with applicable
governmental standards for dietary supplement manufacturing.
"Solids" means, depending on context, either intermediate product,
i.e., retentate obtained by filtering a reflux or wash mixture, or
final product. The phrase "100% pure Reb A" is defined in the HPLC
Method section of the Examples. Reb A purities of starting
materials and final products described herein are determined using
the ChromaDex reference standard for Reb A, as described in the
HPLC Method section of the Examples. "Producer" means the entity
processing Stevia starting material to produce purified Reb A.
"Higher quality Stevia starting material" means Stevia starting
material with approximately 85% or higher Reb A content. "Lower
quality Stevia starting material" means Stevia starting material
with less than approximately 85% Reb A content. One or more reflux
stages and stirred wash stages, as defined below, are used to
obtain 99+% purity Reb A final product. Using RA90 and higher Reb A
purity Stevia starting materials, the combination of a single
reflux stage and a single stirred wash stage typically produces
99+% purity Reb A. Final product Reb A purities, and Reb A yields,
for a given Stevia starting material may be targeted and obtained,
while minimizing processing costs, by use of the starting material
assay with optional mass to solvent ratio assay, as described
below. Obtaining a specified final product Reb A purity ("target
purity") at minimal processing cost is called "process
optimization."
[0013] A single reflux in the selected EtOH solvent followed by one
or two stirred washes in absolute ethanol typically produces 99+%
purity Reb A from higher quality Stevia starting material. Such
high purities of Reb A have never before been achieved, and the use
of a counterintuitive process (Payzant teaches that adding water to
the reflux solvent should be avoided) to achieve such purities is a
radical departure from the prior art. Rather than eliminate water
from the process, a significant, carefully controlled, amount of
water, determined using the starting material assay, is essential
during a reflux stage.
[0014] The number of reflux stages and stirred wash stages required
to reach a target purity reflects the level and type of
contaminants in a given batch of Stevia starting material. In the
industrial embodiment of the invention, a reflux stage is much more
expensive than a stirred wash stage in terms of time, labor,
reagent, heating, and cooling costs. The industrial process for a
given batch of Stevia starting material typically balances target
purity, yield, and cost of production.
[0015] In the invention, typical reflux stages use 1 gram of Stevia
starting material for every 4 ml of the selected EtOH solvent.
Generally speaking, the lower the Reb A content of the Stevia
starting material, (i) the higher the water content (within the 4%
to 15% range) needed in the reflux solvent to produce 99+% purity
Reb A and (ii) the greater the probability that more than one
reflux stage will be needed to produce 99+% purity Reb A final
product. The higher the water content in the reflux solvent, and
the more reflux stages used, however, the lower the yield of Reb A.
In contrast to a reflux stage, a stirred wash stage preferably uses
absolute ethanol as a solvent. Denatured ethanol or other alkanols
can be used as a solvent in the stirred wash stage; using denatured
ethanol or absolute ethanol in the stirred wash stage avoids food
safety issues associated with other alkanols.
[0016] In one laboratory embodiment of the invention, an RA89.95
Stevia starting material (Stevia starting material that is
approximately 90% Reb A by weight is called herein, "RA90 starting
material"; similarly, Stevia starting material that is
approximately 80% Reb A by weight is called herein, "RA80 starting
material"; similarly, Stevia starting material that is
approximately 70% Reb A by weight is called herein, "RA70 starting
material"; etc,) is dissolved in 94% ethanol, 6% water at a ratio
of 1 gram of RA90 starting material to 4 mls of the selected EtOH
solvent. The selected EtOH solvent is also called the "reflux
solvent". A mixture of Stevia starting material and reflux solvent
is called the "reflux mixture". "Starting material" means Stevia
starting material, in the case of an initial reflux, and retentate
obtained by filtering a refluxed mixture or a stirred wash, in the
case of stages after an initial reflux stage. The mixture of RA90
starting material and reflux solvent is heated with vigorous
stirring in a reflux apparatus, equipped with a condenser, from
room temperature to approximately 79.degree. C. to 80.degree. C.
(the boiling point of absolute ethanol is 79.degree. C.; the
boiling point of the reflux mixture is usually not more than
80.degree. C.). As the reflux mixture temperature is increased to
and maintained between 79.degree. C. to 80.degree. C., at one
atmosphere of pressure, the mixture becomes opalescent and viscous
as crystals of rebaudioside A begin to precipitate. The mixture is
typically refluxed for one hour, then filtered (paper or glass
fiber filter media, e.g., Whatman GF/A filter paper (Whatman,
Florham Park, N.J.), in a laboratory embodiment) and the wet solids
placed in 100% absolute ethanol at a ratio of 1 gram of wet solids
to 4 mls of wash solvent, typically absolute ethanol, to create a
"wash mixture". The wash mixture is stirred for 15 minutes at room
temperature. After stirring, the wash mixture is filtered over
filter paper and the retentate solids dried in an oven at
80.degree. C. The dried solids are 100% Reb A and represent a
recovery of approximately 81% of Reb A in the starting material.
The final product has an approximate solubility in water of 33% (1
gram in 3 mls final volume), which is adequate for all commercial
uses.
[0017] The apparatus used in the industrial embodiment of the
invention typically comprises one or more jacketed, stainless steel
pressure vessels ("tanks" or "mixing tanks"), each typically of
multiple thousand liter capacity, equipped for mixing (e.g.,
equipped with flexible scraper blades), heating (steam in jacket),
and cooling (chilled water in jacket), and equipped with inlet and
outlet paths and a reflux condenser. Additional equipment includes
one or more centrifugal decanters (typically continuous flow
models), pumps, valves, holding tanks, pipes that interconnect the
mixing tank(s), decanter(s), and other elements of the apparatus,
and preferably a process control system. The industrial process of
the invention uses the same parameters as the lab process, except
centrifugal decanters are used instead of filter paper, the mass
and volume of starting material and solvent, respectively, is much
greater (limited primarily by tank volume), chilled water cooling
after reflux is typically used, and after the final reflux or
stirred wash stage, an "ethanol purge and drying" stage is used in
which the decanter retentate is dissolved and stirred in 80.degree.
C. water to evaporate residual ethanol before spray drying. The
final product of the industrial process is typically 99+% purity
Reb A with approximately 33% solubility in water. This level of Reb
A purity is a breakthrough in Stevia processing, especially in
industrial quantities.
[0018] Using more than one reflux stage and stirred wash stage, as
defined below, lower quality Stevia starting material can be
processed using the invention to produce a final product of 99+%
purity Reb A. Using lower quality Stevia starting materials reduces
yields and requires more refluxes to obtain 99+% Reb A purity final
product.
BRIEF DESCRIPTION OF THE DRAWINGS
[0019] FIG. 1 shows an overall view of the laboratory embodiment of
the invention.
[0020] FIG. 2 shows the steps in the reflux stage of the laboratory
process.
[0021] FIG. 3 shows the steps in the stirred wash stage of the
laboratory process.
[0022] FIG. 4 shows the first half of the starting material
assay.
[0023] FIG. 5 shows the second half of the starting material
assay.
[0024] FIG. 6 shows the steps in additional reflux stages.
[0025] FIG. 7 shows an overall view of the apparatus used in the
industrial embodiment of the invention.
[0026] FIG. 8 shows the steps in the reflux stage of the industrial
process.
[0027] FIG. 9 shows the steps in the stirred wash stage of the
industrial process.
[0028] FIG. 10 shows the steps in the ethanol purge and drying
stage of the industrial process.
[0029] FIG. 11 shows the first half of the mass to volume ratio
assay.
[0030] FIG. 12 shows the second half of the mass to volume ratio
assay.
[0031] FIG. 13 shows Graph 1, which plots yield vs. water content
of reflux solvent for different Stevia starting materials.
[0032] FIG. 14 shows Graph 2, which plots yield vs. water content
of reflux solvent for different Stevia starting materials.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0033] The invention comprises (i) a high purity, high yield
laboratory process that provides 100% pure rebaudioside A using
Stevia starting material; (ii) a high throughput, high purity, high
yield industrial process, derived directly from the laboratory
process, that provides commercial quantities of 99+% purity
rebaudioside A using Stevia starting material; (iii) the
rebaudioside A made by those processes; (iv) uses of such
rebaudioside A; and (v) a means of maximizing yields of a target
purity of Reb A based on the attributes of a given lot of Stevia
starting material.
[0034] The utility of the laboratory process lies in the ability to
produce 100% pure, water soluble Reb A for research, analytic, and
synthetic purposes. The utility of the industrial process lies in
the ability to produce commercial quantities (e.g., thousands of
kilograms) of 99+% purity Reb A for the beverage, food, baking,
candy, personal care product, pharmaceutical, tobacco, and other
industries, as well as for research, analytics, and synthesis. As a
result of the perceived sweetness of very low concentrations of
pure Reb A, for most uses, including beverages, Reb A water
solubility of less than 1% is acceptable (concentrates require
higher solubility, but within the final product solubility range
provided by the invention).
[0035] Reb A is significantly less expensive to produce using the
invention compared with existing processes, even though the
invention produces Reb A of far higher purity. The means provided
herein of maximizing yields of 99+% purity Reb A based on the
attributes of a given lot of Stevia starting material enables even
greater savings in production costs compared with existing
processes. Stevia starting material with greater than 40% purity
Reb A content is preferred for use with the invention; the
additional reflux stages typically required for Stevia starting
material of less than RA40 quality typically make the process
financially disfavored, and Stevia starting materials with greater
than 40% purity Reb A are readily available. RA40 and higher Stevia
starting materials are available from Shandong Chengwu Aoxing
Stevia Production Co., Ltd., Datianji Town, Chengwu County,
Shandong, People's Republic of China, as well as other sources
known in the art. Stevia starting materials typically do not exceed
RA80; however, a few commercially available, significantly more
expensive Stevia extracts, such as the one sold under the product
name of Rebaudio by Morita Kagaku Kogyou, Ltd., Joto-ku, Osaka,
Japan, have approximately 90% Reb A content. In the reflux and
stirred wash stages of the present invention, denatured ethanol is
typically preferred for cost reasons, particularly for the
industrial embodiment of the invention; absolute ethanol is
approximately ten times as expensive as denatured ethanol in the
U.S., but in locales where both denatured and absolute ethanol have
substantially the same cost, absolute ethanol is preferred over
denatured ethanol for use as a sole solvent and as a reflux solvent
component. Denatured ethanol containing additives other than 1%
ethyl acetate (e.g., ethanol denatured with 5% isopropanol) can be
used in the invention, but typically result in reduced yields and
purity of final product. Alkanols other than ethanol can be used,
but using ethanol avoids food safety issues related to other
alkanols.
[0036] FIG. 1 shows an overview of the invention. A basic
embodiment used with higher quality Stevia starting materials
comprises three or four stages: an EtOH formulation stage, a reflux
stage, optional stirred wash stage (typically required to produce
99+% purity Reb A), and an ethanol purge and drying stage. In
embodiments that use lower quality Stevia starting material, a
second reflux stage is typically added before a stirred wash stage
to maximize purity of the Reb A final product. The decision points
of whether to follow the first reflux with a single stirred wash
stage or with more reflux stages (with or without intermediate and
final stirred wash stages) are shown in FIG. 1 (and certain other
Figures) and labelled as "Iteration Decision". An EtOH formulation
stage always includes a starting material assay and may include a
mass to solvent ratio assay. The term "starting material assay"
includes assays of wet solids retentate following a reflux stage or
a stirred wash stage; Stevia starting material is the starting
material for a first reflux stage; wet solids retentate is the
starting material for a stage subsequent to the first reflux stage.
Starting material assay is abbreviated "SMA" in the Drawings.
[0037] Since there is an inverse relationship between yield and
purity of the Reb A final product, determining the most cost
efficient method to produce a target Reb A purity, e.g., 99+%
purity Reb A, from a given Stevia starting material is important in
process optimization and the economics of production. Process
optimization is performed by first using the laboratory embodiment,
and when the process is optimized at laboratory scale, the results
of optimization are applied at industrial scale. The preferred
embodiment of process optimization comprises using at least one
starting material assay, producing final product Reb A purity equal
to or greater than the target purity, and progressively replacing
any additional reflux stage(s) (an "additional reflux stage" is one
other than the first reflux stage, as described in more detail
below and as shown in FIG. 6) with (i) one or more stirred wash
stages, (ii) use of a different mass to solvent ratio in the
remaining reflux stage(s) and/or stirred wash stage(s), or (iii) a
combination of (i) and (ii). A decision point of whether to perform
an additional reflux or a stirred wash, or whether to proceed to
the ethanol purge and drying stage, are denoted as "Iteration
Decision" in the Drawings. For instance, when processing a given
higher quality Stevia starting material for the first time, if a
single reflux (typically, mass to volume ratio of 1 to 4) stage and
one stirred wash stage does not produce 99+% purity Reb A in the
laboratory embodiment of the invention, a mass to volume ratio
assay of the first reflux stage solvent, and optionally of the
first stirred wash stage solvent, is typically introduced into the
EtOH formulation stage(s) in an effort to obtain the target purity
without a second stirred wash stage or a second reflux stage. The
processing cost of a stirred wash stage is much less than the cost
of a reflux stage, and the processing cost of a single stirred wash
stage is less than that of two stirred wash stages.
[0038] If the Reb A purity of the final product is above the target
purity and if replacing the final additional reflux stage with a
stirred wash stage does not produce the target purity, a higher
mass to solvent ratio in the former penultimate (and earlier)
reflux stage(s) may achieve the target purity and avoid the need
for the final additional reflux stage. A higher mass to solvent
ratio is typically used in this circumstance if the cost of using
and recovering the additional ethanol in the reflux stage(s) is
less than the cost of heating and cooling in an additional reflux
stage using the original mass to solvent ratio, so long as the
target purity is obtained.
