BACKGROUND OF THE INVENTION
The present invention relates to methods for remediating
food and beverage products contaminated with taint compounds which introduce undesirable
flavors or odors into those foods and beverages. In particular, the invention relates
to wine and the contamination of wine with 2,4,6 Trichloroanisole (TCA) which is
also known as "cork taint." TCA imparts a moldy/musty odor to wine and is molecule
produced by a metabolic reaction of mold organisms with chlorine and chlorinated
compounds. TCA taint in wine is typically associated with cork, however, other media
related with wine production and storage can also lead to TCA formation. While the
human detection threshold for TCA is considered to be above 5 nanograms per liter
(part per trillion) of wine it may be capable of suppressing the positive fruit
aroma character in some wines at levels as low as two nanograms per liter. Damage
to the wine industry worldwide from TCA taint is estimated to total $10 billion
The food and beverage industry, especially wineries, need
a cost effective and efficient method for preventing taint compounds from entering
products during manufacturing, packaging, storage, and the distribution processes.
A technology allowing end consumers to remediate foods and beverages immediately
prior to consumption would also be beneficial. A number of solutions have been proposed
to prevent products from becoming tainted and for the remediation of products already
contaminated. These solutions have had limited acceptance principally due to the
quantities of treatment material required, their costs, and their tendencies to
change the flavor profiles of the products being protected or remediated. For example,
technologies capable of removing TCA from contaminated products may also remove
different desired compounds.
Zeolite technologies are among the tools applied to selectively
remove molecules from various matrices. While there are numerous references regarding
the adsorption of halogenated aromatics by zeolites, a literature search indicated
there are no examples of zeolite adsorption of polychlorinated anisoles such as
Andersson et al. (J Appl. Polym. Sci. 95: 583-595, 2005
) placed a proprietary hydrophilic mixture of zeolites in low-density polyethylene
films for use in food packaging. The zeolite was employed in this application to
trap the degradation products of the polymer film preventing them from leaching
into the food product. In the work of Andersson et al., the packaging film was the
origin of the compounds that produce off-flavor in foods. The purpose of the zeolite
in Anderson's technology was to retain these compounds in the film in order to minimize
off-flavor effects. The ultimate goal of this work was to prevent the packaging
material from influencing the aroma profile. A multitude of degradation compounds
were targeted for adsorption instead of the focused targeting described in our novel
U.S. Patent 5,750,611
describes the use of molecular sieves to minimize the odors and flavors
associated with thermoplastic films. Again, this technology focused on non-selectively
retaining the odors and flavors that are created due to the polymer itself or breakdown
products resulting from additives deliberately mixed into the polymer. These references
do not describe a process in which the odors and flavors are selectively prevented
from entering the food product but rather describe a technology in which a nondescript
myriad of odor and flavor molecules are adsorbed.
Of interest to the present invention is a remedial treatment
described in Swan,
US patent 6,610,342
which uses aliphatic synthetic polymers to remove off-flavors and odors
from foods and beverages. This technology requires a comparatively large amount
of synthetic polymer to remove a relatively small amount of taint. Swan's treatment
calls for the use of roughly 150 grams of polymer to treat 1 liter of beverage.
An example given for this technology describes the use of roughly 63 grams of an
ultra-high molecular weight polyethylene (UHMWPE) to reduce TCA levels from 87.5
to 6.4 parts per trillion, in 0.75 liters of wine. The wine was filtered through
a bed made up of the UHMWPE at a rate of 5 ml/min. The total filtration time for
the 0.75 liters was consequently 2.5 hours. This translates to an average TCA removal
rate of approximately 0.54 nanogram (ng) per minute and a capacity of 1.3 ng TCA
removed per gram of treatment material. The novel technology described herein has
taint removal capacities that are orders of magnitude larger than Swan's UHMWPE
polymer. In addition, the presently described adsorbent technology removes these
comparatively large taint amounts in seconds rather than in minutes or hours.
