Welsh 
Scholarship Reports
Nitrogen trifluoride (NF3)
Jon Robson
The Royal Meteorological Society’s undergraduate vacation research scholarship enabled me to join Professor Keith Shine’s Radiation and Climate Processes research group, at the Department of Meteorology at the University of Reading for 8 weeks in the summer. In my time there I carried out a radiative forcing and global warming potential study of a gas called nitrogen triflouride, working closely with Professor Shine and his research assistant Dr Laila Gohar. This report is a summary of the work that I undertook, and also covers the main points of our findings, which are currently being written up for submission to Geophysical Research Letters.
Nitrogen trifluoride (NF3) is an industrial gas that is primarily used within the semiconductor industry, which at the present time is not included in the basket of gases controlled by the Kyoto Protocol. NF3’s usage has increased over the last decade because the semiconductor industry has begun to use it as a substitute for other fluorinated species. Molecules such as sulphur hexafluoride (SF6) and the perfluorocarbons (PFCs), such as hexafluoroethane (C2F6), have been extensively used within the semiconductor industry as plasma etchants and chamber cleaning gases. In the presence of O2 and high temperatures or plasma, the etching gas in converted to fluorine radicals, which react with the silicon oxide. However, a proportion of the gas is not converted and is vented to the atmosphere where it will accumulate due to the long lifetime of these gases. The fluorinated compounds are potent greenhouse gases because they are extremely stable, and strongly absorb in the infrared radiation. These properties mean that on a per kilogram bases, their 100-year global warming potential (GWP) is normally an order of 10,000 times greater than that of CO2 highlighting the need for careful consideration of their usage.
The semiconductor industry is fast growing, with production increases of 15-17% per annum over a long period. The complexity of devices produced has also increased, therefore requiring more steps in their fabrication, and more steps that require the use of etching gases. It had been estimated that if trends continued the volume of PFCs consumed would be 7-10 times higher in 2010 than in 1995. However there are no known substitutes to these fluorinated compounds and in that perspective the global semiconductor industry agreed to voluntary reduce equivalent emissions of PFCs by 10% in 2010 relative to the 1995 levels.
One motivation for the industry’s increasing interest in NF3 as a source of fluorine, as opposed to the PFCs or SF6, is that there is a higher percentage conversion to fluorine in the cleaning process. Between 90 – 95% of the NF3 used is converted and as a result the emissions of NF3 are smaller than that for alternative sources of fluorine. The usage of NF3 in the semiconductor industry has therefore increased over the past decade, with usage now of the order of 2300 metric tonnes. We believe that emissions are now of an order of 200 metric tonnes per year. Previously there has only been one paper which discusses the climatic impact of NF3 which provides a GWP of 8000, but does not provide details on what it is relative to. Therefore, for a balanced assessment on future increased usage, against other gases, it is important that an accurate assessment of NF3’s effect on the climate is made available.
In order to assess NF3’s climatic effects, new laboratory measurements of the infrared absorbtance of NF3 were determined by the Ford Motor Company. The climatic impact of NF3 was then evaluated by myself, using detailed radiative transfer codes to calculate the radiative forcing for a 1 part per billion by volume (ppbv) increase from zero, in the atmospheric concentration of NF3. The Intergovernmental Panel on Climate Change (IPCC) third assessment report, in 2001, defines the radiative forcing as an externally imposed perturbation to the radiation budget of the Earth’s climate system. It is determined by the change in net irradiance in Wm-2 at the tropopause after an external factor is imposed, in this case it is the increase in atmospheric concentration of NF3. The radiative forcing for a 1 ppbv increase in global atmospheric concentration is called the radiative efficiency, with units of Wm-2 ppbv-1, and is used to calculate the GWP of a gas.
The IPCC [2001] calls for all radiative efficiencies to be calculated in terms of “adjusted” cloudy sky radiative forcing calculated at the tropopause. This is where the temperature of the stratosphere is allowed to adjust so that it remains in global radiative equilibrium. This is because the stratosphere’s adjustment timescale is of a matter of months, compared to decades for that of the tropopause because of the thermal inertia of the ocean. Radiative forcing calculated without stratospheric adjustments are referred to as “instantaneous” radiative forcing.
The adjusted cloudy-sky forcing is calculated using a narrowband radiative model, (NBM) which uses spectra averaged over 10 cm-1 bands. Ideally, a more accurate line-by-line (lbl) radiative transfer would be used to calculate the adjusted cloudy sky radiative forcing; however to include cloud and stratospheric adjustment would still prove too computationally expensive at this time. The line by line code is used to produce a scale factor by comparing the instantaneous clear-sky radiative forcing of lbl and NBM. It is then used to correct the adjusted cloudy sky radiative forcing produced by the NBM. Both models agreed very well for the clear-sky instantaneous radiative forcing, with a discrepancy between the results of less than 2%. The radiative efficiency of NF3 was calculated to be 0.211 Wm-2 ppbv-1, and is estimated to be accurate to 10-15%. This value is approximately 62% higher than the current IPCC [2001] value of 0.13 Wm-2 ppbv-1, with most of this difference due to the increase in the absorption cross section found in the newer laboratory measurements.
NF3’s atmospheric lifetime is 740 years, with produces a 100 year GWP 17200 times greater than that of CO2 on a per kilogram basis. At current release rate of 200 tonnes per annum, we calculate the upper limit of NF3’s concentration in the atmosphere to be 0.16 parts per trillion by volume, making the current contribution of NF3 to radiative forcing trivial. However we offer a significant revision in the radiative efficiency and GWP of NF3 and so raise its importance as a greenhouse gas.
In conclusion the study has shown that there should be a significant revision of the radiative efficiency and GWP of NF3 and a need for the IPCC to revise the value it uses. The study was important as it offers an up to date value for the radiative efficiency and GWP so that an assessment of the desirability of the usage of NF3 is a balanced one. I would like to thank the Royal Meteorological Society for giving me this opportunity and for sponsoring me to undertake this work. I believe that I have learnt a lot form my eight weeks undertaking meteorological research, but also about my abilities undertaking such tasks in future and have received some invaluable experience.