Method TO-15A Requirements and Best Practices for Sampling Devices, Canisters, and Analytical Systems

When collecting and analyzing whole air samples, it is critical to ensure that your entire system, from sample collection to sample introduction to GC instrumentation, is as clean and inert as possible. Method TO-15A [1] requirements (and those of similar methods, such as China’s HJ759) define limits on blank interferences and standard recoveries and also provide important guidance on canister cleaning. While both electropolished and silicon-lined (i.e., “coated” or “treated”) canisters are very inert, contaminants can cover the passivated surfaces and degrade performance over time. The consequences of poor system cleanliness include both negative bias (adsorption and breakdown) and positive bias (desorption and non-target interferences). These biases lower the accuracy of air monitoring methods and/or can reduce the acceptable storage time of samples in canisters.

When testing a multi-part system for cleanliness, it is helpful to break it down into components and test backwards from the normal sample flow. This may seem counter-intuitive at first, but because contamination can occur anywhere from the sampling device to the detector, reversing the sequence and eliminating one potential source at a time makes it easier to determine the root cause. Figure 1 shows a general sample flow path from collection to analysis (top arrow); to establish cleanliness, start with the GC-MS and move backward following the bottom arrow. Method TO-15A sections 9.3 through 9.5 follow this order and provide guidance for qualifying the analytical instrumentation, air canisters, and sampling devices. Method TO-15A requirements for qualification include both a blank test and a known standard challenge, which are used to detect positive and negative interferences.

In this article, we will walk through the Method TO-15A (revised Method TO-15) requirements, guidelines, and best practices that are designed to ensure your equipment is clean, inert, properly qualified, and able to produce accurate, reliable results. A detailed troubleshooting table is also provided at the end for quick reference.

Part 1: MS and Leak Checking

To begin the process of meeting Method TO-15A requirements for system qualification, we will start with section 9.3: Qualification of Analytical Instrumentation. We can break this down further by addressing the detector before considering the GC or preconcentrator. First, always check the MS tune on a regular basis; this practice will alert you to many issues that can interfere with accurate air sample analysis.

In particular, look for elevated ions related to nitrogen (m/z 28) or water (m/z 18) because these may indicate a leak in the system. The MSD transfer line nut is a common source of air leaks, so make sure it is fully tightened if elevated levels of m/z 28 or 18 are present (a 28:69 ratio of less than 5% is a good indication the GC-MS is leak tight). Be careful not to overtighten the fitting however, as this can break the GC column. The MS side plate and MS vacuum release knob are also potential sources of leaks, so check that they are fully sealed when the mass spec is pumped down. Leaks are not restricted just to the detector; any connection, often the preconcentrator and transfer lines, can potentially have air leaks. To prevent problems during analysis, use an electronic leak detector (cat.# 28500) to check all connections before starting to ensure a leak-free system.

Elevated amounts of background ions can indicate a dirty ion source or contamination in the GC or carrier gas. Difficulty meeting the tuning criteria is an additional clue that the ion source may be dirty and/or the filament needs to be replaced. Before performing maintenance to resolve these issues, see your instrument maintenance manual for instructions on how to properly vent and clean the MSD. PFTBA tuning solution, filaments, and commercially available GC-MS cleaning kits are readily available and are an essential part of proper instrument maintenance.

Part 2: GC, Carrier Gas, and Transfer Lines

After the detector, the GC and related components (carrier gas lines and transfer lines) are the next step in working through section 9.3 of Method TO-15A. Carrier gas contamination is the root cause of many issues in GC and can lead to qualification failures for air analysis. Compounds in the carrier gas or carrier gas lines can condense on the head of the GC column, especially if the GC is kept idle at lower temperatures. To determine whether contaminants are present in the GC and carrier gas, analyze a blank run with no injection from the preconcentrator. Note that the transfer line from the preconcentrator is normally tied into the GC inlet or connected directly on-column, so it is considered part of the GC system, unless it is disconnected to isolate individual components for troubleshooting.