[0039] If, after replacing a final additional reflux stage with a
stirred wash and/or different mass to solvent ratio in the reflux
stage(s), the Reb A purity of the final product is still above the
target purity, the producer may elect to further reduce processing
costs by substituting a stirred wash stage for a reflux stage,
and/or by using a different mass to solvent ratio in the remaining
reflux stage(s) or stirred wash stage(s). If, after replacing a
final reflux stage with a stirred wash or using a different mass to
solvent ratio, the Reb A purity of the final product is below the
target purity previously achieved using the additional reflux stage
at issue, the producer reverts to using the deleted additional
reflux stage, since the target purity cannot be achieved without
that reflux stage. In this way, guided by EtOH formulation,
different permutations of reflux, and stirred wash stages, the cost
of producing a given target purity using a given Stevia starting
material is minimized. Process optimization is typically performed
using the laboratory embodiment of the invention; the optimal
process for a given Stevia starting material is then applied at
industrial scale. With lower quality Stevia starting material, it
is unlikely that processing using a single reflux and single
stirred wash will produce a 99+% purity Reb A final product.
Processing lower quality Stevia starting materials that contain
"stubborn" contaminants may require at least two EtOH reflux stages
and an absolute ethanol reflux stage, instead of two EtOH reflux
stages and a stirred wash stage, as Example 1 shows (Example 1 used
a blend of Stevia starting material that contained RA40).
[0040] As shown in FIG. 2, step A comprises the EtOH formulation
stage, steps B to E comprise the reflux stage of the laboratory
embodiment of the invention. Mixing and refluxing the Stevia
starting material in the selected EtOH solvent formulated for a
given Stevia starting material and then isolating rebaudioside A by
filtering the refluxed mixture is a "reflux stage". An additional
reflux stage comprises mixing and refluxing retentate (the solids
isolated by filtering the mixture from the immediately previous
stage) in the selected EtOH solvent formulated for a given Stevia
starting material or retentate and then isolating rebaudioside A by
filtering the refluxed mixture.
[0041] As shown in FIG. 3, steps F to I comprise the stirred wash
stage of the laboratory embodiment of the invention. Mixing and
stirring the isolated rebaudioside A in a denatured or absolute
ethanol wash and then isolating rebaudioside A by filtering the
wash mixture is a "stirred wash stage". As discussed above,
absolute ethanol is preferred as the solvent in a stirred wash, but
cost management may dictate using denatured ethanol in the stirred
wash stage. Steps J to K comprise the ethanol purge and drying
stage of the laboratory embodiment of the invention. In the
laboratory embodiment, steps J to K, the ethanol purge and drying
stage, are also known as "purging the ethanol and drying the
isolated retentate."
[0042] As shown in FIG. 2, in an embodiment of the invention for
laboratory use, Stevia starting material is mixed with a solvent
comprising ethanol and water (between 4% to 15%, by volume, of
water). Step A in the laboratory embodiment is a determination of
the water to ethanol balance in the solvent to be used in the
reflux solvent. The preferred means of determination of the water
to ethanol balance is an assay of the Stevia starting material, as
described below ("starting material assay"). An alternative means
of determination of the water to ethanol balance is to use a
guideline, called herein the "reflux solvent approximation," i.e.,
for every 5% decrease in Reb A purity below 90% Reb A purity in the
Stevia starting material, based on the asserted or labeled purity
of a batch of Stevia starting material, the water content of the
reflux solvent should be increased 1% above a baseline of 6% water;
the remaining volume of the reflux solvent is ethanol. For example,
the reflux solvent approximation teaches using a reflux solvent of
8% water, 92% water, volume to volume, to process RA80 starting
material and ultimately produce 100% pure Reb A (obtaining such
purity typically requires at least two reflux stages when using
RA80 Stevia starting material). The reflux solvent approximation
was derived by the first-named inventor from empirical data
collected in perfecting the starting material assay.
[0043] The starting material assay is the preferred means of
determination of the water to ethanol balance in the reflux solvent
since the starting material assay reveals the minimum, not just an
approximation of, water content needed in the reflux solvent to
produce (in conjunction with one or more stirred wash stages) a
target purity, e.g., 99+% purity Reb A. Water content in excess of
the minimum significantly decreases yields of Reb A final product.
As shown in FIGS. 4 and 5, the starting material assay is conducted
by using a coordinated set of solvents in which the water content
is uniformly incremented 0.5% to 15%, or within a subset of that
range, e.g., 2% to 10%,; the balance of each solvent is ethanol
(each combination of water and ethanol is called herein a "test
solvent"). The asserted or labeled purity of the Stevia starting
material is an input in the assay only to the degree that the
asserted or labeled Reb A purity and the reflux solvent
approximation can be used to center the initial set of test
solvents. For instance, a starting material assay for Stevia
starting material labeled as RA80 would typically use a set of test
solvents centered around 8% water and 92% ethanol. In the starting
material assay, a sample of Stevia starting material taken from a
given batch or lot of Stevia starting material ("sampled batch") is
mixed with each in a series of test solvents (e.g., 2% water and
98% ethanol, 4% water and 96% ethanol, 6% water and 94% ethanol,
etc.) in a mass to volume ratio of 1 g Stevia starting material to
4 ml of test solvent.
[0044] As explained in more detail below, the ratio of Stevia
starting material to reflux solvent can be varied over a range of
from 2 to 10 or more parts by volume of EtOH solvent per 1 part of
Stevia starting material by mass, but is preferably from 3 to 6
parts by volume of EtOH solvent per 1 part of Stevia starting
material by mass, and is most preferably approximately 4 parts by
volume of the EtOH solvent per 1 part of Stevia starting material
by mass. As shown in the Example 7, a 1:4 ratio of Stevia starting
material to selected EtOH solvent generally maximizes yield of Reb
A in a single reflux, single stirred wash process using higher
quality Stevia starting material. Each mixture of the Stevia
starting material and test solvent (each, a "test mixture") is
heated (typical heating rate of 4.degree. C./min.) with vigorous
stirring in a reflux apparatus, equipped with a condenser, from
room temperature to approximately 79.degree. C. to 80.degree. C. at
one atmosphere of pressure, refluxed for approximately one hour,
then cooled for approximately one hour (typically, flasks are
placed in an ice bath) and filtered over filter paper. The wet
solids, or retentate, from each test mixture are each placed in
100% absolute ethanol at a ratio of 1 gram of wet solids retentate
to approximately 4 mls of solvent (each, a "test wash
mixture").
[0045] If a ratio of Stevia starting material to EtOH solvent other
than 1:4 was used in the reflux, the same ratio of wet solids
retentate from each test mixture to 100% absolute ethanol can be
used for the test wash mixture. Typically, however, a 1:4 ratio of
Stevia starting material to ethanol in the test wash mixture is
used since higher ratios typically have minimal effect on ultimate
Reb A purity and may decrease Reb A yields.
[0046] Each test wash mixture is stirred for 15 minutes at room
temperature. After stirring, each test wash mixture is filtered
over filter paper and the retentate solids from filtering are dried
in an oven, preferably a blower oven, at or minimally above
79.degree. C. to produce dried samples (each, a "test sample").
Oven temperature should not be too much above 80.degree. C. to
avoid degrading the test sample or igniting the residual ethanol.
Each test sample is tested for Reb A purity using the HPLC analysis
method described in the HPLC Method section of the Examples below.
For a target purity of 100% pure Reb A final product, the test
solvent (i) corresponding to the test sample indicated as 100% pure
Reb A (or highest Reb A purity in the event a single reflux, single
stirred wash fails to produce 100% pure Reb A) in the HPLC analysis
and (ii) containing the lowest water content is selected ("first
iteration result") as the solvent for processing the remainder of
the sampled batch. For instance, if test solvents containing 6%,
8%, and 10% water (the balance of each test solvent is ethanol)
each produces a test sample of 100% pure Reb A, then a 6% water,
94% ethanol solvent is the first iteration result and (unless a
second iteration of the starting materials assay is used to refine
the result) would be used to process the remainder of the sampled
batch.
[0047] The starting material assay may be multiply iterated to
select a solvent that further optimizes the yield of 100% pure Reb
A. In a second iteration of the starting material assay, the water
content in the coordinated set of test solvents is uniformly
decremented below the water content of the first iteration result,
but in smaller increments, e.g., decrements of 0.5% water. For
example, if the first iteration result were 6% water, 94% ethanol,
the test solvents in the second iteration could be 6.0%, 5.5%,
5.0%, and 4.5% water. The remainder of the process as described
above (first reflux, stirred wash, drying, and HPLC analysis) is
performed using second iteration test solvents, test mixtures, and
test wash mixtures. The second iteration test solvent (i)
corresponding to the test sample indicated as 100% pure Reb A (or
highest Reb A purity in the event a single reflux, single stirred
wash fails to produce 100% pure Reb A) in the HPLC analysis and
(ii) containing the lowest water content is selected as the solvent
for processing the remainder of the sampled batch and replaces the
first iteration result. For instance, if test solvents containing
6.0%, 5.5%, and 5.0% water (the balance of each test solvent is
ethanol) each produces a test sample of 100% pure Reb A, then a 5%
water, 95% ethanol solvent would be used to process the remainder
of the sampled batch. More than two iterations of the starting
material assay can be performed to more finely optimize the
selection of reflux solvent, e.g., a third iteration could use 0.1%
decrements of water content below the second iteration result.
[0048] The water and ethanol solvent selected using the reflux
solvent approximation or the starting material assay is called the
"selected EtOH solvent." "EtOH" is a trivial name for a blend of
ethanol and water. For convenience, the Stevia starting material to
be processed using the selected EtOH solvent (even if selected
using the reflux solvent approximation) is called the "sampled
batch". The starting material assay typically produces a much more
accurate result than the reflux solvent approximation, and is the
preferred method of determining the selected EtOH solvent.
[0049] Using the starting material assay or the reflux solvent
approximation, including multiple iterations thereof, is called
"formulating a selected EtOH solvent for the starting material" and
is also called the "EtOH formulation stage". By substituting the
wet solids retentate obtained by filtering a reflux mixture (or a
stirred wash mixture) in place of Stevia starting material in the
starting material assay, an EtOH formulation stage can be performed
before an additional reflux stage (or stirred wash stage).
Performing an EtOH formulation stage before each reflux is usually
done only if very large quantities of wet solids retentate are
being processed or if yields and purity of a previous reflux stage
did not meet expectations.
[0050] An EtOH formulation stage can optionally include, after
determining the selected EtOH solvent, varying the ratio of EtOH
solvent to Stevia starting material (before an initial reflux
stage) or varying the ratio of EtOH solvent to wet solids retentate
(before an additional reflux stage or a stirred wash stage) to
determine the ratio that maximizes Reb A yield. This assay is
called the "optional mass to volume ratio assay." The ratio of the
mass of Stevia starting material or wet solids retentate to the
volume of the selected EtOH solvent can be varied over an
open-ended range starting from at least 3 parts by volume of the
selected EtOH solvent per 1 part of starting material by mass, but
is preferably from 3 to 10 parts by volume of the selected EtOH
solvent per 1 part of starting material by mass, more preferably
from 4 to 6 parts by volume of the selected EtOH solvent per 1 part
of starting material by mass, and is most preferably approximately
4 parts by volume of the selected EtOH solvent per 1 part of
starting material by mass. Using higher quality Stevia starting
material, maximum Reb A yield is typically obtained at, or very
close to, a ratio of 4 parts by volume of the selected EtOH solvent
per 1 part of Stevia starting material (before an initial reflux
stage), or of wet solids retentate (before additional reflux
stages) by mass. If very large quantities of wet solids retentate
are being processed, or if yields and purity of a previous reflux
stage did not meet expectations, the EtOH formulation stage and
optional mass to volume ratio assay can be repeated before an
additional reflux. For processing higher quality Stevia starting
materials, the first optional mass to volume ratio assay is usually
performed using small increments of ratios near 1:4 to determine
the mass to solvent ratio that produces maximum Reb A yield. For
instance, ratios of 1:3.8, 1:3.9, 1:4.0, 1:4.1, and 1:4.2 (m/v) may
be used in a first optional mass to volume ratio assay when
processing higher quality Stevia starting materials.
[0051] Higher ratios of Stevia starting material to EtOH solvent in
the reflux, such as 1:6 or 1:8 (m/v), are sometimes needed to
process lower quality Stevia starting material (e.g., RA40 to RA85)
when the target purity is 99+% purity or 100% pure Reb A. As shown
in the Example 7, Reb A yields typically decrease as higher ratios
of Stevia starting material to the selected EtOH solvent in a
reflux, and of wet solids retentate to ethanol in a stirred wash or
in a second or subsequent reflux, are used; costs also increase as
higher ratios are used, since more reagents and bigger tanks are
used. As shown in Example 7, such higher ratios may be necessary to
produce 100% pure Reb A final product from lower quality Stevia
starting materials using two reflux stages. Ratios of Stevia
starting material to the selected EtOH solvent in a first reflux
stage, or of wet solids retentate to the selected EtOH solvent in
an additional reflux stage, below 1:3.5 (m/v) typically produce
very viscous mixtures that are difficult to process, produce lower
yield of Reb A, and produce lower purity Reb A. Ratios of Stevia
starting material to the selected EtOH solvent in a first reflux
stage, or of wet solids retentate to the selected EtOH solvent in a
subsequent reflux stage, below 1:3 (m/v) may produce mixtures that
crust or solidify when cooled, which greatly complicates
processing.