Another method used to remove taint from wine utilizes
milk or the combination of milk and diary cream known as "half-and-half." The application
of this technique has been approved by regulatory authorities in the past at a usage
rate of 2 liters of milk or half-and-half per 1000 liters of wine. This method is
stated to be capable of removing TCA without significantly changing the phenolic
profile of the wine but no quantitative information regarding the aroma profile
was provided. Half-in-half is said to perform better at removing TCA due to its
higher fat content. Milk use at this rate is approved by the federal government
(27 CFR 24.246 Materials authorized for treatment of wine and juice.) for the fining
of specific types of wine. A usage rate for taint removal of 10 liters of milk or
half-in-half per 1000 liters of wine has been proposed.
The capacity, selectivity, and speed of half-and-half for
TCA removal have not been specifically reported for these studies and the efficacy
of this method is unclear. Nevertheless, the addition of milk products to wine and
the subsequent removal might be relatively expensive. Moreover, the use of milk
products could also raise questions regarding the acceptability for consumption
by individuals with allergies to dairy products.
Although the wine industry has made strides in minimizing
the potential for TCA contamination, the problem remains. Accordingly, there remains
a need for methods allowing for the fast, selective, and economical removal of TCA
in a manner which do not remove other desirable components or otherwise change the
contaminated product's flavor and aroma profiles.
Brief Summary of the Invention
The present invention relates to the discovery that Zeolite-Y
(Faujasite) molecular sieves are capable of selectively removing TCA from solutions
including liquid food and beverage products (hereinafter collectively "beverage
products"). In particular, the invention is directed to the use of Faujasite molecular
sieves for the removal of TCA from wine and the discovery that TCA may be thus removed
from wine without significant negative impacts on the desirable flavor and aroma
components of wine.
While the invention is particularly directed to wine during
its manufacture and storage the methods of the invention may also be used for the
rapid and selective remediation of the raw materials used to make the beverage products,
the intermediate products created during manufacturing, and the final products.
Additionally, this technology is beneficial for direct incorporation into packaging
materials, and package containers.
A particularly preferred Faujasite for use according to
the invention is one having a silica-to-alumina (Si/Al) ratio greater than 5 with
Si/Al ratios of 10 and greater being particularly useful according to the invention.
Faujasites with compensating cations selected from the group consisting of hydrogen,
ammonium, alkali metals (such as sodium), rare earth elements, and organic can be
used for this tuning may be used according to the invention with H+ and
Na+ Faujasites being particularly preferred.
The Faujasites may be used at various concentrations as
determined empirically based on the level of TCA contamination of the beverage product
and the composition of the beverage product itself. In general, concentrations of
from 0.001 to 1 gram per liter have been found to be useful with concentrations
ranging from 0.02 to 0.1 gram per liter of H+ and Na+ Faujasites
being generally sufficient to remove TCA present in wine at concentrations approaching
TCA may be removed from beverages by the steps of mixing
the molecular sieve with the beverage, incubating the sieve with the beverage, and
then separating the molecular sieve from the beverage. According to one such method,
the beverage is passed through a bed comprising the molecular sieve. At a point
when the molecular sieve becomes saturated with TCA it can be regenerated by various
means known to the art and then reused for purification of foods and beverages.
According to a further aspect of the invention, faujasites
may be incorporated into closure or container devices as a prophylactic against
new contamination as well as to remediate past contamination. In particular, the
faujasites can be incorporated into a closure device such as a cork to remove TCA
present in the closure device or otherwise in the beverage container.
The invention also provides devices such as sachets or
dipsticks for the remediation of TCA contamination in a beverage comprising Zeolite-Y
(Faujasite) having a Si/Al ratio greater than 5 in an amount effective to remediate
TCA contamination in said beverage. The device can comprise a sachet or dipstick
or can be incorporated into a beverage product's packaging.
