The CQC Sampling Protocol is a method for analyzing corks for TCA using an objective chemical analysis based on the technique referred to as SPME. 

The procedure allows CQC members to screen more lots with larger samples than current ISO guidelines. It also provides quantified results, measuring "Releasable TCA" to concentrations as low as 1.0 parts per trillion (well below sensory recognition levels). 

CQC and ETS Laboratories designed the methodology for the SPME analysis - which is presently conducted under commercial conditions by ETS Laboratories on behalf of CQC members.  The tests are used to evaluate incoming cork shipments, and the expected benefit is a greatly improved ability to identify cork bales with high levels of TCA.  The CQC believes that less than 5% of cork bales are responsible for as much as 50% of the cork related TCA taint affecting bottled wine.

Though there are several types of chloroanisoles that cause wine to be "tainted", TCA (2,4,6-Trichloroanisole) is the chemical most closely associated with the cork. It is the compound with the lowest sensory threshold and has the typical aroma of cork taint. TCA is reported to occur in more than 70% of corks that contain significant "off" aromas. 

Some other chemical compounds, like methyliso- borneol and geosmin, may also be present in problem corks, but they have higher sensory thresholds, are much less common than TCA, and are not commonly linked to bottled wine described as "tainted".

The SPME analysis is based on a combination of solid phase micro-extraction (SPME) and gas chromatography (GC) with mass spectrometry detection (MS).

SPME technology extracts TCA from the "headspace" above cork soaks using an absorbent fiber. The fiber transfers TCA to the GC/MS where TCA is measured using Single Ion Monitoring (SIM). 

Precision at levels of 10ppt is excellent with an analytical variance less than 5%. This precision is enabled by use of a deuterated internal standard (a TCA analog). The combination of techniques has created an accurate system for measuring TCA that is more sensitive and more dependable than human sensory screening.

Both the human nose and SPME "capture" volatile TCA above a liquid sample. In cork quality control programs, the liquid is usually a cork soak containing an unknown amount of TCA.

The results of sensory screening are Pass/Fail. Either a cork soak is deemed acceptable or it fails based on a defined defect, typically TCA or "other" aromas. The final result is commonly expressed as a % failure rate, for example 2% TCA. Sensory analysis is hindered by several factors, including individual sensitivities of the panelists, masking by other normal cork aromas, and sensory fatigue. These problems result in false positives (lots that are unnecessarily rejected) and false negatives (lots that are bad but not detected). 

The results from SPME/GC/MS analysis are quantitative and consequently reported as a specific TCA concentration. Results are consistent, performance of the analysis is carefully tested, monitored and controlled, and results are reportable to levels that are well below sensory thresholds.

 

Releasable TCA is defined as the concentration of TCA in a cork soak after it reaches equilibrium. Only a small portion of the cork's TCA is transferred to a soak solution. The remainder is bound to the cork. Typically, less than 1% of the TCA in a cork moves into the soak. There is not a good correlation between releasable TCA and total cork TCA.

Bottling trials conducted with corks containing known releasable TCA levels found a strong correlation between releasable TCA and bottled wine TCA (see below). Previous research has already shown that there is a poor correlation between total cork TCA and wine TCA.

 

Releasable TCA is determined by a combination of factors: the total amount of TCA in a cork, its location in the cork (is it on the surface or deep inside the cork?), and other physical or chemical properties of the cork that influence the affinity of the cork for TCA.

The amount of TCA in a cork soak is highly dependent on the soak conditions. One of the most significant factors influencing releasable TCA is the alcohol content of the soak solution. CQC protocol calls for a neutral white wine with ethanol of 10%. 

Releasable TCA in cork soaks is reported in parts per trillion (ppt). This corresponds to nanograms per liter (ng/L). These same units are used to report TCA levels in wine.

A cork Bale normally consists of a large bag of corks (typically 10,000) sent from the manufacturer to its US importer. Bales normally contain corks of the same grade that have been washed and dried concurrently.

The CQC defines a lot as a single grade of corks, with the same wash, delivered from the same supplier, in the same container.  A single lot usually contains between 20,000 and 200,000 corks. 

