14
3. Heat the pressurized canisters to 120 - 250°C, depending on the type of
canister being cleaned. Do not allow the temperature of a SUMMA® can-
ister to exceed 155°C, because the Nupro valve it employs has Viton® O-
rings and requires greases that cannot be exposed to high temperatures.
Many commercial cleaning systems avoid this problem by ensuring the
valve is not within the heated zone. The canister below the valve is heated
but the valve receives only radiant heat. In contrast, the Parker Hannifin
diaphragm valve in a SilcoCan™ canister is far less heat sensitive, allow-
ing the canister to be cleaned at temperatures up to 250°C, to help
remove less labile impurities.
Heat the canisters filled with humidified air for at least 1 hour.
4. Re-evacuate the canisters to remove the desorbed impurities. Allow the
canisters to equilibrate for 1 hour.
5. Determine if the canisters have been cleaned effectively by following the
procedure in Certifying the Canister, below. US EPA methods recom-
mend testing every canister until a reliable procedure is developed.
Repeat steps 1-5 as necessary; the number of cycles will be determined by
how dirty the canisters are and how easily they are cleaned. We recom-
mend developing a cleaning procedure that matches your specific sam-
pling procedure, by testing the canisters for cleanliness after each cycle
and determining the number of cycles necessary for proper cleaning. If
the canisters are not heated, the number of cycles required to clean the
canisters might be higher.
6. Once a canister is clean, prepare it for collecting a sample by evacuating it
to 10-50 mTorr. If your system is leak-tight, you can do this by using a
roughing pump, but many commercial systems include a molecular drag
pump to reach final vacuum quickly.
IX. Certifying the Canister
We recommend certifying canisters for both cleanliness and for analyte stabil-
ity. To certify a canister clean, pressurize the canister to 14.7 psig with humid-
ified ultra-high purity air or nitrogen after it has gone through the cleaning
cycles. The humid air or nitrogen stream must be certified clean before it can
be used for canister certification. Analyze an aliquot of the canister content by
GC/MS or GC/FID/ECD. US EPA Method TO-14A/15 specifies a canister
must contain less than 0.2 ppbv of any target VOC compound (Figure 9); EPA
Method TO-12 specifies less than 0.02 ppmC, as detected by GC/FID. If a can-
ister does not meet specification, it must be re-cleaned and re-tested for certi-
fication.
To certify a canister for analyte stability, introduce a low working concentra-
tion of a characterized test mix into the canister. Analyze an aliquot of the con-
tents of the canister immediately after introducing the test mixture and at
periodic intervals. We recommend monitoring for changes for a minimum of
2 weeks, or for a timeframe similar to your anticipated holding period.
Responses should not decrease more than 20% over this period.
Commercially available standards are available for stability testing, but we rec-
ommend you make your own test mixture that is comparable to the target
compound list that the canister will hold. For example, if you are analyzing
sulfur compound content in ambient air, prepare a sulfur-specific test mix and
evaluate the canister's performance for sulfurs. Maintain a log sheet for each
canister, and record the test results and certification. This will be a permanent
record for each canister. Some labs certify canisters for certain compounds
and use a canister only for this specific application.
Figure 8
User-designed system for
cleaning 24 six-liter canisters.