[0052] As a general rule, a mass to volume ratio of Stevia starting
material to the selected EtOH solvent in a first reflux stage, or
of wet solids retentate to the selected EtOH solvent in an
additional reflux stage, of 1:4 or very close thereto is the most
preferred ratio, especially for higher quality Stevia starting
material, since the maximum Reb A yield will be at or very close to
a 1:4 (m/v) ratio.
[0053] As shown in FIGS. 11 and 12, the steps used in a mass to
solvent ratio assay are very similar to those in the starting
materials assay, except that, for a given mass of Stevia starting
material (or following a first reflux, the mass of wet solids
retentate), in a mass to solvent ratio assay, for a given mass of
Stevia starting material (or following a first reflux, the mass of
wet solids retentate), the volume of the selected EtOH solvent is
varied and the mass to volume ratio that maximizes Reb A purity of
the final product is typically chosen for processing the sampled
batch. The mass to volume ratio assay is optional and is typically
only used with lower quality Stevia starting material; as shown in
Example 7, a higher mass to volume ratio (e.g., 1:6) of starting
material to selected EtOH solvent in a reflux can sometimes produce
100% pure Reb A with fewer refluxes. The mass to volume ratio assay
can be reiterated (using smaller increments and decrements of
selected EtOH solvent volume) to optimize the accuracy of the mass
to volume ratio for a given Stevia starting material or
retentate.
[0054] As the data in Example 1 shows, 1% more water in the reflux
solvent above optimum water content can reduce the yield of 100%
pure Reb A by 1% or more. The fine tuning of optimum water content
is very important in the economics of the industrial process. The
asserted or labeled Reb A content of Stevia starting materials may
differ substantially from actual Reb A content, and a given batch
of Stevia starting material often is a blend of Stevia extract from
many sources. The selected EtOH solvent determined by the starting
material assay is specific to a given batch of Stevia starting
material. One of the primary uses of the laboratory embodiment of
the invention is to perform the starting material assay for the
industrial embodiment of the invention.
[0055] As shown in FIG. 2, after the EtOH solvent is selected, in a
laboratory embodiment of the invention, a producer-determined
amount of the sampled batch is processed using the reflux and
stirred wash stages, as follows. Stevia starting material of the
sampled batch is mixed with the selected EtOH solvent in a mass to
volume ratio of 1 g Stevia starting material to 4 ml of selected
EtOH solvent (or other ratio determined using the EtOH formulation
stage, including any optional mass to volume ratio assay). The
mixture of the Stevia starting material and the selected EtOH
solvent ("reflux mixture") is heated (typically, 4.degree. C./min.)
with vigorous stirring in a reflux apparatus, equipped with a
condenser, from room temperature to approximately 79.degree. C. to
80.degree. C. at one atmosphere of pressure, and maintained in
reflux typically for one hour; the reflux mixture is then cooled
for one hour (typical cooling rate of -1.8.degree. C./min., e.g.,
by placing flasks in an ice bath) and then filtered over filter
paper. Allowing the solvent mixture to stand after completion of
reflux and before filtering marginally increases Reb A yield when
processing higher quality Stevia starting material, but
significantly increases Reb A yield when processing lower quality
Stevia starting material.
[0056] As shown in FIG. 3, the wet solids retentate from the first
reflux stage are placed in 100% absolute ethanol at a ratio of 1
gram of wet solids to 4 mls of solvent ("wash mixture") (or other
ratio determined using the EtOH formulation stage, including any
optional mass to volume ratio assay). The wash mixture is stirred
for 15 minutes at room temperature. Ratios of wet solids retentate
to ethanol below 1:3.5 (m/v) in a stirred wash stage may produce
mixture viscosity that impairs processing. Using higher quality
Stevia starting materials, ratios of wet solids retentate to
ethanol above 1:4 (m/v) in a stirred wash stage produce negligible
improvement in Reb A purity and typically decrease Reb A yield.
Therefore, a mass to volume ratio of wet solids retentate to
ethanol in a stirred wash stage of 1:4 or very close thereto is the
most preferred ratio when using higher quality Stevia starting
materials. After stirring, the wash mixture is filtered over filter
paper and the retentate solids from filtering are dried in an oven,
preferably a blower oven at or minimally above 79.degree. C. for up
to two hours to produce dried solids. An oven temperature of
80.degree. C. safely vaporizes the ethanol in the wash mixture
without igniting the ethanol, and thus purges the retentate solids
of ethanol. Allowing the wash mixture to stand after completion of
stirred wash and before filtering marginally increases yields when
processing higher quality Stevia starting material, but
significantly increases yields when processing lower quality Stevia
starting material. If a powdered final product is desired, the
dried solids are pushed through a 30-mesh screen, resulting in a
final product of coarse white powder (Reb A final product).
[0057] The final product, dried solids, produced using higher
quality Stevia starting material and a single reflux, single
stirred wash process is typically 100% Reb A and typically
represents a recovery of approximately 81% of Reb A in the starting
material. The final product has solubility in water of 33% (1 gram
in 3 mls final volume), which is adequate for all commercial uses,
including concentrates. As shown in Example 3, the stirred wash
stage typically increases Reb A purity from approximately 96%
purity Reb A (purity after the first reflux stage and before the
stirred wash stage) to 100% pure Reb A (after the stirred wash
stage). If the uses of the final product can tolerate the residual
bitterness of Reb A of less than 99% purity, the stirred wash stage
can be omitted and the retentate from filtering the reflux mixture
can be dried and used commercially. However, a stirred wash stage
is preferred to eliminate all perceptible bitterness.
[0058] As shown in Examples 1 and 2, if the Stevia starting
material is less than 85% Reb A, at least one additional reflux
stage is typically required to produce 100% pure Reb A. To process
Stevia starting material of less than RA85 purity, a first reflux
stage is performed using the selected EtOH solvent, but the wet
solids retentate from filtering the reflux mixture is typically
used as starting material for a second reflux instead of being
placed in a stirred wash. The processing parameters for the
additional second reflux stage (and more additional reflux stages,
as needed to achieve a target Reb A purity) are the same as for the
first reflux stage described, including the EtOH formulation stage.
Optionally, a stirred wash stage (but without drying the wet
solids) can be performed between reflux stages, in which case the
wet solids retentate from filtering the wash mixture ("intermediate
stirred wash stage") is used as starting material for the following
reflux stage. Intermediate stirred wash stages are sometimes needed
when processing Stevia starting material of RA80 or lower
quality.
[0059] For expediency, the selected EtOH solvent used in the first
reflux can be used in the additional reflux stages. However, to
further optimize yields, a starting material assay can be performed
on samples taken from the wet solids retentate produced in the
filtering step that concludes each reflux stage (if an intermediate
stirred wash stage is inserted between reflux stages, the wet
solids retentate from filtering an intermediate stirred wash stage
wash mixture is used as starting material for the starting material
assay). After the final reflux stage, a stirred wash stage is
typically conducted (and then the ethanol purge and the drying
stage), as described above, to produce up to 100% pure Reb A. Each
additional reflux stage, however, decreases yields of final
product. If the uses of the final product can tolerate the residual
bitterness of Reb A of less than 99% purity, the number of reflux
stages (and optional intermediate stirred wash stages) is selected
to produce the target purity and yield required for the given
uses.
[0060] Although a heating rate of 4.degree. C./min. and a cooling
rate of -1.8.degree. C./min. are described, other cooling and
heating rates obtainable with typical laboratory equipment may be
used without significantly decreasing yield or purity of Reb A in
the final product. Very high heating rates require vigorous
agitation of the mixture to prevent burning or hydrolysis of the
starting material. Heating and cooling rates obtainable may depend
on batch size. For instance, force-cooling in a laboratory may be
accomplished by setting a reaction flask on ice or in ice water;
larger reaction volumes in a given flask size take longer to cool
on ice or in ice water. Generally speaking, following a reflux,
during optional standing steps, mixture temperatures that are lower
than room temperature but higher than 0.degree. C. (for mixtures
containing water) and long standing times, will improve yields of
Reb A, especially when processing lower quality Stevia starting
material. During long standing times, especially with post-reflux
mixture temperatures below room temperature, it is critical that
the mixture be stirred to avoid gelling, crusting, or
solidification of the mixture; if gelling, crusting, or
solidification occurs, yield and/or purity may be adversely
affected. If the post-reflux mixture is cooled and inadequately
stirred during a standing time in excess of 12 hours, the mixture
may gel, crust, or solidify; if this happens, some or all of the
mixture may not thereafter be easily convertible into a pumpable
slurry for further processing. Following the stirred wash, mixture
temperatures below 0.degree. C. and approaching the freezing point
of ethanol can be used, Even when processing RA40 Stevia starting
material, standing times after reflux or stirred wash that are
longer than 96 hours provide little improvement in yield.
[0061] With a solvent of 5% to 15% water and the remainder ethanol,
Stevia starting material immediately goes into solution with
stirring. In an anhydrous ethanol solvent, however, Stevia starting
material only goes into solution with heating, and both the yields
and purity of Reb A in the final product are lower using anhydrous
ethanol as a reflux solvent. With 5% to 15% water in the solvent,
when the mixture of solvent and Stevia starting material reaches
50.degree. C., Reb A begins to precipitate out of solution; in
contrast, in an anhydrous ethanol solvent, however, when the
mixture of solvent and Stevia starting material reaches 50.degree.
C., the Stevia starting material is still dissolving and
precipitation of Reb A has not begun. After each reflux stage,
rather than force-cooling the mixture after removal of heat, the
mixture can simply be allowed to air cool to room temperature, but
yields of Reb A are lower compared with force-cooling (even with
longer standing times, e.g., 96 hours, after reaching room
temperature). In other words, force cooling of a reflux mixture
improves yields for lower quality Stevia starting materials.
[0062] FIG. 7 shows an overview of the apparatus used in the
industrial embodiment of the invention. The basic industrial
embodiment for processing higher quality Stevia starting material
comprises three or four stages: an EtOH formulation stage, a reflux
stage, optional stirred wash stage (typically required to produce
99+% purity Reb A), and an ethanol purge and drying stage. The
decision points of whether to follow the first reflux with a single
stirred wash stage or with one or more additional reflux stages
(with or without more stirred wash stages) are shown in FIG. 1 (and
other Figures) and denoted as "Iteration Decision". In embodiments
that use lower quality Stevia starting material, at least one
additional reflux stage is typically added before a final stirred
wash stage to maximize purity of the Reb A final product. As shown
in FIGS. 8 to 10 respectively, step I comprises the EtOH
formulation stage, steps II to V comprise the reflux stage; steps
VI to IX comprise the stirred wash stage; and steps X to XIII
comprise the ethanol purge and drying stage of the industrial
embodiment of the invention. In the industrial embodiment, steps X
to XIII, the ethanol purge and drying stage, are also known as
"purging the ethanol and drying the retentate."
[0063] As shown in FIG. 7, the typical apparatus to practice the
industrial embodiment includes one or more jacketed, stainless
steel pressure vessels ("tanks" or "mixing tanks"), each typically
of 3,000 liter capacity, equipped for mixing (flexible scrapper
blades), heating (steam in jacket), and cooling (chilled water in
jacket), and equipped with inlet and outlet paths and a reflux
condenser (not separately shown in FIG. 7). In all tank process
steps other than step XI below, the outlet path of the reflux
condenser attached to a mixing tank is connected back to the mixing
tank. An adequately rated pressure vessel without a reflux
condenser can be used in the industrial embodiment, but a reflux
condenser is preferred because of the superior mixture agitation
and reduction of mixture superheating. All tanks, shells, tubes,
valves, motors, etc., are ASTM pressure vessel certified for use
with explosive liquids and vapors. Each tank has: pressure and
temperature sensors and flow and/or quantity indicators; powder
hopper and feeder, with inlet flow metering; ethanol and water
feeds with inlet flow metering; shut-off valves; and customary
other fixtures known in the art of processing in pressure vessels,
such as one or more centrifugal decanters (typically continuous
flow models), membrane filters (optional), pumps, holding tanks,
pumps, spray dryer, hopper feeder, pipes that interconnect the
elements of the apparatus, filters, etc., and a process control
system.
[0064] As shown in FIG. 8, step I in the industrial embodiment of
the invention is a determination of the water to ethanol balance in
the reflux solvent. The preferred means of determination of the
selected EtOH solvent is to perform one to three iterations of the
starting material assay, as described above and illustrated in
FIGS. 4 and 5. An alternative means of determining the selected
EtOH solvent is to use the reflux solvent approximation, as
described above. The starting material assay is preferred, since
the starting material assay provides the minimum, not just an
approximation of, water content needed in the reflux solvent to
produce (after the final stirred wash stage) 100% pure Reb A when
using higher quality Stevia starting material. Given the large
quantities of Stevia starting material, ethanol, and energy used in
the industrial process, optimizing yields by optimizing the
selected EtOH solvent is often critical to the profitability of a
producer. Accordingly, the starting material assay is typically
multiply iterated, as described above, to optimize the selected
EtOH solvent for a given Stevia starting material (and optionally,
for a given retentate before one or more additional refluxes). As
discussed above, each starting material assay can optionally
include, after the selected EtOH solvent is determined, a mass to
volume ratio assay to determine the ratio of mass to volume of
selected EtOH solvent that maximizes Reb A yield.