The methods of the invention utilize the molecular sieve
as a taint barrier or direct remediation product for the beverage product. The invention
thus addresses the removal of taint from liquid foods and beverages in the production
stages, the prevention of taint during the beverage product's distribution and provides
methods for removing taint compounds from products immediately prior to consumption.
The latter being especially important to consumers who have purchased expensive
wines only to discover that the wine was tainted.
Brief Description of the Drawing
FIG. 1 is a bar chart comparing the aroma profiles of an
untreated control with an H+ Faujasite treated Cabernet Sauvignon: one,
an untreated control sample; the second, an H+ faujasite-treated sample;
FIG. 2 is a two-dimensional graph depicting TCA concentrations
in tainted wine over various contact times and amounts of H+ Faujasite;
FIG. 3 is a three dimensional graph depicting TCA concentrations
versus H+ Faujasite Si/Al ratios and adsorbent concentrations; and
FIG. 4 is a three-dimensional graph depicting Ethyl Hexanoate
concentrations versus H+ Faujasite Si/Al ratios and adsorbent concentrations;
FIG. 5 is a bar chart depicting Aroma molecule concentrations
in an untreated French Colombard control sample and a mathematically-optimized French
Colombard sample treated with H+ Faujasite.
DETAILED DESCRIPTION OF THE INVENTION
The invention provides the use of Faujasite molecular sieves
for the selective removal of 2,4,6 Trichloroanisole (TCA) from beverage compounds.
Zeolites are a class of molecular sieves that have a crystalline structure made
up of aluminosilicates. The Faujasite molecular sieves of the invention have pore
sizes less than 2 nm and are able to preferentially adsorb one or more molecules
from a mixture by utilizing differences in various properties of the molecules.
For example, molecules of differing sizes can be separated using a molecular sieve
having pore sizes too small to allow access for the desirable molecules in a mixture
but large enough for the undesirable molecules to enter. Other properties such as
hydrophobicity and charge compensating cation related physical and electrical charges
exist and can be exploited to produce physicochemical separation.
Crystalline molecular sieves with pores constructed of
8, 10, or 12 oxygen atoms are called small, medium, and large pore materials, respectively.
The methods of the invention using large pore molecular sieves that possess hydrophobic
tendencies have been successfully demonstrated. In order to capture taint molecules
such as TCA, it has been found that large pore materials are necessary for their
adsorption (medium pore materials have pore sizes that do not allow fast, effective
adsorption). It has also been found that the molecular sieves need to be relatively
hydrophobic in order to selectively attract the comparatively water insoluble TCA.
Further, Faujasites characterized by Si/AL ratios above 10 are preferred.
The adsorbents of the invention can be employed in a variety
of ways. For example, low concentrations of the selective adsorbent can be mixed
directly into a contaminated beverage product and subsequently filtered out prior
to packaging. In another configuration, the contaminated liquid could be run through
a fixed bed of selective adsorbent preceding packaging. A particular advantage of
zeolites is that they can be regenerated to remove adsorbed TCA and then reused.
Those of ordinary skill would therefore recognize that a semi-continuous process
could be carried out wherein active beds are used to remove TCA from a beverage
while other beds are being regenerated. The regenerated beds can then be cycled
into use upon exhaustion of the active "on-line" bed.
In a case where the beverage product also contained solids,
this technology could employ the use of sachet or "stick" containing the adsorbent.
The sachet or stick could be inserted into the tainted mixture of liquid and solids
and allowed to contact the product for a sufficient time to remove the offensive
molecules. Once the taint molecules were adsorbed, the sachet or stick would be
simply removed. In all cases, the beverage product tainted prior to packaging would
be suitable for further processing with minimal changes to their intended odor and
A number of solutions have been proposed to prevent products
from becoming tainted after packaging. Many of these technologies focus on taint
compounds originating from the packaging materials themselves. Degradation products
of molecules purposefully placed in the packaging material for some perceived benefit,
such as oxygen scavenging, are typically the target. The methods of the invention
can be used to prevent introduction into a finished product of both taint compounds
found within the packaging materials and ubiquitous environmental taint molecules.