If contaminants are present, they must be removed so that Method TO-15A requirements for cleanliness can be met. This can be accomplished by baking out the GC column and replacing the carrier gas line filters. To bake out the column, condition it at 20 °C above the final analysis temperature of the method for an hour. Do not exceed the column’s maximum operating temperature. If this does not eliminate the contamination, then it is possible that the preconcentrator transfer line is contaminated. If possible, the transfer line temperature can be elevated to drive off contaminants, but make sure that downstream temperatures are not lower because this can trap the contaminants further in the system. Refer to your preconcentrator manual for appropriate temperature limits and further instructions for cleaning or replacing the transfer line.

Part 3: Preconcentrator, Autosampler, and “Zero” Air Source

After the GC-MS, the final part in the analytical system, as defined in section 9.3 of Method TO-15A, is the preconcentrator and autosampler. The first step in meeting Method TO-15 requirements here, before any blank checks are run, is to make sure the preconcentrator is leak free. If the control software does not have a leak check built in, then pressurizing the system and watching for a pressure drop or using an electronic leak detector are good alternatives.

To check the preconcentrator for cleanliness, first test the focusing trap by running a blank with no sample injection. If the blank fails, the trap may need to be baked out or replaced. Next, connect a clean zero air source, ideally a zero air generator, directly to the sample inlet and run it as you would with a sample canister. Note that while TO-15A states to use humidified zero air, it is best practice to first use dry air to verify its purity, and then add humidification to verify the water source is clean once you are satisfied that your air is clean.

Contamination in zero air can come from several potential sources. The first step in isolating the cause is to replace the gas filters or perform maintenance on the zero air generator. If this does not help, then the contamination may be built up in the gas lines and they may need to be replaced. If the lab environment has heavy solvent use, try to keep the gas lines away from areas of high solvent use to avoid diffusion into the gas lines. If neither maintenance nor replumbing fixes the issue, it may help to change gas sources (i.e., switching from an air compressor to UHP cylinders, or even just using a new air cylinder. If your lab is not using a zero air generator, installing one can greatly improve air quality [2].

Once the dry air source has been determined to be clean, the next step is testing your system with humidified air. Method TO-15A recommends humidification to 40-50% relative humidity (RH). If humidifying your air results in blank interferences, maintenance on your deionized (DI) or reverse osmosis (RO) water system may be required. If the water system is located in an area that uses volatile solvents, then it may be necessary to relocate the system, boil or sparge the water to remove volatile contamination, or find a new source of humidification water.

Part 4: Analytical System Known Standard Challenge

Once your entire analytical system has been demonstrated to be clean enough for blank analysis, the next step is performing a known standard challenge to test for sources of negative bias, such as cold spots or active sites. Method TO-15A section 9.3.2 recommends testing with known standards at 100-500 pptv (40-50% RH), with consideration given to the expected levels in your samples. Note that while canister cleanliness will not be formally covered until Part 5, you still need to be confident in the cleanliness of the canisters used here for preparing the known standard in order to get the initial results needed to certify your system and canisters. If possible, canisters that are new or have not been used for field sampling (i.e., have only been used for in-house standards) are highly recommended for initial calibration and internal standards. Allow the canisters containing the known standards to equilibrate for 24 hours as per TO-15A section 9.4.3, but use them within 48 hours to reduce the effects of any active sites that may be present.

If compounds show poor response or peaks are missing entirely, this is might be caused by active sites. Active sites are usually caused by the buildup of less volatile material, which can be cleaned out by baking out your preconcentrator at higher temperatures. Check your instrument manual for instructions on how to properly clean or replace potentially contaminated parts and lines.

Other possible causes of low responses include system leaks, cold spots, or problems with the focusing trap. Make sure to fully leak check all parts of the system, verify that all heated zones are set correctly and working properly, and that your trap is properly installed and conditioned.

If your preconcentrator has an autosampler, then the blank and known standard challenge test should be applied to all autosampler ports.