[0065] As shown in FIG. 8, after the EtOH solvent is selected, in
an industrial embodiment of the invention, a producer-determined
amount of the sampled batch is processed using the reflux and
stirred wash stages (or, when processing lower quality Stevia
starting materials, a one or more additional refluxes, each
additional reflux stage with or without an intermediate stirred
wash stage). Step II in the industrial process is to supply and mix
ethanol and water in a first tank ("Tank 1") to produce a reflux
solvent that is the selected EtOH solvent. The ethanol content is
maintained throughout process steps II to IV at approximately +/-2%
of the ethanol content in the selected EtOH solvent by capturing
and cooling reaction vapors using a reflux condenser and returning
the condensate to Tank 1. In step II, the Stevia starting material
is added to and mixed with the selected EtOH solvent at a mass to
volume ratio of 1 to 4 (Stevia starting material to selected EtOH
solvent) or at the mass to volume ratio determined using the
optional mass to volume ratio assay, as described above. In step
III, while being continuously mixed in Tank 1, the mixture is
heated (steam fed to jacket of Tank 1, typical heating rate of
1.degree. C./min.) to 79.degree. C. All reactions in the typical
industrial embodiment are at one atmosphere of pressure.
[0066] After the Tank 1 mixture has reached 79.degree. C., mixing
during reflux is continued for 1 hour while maintaining the reflux
mixture temperature between 79.degree. C. to 80.degree. C. (the
ethanol in Tank 1 boils and is recovered through the reflux
condenser). In Step III, as the reflux mixture temperature
increases above 50.degree. C., Reb A typically begins to
precipitate out of solution. If Tank 1 is not filled to maximum
mixing capacity, a thick creamy substance (primarily precipitating
Reb A crystal) may form and line the interior surface of Tank 1;
scraping blades are preferably used to return the creamy substance
to the mixture. In step IV, the refluxed mixture is cooled (chilled
water (4.degree. C). fed to Tank 1 jacket, typical cooling rate of
-0.3.degree. C./min.) to a temperature between 4.degree. C. to
20.degree. C., typically 14.degree. C. In step IV, when processing
Stevia starting material of less than RA80 quality, the mixture is
typically held chilled at constant temperature (typically
14.degree. C.), while being mixed, to improve yield. In such cases,
the longer the chilled standing period, the greater is the recovery
of precipitated Reb A, up to approximately 96 hours of standing
time. During this standing time, mixing blades scrape the interior
surface of Tank 1 to maintain a slurry in the bottom of Tank 1. If
cooled and unstirred during standing time, the mixture may become
gelatinous or even crusty; if the mixture become gelatinous or
crusty, the mixture may not thereafter be convertible into a
pumpable slurry for further processing without adversely affecting
yield and/or purity. In step IV, when processing higher quality
Stevia starting material, standing time is typically not necessary
after cooling the reflux mixture (typically to 14.degree. C.).
Batch processing of Stevia starting materials, as described above
and below for Tank 1, can be performed concurrently in other
available tanks Typically two or more tanks are used in a
commercial production facility, with batches phased based on
workforce and product demand
[0067] Next, enough absolute ethanol (anhydrous ethanol) is added
to the contents of Tank 1 to create a pumpable slurry. Once the
chilled slurry in Tank 1 becomes pumpable (pumpability depends upon
the type of pump; typically a diaphragm pump handles a more viscous
fluid than an impeller pump), the slurry in Tank 1 is pumped to a
continuous-flow, centrifugal decanter. The centrifugal decanter is
supplied with the slurry stream, and in step V, the ethanol is
decanted to the "lights" discharge outlet, and the retentate is fed
(typically, a gravity feed) out the retentate, or "heavies",
discharge outlet. Typically in large scale production, the "lights"
discharge stream is pumped through a means (e.g., a membrane filter
or falling film evaporator) of separating the small solids from the
ethanol in the "lights" discharge stream. The recovered ethanol is
typically distilled and reused. The small solids captured by the
separating means are typically returned to a mixing tank for
further processing (e.g., in a stirred wash stage or additional
reflux stage).
[0068] Even though a centrifugal decanter is not a true filter, it
performs a filter-like function of separating "lights" and
"heavies" (i.e., separating liquid and small solids, or lights,
from heavier particles, or heavies). Therefore, the use of a
centrifugal decanter in the invention is called "filtering".
Filtering requires that the filtering means separate particles with
an axis or diameter approximately greater than 2.0 .mu.m, more
preferably greater than 1.5 .mu.m; and most preferably equal to or
greater than 0.1 .mu.m from the mixture being filtered. Other
apparatus that separate liquids from solids, such as membrane
filters and falling film evaporators, can be used instead of a
centrifugal decanter, but for a given investment in apparatus, the
throughput of alternative technology is often significantly lower
compared with the throughput of a centrifugal decanter. The
filtering means is selected must be appropriate for the volume of
mixture to be filtered and throughput required. For this reason,
centrifugal decanters are typically favored over membrane filters
and falling film evaporators for use in filtering refluxed mixtures
and stirred wash mixtures in the industrial embodiment of the
invention, since centrifugal decanters have much higher throughput,
far fewer clogging problems, and cost less compared with
alternative technologies.
[0069] As shown in FIG. 9, in step VI, the wash stage commences by
supplying an available tank ("Tank 2") with the wet solids from the
retentate discharge outlet of the centrifugal decanter and 100%
absolute ethanol typically at or near a ratio of 1 gram of wet
solids to 4 mls of ethanol ("wash mixture"). Ratios of wet solids
retentate to ethanol below 1:3.5 (m/v) in a stirred wash stage may
be so viscous that processing is impaired. Ratios of wet solids
retentate to ethanol above 1:4 (m/v) in a stirred wash stage
typically produce negligible improvement in Reb A purity and often
decrease Reb A yield. Therefore, a mass to volume ratio of wet
solids retentate to ethanol of 1:4 or very close thereto in a
stirred wash stage is the most preferred ratio when using higher
quality Stevia starting materials. In Step VII, the wash mixture is
stirred, using the mixing blades of Tank 2, for 15 minutes without
further heating or cooling of the wash mixture (i.e., at ambient
temperature). The mixture may be allowed to stand (step VIII) to
improve yields when using lower quality Stevia starting materials.
After stirring and optional standing time, in step IX, the wash
mixture in Tank 2 is pumped to a continuous-flow, centrifugal
decanter. The centrifugal decanter is supplied with the wash
mixture as input, the ethanol is decanted to the "lights" discharge
outlet, and the retentate is fed (typically, a gravity feed) out
the "heavies" (retentate) discharge outlet. Chilling the wash
mixture during the stirred wash, and allowing the wash mixture to
stand (step VIII), typically with mixing, after completion of
stirred wash and before decanting, marginally increases yields when
processing higher quality Stevia starting material, but
significantly increases yields when processing lower quality Stevia
starting material.
[0070] As shown in FIG. 10, the ethanol purge and drying stage, in
order to produce final product as a powder, in step X retentate
from the decanter is fed (typically, a gravity feed) into an
available tank ("Tank 3") and combined with water such that the
ultimate volume is approximately twice the mass of the retentate
supplied to Tank 3. For example, to 500 kg of retentate, enough
potable (preferably, filtered) water is added to produce
approximately 1,000 liters final volume of mixture ("water
mixture"). In step XI, the contents of Tank 3 are heated (heating
rate of 1.degree. C./min.) to 79.degree. C. in a vessel equipped
with a vapor discharge outlet, then maintained at between
79.degree. C. to 80.degree. C., and mixed to evaporate residual
ethanol in the mixture. Mixing the contents of Tank 3 at that
temperature (the boiling point of ethanol) causes any ethanol
remaining in the mixture to evaporate; the ethanol in the
discharged vapor is preferably captured and recycled (but can be
vented as a waste stream out of Tank 3). The mixing continues until
the ethanol content of the mixture is safe for spray-drying. The
manufacturer of the spray dryer used in step XIII specifies a
maximum content of flammables in liquids to be spray-dried, which
is the upper limit of ethanol content fed from Tank 3 to a spray
dryer. Alternate means of drying that do not require pre-removal of
the solvent can be used, such as an explosion-proof spray dryer,
but such alternate means may require more operator attention. An
explosion-proof spray dryer would shorten production time and
reduce energy consumption. As described in Example 6, at industrial
scale (e.g., 500 kgs of Stevia starting material), boiling off the
ethanol before spray-drying can require ten or more hours to purge
the ethanol.
[0071] In step XII, the contents of Tank 3 are maintained between
60.degree. C. and 99.degree. C. (preferably between 79.degree. C.
to 80.degree. C.) while pumping the contents of Tank 3 to a spray
dryer (step XIII). Since the Stevia starting material is refluxed
in EtOH solvent, it is doubtful that any pathogens would survive
the reflux (microbial content in final product is undetectable,
based on third party lab tests). Before spray-drying, the mixture
can optionally be pumped trough a membrane filter or equivalent
means of separation to remove any particulate contaminants. The
spray dryer typically discharges through an automatically
controlled airlock valve into a feed hopper (typically equipped
with a screw-feed bag loader). The solids in the feed hopper are
the final product, water soluble, 99+% purity Reb A, in powder
form. As shown in Example 4 (a two reflux, industrial process)
below, 68% of the Stevia starting material by dry weight was
recovered when processing RA89.95 Stevia starting material, and the
recovered solids were 99+% Reb A. Example 4 represents a recovery
of 75.25% of the Reb A in the Stevia starting material. As shown in
Example 6 (a single reflux with single stirred wash process using
RA83.14 Stevia starting material) below, 59.04% of the Stevia
starting material by dry weight was recovered, and the recovered
solids were 99+% Reb A. Example 6 represents a recovery of 70.59%
of the Reb A in the Stevia starting material.
[0072] The 99+% purity Reb A final product of the industrial
embodiment typically has solubility in water of 33% (1 gram in 3
mls final volume). As shown in Example 4, the stirred wash stage
typically increases Reb A purity from approximately 96% purity Reb
A (purity after the first reflux stage and before the stirred wash
stage) to 100% pure Reb A (after the stirred wash stage). If the
uses of the final product can tolerate the residual bitterness of
Reb A of less than 99% purity, e.g., for use in pickling or in
flavor modification below the sweetness threshold, the stirred wash
stage can be omitted and the retentate from filtering the reflux
mixture can be dried and used commercially. However, a stirred wash
stage is preferred to eliminate all perceptible bitterness.
Solubility in water of approximately 33% is adequate for all
commercial uses, including concentrates.
[0073] As shown in Examples 1, 5, and 7 below, if lower quality
Stevia starting material is used, at least one additional reflux
stage is typically required to produce 99+% purity Reb A (final
product of the stirred wash stage). To process Stevia starting
material using two reflux stages, a first reflux stage is performed
using the selected EtOH solvent, but the wet solids retentate from
filtering the reflux mixture is used as starting material for a
second reflux instead of being placed in a stirred wash. A decision
to conduct one or more additional refluxes and/or stirred washes is
called an "iteration decision", and denoted in several of the
Figures as "Iteration Decision". The processing steps for an
additional reflux stage (and any more additional reflux stages, as
needed to achieve a target Reb A purity), as shown in FIG. 6, are
described above, including one or more EtOH formulation stages.
[0074] In FIG. 6, the mix, heat, reflux, cool, and stand steps
(i.e., steps C' to E') of an additional reflux stage would
typically be performed in Tank 1A (as shown in FIG. 7). Optionally,
an intermediate stirred wash stage can be performed between reflux
stages, in which case the wet solids retentate from decanting the
intermediate stirred wash mixture is used in an optional EtOH
formulation stage (step B') or otherwise as starting material for
an additional reflux stage (steps C' to E'). After cooling, the
mixture is filtered (step F'). An additional reflux stage is
followed either by another additional reflux stage, a stirred wash
stage, or by an ethanol purge and drying stage. Iterations of the
starting material assay are preferably done at the beginning of
each additional reflux stage to monitor the improvement in Reb A
purity. Generally, if retentate Reb A purity is over 96%, after a
reflux stage, then a stirred wash stage and an ethanol purge and
drying stage will produce 99+% purity Reb A final product.
[0075] Typically, the selected EtOH solvent used in the first
reflux is also used in any additional reflux stages. However, to
further optimize yields, an EtOH formulation stage can be performed
on samples taken from the wet solids retentate produced in the
decanting step that concludes each reflux stage (if an intermediate
stirred wash stage is inserted between reflux stages, the wet
solids retentate from decanting an intermediate stirred wash stage
is used as starting material for the EtOH formulation stage). After
the final reflux stage in the industrial process, a stirred wash
stage and the ethanol purge and drying stage are conducted, as
described above, to produce 99+% purity Reb A. Each additional
reflux stage, however, decreases yields of final product.
[0076] Although a heating rate of 1.degree. C./min. and a cooling
rate of -0.3.degree. C./min. are described in the industrial
process, other cooling and heating rates obtainable with typical
industrial equipment may be used without significantly decreasing
yield or purity of Reb A in the final product. Very high heating
rates require vigorous agitation of the mixture to prevent burning
or hydrolysis of the starting material. Heating and cooling rates
obtainable may depend on batch size, pressurized heating and
cooling, and (in the case of cooling) the use of refrigerants
instead of chilled water. For instance, force-cooling may be
accomplished by using refrigerants with temperatures below
0.degree. C. (at least during the initial cooling period with
vigorous mixing while the mixture temperature is considerably
higher than 0.degree. C.) in the pressure vessel jackets. Mixture
temperatures during optional standing steps after the first or
additional reflux steps that are lower than 4.degree. C. but higher
than 0.degree. C. (which avoids ice formation in the mixture) and
long standing times will improve yields of Reb A slightly,
especially when processing Stevia starting material of less then
RA80 purity. However, even when processing RA40 Stevia starting
material, standing times after reflux or stirred wash that are
longer than 96 hours provide little improvement in yield.
[0077] Rather than use reflux condensers in the laboratory and
industrial processes, other methods (e.g., venting the vapor and
supplying fresh ethanol and water, or other apparatus to capture
and condense solvent vapor and return it to the mixture) can be
employed to maintain the approximate desired concentration of mixed
components.
[0078] Instead of EtOH, the reflux solvent used in the invention
can be hydrous alkanols other than EtOH, but purities and yields
decrease if other hydrous alkanols are substituted for EtOH.
