Factors concerning NO2 uptake by the absorbent triethanolamine (TEA) in NO2 diffusion tubes are examined. Although the nominal freezing point of TEA is 17.9-21.2 degrees C, we show that, for a range of aqueous TEA solutions (0-20%, H2O), no freezing occurs even at -10 degrees C. Therefore NO2 collection efficiency is unlikely to be impaired by low temperature exposure. The recovery of TEA from the meshes of exposed samplers is determined as approximately 98%, even after 42 days, showing that the stability in situ of TEA is unaffected by long-term exposure. A model of a diffusion tube sampling array for simultaneous exposures, with a 0.1 m sampler spacing, shows that NO2 uptake by individual samplers is not affected by the presence of neighbouring tubes in the array. This is confirmed by sampler precision at two Cambridge sites. Four sampler preparation methods are compared for differences in NO2 uptake of exposed samplers. All methods employ TEA as absorbent, transferred by either dipping meshes in a TEA-acetone solution or pipetting aliquots of a TEA-H2O solution onto the meshes. For samplers prepared by three of the methods, no difference in NO2 uptake is found, but for samplers prepared using a 50% v/v TEA-H2O solution, a mean reduction of 18% is found. Student's t-tests show that the difference is highly significant (P < or = 0.001). Reasons for the difference are discussed.