Part 5: Air Canisters

Cleaning

After the system has been shown to be clean, the next step in meeting Method TO-15A requirements is to prepare and test the air sampling canisters as described in section 9.4. Prior to testing, canisters should be cleaned using multiple cycles of evacuation and pressurization with clean, humidified air (as is recommended in TO-15A) or nitrogen. The use of humidified air is preferable for two reasons. First, the oxygen present in air can oxidize compounds that may have adhered to canister walls, breaking them down and helping remove them from the walls so they can be evacuated from the canister. Second, the water from humidification can hydrolyze compounds on canister walls, allowing them to be removed from the canister. TO-15A section 10.1.4 recommends five cycles of pressurization to ≤30 psig and evacuation to 28 in. Hg with a 1-minute hold time on each cycle. The final evacuation should be to ≤50 mTorr. The recommended temperatures for canister cleaning from Method TO-15A are presented in Table I; always consider the presence of components and use caution not to exceed maximum temperatures. See Restek’s Frequently Asked Questions: Air Sampling [3] for more information on canister temperature and pressure limits.

Table I: Canister Cleaning and Maximum Temperatures.

	Canister Cleaning Temperature*		Maximum Temperature
Canister	Air Purge Gas	Nitrogen Purge Gas	Air Purge Gas	Nitrogen Purge Gas
SilcoCan (with gauge)	≤80 °C	≤100 °C	≤80 °C	≤120 °C
SilcoCan (without gauge)	≤80 °C	≤100 °C	≤80 °C	≤140 °C
TO-Can (with gauge)	≤100 °C	≤100 °C	≤120 °C	≤120 °C
TO-Can (without gauge)	≤100 °C	≤100 °C	≤120 °C	≤140 °C

*Never exceed maximum temperatures when cleaning equipment.

Post-Cleaning Testing

To test performance after cleaning, the canisters are first blank tested with 50% RH air. After filling, allow the canisters to equilibrate for 24 hours before analysis. Results should show no target analytes above 20 pptv in the canisters. In addition, the chromatogram should be evaluated for the presence of non-target compounds, as these may interfere with target compounds or be indications of possible active sites or compounds.

To verify the canisters are non-reactive, they also should undergo the known standard challenge, as described in section 9.4.3. Canisters that show low recoveries (70% or less of the nominal concentration) indicate the presence of active sites in the canister. High recoveries (130% or higher of the nominal concentration) indicate potential activity that is creating interferences for target compounds. Canisters that fail the zero air or known standard challenges should be cleaned again. If the recommended cleaning procedure is not sufficient to clean the canisters, check that your canister cleaning oven is reaching the desired temperature and humidity levels. The source of humid air used by the cleaning oven must be clean to avoid contamination of the canisters. The zero air and water sources validated in the previous steps would be suitable for use. Increasing the number of cleaning cycles may be needed to remove higher levels of contamination.

After cleaning, canisters can be stored under vacuum or pressurized with dry air or nitrogen (TO-15A section 10.1.4). Over time, compounds on canister walls may start to off-gas, so canisters that have been stored for some time should be given a quick pressurization/evacuation cycle with clean, humid air before being sent to the field.

Part 6: Sampling Devices

The final components of your system to be tested are the sampling devices, which are covered in section 9.5. To meet Method TO-15A requirements, sampling devices should undergo both a humid zero air challenge to verify cleanliness, as well as a known standard challenge to verify inertness. Samplers may be connected either directly to the analytical system or to a certified clean canister to be analyzed later. Passive air samplers, or other sampling devices, should be tested at the flow rates at which they will be used in the field. This means the incoming humid zero air and humid challenge standards must be supplied to the sampling device at near ambient pressure. This can be done using regulators (cat.# 21666 or 22452) and flow controllers to reduce the pressure of the incoming gas, or with a manifold using a vent to keep the sampling system at ambient pressure.

If air samplers do not pass the blank or known standard challenge, they may be cleaned by flushing with humidified air under low heat (maximum temperature of the assembled sampler is 94 °C) or by disassembling the unit and then sonicating with methanol or baking at higher temperatures. Be sure to check the manual of your specific sampling device to determine which parts should not be heated or exposed to methanol.

Summary

The process of meeting Method TO-15A requirements and qualifying your full analytical system as clean and inert enough for air analysis can be a long one, but by breaking it down into its component parts it can become a manageable task. By first verifying the cleanliness and inertness of your analytical system using the approach described here, you can have high confidence in the results for your canisters and sampling devices. The 20 pptv cleanliness requirement and known standard challenge for Method TO-15A may be difficult target for all compounds, but some deviations may be acceptable depending on customer or regulatory requirements. By establishing initial qualification and performing any ongoing qualifications required by Method TO-15A, Method TO-15, or other methods, the data generated by your air lab should be robust, reliable, and accurate.