Likewise, the stirred wash solvent can be alkanols other than
ethanol. The use of other hydrous alkanols as the reflux solvent,
and other alkanols as the stirred wash solvent, may result in
akylated contaminants incompatible with use in compositions for
human consumption. If other hydrous alkanols are used, the starting
material assay (or reflux solvent approximation), and optional mass
to volume ration assay, are performed to determine the water
content of the reflux solvent for a given reflux and Stevia
starting material and the mass to solvent ratio, respectively.
[0079] As described above, the minimum number of reflux stages and
stirred wash stages needed to achieve a target Reb A purity is
determined empirically using permutations of refluxes (with mass to
solvent ratios sometimes higher than 1 to 4 for lower quality
Stevia starting material) and stirred washes for a given Stevia
starting material and target Reb A purity. As a general rule, when
a single reflux stage and one or two stirred wash stages do not
produce 99+% purity Reb A, a mass to volume ratio assay is included
in one or more EtOH formulation stages. As noted above, using
higher quality Stevia starting material in the invention typically
provides 99+% purity Reb A using a single reflux and single stirred
wash and a mass to volume ratio of 1 to 4. For lower quality Stevia
starting material, with a target purity of 99+% purity Reb A final
product, a preferred approach is to use a mass to volume ratio of 1
to 4 and as many refluxes as necessary ("selected reflux iteration
count") to obtain about a 96% to 98% purity Reb A final product in
an EtOH formulation stage. An EtOH formulation stage with the
selected EtOH iteration count, plus a stirred wash stage, is then
performed to determine if adding a final stirred wash will provide
the target purity; if not, the options are to add another reflux
stage or to adjust the mass to solvent ratio (using a mass to
solvent ratio assay) in one or more EtOH formulation stages, and
again determine if the target purity has been reached. As Example 7
shows, if the target purity is 100% purity Reb A final product
using RA82.3 Stevia starting material, and the target purity cannot
be reached using two refluxes, a single stirred wash, and a mass to
volume ratio of 1:4, increasing the mass to volume ratio in the
refluxes to 1:6 can provide the target purity. Lower quality Stevia
starting material is considerably cheaper than higher quality
Stevia starting material. In locales in which the cost of ethanol
and energy is low, lower quality Stevia starting material can be
processed to 99+% purity Reb A cost effectively. In countries in
which the cost of ethanol and energy is high, higher quality Stevia
starting materials are typically required for cost
effectiveness.
[0080] A reflux mixture temperature of between 79.degree. C. to
80.degree. C. typically produces the shortest acceptable reflux
time (typically one hour). Reflux mixture temperatures lower than
79.degree. C. can be used; however, such lower reflux mixture
temperatures typically require longer durations of reflux and
produce considerably lower yields of Reb A. For example, as the
reflux mixture temperature is increased above 50.degree. C., the
reflux mixture becomes opalescent as Reb A begins to precipitate;
the reflux mixture could be maintained in the 50.degree. C. to
60.degree. C. range, and depending on duration of reflux, could
reduce heating costs, but the yield of Reb A would be considerably
reduced compared with a reflux mixture temperature of between
79.degree. C. to 80.degree. C. In some locales, however, e.g.,
locales using solar heating, only lower reflux mixture temperatures
may be feasible.
[0081] Uses of the Reb A Product. The 99+% purity Reb A produced by
the invention ("Reb A Product"), can be used as a sole sweetener of
food, beverage, medicine, tobacco, pharmaceutical, and personal
care products, or mixed with other sweeteners in such products
(i.e., as a "co-sweetener"). More than one sweetener is typically
used to obtain specific taste profiles and/or physical properties.
Such sweeteners include conventional sweeteners (cane sugar, beet
sugar, honey, syrups, and other "natural" sweeteners) and
high-intensity sweeteners (cyclamates, saccharin, sucralose,
aspartame, Stevia, and other chemically produced and/or natural
high-intensity sweeteners).
[0082] To modify the perceived sweetness of orally consumable
compositions containing the Reb A Product, Reb A Product can be
modified by the addition of taste modifying moieties, such as
galactosides. For instance, .beta.-1,4-galactosyl can be
substituted on the Reb A Product using a .beta.-1,4-galactosyl
transferase enzyme in reactions known in the art. The 99+% purity
Reb A produced by the invention that has been modified by one or
more functional groups to produce a compound with pleasant sensory
perceptions when used in orally consumed compositions is included
in the term "Reb A Product".
[0083] For use as a co-sweetener, the Reb A Product can be used in
ways known in the art of sweeteners (e.g., steam, ethanol, or
alkanol aerosolized Reb A Product vapor-deposited on a
co-sweetener) to coat or permeate other solid sweeteners, such
granular and powdered sugar and artificial sweeteners, to be mixed
as a separate powder with such solid sweeteners, to be
co-crystallized with other solid sweeteners, or to be suspended or
dissolved in liquid sweeteners, such as corn syrup and honey.
Commercially available spray dryers used in the ethanol purge and
drying stage of the industrial embodiment can typically be
configured to produce a particulate size of Reb A Product
appropriate for an intended use.
[0084] The term "flavor" or "flavor characteristic", as used
herein, is the combined sensory perception of the components of
taste, odor, and/or texture. The term "enhance", as used herein,
includes augmenting, intensifying, accentuating, magnifying, and
potentiating the sensory perception of a flavor characteristic
without changing the nature or quality thereof The term "modify",
as used herein, includes altering, varying, suppressing,
depressing, fortifying and supplementing the sensory perception of
a flavor characteristic where the quality or duration of such
characteristic was deficient.
[0085] In the art of flavoring foodstuffs and medicinal
compositions, there is a continuing need for compositions which can
modify and improve the flavor of such materials, because acceptance
and demand for foodstuffs and medicinal products is generally
related to the sensory perception of them. In the art of flavoring
oral hygiene compositions, such as mouthwash and toothpaste, and in
the art of flavoring chewing compositions, such as chewing tobacco,
snuffs and chewing gum, there is a need to improve the flavor
characteristics of such chewing compositions with flavor modifiers
or enhancers which are non-cariogenic and do not support the growth
of tooth decay producing streptococci, lactobaccilli, or the like.
Likewise, there is need to improve the flavor characteristics of
smoking compositions.
[0086] The term "orally consumable composition" includes
foodstuffs, medicinal compositions, smoking compositions, chewing
compositions and oral hygiene compositions, including mouthwashes
and toothpastes. The term "foodstuff" includes both solid and
liquid ingestible materials which usually do, but need not, have a
nutritional value and are intended for consumption by man or
animal. Representative examples of foodstuff include coffee, teas,
herbal teas, baked goods, natural and synthetic flavors, spices,
condiments, soups, stews, convenience foods, beverages (both
carbonated and non-carbonated), dairy products, candies,
vegetables, cereals, fruits, fruit drinks, snacks, cocoa products,
chocolates, animal feed, and the like. The term "medicinal
composition" includes solids, gases and liquids which are
ingestible materials having medicinal value, such as cough syrups,
cough drops, medicinal sprays, vitamins, and chewable medicinal
tablets. The term "chewing compositions" include chewing tobacco,
smokeless tobacco, snuff, chewing gum and other compositions which
are masticated and subsequently expectorated. Chewing gum includes
compositions which comprise a substantially water-insoluble,
chewable gum base, such as chicle or substitutes therefor,
including jetulong, guttakay rubber or certain comestible natural
synthetic resins or waxes. The term "oral hygiene compositions"
includes mouthwashes, mouth rinses, toothpastes, tooth polishes,
dentifrices, mouth sprays, and mouth refreshers. The term "smoking
composition", as used herein, includes cigarette, pipe and cigar
tobacco, and all forms of tobacco such as shredded filler, leaf,
stem, stalk, homogenized leaf cured, reconstituted binders, and
reconstituted tobacco from tobacco dust, fines, or other sources in
sheet, pellet or other forms. "Smoking compositions" also include
tobacco substitutes formulated from non-tobacco materials, such as
representative tobacco substitutes described in U.S. Pat. Nos.
3,529,602, 3,703,177 and 4,079,742 and references cited
therein.
[0087] In accordance with one embodiment of this invention, an
orally consumable composition having flavor enhanced or modified by
the Reb A Product is provided. The Reb A Product can modify or
enhance flavor characteristics that are sweet, fruity, floral,
herbaceous, spicy, aromatic, pungent, "nut-like" (e.g., almond,
pecan), "spicy" (e.g., cinnamon, clove, nutmeg, anise and
wintergreen), "non-citrus fruit" flavor (e.g., strawberry, cherry,
apple, grape, currant, tomato, gooseberry and blackberry), "citrus
fruit" flavor (e.g., orange, lemon and grapefruit), and other
useful flavors, including coffee, cocoa, peppermint, spearmint,
vanilla and maple.
[0088] In accordance with one variation of this embodiment, an
orally consumable composition comprises a Reb A Product in an
amount effective to sweeten or to modify or enhance the taste, odor
and/or texture of the orally consumable composition.
[0089] The terminology "amount effective" or "effective amount"
means an amount that produces a sensory perception. The use of an
excessive amount of a Reb A Product will produce sweetness that may
not be desired for flavor modification or enhancement, just as too
much sugar can be added to a foodstuff or beverage. The amount of
Reb A Product employed can vary over a relatively wide range,
depending upon the desired sensory effect to be achieved with the
orally consumable composition and the nature of the initial
composition.
[0090] Reb A Product can be added to an orally consumable
composition by admixing the Reb A Product with the orally
consumable composition or admixing the Reb A Product with a
component of the orally consumable composition.
[0091] The Reb A Product can be used in tobacco and tobacco-related
products selected from the group comprising cigarettes, cigars,
snuffs, chewing tobacco, other tobacco goods, filters, smoking
papers, and other smoking compositions. A smoking composition
having a sweetened, enhanced, or modified flavor comprises a
smoking filler material selected from the group consisting of
tobacco, reconstituted tobacco, non-tobacco substitutes and
mixtures thereof, and containing an effective amount of Reb A
Product. "Containing" means both being included as an ingredient
and being adsorbed to a material. In one variation of this
embodiment, the smoking composition comprises a filter means
containing a Reb A Product. The term "filter means", as used
herein, includes a smoking device means such as a cigar or
cigarette holder having a filtering or flavoring module
incorporated therein and includes acetate, cotton, charcoal and
other fiber, flake or particle filtering means. In another
variation of this embodiment, the smoking composition comprises a
wrapper means containing a Reb A Product. In one variation of this
embodiment of this invention, 0.003 to 0.30 parts by weight of a
Reb A Product is added to 100 parts by weight of the smoking filler
material. In a preferred variation of this embodiment of this
invention, 0.015 to 0.30 parts by weight of a Reb A Product is
added to 100 parts of a weight of a smoking filler material.
[0092] Those skilled in the art of flavoring tobacco understand
that the effective amount of the Reb A Product added to a smoking
composition may depend upon the method in which the Reb A Product
is added to the smoking composition and to which portion of the
smoking composition Reb A Product is added. Reb A Product can be
added directly to the smoking filler material, to the filter means,
or to the wrapper means of a smoking composition. Reb A Product can
be added to a filter means of a smoking composition by any manner
known to those skilled in the art of flavoring filter means,
including but not limited to, incorporating the Reb A Product among
the fibers, flakes or particles of a filter means, filling the Reb
A Product between two or more layers of fibers of a fiber filter
means to form a triple filter means, or inserting the Reb A Product
into a smoking device means, such as a cigarette holder.
[0093] It is apparent to those skilled in the art that only a
portion of the smoking filler material or filter means need be
treated with a Reb A Product, since blending or other operations
may be used to adjust the final or ultimate smoking composition
within the effective or desired ranges of concentration of Reb A
Product. In addition to Reb A Product, other flavorings or aroma
additives known in the smoking composition flavoring art may be
used with Reb A Product and added along with Reb A Product to the
smoking composition. Representative flavorings used in the smoking
composition flavoring art include ethyl acetate, isoamyl acetate,
propyl isobutyrate, isobutyl butyrate, ethyl butyrate, ethyl
valerate, benzyl formate, menthol, limonene, cymene, pinene,
linalool, geraniol, citroneilol, citral, peppermint oil, orange
oil, coriander oil, lemon oil, borneol, cocoa extract, tobacco
extract, licorice extract and fruit extractives.
[0094] Reb A Product, in its purified state after spray drying, is
generally a fine powder, having a particle size in the range of
about 1 to 100 microns. Fine powders are difficult to handle and
difficult to admix with orally consumable compositions, such as tea
leaves, tobacco products, herb leaves, coffees and other orally
consumable compositions. Also, generally, only a relatively small
amount of Reb A Product is used with an orally consumable
composition when the Reb A Product is used as a flavor modifier or
enhancer, sweetener, or co-sweetener.
[0095] In accordance with another embodiment of this invention, a
process for adding Reb A Product to an orally consumable
composition comprises admixing Reb A Product with a carrier to form
a Reb A Product-carrier mixture. Preferred carriers include water,
ethanol, other alkanols used in food processing, or mixtures
thereof. The Reb A Product solution so formed is contacted with an
orally consumable composition, and the carrier is removed from the
orally consumable composition by evaporation, or otherwise, and the
Reb A Product residues deposited with the orally consumable
composition. This process is particularly useful for adding Reb A
Product to tea leaves, herbal plant leaves, and other sweeteners,
particularly granular sucrose (table sugar).