In preliminary tests, 4 bales from 40 cork lots were tested individually. These results indicated that bales within a lot generally had similar TCA levels. In lots with high TCA levels, all the bales tested had similar results. 

Current CQC protocol recommends testing at least 3 out of every 10 bales.

In a CQC applied research project, 100 corks were pulled from 14 different bales of 10,000. Each cork was analyzed for releasable TCA. A second 100-cork sample was pulled from the same bales and soaked as a composite or group sample (all 100 corks from a bale soaked together). For each bale, the releasable TCA from the group soak was equal to the average releasable TCA for the individually analyzed corks. 

 

Research to date involving tests of over 2,000 individual corks from 14 separate bales indicates that a sample of 100 corks provides a reliable and representative indication of the average of the individual corks in those bales. Furthermore, statistical analysis suggests that a sample of 50 corks is suitable to identify lots with above average TCA concentrations.

Preliminary results from the 14 bales studied by CQC and ETS suggest a significant relationship between the releasable TCA of a group soak and population distribution. In all 14 populations, most of the corks were below the group soak average. Typically, 75% of the population had lower TCA values than the average. 

In the bales we have studied with high group soak TCA concentrations, there were very few corks with low TCA levels. In bales with low group Soak TCA levels, there were very few corks with high TCA levels. 

The 14 bales used in this study were selected to represent a wide range of TCA levels from very low to very high. They were not selected as examples of typical commercial bales . CQC is planning additional population research to better understand TCA distribution within lots and the relationship of population distributions to group soaks. 

 

There is a strong correlation between releasable TCA measured in individual corks and TCA found in wines after bottling.

Three bales were selected for the bottling of a White Zinfandel with a 10% alcohol. Corks were measured for releasable TCA prior to bottling, dried and coated with standard cork coating materials, and then bottled and stored (neck down) using common industry standards. The bottled wine was analyzed for TCA at 1 month, 3 months, 8 months, and 14 months. 

By 3 months, there was a strong relationship between the wine TCA and pre-analysis of releasable TCA in the corks. By 8 months, this correlation was very strong; pre-analysis of corks for releasable TCA was able to describe more than 90% of the variation in wine TCA.

TCA moved slowly from the cork to the bottled wine. The average of all samples showed that the wine TCA was less than 15% of cork releasable TCA at 1 month. At eight months this had increased to 32%. At 14 months the average level of TCA in the bottle was equal to 50% of the original releasable TCA. 

The SPME method can detect trace levels of TCA, well below sensory thresholds. Experience with the technique shows that TCA is far more ubiquitous than sensory experiences have indicated. Where it was commonly assumed that the acceptable level of TCA was "zero", it is now apparent that TCA can be found at some level in a majority of wines. It is not clear, however, at what level of TCA causes taint.

There are many factors that will affect the sensory presentation of TCA. Literature cites examples of thresholds ranging from 6 ppt for dry white wines - to twice that level in some reds. Alcohol can repress the volatility of TCA and mask sensory perception. 

There is no absolute point where a group TCA score can conclusively identify whether a bale should be accepted or rejected. 

Current research has demonstrated a generalized population curve of TCA within a bale. That generalized curve is not capable of accurately predicting the number of defective corks based on the group TCA average.

Furthermore, the sensory effects of TCA in bottled wine show tremendous variation based upon the characteristics of the individual wine.

Clearly, a low TCA measurement is preferred. Likewise it is easy to identify and reject bales with group TCA scores that are well above average. Early results from commercial screening suggest that approximately 5% of incoming cork bales may contain as much as half of the total number of defective corks in the total population. It is also apparent that there is a strong probability that individual bales within a cork lot are likely to have similar TCA results.

Research to date has been conducted strictly on traditional corks. There has been no work performed on agglomerated corks for still or sparkling wines. It is not clear if the same relationship between releasable TCA and its subsequent transfer to bottled wines exists in these products.

Further research must be performed in the immediate future.
 
 

END OF FAQ SECTION