TROUBLESHOOTING GUIDE

MS Troubleshooting

Symptom	Possible Cause	Solution
Elevated amounts of m/z 28 and 18	Leak in MS system	Check MSD transfer line nut, vacuum release knob/O-ring, and side plate for leaks Check inlet or column connections for leaks
Instrument does not meet tuning criteria	Dirty ion source	Vent MSD and clean source Replace worn filaments
Presence of ions not found in PFTBA tuning solution	Dirty ion source Low or empty PFTBA tuning solution PFTBA valve not open Contamination from GC or carrier gas	Vent MSD and clean source Make sure tuning solution is full and valve is open Troubleshoot GC system (see Part 2)

GC and Carrier Gas Troubleshooting

Symptom	Possible Cause	Solution
Presence of target compounds, non-target interferences, or poor baseline	Contamination at head of GC column	Bake out column Replace column
	Carrier gas contamination	Replace carrier gas filters
	Transfer line contamination	Bake out, clean or replace transfer line

Preconcentrator/Autosampler Troubleshooting

Symptom	Possible Cause	Solution
Focusing trap blank with no sample injection shows target or non-target interferences	Contaminated focusing trap	Bake out or replace focusing trap
Blank with direct air connection shows target or non-target interferences	Contaminated sample lines	Clean or replace sample inlet lines
	Contaminated fill gas source	Perform maintenance on zero air generators Replace gas filters Try different gas source (i.e., air cylinder instead of compressor)
	Contaminated ISTD	Run without ISTD Remake ISTD in clean canister
Blank with humid air shows target or non-target interferences	Contaminated water source	Perform maintenance on DI water system Obtain water from new source, ideally isolated from areas of solvent use
Failure of known standard challenge	Active sites in system caused by contaminants	Bake out system Replace or clean sources of active sites (column, transfer lines, autosampler lines, etc.)
	Cold spots in system	Make sure all heated zones in GC-MS and preconcentrator are set and working properly
	Leak in system	Leak check GC-MS and preconcentrator/autosampler

Air Canister Troubleshooting

Symptom	Possible Cause	Solution
Humid canister blank shows target or non-target interferences	Contamination present in air canister	Make sure canister cleaning oven is reaching desired temperature and humidity levels Increase number of cleaning cycles Replace canister
	Contamination present in water source	Perform maintenance on DI water system Obtain water from new source, ideally isolated from areas of solvent use
	Contaminated fill gas	Perform maintenance on zero air generators Replace gas filters Try different gas source (i.e., air cylinder instead of compressor)
Failure of known standard challenge	Active sites in canister	Make sure canister cleaning oven is reaching desired temperature and humidity levels Increase number of cleaning cycles Replace canister

Sampling Device Troubleshooting

Symptom	Possible Cause	Solution
Blanks show target or non-target interferences	Contamination present in sampling device	Flush assembled unit with humidified air while heating (maximum temperature of assembled unit is 94 °C) Disassemble unit and sonicate with methanol (do not sonicate O-rings or orifice) Bake disassembled unit at 100-150 °C (do not bake O-rings or gauges)
Failure of known standard challenge	Active sites present in sampling device	Flush assembled unit with humidified air while heating (maximum temperature of assembled unit is 94 °C) Disassemble unit and sonicate with methanol (do not sonicate O-rings or orifice) Bake disassembled unit at 100-150 °C (do not bake O-rings or gauges)

References

1. Method TO-15A Determination of Volatile Organic Compounds (VOCs) in Air Collected in Specially Prepared Canisters and Analyzed by Gas Chromatography–Mass Spectrometry (GC-MS), U.S. Environmental Protection Agency, September 2019. https://www.epa.gov/sites/production/files/2019-12/documents/to-15a_vocs.pdf">https://www.epa.gov/sites/production/files/2019-12/documents/to-15a_vocs.pdf
2. J.S. Herrington, The new U.S. EPA Method TO-15A blog series—part 3: use clean air on a clean analytical system, Restek Corporation (2020)
3. Frequently Asked Questions: Air Sampling, Restek Corporation