[0096] In accordance with still another embodiment of this
invention, a liquid filter material, suitable for use with an
orally consumable composition, is prepared with Reb A Product. The
term "liquid filter", as used herein, refers to a porous or
semi-porous filter material used for preparation of an orally
consumable composition such as a tea bag, a coffee filter or a
filter disk. The term "filter disk" refers to a porous or
semi-porous inactive article added to an orally consumable
composition for the purposes of acting as a vehicle for the
addition of a flavoring or sweetening composition to the orally
consumable composition. A process for preparing a liquid filter
comprising a filter material and Reb A Product is typically by
admixing Reb A Product with a carrier to form a Reb A
Product-carrier mixture; contacting the Reb A Product-carrier
mixture with the filter material; and removing the carrier from the
filter material thereby depositing a Reb A Product residue on the
filter material.
[0097] The Reb A Product can be used in beverages, broths, and
beverage preparations selected from the group comprising
carbonated, non-carbonated, frozen, semi-frozen ("slush"),
non-frozen, ready-to-drink, concentrated (powdered, frozen, or
syrup), dairy, non-dairy, herbal, non-herbal, caffeinated,
non-caffeinated, alcoholic, non-alcoholic, flavored, non-flavored,
vegetable-based, fruit-based, root/tuber/corm-based, nut-based,
other plant-based, cola-based, chocolate-based, meat-based,
seafood-based, other animal-based, algae-based, calorie enhanced,
calorie-reduced, and calorie-free products, optionally dispensed in
open containers, cans, bottles or other packaging. Such beverages
and beverage preparations can be in ready-to-drink, ready-to-cook,
ready-to-mix, raw, or ingredient form and can use the Reb A Product
as a sole sweetener or as a co-sweetener.
[0098] The Reb A Product can be used in foods and food preparations
(e.g., sweeteners, soups, sauces, flavorings, spices, oils, fats,
and condiments) selected from the group comprising dairy-based,
cereal-based, baked, vegetable-based, fruit-based,
root/tuber/corm-based, nut-based, other plant-based, egg-based,
meat-based, seafood-based, other animal-based, algae-based,
processed (e.g., spreads), preserved (e.g., meals-ready-to-eat
rations), and synthesized (e.g., gels) products. Such foods and
food preparations can be in ready-to-eat, ready-to-cook,
ready-to-mix, raw, or ingredient form and can use the Reb A Product
as a sole sweetener or as a co-sweetener.
[0099] The Reb A Product can be used in candies, confections,
desserts, and snacks selected from the group comprising
dairy-based, cereal-based, baked, vegetable-based, fruit-based,
root/tuber/corm-based, nut-based, gum-based, other plant-based,
egg-based, meat-based, seafood-based, other animal-based,
algae-based, processed (e.g., spreads), preserved (e.g.,
meals-ready-to-eat rations), and synthesized (e.g., gels) products.
Such candies, confections, desserts, and snacks can be in
ready-to-eat, ready-to-cook, ready-to-mix, raw, or ingredient form,
and can use the Reb A Product as a sole sweetener or as a
co-sweetener.
[0100] The Reb A Product can be used in prescription and
over-the-counter pharmaceuticals, assays, diagnostic kits, and
therapies selected from the group comprising weight control,
nutritional supplement, vitamins, infant diet, diabetic diet,
athlete diet, geriatric diet, low carbohydrate diet, low fat diet,
low protein diet, high carbohydrate diet, high fat diet, high
protein diet, low calorie diet, non-caloric diet, oral hygiene
products (e.g., toothpaste, mouthwash, rinses, floss, toothbrushes,
other implements), personal care products (e.g., soaps, shampoos,
rinses, lotions, balms, salves, ointments, paper goods, perfumes,
lipstick, other cosmetics), professional dentistry products in
which taste or smell is a factor (e.g., liquids, chewables,
inhalables, injectables, salves, resins, rinses, pads, floss,
implements), medical, veterinarian, and surgical products in which
taste or smell is a factor (e.g., liquids, chewables, inhalables,
injectables, salves, resins, rinses, pads, floss, implements), and
pharmaceutical compounding fillers, syrups, capsules, gels, and
coating products.
[0101] The Reb A Product can be used in consumer goods packaging
materials and containers selected from the group comprising plastic
film, thermoset and thermoplastic resin, gum, foil, paper, bottle,
box, ink, paint, adhesive, and packaging coating products.
[0102] The Reb A Product can be used in goods selected from the
group comprising sweeteners, co-sweeteners, coated sweetener
sticks, frozen confection sticks, medicine spoons (human and
veterinary uses), dental instruments, pre-sweetened disposable
tableware and utensils, sachets, edible sachets, pot pourris,
edible pot pourris, hotch potches, edible hotch potches, artificial
flowers, edible artificial flowers, clothing, edible clothing,
massage oils, and edible massage oils.
EXAMPLES
[0103] HPLC Method
[0104] High performance liquid chromatography ("HPLC") was used to
determine the purity of the final products produced by the methods
of the invention. In the HPLC analysis method used in the Examples
and in the description above, 0.1 grams of dried solids (e.g., the
final product produced using the invention) is dissolved in 25 mls
of 50% acetonitrile/50% water to create a sample. The sample is
passed through a 0.45 .mu.m nylon filter and 10 .mu.l of filtrate
injected into the HPLC injection port. The HPLC apparatus comprises
a Shimadzu LC-10AS pump (Shimadzu North America,
www.ssi.shimadzu.com), a Shimadzu SCL-10A controller and a Shimadzu
SPD-10A variable wavelength detector. Raw data was collected on a
Shimadzu CR501 Chromatopac integrator. A 4.2 mm i.d..times.250 mm,
10 .mu.m amine column was used (Alltech and Associates Inc.,
www.alltech.com) with an isocratic mobile phase consisting of 72.5%
acetonitrile/27.5% water and a flow rate of 0.8 ml/min. Peaks were
detected at 210 nm. Glycosides (stevioside and rebaudioside A) were
quantified by comparison with a standard linear regression curve
(y=mx+b) constructed from known concentrations of either stevioside
(Chromadex Catalog No. ASB-00019351) or rebaudioside A (Chromadex
Catalog No. ASB-00018226), respectively. The reference standards
were used to analyze Stevia starting materials and solids produced
by the invention. Stevioside and rebaudioside C were at the lower
limits of detection, as described below, in the final product
produced using the invention. The final product described in the
invention has a higher purity than the current primary analytical
standard for rebaudioside A, which has a certified purity of 98.7%.
(Chromadex, Inc., Santa Ana, Calif., www.chromadex.com) The
Chromadex standard was established in approximately 2002. The
higher purity of the Reb A produced by the invention arises
primarily from use of the EtOH formulation stage and a selected
EtOH solvent, as described above. The phrase "100% pure Reb A"
means that the chromatograph from the above-described high
performance liquid chromatographic ("HPLC") analysis of a sample of
final product produced using the invention shows miniscule blips
(by retention time) of stevioside and rebaudioside C, a single
strong peak of Reb A, and the HPLC integrator reports a value equal
to or greater than the lowest value (typically, 98.7%) at 210 nm
reported by the HPLC integrator for a sample of only the Chromadex
reference standard for pure Reb A. Many Reb A reference standards
for purity in general use prior to 2002 have been found to
overstate Reb A purity; for instance, it is believed that the
standard in common use in Canada before 2002 overstates Reb A
purity by approximately 6% compared with analytic results using the
Chromadex standard.
[0105] Since absorption at 210 nm in an HPLC apparatus varies based
on the nature of the eluted compound and the compound's
chromophore(s), using analytic methods other than standard linear
regression curves constructed from reference compounds of known
purity (e.g., Chromadex Catalog No. ASB-00018226) will produce
erratic results.
Example 1
[0106] Example 1. Two reflux stages using a series of test solvents
followed by an absolute ethanol reflux.
[0107] Example 1 demonstrates the effect of water content in reflux
solvent on solids yield and rebaudioside A purity. In Example 1, a
series of six solvents was formulated containing 0%, 2%, 4%, 6%,
8%, and 10% water and the remaining percentage ethanol
(volume/volume); each blend of solvent is called a "test solvent".
Three sets of six, 5 gram samples of Stevia starting material
containing (i) 71.85% RA, 13.55% stevioside; (ii) 76.13% RA, 6.96%
stevioside; and (iii) 81.58% rebaudioside A, 4.49% stevioside (the
remainder in each consisted of various concentrations of
dulcosides, rebaudiosides, and sterebin compounds) were prepared by
blending two commercially available Stevia extracts, RA81.58 and
RA40.27. All three sets of blends of Stevia starting material
(each, a "Stevia blend") were dried in a blower oven at 80.degree.
C. overnight and were assumed to contain 0% moisture before
processing.
[0108] 5 grams of each Stevia blend were each placed in three sets
of six, 100 ml glass round-bottomed flasks and 20 mls of each test
solvent added, respectively, to the sets of flasks to create a
100Et/00W, 98Et/02W, 96Et/04W, 94Et/06W, 92Et/08W, and 90Et/10W
"test mixtures" of each Stevia blend. The test mixtures were mixed
at room temperature. With the exception of the 100Et/00W test
mixtures, all of the Stevia starting material went into solution.
Each test mixture was refluxed at 80.degree. C. for 1 hour. Each
test mixture was then cooled on ice for 1 hour. Each test mixture
was then filtered over a Whatman GF/A glass microfiber filter (1.6
.mu.m pore size). The solids from filtration of each test mixture
("first retentate") were collected and each first retentate mixed
with a further 20 mls. of the same test solvent used for the first
reflux of the corresponding test mixture to produce a "second test
mixture". Each second test mixture was refluxed at 80.degree. C.
for one hour. Each second test mixture was than stored at 4.degree.
C. for 96 hours. Each second test mixture was then filtered over a
Whatman GF/A glass microfiber filter to produce a retentate
("second retentate") of wet solids and each second retentate was
dried in a blower oven at 80.degree. C. until a constant weight was
reached. Each gram of the dried second retentate from a given
second test mixture was mixed with 4 mls of absolute ethanol in a
100 ml glass round-bottomed flask to create a "third test mixture".
Each third test mixture was refluxed for one hour at 80.degree. C.
Each third test mixture was then filtered over a Whatman GF/A glass
microfiber filter, the solids ("third retentate") collected and
then dried in a blower oven at 80.degree. C. until a constant
weight was observed. To measure yield, the weight of each dried
third retentate was determined. To measure purity, a 0.1 gram
sample of each third retentate was analyzed for rebaudioside A and
stevioside content using the HPLC Method described above.
[0109] Results:
[0110] Table 1, shows the effect of water content in test solvent
on solids yield from RA71.85, RA76.13, and RA81.58 Stevia starting
materials.
[0111] Graph 1 in FIG. 13 plots the effect of varying water content
in test solvent on solids yield from RA71.85, RA76.13, and RA81.58
Stevia starting materials.
[0112] It can be seen from Table 1 that as the water content in the
test solvent increased up to 8%, the solids yield declined
gradually. However, when water content in the test solvent exceeded
8%, in all cases the solids yield declined significantly and at a
much greater rate than with lower water contents. This suggests
that to maintain yield with RA70 or higher Stevia starting
material, the maximum water content should not exceed 8% by volume.
Reb A yield was positively correlated to rebaudioside A content of
the Stevia starting material: a higher initial Reb A content gave a
higher solids yield in all test solvents. Moisture in Stevia
starting material contributes to the total water content of a test
solvent and will consequently have an effect on solids yield.
Therefore, to control total water content in the text mixtures, all
Stevia starting material was dried before formulating the test
mixtures.
[0113] Table 2 shows the effect of water content in test solvent on
Rebaudioside A purity from RA71.85, RA76.13, and RA81.58 Stevia
starting materials.
[0114] Graph 2 in FIG. 14 plots the effect of water content in test
solvent on rebaudioside A purity from RA71.85, RA76.13, and RA81.58
Stevia starting materials.
[0115] It can seen from Table 2 and Graph 2 that Reb A purity of
the final product generally increased with increasing water content
in the test solvent. The exception was 2% water content, where
there was either a decrease in Reb A purity or only a very small
increase in Reb A purity compared with having no water in the test
solvent. To obtain final product containing 99%+ purity Reb A,
Table 2 shows (bold figures) that for RA81.58 Stevia starting
material, 6% moisture in the test solvent gave the desired 99+% Reb
A purity with the highest yield. With RA76.13 Stevia starting
material, 8% water content gave the desired 99+% Reb A purity with
the highest yield. With RA71.85 Stevia starting material, 10% water
content gave the desired 99+% Reb A purity with the highest yield.
However, as shown by the data from Table 1, reaching the desired
99+% Reb A purity with lower quality Stevia starting material
results in lower absolute yield.
[0116] Example 1 illustrates the need to empirically determine the
optimal water content in the test solvent for each Stevia starting
material used. These empirical observations led to the starting
material assay and the reflux solvent approximation described
above.
Example 2
[0117] Example 2. One reflux stage using a series of test solvents
followed by a stirred wash stage.
[0118] Example 2 demonstrates that the second reflux in Example 1
can be eliminated and reflux with absolute ethanol can be replaced
with a shorter duration stirred wash in absolute ethanol when using
RA80 or higher Stevia starting material (if the Stevia starting
material does not contain "stubborn" contaminants that require an
additional reflux stage). The results of Example 2 show that using
RA90, a single reflux stage and single wash stage typically produce
99+% purity Reb A.
