Silicon-Lined Canister Cleaning Practices and Blank VOC Concentrations Jason S. Herrington, Gary Stidsen, Trent Sprenkle, Steve Kozel

U.S. EPA Method TO-15 A “guidance document” for a Performance Based Method (PBM) Means a laboratory can meet performance criteria by hook or crook

NEMC Canister Blank/Cleaning “Trilogy” 2015 Evaluated a proprietary cleaning procedure for stubborn canisters 2016 Evaluated storage time, sweep gas, and temperature impacts on blank concentrations 2017 Fill gas impacts Un-Coated vs. Silicon-Lined

Why all the focus on canister blanks/cleaning? Compound 1-in-1-million cancer risk (pptv) Noncancer effects (pptv) Acrolein NA 9 Naphthalene 5.6 570 Propylene Dichloride 11 870 1,3-Butadiene 15 900 Acrylonitrile 6.8 920 Ethylene Dibromide 0.22 1200 Methyl Bromide 1300 Benzene 39 9200 Carbon Tetrachloride 26 16000 Vinyl Chloride 44 38000 1,4-Dichlorobenzene 133000 Ethylbenzene 92 230000 1,1,2,2-Tetrachloroethane 2.5 Ethylene Dichloride 9.5 Tetrachloroethylene 25 Trichloroethylene 93 GC-MS systems are getting more sensitive Preconcentrators are advancing Canister cleaning remains unchanged…

TO-15 Blank Criteria 8.4.1.6 At the end of the evacuation/pressurization cycle, the canister is pressurized to 206 kPa (30 psig) with humid zero air. The canister is then analyzed by a GC/MS analytical system. Any canister that has not tested clean (compared to direct analysis of humidified zero air of less than 0.2 ppbv of targeted VOCs) should not be used.

“Guidance” for Canister Cleaning 8.4.1 Evacuate down to 500 mTorr Hold under vacuum for 60 minutes Fill with humidified “zero air” to 30 psig Repeat cycle two additional times for a total of three cycles

Blank Contributions Canister Cleaning Oven Fill Gas Storage Time Uncoated Coated Cleaning Oven Cycle Time Temperature Cycles Sweep Gas DI Water for Humidification Fill Gas Storage Time Connection Lines Preconcentrator GC-MS

Storage Time Recap from 2016

Sweep Gas Recap from 2016

Cleaning Cycles Last year I used 12 cleaning cycles, so why not 36? 12 canisters Cleaned 12 times and analyzed Cleaned 24 more times and analyzed All canisters filled to 30 psig with 50% RH air 1000 mL sample volumes SIM calibrated from 10 to 300 pptv

Ethylene to Hexachlorobutadiene Rxi-1 105 m x 0.25 mm x 1.0 µm -45 ⁰C for 4.0 min then 4.5 °C/min to 125 °C for 0 min then 9.0 °C/min to 250 °C for 0 min Only 2 critical co-elutions. 2-Methylpentane and vinyl acetate; and n-hexane and ethyl acetate. 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00

36 Cleaning Cycles Results Un-Filtered House Air 12 Cycles 36 Cycles Average n = 12 Propylene 246 199 Dichlorodifluoromethane (Freon® 12) 156 110 n-Butane 281 327 n-Pentane 167 Trichlorofluoromethane (Freon® 11) 26 22 Ethanol 11 49 Carbon disulfide 218 363 Acrolein 41 30 Isopropyl alcohol 111 Methylene chloride 1069 227 Acetone 14 Hexane 536 136 Tetrahydrofuran 36 39 2-Butanone (MEK) 297 38 Heptane 79 Benzene 150 245 Toluene 325 58 2-Hexanone (MBK) Naphthalene 7 12

Newly Renovated Innovations Lab

The “Fix”

Fill Gas Fix House Air 12 Cycles S-Trap Average n = 12 Propylene 246 ND Dichlorodifluoromethane (Freon® 12) 156 n-Butane 281 n-Pentane 167 Trichlorofluoromethane (Freon® 11) 26 Ethanol 11 13 Carbon disulfide 218 177 Acrolein 41 Isopropyl alcohol 21 Methylene chloride 1069 16 Acetone 111 19 Hexane 536 Tetrahydrofuran 36 12 2-Butanone (MEK) 297 56 Heptane Benzene 150 18 Toluene 325 9 2-Hexanone (MBK) Naphthalene 7

36 Cleaning Cycles Conclusions “IDK” as I got derailed by the lab line contamination Lesson learned: “Garbage in… garbage out” Fill gas has more of bearing on blank concentrations than probably all the other variables put together The zero in “zero air” is a relative term

Fill Gas Evaluation Since the fill gas seemed almost more important than the cleaning itself I took 6 new canisters and cleaned them and 6 new canister as is Aged appropriately and analyzed

Fill Gas Evaluation Un-Filtered Fill Gas Uncleaned Cleaned Uncleaned2 Average He Average Air n = 6 Propylene 222 150 60 79 Dichlorodifluoromethane (Freon® 12) ND 87 89 n-Butane 38 17 244 202 n-Pentane 3121 3142 Trichlorofluoromethane (Freon® 11) 37 35 Ethanol 4 3 14 18 Carbon disulfide 163 140 496 Acrolein 50 28 86 Isopropyl alcohol 95 72 152 153 Methylene chloride 12 123 Acetone 70 57 78 Hexane 737 726 Tetrahydrofuran 147 93 208 137 2-Butanone (MEK) 203 174 165 158 Heptane 27 Benzene 179 Toluene 22 53 96 2-Hexanone (MBK) 111 75 1420 102 Naphthalene 15 5 25 30

Fill Gas Conclusion The “zero-ness” of the zero air has more bearing on blank concentrations than anything else This statement applies to clean canisters The canister cleaning practices play a more significant role for dirty canisters However, once a dirty canister becomes clean… zero air takes over again

Silicon-Lined Canisters Customers always ask “what is the difference between coated and un-coated canisters?” The response usually has focused on sample stability and not blank cleanliness Obviously I wanted to generated the data so as to provide a reasonable response

Un-Coated vs Silicon-Lined 6 un-coated canisters and 6 silicon-lined canisters Filled with He, Filtered He, Air, Filtered Air and analyzed Trying to evaluate the actual difference in the canisters, not just the impact of the fill gas

Un-Coated vs Silicon-Lined Blanks Un-Coated Silicon-Lined Average n=24 Propylene 62 46 Dichlorodifluoromethane (Freon® 12) 75 68 n-Butane 277 201 Trichlorofluoromethane (Freon® 11) 36 35 Ethanol 170 ND Acrolein Isopropyl alcohol 183 135 Methylene chloride 110 Acetone 52 22 Tetrahydrofuran 382 326 2-Butanone (MEK) 56 29 Heptane 50 61 Benzene 146 129 Toluene 63 26 2-Hexanone (MBK) 65 64 Naphthalene 156 16

Un-Coated vs Silicon-Lined Conclusions Virtually indistinguishable, with a slight nod toward silicon-lined This is intuitive, considering the surface differences Would it even be statistically significantly different? Ethanol and naphthalene clearly show differences

Thank You! Co-authors Markes International