We use a regional-scale, three-dimensional atmospheric model to evaluate U.S. air quality effects that would result from replacing HFC-134a in automobile air conditioners in the U.S. with HFO-1234yf. Although HFO-1234yf produces tropospheric ozone, the incremental amount is small, averaging less than 0.01% of total ozone formed during the simulation. We show that this production of ozone could be compensated for by a modest improvement in air conditioner efficiency. Atmospheric decomposition of HFO-1234yf produces trifluoroacetic acid (TFA), which is subject to wet and dry deposition. Deposition and concentrations of TFA are spatially variable due to HFO-1234yf's short atmospheric lifetime, with more localized peaks and less global transport when compared to HFC-134a. Over the 2.5 month simulation, deposition of TFA in the continental U.S. from mobile air conditioners averages 0.24 kg km(-2), substantially higher than previous estimates from all sources of current hydrofluorocarbons. Automobile air conditioning HFO-1234yf emissions are predicted to produce concentrations of TFA in Eastern U.S. rainfall at least double the values currently observed from all sources, natural and man-made. Our model predicts peak concentrations in rainfall of 1264 ng L(-1), a level that is 80x lower than the lowest level considered safe for the most sensitive aquatic organisms.

Vapor pressure and aqueous solubility are important parameters used to estimate the potential for transport of chemical substances in the atmosphere. For fluorotelomer alcohols (FTOHs), currently under scrutiny by environmental scientists as potential precursors of persistent perfluorocarboxylates (PFCAs), vapor pressure is the more significant property since these compounds are only very sparingly soluble in water. We have measured the vapor pressures of a homologous series of fluorotelomer alcohols, F(CF2CF2)nCH2CH2OH (n = 2-5), in the temperature range 21-250 degrees C by three independent methods: (a) a method suitable for very low vapor pressures at ambient temperatures (gas-saturation method), (b) an improved boiling point method at controlled pressures (Scott method), and (c) a novel method, requiring milligram quantities of substance, based on gas-phase NMR, a technique largely unfamiliar to chemists and holding promise for studies of relevance to environmental chemistry. The concordant values obtained indicate that recently published vapor pressure data overestimate the vapor pressure at ambient temperature, and therefore the volatility, of this series of fluorinated compounds. It was suggested that substantial intramolecular -O-H...F- hydrogen bonding between the hydroxylic proton and the two fluorines next to the ethanol moiety was responsible for their putative high volatility. Therefore, we have used gas-phase NMR, gas-phase FTIR, 2D NMR heteronuclear Overhauser effect measurements, and high-level ab initio computations to investigate the intramolecular hydrogen bonding in fluorotelomer alcohols. Our studies unequivocally show that hydrogen bonding of this type is not significant and cannot contribute to and cause unusual volatility. The substantially lower vapor pressure at ambient temperatures than previously reported resulting from our work is important in developing a valid understanding of the environmental transport behavior of this class of compounds.