[0119] In Example 2, as in Example 1, a series of six solvents was
formulated containing 0%, 2%, 4%, 6%, 8%, and 10% water and the
remaining percentage ethanol (volume/volume); each blend of solvent
is called a "test solvent". 5 grams of RA89.95 Stevia starting
material were placed in each of six 100 ml glass round-bottomed
flasks and mixed with 20 mls of 100Et/00W, 98Et/02W, 96Et/04W,
94Et/06W, 92Et/08W, and 90Et/10W test solvents to create 100Et/00W,
98Et/02W, 96Et/04W, 94Et/06W, 92Et/08W, and 90Et/10W "test
mixtures" of the Stevia starting material went into solution. Each
test mixture was refluxed at 80.degree. C. for 1 hour. Each test
mixture was then held at room temperature for 96 hours and then
filtered over a Whatman GF/A glass microfiber filter (1.6 .mu.m
pore size). The solids in each test mixture ("prime retentate")
were collected and each prime retentate mixed with 16 mls of
absolute ethanol and stirred (stir bar) for 15 minutes at room
temperature ("stirred wash"). The assumption was made that 20%
(i.e., 1 gram) of the original solids remained in solution after
the first reflux of Example 2 and 96 hour stand time, therefore 16
mls of EtOH solvent were use for each stirred wash to maintain the
1:4 mass to volume ratio of the test mixtures. After stirring, each
test mixture was filtered over a Whatman GF/A glass microfiber
filter and the solids ("stirred wash retentate") collected and then
dried in a blower oven at 80.degree. C. until the stirred wash
retentates maintained a constant weight. Each dried, stirred wash
retentate was weighed and a 0.1 g sample analyzed for rebaudioside
A content using the HPLC Method described above.
[0120] Results:
[0121] Table 3 shows the final solids (dried, stirred wash
retentate) weight, rebaudioside A recovery, and purity after a
single EtOH reflux and stirred wash using RA89.95 Stevia starting
material.
[0122] The method of Example 2 was repeated using RA81.67 Stevia
starting material instead of RA89.95 Stevia starting material,
ceteris paribus, with results shown in Table 4.
[0123] Table 4 shows the final solids (dried retentate from stirred
wash stage) weights, rebaudioside A recovery, and Reb A purity
after a single EtOH reflux and stirred wash using RA81.67 Stevia
starting material. It can be seen in Table 3 that for higher
quality Stevia starting material (RA90), a single EtOH reflux,
followed by a short wash in absolute ethanol, is sufficient to
obtain 100% pure Reb A final product. With higher quality Stevia
starting material, there is no need for a second EtOH reflux or a
reflux in absolute ethanol. A 15 minute wash in absolute ethanol is
sufficient.
[0124] However, the same methodology (one EtOH reflux, one stirred
wash) with approximately 10% lower Reb A purity Stevia starting
material (i.e., RA80) does not result in a final product with
rebaudioside A content greater than 96.24%. Moreover, the yield
using RA80 Stevia starting material is approximately 10% less
(Table 4). Therefore for lower quality Stevia starting material, it
is necessary to use an additional reflux stage to produce 99+%
purity Reb A.
Example 3
[0125] Example 3. One reflux stage, without stirred wash stage.
[0126] Example 3 demonstrates the effect of omitting the stirred
wash stage on rebaudioside A purity of the final product.
[0127] In Example 3, two, 5 gram samples of RA89.95 Stevia starting
material were placed in two 100 ml glass round-bottomed flasks,
respectively, and each mixed with 20 mls of a 92Et/08W test solvent
to form a test mixture (this experiment was conducted while the
starting material assay was still being perfected). Each test
mixture was refluxed for 1 hour at 80.degree. C. Both test mixtures
were left at room temperature for 96 hours and then filtered over a
Whatman GF/A glass microfiber filter (1.6 .mu.m pore size). The
solids for each test mixture ("reflux retentate") were collected.
One reflux retentate was dried in a blower oven at 80.degree. C.
("non-stirred retentate") and the other reflux retentate was mixed
with 16 mls of 100% ethanol, stirred (stir bar) for 15 minutes at
room temperature, then filtered over a Whatman GF/A glass
microfiber filter, and the solids ("stirred retentate") collected.
The non-stirred retentate and the stirred retentate were dried in a
blower oven at 80.degree. C. until the retentates maintained a
constant weight. 0.1 gram samples from both retentates were then
analyzed for rebaudioside A content using the HPLC Method described
above.
[0128] Results:
[0129] The dried, non-stirred retentate contained 96.32%
rebaudioside A. The dried, stirred retentate contained 100%
rebaudioside A. The 15 minute stirred wash step in absolute
ethanol, in addition to a single EtOH reflux, is therefore
necessary to obtain the desired 99+% Reb A purity of final product
using an RA90 Stevia starting material.
Example 4
[0130] Example 4. Two reflux stages with ethanol purge and drying
stage of industrial process in tandem with stirred wash stage of
laboratory process.
[0131] Example 4 demonstrates that, using higher quality Stevia
starting material, a 15 minute stirred wash in absolute ethanol is
sufficient to obtain 99+% Reb A final product with desired
solubility. In Example 4, RA89.95 Stevia starting material was
processed through two reflux stages of the industrial embodiment of
the invention, and a sample of the final product of the industrial
process used as starting material in a stirred wash stage of the
laboratory embodiment of the invention. Specifically, 100 kilograms
of RA90 Stevia starting material were mixed with 400 liters of
90Et/10W solvent (this experiment was conducted while the starting
material assay was still being perfected) and refluxed for 1 hour
at 80.degree. C., then cooled (chilled water in tank jacket) and
allowed to stand for a minimum of 2 hours with mixture temperature
maintained between 4.degree. C. and 20.degree. C. The mixture was
filtered through a centrifugal decanter, and the retentate added to
400 liters of 90Et/10W solvent and refluxed for 1 hour at
80.degree. C. The mixture from the second reflux was filtered
through a centrifugal decanter. The retentate of the filtering of
the second reflux mixture was mixed with water until the volume of
the mixture was approximately twice the mass of the retentate
(e.g., 350 kg retentate in a total mixture volume of 700 liters).
The mixture in water was heated to 80.degree. C. for approximately
one hour to purge any residual ethanol from the mixture, then the
mixture was spray-dried. There was no reflux in absolute ethanol
prior to spray drying. The spray-dried product was weighed and a
0.1 gram sample was analyzed for rebaudioside A content using the
HPLC Method described above. Five grams of the spray-dried material
was mixed with 20 mls of absolute ethanol in a 100 ml
round-bottomed flask and refluxed for 1 hour at 80.degree. C.
("refluxed sample") and then allowed to cool to room temperature. A
second 5 gram sample of the spray-dried material was mixed with 20
mls of absolute ethanol and stirred for 15 minutes at room
temperature ("stirred sample"). Both samples were filtered over a
Whatman GF/A glass fiber filter (1.6 .mu.m pore size) and the
retained solids of both samples were dried in a blower oven at
80.degree. C. until the solids maintained a constant weight. 1 gram
each of the refluxed sample and of the stirred sample were placed
in 15 ml glass centrifuge tubes. Water was added to each tube in
0.1 ml increments and the mixture thoroughly shaken to ensure
maximum solubilization. Incremental aliquots of water were added
until complete solubilization was achieved as evidenced by the lack
of any observable suspended solids in the solution. The final
volume at which complete solubilization occurred in each sample was
noted.
[0132] Results:
[0133] Table 5 shows the rebaudioside A purity and solids yield
achieved using two reflux stages, with ethanol purge and drying
stage, of the industrial embodiment of the invention.
[0134] Table 6 shows the solubility in water of the refluxed sample
and of the stirred sample.
[0135] Using two reflux stages, in tandem with an ethanol purge and
drying stage, of the industrial embodiment of the invention
produced a 99.6% Reb A purity product. However, solubility of that
product in water at room temperature was only 2%. This solubility
may be useful for some commercial applications. The increased
solubility (40%, at room temperature) afforded by the 1 hour reflux
in absolute ethanol would meet the solubility requirements of most,
if not all, commercial applications. It was found, however, that
replacing the 1 hour reflux with a 15 minute stirred wash in
absolute ethanol at room temperature resulted in a product with 33%
solubility (at room temperature), which is acceptable for most
commercial needs, including concentrates.
Example 5
[0136] Example 5. One reflux stage and stirred wash stage, with
reduced standing times and without cooling.
[0137] Example 5 demonstrates the effects of a reduction in
standing time and omission of cooling on purity and yield of
rebaudioside A.
[0138] In Example 5, five, 5 gram samples of RA81.58 Stevia
starting material and five, 5 gram samples of RA89.95 Stevia
starting material were placed in two sets of five, 100 ml
round-bottomed glass flasks, respectively, and each sample mixed
with 20 mls of 92Et/08W solvent (each, a "test mixture"). The test
mixtures were refluxed for 1 hour at 80.degree. C. After the
reflux, the five test mixtures in a given set were allowed to stand
at room temperature (without stirring) for 0, 18, 24, 36, and 160
hours, one test mixture per set per standing time. Each test
mixtures was filtered over a Whatman GF/A glass microfiber filter
(1.6 .mu.m pore size) and each retentate collected and mixed with
16 mls of absolute ethanol (each, a "test sample") and stirred
(stir bar) at room temperature for 15 minutes. After the relevant
standing time, a given test sample was then filtered over a Whatman
GF/A glass microfiber filter and the retentate collected and dried
at 80.degree. C. until the solids maintained a constant weight.
Each sample was then weighed and a 0.1 gram sample analyzed for Reb
A content using the HPLC Method described above.
[0139] Results:
[0140] Table 7 shows yield and rebaudioside A purity of final
product after various standing times.
[0141] For the RA81.58 purity Stevia starting material, a minimum
standing time of 18 hours allowed maximum solids yield to be
achieved. For the RA89.95 purity Stevia starting material, zero
standing time produced 99+% purity Reb A final product. For lower
RA purity Stevia starting material, increasing the standing time
after a reflux stage increases the yield of final product (multiple
refluxes are typically required to produce 99+% purity Reb A when
using lower quality Stevia starting material).
Example 6
[0142] Example 6. One reflux stage and two stirred wash stages with
ethanol purge and drying stage, industrial scale.
[0143] Example 6 demonstrates that a single reflux and two stirred
washes, with spray drying, can produce 99+% purity rebaudioside A
with acceptable water solubility at industrial scale when using
lower quality Stevia starting material.
[0144] In Example 6, five hundred (500) kilograms of RA83.14 Stevia
starting material were mixed with 2000 liters of 92Et/08W EtOH
solvent and refluxed for 1 hour at 79.degree. C., then cooled
(chilled water in tank jacket) to 29.degree. C. (elapsed time from
end of reflux to start of decanting, 7 hrs., 15 min.). The cooled,
refluxed mixture was filtered through a centrifugal decanter, and
the retentate added to 1,600 liters of ethanol (retentate mass of
400 kg, 1:4 mass to volume ratio) at room temperature (29.degree.
C). and stirred for 15 minutes (stirred wash stage). The stirred
wash mixture was filtered through a centrifugal decanter, and the
stirred wash stage was then repeated, i.e., the retentate from the
centrifugal decanter was added to 1,600 liters of ethanol (1:4 mass
to solvent ratio) at room temperature (29.degree. C). and stirred
for 15 minutes (second stirred wash stage). The second stirred wash
mixture was filtered through a centrifugal decanter, and the
retentate was mixed with water until the volume of the mixture was
approximately twice the mass of the retentate (e.g., 800 kg
retentate in a total mixture volume of 1,600 liters). In the
ethanol purge and drying step, the aqueous mixture was heated to
80.degree. C. and maintained at that temperature for approximately
ten hours to purge any residual ethanol from the mixture, then the
mixture was spray-dried. Due to limited throughput of the spray
dryer, the spray drying process required 6.5 hours. From the 500 kg
of Stevia starting material, 302.3 kg of solids were recovered, or
60.46% of the Stevia starting material by dry weight. The 500 kg of
Stevia starting material were RA83.14, i.e., contained 415.7 kg of
Reb A. A recovery of 302.3 kg of solids represents a recovery of
72.3% of the Reb A in the Stevia starting material. The final
product was 99.4% purity Reb A, determined using the HPLC Method
described above. The solubility was 59.0% (at 25.degree. C.).
Example 7
[0145] Example 7. Effect of varying ratio of solids weight to
solvent volume on purity and yield.
[0146] Example 7 demonstrates the optional mass to solvent ratio
analysis.
[0147] Ten gram samples of oven-dried Stevia starting material
containing 82.3% rebaudioside A were placed in each of five, 100
ml, glass, round-bottomed flasks and 20 mls, 30 mls, 40 mls, 50
mls, and 60 mls of 92Et/08W selected EtOH solvent were added,
respectively, to each flask, which resulted in mass to solvent
ratios of 1:2 through 1:6.
[0148] A magnetic stir bar was added to each flask and the mixtures
refluxed for 1 hour at 80.degree. C. with stirring. The samples
were then cooled by placing them on ice for 1 hour. Each mixture
was then filtered over a Whatman GF/A glass fiber filter. The wet
solids retentate was then washed by adding 40 mls of 100% absolute
ethanol and stirring at room temperature for 15 minutes. The
mixture was then filtered over a Whatman GF/A glass fiber filter
and the wet solids retentate dried at 80.degree. C. in a blower
oven. The resulting dried solids were then weighed and analyzed by
the HPLC method, as described above, for rebaudioside A
content.
[0149] A duplicate set of five, 100 ml, glass, round-bottomed
flasks, each containing 10 grams of RA82.3 Stevia starting material
and the respective volumes of 92Et/08W solvent described above per
flask, were similarly mixed and refluxed for 1 hour at 80.degree.
C., cooled, then filtered. The wet solids retentate was placed back
in the set of five, 100 ml, glass, round-bottomed flasks and the
respective volumes of 92Et/08W solvent per flask added and mixed as
described above. The mixtures were refluxed for 1 hour at
80.degree. C., cooled, filtered, and then washed in 40 mls of
absolute ethanol. Each wash mixture was then filtered and the
resulting wet solids retentate dried in a blower oven at 80.degree.