The materials of the invention can also be fashioned to act as a remedial packaging
component that removes taint from foods and beverages existing prior to or introduced
during the packaging step. These adsorbents can be added to natural bark wine corks,
agglomerated, composite, and synthetic wine corks, screw cap liners, packaging films,
coatings, sachets and the like.
Finally, the methods and devices of the invention can be
used to remediate beverage products at the product's site of consumption. A retail
customer, for example, could remediate a tainted bottle of wine by dipping a device
containing the molecular sieve into the liquid and swirling it around for a specified
period of time. Another configuration of this type of remedial device could be a
strainer containing a disposable insert securing the molecular sieve. Retail customers
or restaurants could simply pour tainted wine through the strainer resulting in
a taint remediated product.
According to this example, various commonly available zeolites
were tested for the selective removal of TCA from wine. These zeolites included
Zeolite Y (FAU), Beta (BEA), and ZSM-5 (MFI). In order to determine the adsorbent's
efficacy, testing was conducted using commercially available red and white wines
contaminated with TCA concentrations ranging from approximately 10 to more than
400 part per trillion. The treatment was performed by directly mixing the various
zeolite powders into the contaminated wine, allowing the wine to contact the powders
for a prescribed amount of time (from seconds to hours), filtering the wine to remove
the zeolite, and analyzing the wine for TCA concentration. These concentrations
were determined using either a sensory panel trained to detect and judge the TCA
odor's intensity or a Gas Chromatograph Spectrophotometer (GCMS) Solid Phase Micro-extraction
(SPME) method. Table 1 provides a qualitative indication of the removal efficacy
of the three zeolite types along with silica to alumina (Si/Al) ratios and charge
TCA Removal from Wine
The faujasite zeolites performed best for this application,
BEA was a distant second, and MFI showed no effectiveness. Unexpectedly, the faujasite
molecular sieves performed well compared to the BEA sieves which had a similar pore
opening size and a higher Si/Al ratio.
In conclusion, zeolites having a faujasite structure were
successful in rapidly removing TCA from wine. Further, the performance of the faujasite
appeared dependent on the amount of silica in relation to alumina in its framework.
Specifically, faujasite adsorbents containing higher Si/Al ratios adsorbed more
TCA from wine per unit of adsorbent applied.
According to this example, a qualitative test of TCA removal
was conducted using various faujasites and a ZSM-5. Specifically, a sensory panel
trained and skilled in identifying the relative intensity of the TCA odor in wine
was used to carry out an initial qualitative testing. The samples for treatment
and evaluation were made using a non-sorbated French Colombard wine spiked with
TCA at concentrations ranging from zero to 280 nanograms per liter. The various
adsorbents, at concentrations ranging from zero to 0.08 milligrams per gram of wine,
were added to the respective sensory jars and mixed by shaking.
The experiment was designed to evaluate the various adsorbent's
effectiveness at removing the TCA odor from the wine samples and to determine if
the adsorbent appreciably changed the wine samples' aroma. To achieve this objective,
some of the samples containing no TCA taint were treated with the adsorbent and
several of the samples that were spiked with TCA received no adsorbent. The panel
members were also given a control wine that was neither spiked with TCA nor treated
with adsorbent, as a reference.
The panel members were instructed to rank the samples for
TCA odor intensity and aroma reduction (scalping). The control wine was used to
provide a baseline aroma level for the panel members to allow an evaluation (ranking)
of the amount of aroma scalping that occurred in treated wine samples.
The sensory panel for this test consisted of five members.
In order for a member's TCA intensity or scalping value to be counted, at least
two of the five members needed to register a response. If two or more members recorded
a response, the responses were averaged and recorded. The results are shown in Table
Qualitative TCA Removal Study- Wine
Adsorbent Level (mg/g wine
TCA Concentration (ng/L wine)
TCA Target Sensory Ranking (Panel;
TCA Sensory Ranking (0-no TCA, 7 - max TCA)
Flavor Scalping (0-no scalping, 7- no flavor)
1 Average of 5 Sensory Panel members. Criteria minimum of 2 members required
to respond prior to response being averaged.
