Laboratory Response to Acrylonitrile Problem

1
Laboratory Response to Acrylonitrile Problem

• AMTAC
• May 6, 2009
• John Bricarello
• California Air Resources Board

2
Requirements

• Need to find alternate to 910 sampler
• Cannot pass air through an instrument prior to
  capturing the acrylonitrile
• Must be more sensitive than current method
  (MLD066 DL 0.3 ppb)
• Must be easily integrated in existing sample
  scheme (i.e. cost effective)

3
Alternate Methodology to Canisters

• Impinger Sampling
• Trap sample into organic solvent
• Requires constant attention
• Involves shipping flammable solvents
• Solid Sorbent Sampling
• Collect sample using sorbent packed tubes
• Elute packing with organic solvent or by thermal
  desorption

4
Experimental ResultsSolid Sorbent / Solvent
Extraction

• Charcoal Tubes
• Low concentration recoveries generally less than
  20, mostly zero
• Variable recoveries
• Dilution with extracting solvent led to low
  sensitivities
• Larger air volume Lower recoveries
• Other sorbents tested
• Chromosorb 106, Porapak N, Tenax
• Similar poor results
• Solvent degraded GC column

5
Thermal Desorption (TD)

• Evaluated Thermal Desorption
• Good sensitivity no solvent dilution
• Easy sample preparation
• Placing tubes onto the front of flow controlling
  device eliminates sampler effects
• Shipping stainless steel tubes should be easy

6
Solid Sorbent Evaluation

• Porpak N
• Carbopack B
• Carboxen S-III
• Selection criteria
• Absorbs low molecular weight compounds
• Cited in literature
• Available in lab
• Each had its problems
• Some had long dry and trap desorption times
• Some had problems with entrained water
• Some required high desorb temperatures, producing
  GC artifacts
• Poor recoveries with high (gt50 cc/min) flow rates

7
Proposed Sorbent

• Porapak N
• Good reproducibility for standards
• Low dry and trap desorption times
• Low trap desorption temperature, few artifacts
• Needs low sampling flows
• Detection on tube lt0.1ng therefore combined
  with 24 hour low flow sampling gives estimated
  LOD lt 10 ppt

8
Where We Are

• Have thermal desorption unit operational
• Perkin Elmer TDA650
• Have TD instrumental parameters established
• Flows, Times, and Temperatures
• GC /Column/Detector Parameters
• Varian 3800 equipped with DB-624 (60 m x 0.25 mm
  x 1.4um) connected to Varian 4000 ion trap
• Analytical scheme
• Most consistent air sample results are produced
  using a standard addition technique

9
Where We Are Going

• Continue to develop sampling method
• Low flow pump and manifold on order

10
Where We Are Going (cont.)

• Refine LOD
• Develop in lab method to emulate air sampling
• Investigate appropriate internal standards
• Deploy to field

11
Contact

• John Bricarello, APS
• California Air Resources Board
• Monitoring and Laboratory Division
• jbricare_at_arb.ca.gov
• [phone]