C. The dried solids were weighed and analyzed by the HPLC method
described above for rebaudioside A content.
[0150] Table 7 below shows the yield and purity data obtained for
each sample that was processed using one reflux stage and a stirred
wash stage.
TABLE-US-00001 TABLE 7 Using 92Et/08W selected EtOH solvent Number
of Refluxes/ Mass:Solvent Ratio % Reb A % Solids Yield % Reb A
Yield 1 Reflux/1:2 ratio 88.88 79.7 70.8 1 Reflux/1:3 ratio 90.85
79.7 72.4 1 Reflux/1:4 ratio 93.67 78.6 73.6 1 Reflux/1:5 ratio
96.51 74.8 72.2 1 Reflux/1:6 ratio 97.15 72.8 70.7
[0151] Table 8 below shows the yield and purity data obtained for
each sample that was processed using two reflux stages and a
stirred wash stage.
TABLE-US-00002 TABLE 8 Using 92Et/08W selected EtOH solvent Number
of Refluxes/ Mass:Solvent Ratio % Reb A % Solids Yield % Reb A
Yield 2 Refluxes/1:2 ratio 92.74 65.3 60.6 2 Refluxes/1:3 ratio
97.42 62.4 60.8 2 Refluxes/1:4 ratio 99.17 59.9 59.4 2 Refluxes/1:5
ratio 99.03 56.1 55.6 2 Refluxes/1:6 ratio 100.00 51.9 52.2
[0152] During the experimental procedure described above, it was
noted that using a mass to solvent ratio of less than 1:4 resulted
in solidification of the mixture in the flask after reflux. Using a
ratio of less than 1:4 on an industrial scale would therefore
significantly hamper transfer of material from a mixing tank after
reflux. It can be seen from Table 7 above that while a single
reflux and stirred wash generally raised rebaudioside A purity,
even a 1:6 mass to solvent ratio (m/v) was not sufficient to obtain
a final product of 99%+ purity Reb A using this particular RA82.3
Stevia starting material. In addition, with the single reflux and
stirred wash, solids yield declined only 7.1% as the ratio of mass
to solvent ratio increased 3-fold. Reb A yield did not appear to
change linearly with the increasing ratio of solids mass to reflux
solvent. This suggests that increasing the amount of reflux solvent
does not simply dilute the sample and allow a directly
proportionate increase in the amount of solids remaining in
solution after reflux, but that the chemical interactions during
reflux, which allow for preferential precipitation of Reb A, are
more directly linked to the ratio of water to ethanol in the
selected EtOH solvent, as shown in Examples 1 and 2.
[0153] Table 8 shows the results obtained using RA82.3 Stevia
starting material, two refluxes, and a stirred wash. Rebaudioside A
purity in the final product only exceeded 99% when a ratio of 1:4
or higher was used. Solid recovery dropped 13.4% as the mass to
solvent ratio (m/v) increased threefold. This drop in yield was
approximately twice that seen with a single reflux and single
stirred wash, which reflects the use of an additional reflux stage.
In addition, rebaudioside A yield dropped 8.4% over the range of
mass to solvent ratios used (1:2 through 1:6. (m/v)). Notably, a
1:6 mass to solvent ratio gave a 100% pure rebaudioside A final
product from an RA82.3 Stevia starting material.
[0154] Based upon the data in Example 7, a minimum 1:4 mass to
solvent (m/v) ratio is preferred for a target purity of 99+% Reb A.
Higher mass to solvent ratios can be used, but the increased cost
in terms of total solvent consumed, and heating and cooling larger
mixing tanks, must be balanced against the desired target purity.
Example 7 also shows that two reflux stages and an ethanol purge
and drying stage was required to obtain 99+% purity Reb A product
using this particular lower quality Stevia starting material.
[0155] The experience in Example 6 of obtaining 99+% purity Reb A
final product using RA83.14 Stevia starting material, a single
reflux, and two stirred washes, compared with two reflux stages and
single stirred wash stage required in Example 7 using RA82.3 Stevia
starting material, illustrates the importance of the EtOH
formulation stage, with optional mass to solvent ratio assay, in
process optimization (usually, maximizing Reb A yield at 99+%
purity final product while minimizing processing costs).
Example 8
[0156] Example 8 demonstrates processing an RA82.3 Stevia starting
material (the same batch as used in Example 7) using one reflux
stage, in conjunction with two stirred wash stages in which the
mass to solvent ratio is varied.
[0157] To determine the effect of varying the solvent ratio in two
stirred wash stages following a single reflux stage, 10 gram
samples of RA82.3 Stevia starting material were placed in each of
four, 100 ml round-bottomed flasks (magnetic stir bar added). 40
mls of 92Et/08W were added to each flask of the four flasks and the
mixtures stirred to dissolve solids. Each mixture was refluxed for
1 hour at 80.degree. C. with stirring. After reflux, each mixture
was placed on ice for 1 hour and then filtered over Whatman GF/A
glass microfiber filter paper. The wet solids retentate from each
mixture was placed in a 100 ml round-bottomed flask (magnetic stir
bar added) with 30 mls, 40 mls, 50 mls, or 60 mls, respectively, of
100% ethanol and stirred with a magnetic stirrer at room
temperature for 15 minutes. Each mixture then filtered over Whatman
GF/A glass microfiber filter paper to produce four samples of wet
solids retentate. A small sample (approximately 0.1 gram) of each
sample was dried in a blower oven at 80.degree. C. until a constant
weight was reached and then analyzed for rebaudioside A content
using the HPLC Method described above. The remaining wet solids
retentate from each mixture was placed in a 100 ml round-bottomed
flask (magnetic stir bar added) with 30 mls, 40 mls, 50 mls, or 60
mls, respectively, of 100% ethanol and stirred with a magnetic
stirrer at room temperature for 15 minutes. Finally, each mixture
was filtered, and the wet solids retentate of each sample dried at
80.degree. C. until a constant weight was reached. Each sample of
dried solid was weighed and a sample of the dried solids assayed
for rebaudioside A content using the HPLC Method described
above.
TABLE-US-00003 TABLE 9 Number of Refluxes/ Mass:Solvent Ratio % Reb
A % Solids Yield % Reb A Yield 1 reflux 1:4, 1 wash 1:3 89.90 nd*
nd* 1 reflux 1:4, 1 wash 1:4 89.28 nd* nd* 1 reflux 1:4, 1 wash 1:5
90.74 nd* nd* 1 reflux 1:4, 1 wash 1:6 92.85 nd* nd* 1 reflux 1:4,
2 washes 1:3 95.25 64.4 74.54 1 reflux 1:4, 2 washes 1:4 96.99 64.2
75.66 1 reflux 1:4, 2 washes 1:5 95.01 64.0 73.88 1 reflux 1:4, 2
washes 1:6 95.16 63.7 73.65 * no data
[0158] Based on the data in Table 9, inclusion of a second wash
step raised the final rebaudioside A purity between 2.3 and 7.7%,
with the largest increase in purity seen with an additional 1:4
(mass to solvent ratio) wash. However, the target product of 99+%
purity Reb A was not obtained using a single reflux and two stirred
washes, regardless of the mass to solvent ratio in the stirred wash
stages. The final purities after the second stirred wash stage,
regardless of the ratio used, only varied by 1.98%. This suggests
that simply raising the mass to solvent ratio in the stirred wash
stage(s) does not have a significant effect on raising Reb A
purity. Similarly, yield was not significantly affected by the
increasing the mass to solvent ratio in the stirred wash
stages.
Example 9
[0159] Example 9 demonstrates processing an RA82.3 Stevia starting
material (the same batch as used in Example 7) using two reflux
stages in which the mass to solvent ratio was varied, in
conjunction with a final stirred wash stage (1:4 mass to solvent
ratio).
[0160] To determine the effect of varying the solvent ratio in two
reflux stages, a 10 gram sample of RA82.3 Stevia starting material
was placed each of eleven, 100 ml round-bottomed flasks (magnetic
stir bar added). To three of the eleven flasks, 40 mls of 92Et/08W
solvent were added (collectively, the "4x/series"), respectively,
to the flasks and the mixtures stirred to dissolve solids. To four
of the eleven flasks, 50 mls of 92Et/08W solvent were added
(collectively, the "5x/series"), respectively, to the flasks and
the mixtures stirred to dissolve solids. To the remaining four of
the eleven flasks, 60 mls of 92Et/08W solvent were added
(collectively, the "6x/series"), respectively, to the flasks and
the mixtures stirred to dissolve solids. Each mixture was refluxed
for 1 hour at 80.degree. C. with stirring. After reflux, each
mixture was placed on ice for 1 hour and then filtered over Whatman
GF/A glass microfiber filter paper. The wet solids retentate in the
three samples in the 4x/series was mixed with 40 mls, 50 mls, or 60
mls, respectively, of 92Et/08W solvent and the samples refluxed for
a second time at 80.degree. C. with stirring. The wet solids
retentate in the four samples in the 5x/series was mixed with 30
mls, 40 mls, 50 mls, or 60 mls, respectively, of 92Et/08W solvent
and the samples refluxed for a second time at 80.degree. C. with
stirring. The wet solids retentate in the four samples in the
6x/series was mixed with 30 mls, 40 mls, 50 mls, or 60 mls,
respectively, of 92Et/08W solvent and the samples refluxed for a
second time at 80.degree. C. with stirring. Each mixture was placed
on ice for 1 hour and then filtered over Whatman GF/A glass
microfiber filter paper. Finally, the wet solids retentate of each
sample was mixed with 40 mls of absolute ethanol and stirred for 15
minutes at room temperature, then filtered over Whatman GF/A glass
microfiber filter paper. The wet solids retentate from each sample
was then dried to a constant weight at 80.degree. C. in a blower
oven. All samples were then weighed and analyzed for rebaudioside A
using the HPLC Method described above.
TABLE-US-00004 TABLE 10 Sample* % RA % Solids yield % RA yield
4x/4x 100 58.6 71.26 4x/5x 100 56.7 68.85 4x/6x 99.5 44.7 53.99
5x/3x 99.3 49.0 59.12 5x/4x 99.9 60.2 73.03 5x/5x 99.9 48.7 59.10
5x/6x 100 56.9 69.14 6x/3x 99.0 53.4 64.29 6x/4x 98.4 60.3 72.10
6x/5x 100 46.2 56.10 6x/6x 100 42.9 52.10 *"4x/4x" means a first
reflux stage using a mass to solvent ratio of 1:4 followed by a
second reflux stage using a mass to solvent ratio of 1:4; "4x/5x"
means a first reflux stage using a mass to solvent ratio of 1:4
followed by a second reflux stage using a mass to solvent ratio of
1:5; and so on. In all samples, the second reflux stage was
followed by a stirred wash stage.
[0161] The data in Table 10 shows that varying the mass to solvent
ratio in each of two reflux steps did not have a large effect on
purity, with virtually all final products reaching the desired
target purity of 99+% purity Reb A. However, when solids yield was
taken into account, the rebaudioside A recovery was affected by
varying the ratio. The samples processed using 1:4 and 1:4, and 1:5
and 1:4, mass to solvent ratios gave the highest recovery of
rebaudioside A at the target purity of 99+% purity Reb A.
Therefore, the most economical means of achieving maximum yield and
target purity is by using a first reflux stage with a mass to
solvent ratio of 1:4 followed by a second reflux stage with a mass
to solvent ratio of 1:4.
TABLE-US-00005 TABLE 1 % water in 81.58% RA start 76.13% RA start
71.85% RA start extraction % recovery % recovery % recovery solvent
of solids of solids of solids 0 74.2 72.8 69.6 2 71.8 69.4 64.8 4
65.2 61.8 56.2 6 58.4 52 45.6 8 51.8 44.8 37.8 10 31.2 29 13.8
TABLE-US-00006 TABLE 2 % water in 81.58% RA start 76.13% RA start
71.85% RA start extraction % purity of final % purity of final %
purity of final solvent product product product 0 88.49 87.66 82.68
2 88.34 88.29 82.38 4 98.56 95.65 85.99 6 99.62 98.55 87.59 8 99.88
99.72 93.87 10 100 100 100
TABLE-US-00007 TABLE 3 % water in Start End Initial weight weight %
Rebaudioside % Recovery Extraction Grams grams A content
Rebaudioside A 0 5 3.970 95.06 83.91 2 5 3.929 95.82 83.72 4 5
3.635 98.04 79.24 6 5 3.647 100 81.09 8 5 3.609 100 80.24 10 5
3.422 100 76.08
TABLE-US-00008 TABLE 4 % water in Start End Initial weight weight %
Rebaudioside % Recovery Extraction Grams grams A content
Rebaudioside A 0 5 3.280 89.76 72.10 2 5 3.609 88.46 78.18 4 5
3.106 92.54 70.39 6 5 2.834 90.59 62.87 8 5 2.999 94.58 69.46 10 5
2.882 96.24 67.92
TABLE-US-00009 TABLE 5 Process Process end end Reb A Stevioside
Process % Start dry content (% content (% end dry wt. Recovery wt.
(grams) dry wt.) dry wt.) (grams) of Solids 100,000 99.6 ND* 68,000
68
TABLE-US-00010 TABLE 6 Sample % Solubility (solids (g)/final volume
(ml)) No 100% ethanol treatment 2 1 hour reflux in 100% ethanol 40
15 minute stir in 100% ethanol 33 81.58% Rebaudioside A Start
89.95% Rebaudioside A Start % solids % Rebaudioside % solids %
Rebaudioside Stand time (hrs) yield A Purity yield A Purity 0 31.6
99.44 73.4 99.87 18 38.7 99.68 71.8 100 24 40.5 99.55 72.9 99.84 36
38.5 99.45 72.8 100 160 40.1 99.48 72.2 99.77
* * * * *
References