Natural Abundance of Mass 47 in CO2 Emitted in Car Exhaust and Human Breath

Atmospheric CO₂ is widely studied using records of concentration, δ ¹³C and δ ¹⁸O, although the number and variability of sources and sinks prevents these alone from uniquely defining the budget. CO₂ of mass 47 (mainly ¹³C¹⁸O¹⁶O) provides an additional potential tracer, but little is known about its ability to differentiate among various budget components. We present study of differences in ¹³C¹⁸O¹⁶O abundance between combustion and respiration. We define Δ 47 as the difference in permil between the measured R47 (=[mass 47]/[mass 44]) and R47 expected for CO₂ whose isotopes are distributed randomly among all isotopologues. Previous studies have shown that Δ 47 values at thermodynamic equilibrium vary between zero at 1000\deg C and 0.9\permil at room temperature, raising the possibility that it could differentiate between CO₂ produced by high temperature processes, such as combustion, and that produced in respiration. Values of Δ 47 are non-linear in mixing. Therefore, it is useful to discuss the δ 47=(R47/R47_ST-1)1000, where R47_ST is the R47 expected for CO₂ having δ ¹³C-VPDB=0, δ ¹⁸O-VSMOW=0 and Δ 47=0. We used a Keeling plot approach to estimate δ ¹³C, δ ¹⁸O, δ 47 and Δ 47 in CO₂ from car exhaust and from human breath. Air sampled at 10am in the Caltech campus in Pasadena, CA, varied in CO₂ concentration from 383 to 404ppm, in δ ¹³C and δ ¹⁸O from -9.2 to -10.2\permil and from 40.7 to 42.0\permil, respectively, in δ 47 of from 32.6 to 34.0\permil, and in Δ 47 from 0.73 to 0.96\permil. We then sampled at varying distances from a car exhaust pipe. The intercepts in Keeling plots defined by these data, reflecting the car exhaust end-member, were similar to the values obtained very close to the exhaust pipe: δ ¹³C was found to equal -24.4±0.2\permil, similar to the measured value of the gasoline used; δ ¹⁸O =30.0±0.4\permil; δ 47=6.7±0.6\permil; and Δ 47=0.41±0.03\permil. Both δ ¹⁸O and Δ 47 are consistent with that expected for thermodynamic equilibrium at 200\deg C between water and CO₂ generated by combustion of gasoline-air mixtures. This temperature is lower than that of the catalytic converter, suggesting re-equilibration in the cooling exhaust as it travels through the tail pipe. This can explain why the δ ¹⁸O of CO₂ from car exhaust is substantially greater than that of O₂ in air. Samples of CO₂ in human breath had δ ¹³C and δ ¹⁸O values broadly similar to those of car exhaust (-22.3±0.2 and 34.4±0.3\permil, respectively), δ 47 of 13.5±0.4\permil, but Δ 47 of 0.74±0.02\permil, far higher than exhaust and similar to that of background Pasadena air. δ ¹³C of human breath is similar to that of car exhaust, much as other respiration and fossil-fuel sources of CO₂ generally overlap. Similarly, δ ¹⁸O of human breath and soil respiration are close to that of car exhaust. Therefore, conventional stable isotope constraints do not easily differentiate fossil-fuel and respiratory sources. In contrast, the Δ 47 value of CO₂ from car exhaust is easily differentiated from those of CO₂ in human breath, largely due to enhanced thermodynamic stability of ¹³C¹⁸O¹⁶O at the low temperatures characteristic of respiration. Hence, Δ 47 is a potentially useful tracer to distinguish anthropogenic, mostly combustion, CO₂ sources from natural, low temperature, sources.