Appendix 2. Sampling and analytical methodology

1. Sampling
   Samples were collected by passive diffusion onto Perkin Elmer stainless steel tubes packed with Chromosorb 106, fitted with diffusive heads containing a silicone membrane. Swagelok caps were fitted to both ends of each tube during transit and to the end opposite the diffusive head during sampling.
   
   
2. Analysis
   Tubes were analysed by thermal desorption scanning gas chromatography mass spectrometry (GCMS) for benzene. Benzene was determined by the internal standard technique with formal native compound calibration.
   
   Each sample was spiked with 2 ml of a 40 ng/ml internal standard spiking solution and then loaded into a Perkin Elmer ATD-50 automated thermal desorption system coupled directly to a DB-624 fused silica capillary column, and to the ion source of a VG 7070H mass spectrometer. The mass spectrometer was operated in the repetitive scanning mode at 500 resolution (10% valley definition) in electron impact ionisation mode.
   
   The thermal desorption conditions were:
   

   	Desorption temperature	225°
   	Desorb time	10 minutes
   	Cold trap temperature	-30°C
   	Trap desorption temperature	300°C

   
   The GC Conditions were:
   
   

   	Head pressure	10 psi
   	Initial temperature	40°C
   	Hold time	4 minutes
   	Program rate 1	6°C/minute
   	Temperature 2	70°C
   	Program rate 2	30°C/minute
   	Final temperature	220°C

   
   The mass spectrometer conditions were:
   
   

   	Resolution	500
   	Trap current	500 mA
   	Electron energy	70 eV
   	Source temperature	250°C
   	Interface temperatures	250°C
   	Scan time	1 second per decade from 350 to 35 daltons

   Benzene was initially identified by the generation of the appropriate mass chromatogram, and where a positive response was found, confirmed by a full comparison with a standard spectrum.
   Following analysis 10% of all tubes were checked for residual traces of benzene. Any tubes containing greater than 10 ng of benzene were rejected and re-cleaned.
   
   
3. Calculation of Benzene Concentrations
   
   The mass of benzene collected in the diffusion tubes is expressed as an average airborne concentration (ppb) over the monitoring period. as shown below:
   

   Average Benzene Concentration (ppb)   =	{M (ng) x1000
   	T (mins) x Dc

   Where:
   M is the amount of benzene adsorbed by each tube
   T is the period during which the tube was exposed
   Dc is the diffusion coefficient
   

   Diffusion coefficient   =	D(v) x F x 1000
   	T x C

   Where:
   D(v)= uptake of benzene (ng)
   F = 3.25 = ppm to µg m^-3 conversion factor at 20°C
   T = Time in mins
   C = Concentration (µg m^-3)
   1 ppb = 3.25 µg m^-³
   
   The diffusion coefficient is determined empirically as detailed below.
   
   
4. Determination of the Benzene Diffusion Coefficient
   Benzene tubes were exposed to a known benzene concentration in air generated using a permeation vial held at 50^oC in a glass oven, in turn held in a thermostatic water bath. The purge flow of pure air was generated through a glass ball filled heat exchanger from an Aadco Model 737 Pure Air Generator at a rate of 1 litre/minute.
   
   The generated benzene/air mix was fed to a 30 cm diameter spherical glass reaction vessel. Diffusion tubes were mounted on a carousel turning at approximately 1.2 revs per minute.
   
   Tubes were exposed over a period of two weeks and benzene uptake was determined by thermal desorption and detection with FID using internal standards. The diffusion coefficient was calculated according to the equation shown in Appendix B. A Photovac PID GC was used to determine any losses of benzene in the diffusion coefficient test rig.
   
   
5. Quality Assurance and Control
   In order to validate the experimental method the recovery of benzene from diffusion tubes exposed in a standard atmosphere has been investigated. Tubes are exposed (or 'spiked') for different time periods in a standard atmosphere to give a range of loadings. These tests indicate good recovery of benzene on the exposed tubes, with a mean ratio of 0.9 between the benzene loaded and benzene recovered.
   
   Prior to each analytical run blank diffusion tubes are spiked with concentrations of 10, 20, 50, 100 and 200 ng benzene. Recovery of benzene is excellent, achieving linearity over the range with a standard deviation of less than 5%.
   
   

   ---

   Report prepared by Stanger Science and Environment
   Site prepared by the National Environmental Technology Centre, part of AEA Technology, on behalf of the UK Department of the Environment, Transport and the Regions