The results shown in Table 2 demonstrate that the Faujasite
adsorbent having a silica/alumina ratio of 80 performed well with respect to TCA
removal without flavor scalping. No TCA aroma was detected by the panel from either
the 62.5 or 250 ng/L samples treated with this zeolite. Further, no aroma scalping
was detected in any samples.
According to this example, the quantitative change in the
aroma profile of a H+ Faujasite treated wine was studied. A Cabernet Sauvignon was
chosen for this work and a baseline aroma profile from forty-eight aroma molecules
(shown in Table 3), commonly found in wine and deemed important by a number of winemakers,
was determined using quantitative analytical methods. In the untreated wine, forty-one
out of the forty eight aroma molecules were at detectable concentrations in the
Aroma Molecules Evaluated
A one liter aliquot of the baseline Cabernet Sauvignon
was extracted and prepared by adding 0.1 grams of H+ Faujasite powder
having a Si/Al ratio of 80. The sample was then thoroughly mixed and the powder
allowed to settle. Once settled, the treated wine was filtered through a glass fiber
pad to remove the zeolite. From each filtered one-liter sample, 100 milliliters
were extracted and delivered to the lab for aroma quantification. Figure 1 presents
the results of the aroma profiles of the zeolite-treated and untreated wine samples.
As can be seen, the aroma profiles of the two samples are quite similar. These data
support the qualitative results regarding the lack of aroma scalping found in Example
2. Out of the forty-one aroma molecules found in the baseline wine, only three (ethyl-octanoate,
ethyl-decanoate, and cis-oak lactone) may have been reduced by an organoleptically
According to this example, the capacity of an H+
Faujasite to remove TCA from wine and the rate of such removal was measured quantitatively.
A sample of an untainted non-sorbated French Colombard wine was split equally with
one half serving as a reference and the other half spiked with TCA at a concentration
of 500 ng per liter (parts per trillion). Each bottle was shaken and allowed to
sit for 24 hours before continuing with the experiment.
Five levels of the Faujasite powder (0.0, 0.25, 0.50, 0.75,
and 1.00 grams per liter of wine) were evaluated for TCA-removal capability. Each
treatment consisted of placing 75 ml of the 500 ng-TCA per liter-tainted tainted
wine into a 40- x 80- mm glass bottle containing the appropriate amount of adsorbent
To determine the rate of TCA adsorption, aliquots of each
treated wine were analyzed for TCA using, GCMS/SPME after wine/adsorbent contact
times of 0, 1, 5, and 25 hours. For the 0 hour contact time, the samples were immediately
filtered through a glass-fiber filter pad to remove the zeolite powder. The filtrate
was immediately placed into a 30 ml GC/MS headspace vial, tightly sealed with a
septa cap and inverted 4 times. All of the treatments, including the no adsorbent
sample, were filtered and prepared in this way. After preparing all of the headspace
vials for the 0-time sample, the vials were quickly placed on the GC/MS auto-sampler
for TCA analysis by SPME technique.
This sequence was followed for the 1-, 5-, and 25- hour
samples. The results of this testing are depicted in Fig. 1 and indicate that the
TCA removal occurred very rapidly after the application of the zeolite powders.
The results indicate that an average of over 97% of the TCA was removed immediately.
After this immediate, rapid removal, the TCA removal rate appeared to taper off
and approach an asymptote over the 25- hour treatment period. The speed of TCA removal
and the quantity removed were substantially higher than the rates and quantities
removed using competitive remediation technologies. Even the lowest treatment amount
of 0.25 gram per liter Faujasite removed over 310 nanograms of TCA per liter of
According to this example, a test was carried out to quantify
how well various adsorbent could remove TCA from wine without affecting the wine's
desirable aromas. As with Example 3, forty-eight aroma molecules (shown in Table
4), commonly found in wine and deemed important by a number of winemakers, were
selected for analyses.
Faujasite adsorbents having Si/Al ratios ranging from 5.1
to 80 were used for this investigation with concentrations of adsorbents ranging
from 0.01 to 0.10 grams per liter of wine. The samples were created as they were
in the previous example, with the contact time and TCA spike concentration held
constant at 24 hours and 20.4 ng/L, respectively. The analysis for the TCA and aroma
molecule concentrations was performed using GC/MS methods with the TCA results shown
in Table 4 below.
Adsorbent (g Adsorbent/L wine)
An analysis of these results shows that Faujasite zeolites
having Si/Al ratios greater than about 5 were superior in removing TCA from wine.
The results also showed that the extent of TCA removal was impacted very little
by the adsorbent concentrations as long as the Faujasite had an Si/Al ratio of 20
or greater (Figure 3).
Of the 48 compounds deemed important to the aroma of wine,
30 were detectable in the French Colombard used in this study. Of these 30, six
(6) appeared to be reduced in concentration due to the adsorbent additions (see
Aroma Compound (ppb)
Adsorbent (g Adsorbent/L wine)
Out of these six compounds, the concentrations of the ethyl
esters appeared to be reduced the most. The removal of ethyl hexanoate (a compound
described as having a fresh fruity, pineapple odor) is graphed in Figure 4 and illustrates
how differently a desirable aroma compound was reduced in relation to the Faujasite's
concentration and Si/Al ratio when compared to the extent of TCA reduction. Significantly,
this difference in the removal response of the aroma compound versus TCA allows
one of skill in the art to maximize the removal of TCA while minimizing the removal
of a desired aroma molecule. Those of ordinary skill in the art would thus be able
to compare the TCA remediation graph (Figure 3) and the ethyl hexanoate removal
response (Figure 4) to determine adsorbent amounts and Si/Al ratios needed to optimize
the retention of desired aroma compounds and the removal of undesirable flavor/aroma
Using the information from Example 5, this example demonstrates
a mathematical modeling approach for minimizing the removal of agreeable aroma molecules
while maximizing the removal of TCA by way of desirability profiling for multiple
responses. Using JMP version 5.1.1 (a commercial statistical package created by
SAS Institute Inc., 100 SAS Campus Drive, Cary, NC 2751302414), each aroma and TCA
reduction was deemed a response potentially driven by the Si/Al ratio and Faujasite
concentration factors. The aroma's desirability functions were set as targets of
the original wine's aroma concentration, or set up as higher-the-better (HTB) when
appropriate. The TCA's desirability function was set up as lower-the-better (LTB).
Prediction profile plots for all aroma and TCA data were created along with their
associated desirability functions and the overall desirability was maximized. A
Is/AL ration of approximately 11 and adsorbent amount of 0.02 grams per liter was
predicted "optimal" at the maximized overall desirability. Further, a theoretical
reduction of TCA from 20.4 to below 2.5 ng/L (i.e., >87% TCA removal) was predicted
while the aroma profile was substantially preserved. As can be seen in Table 6,
the maximized overall desirability model resulted in good theoretical retention
of the six adsorbent-sensitive molecule concentrations.
Desired Aroma Molecule
H+ Faujasite Treated
all values in µg/L
Clearly, the combined use of zeolite adsorption differences
between molecules and this multivariate technique can minimize the negative impact
of the TCA-reduction treatment on the desirable aromas. The ability to maintain
the product's original aroma profile is further illustrated in Figure 5 where all
30 of the detectable aroma molecule concentrations in the French Colombard study
are displayed for the untreated control and the multivariate-optimized Faujasite-treated
Numerous modifications and variations in the practice of
the invention are expected to occur to those skilled in the art upon consideration
of the presently preferred embodiments thereof. Consequently, the only limitations
which should be placed upon the scope of the invention are those which appear in
the appended claims.