ASSIGNMENT OF INFRARED AMMONIA SPECTRA Jonathan Tennyson Physics

ASSIGNMENT OF INFRARED AMMONIA SPECTRA Jonathan Tennyson Physics and Astronomy, University College London Ohio, June 2011

Two projects: 1. (Re-) analysis of data in HITRAN Michael Down, Christian Hill, RJ (Bob) Barber, Sergei Yurchenko 2. Assignment of hot emission spectrum from University of York Analysis of high temperature ammonia spectra from 780 to 2100 cm− 1 Journal of Molecular Spectroscopy (in press) N. F. Zobov, S. V. Shirin, R. I. Ovsyannikov, O. L. Polyansky, S. N. Yurchenko, R. J. Barber, J. Tennyson, R. J. Hargreaves, P. F. Bernath See RD 09 Peter Bernath (Thursday)

The HITRAN database●NH 3 spectrum in 29, 084 lines. 111 0. 083423 9. 385 E-34 1. 991 E-11. 07940. 397 3963. 68610. 910. 000000 0 0 a 0 0 s 20 5 0 a 20 5 0 s 542000 8 8 2 0 0 0 246. 0 111 0. 084660 1. 349 E-34 1. 863 E-11. 07940. 397 4355. 84730. 910. 000000 0 0 a 0 0 s 21 5 0 a 21 5 0 s 542000 8 8 2 0 0 0 258. 0 111 0. 084896 4. 615 E-34 2. 698 E-11. 08680. 414 4320. 95880. 820. 000000 0 0 a 0 0 s 21 6 0 a 21 6 0 s 542000 8 8 2 0 0 0 516. 0 1. G. Guelachvili, A. H. Abdullah, N. Tu, K. Narahari Rao, Š. Urban, and D. Papoušek, J. Mol. Spectrosc. 133, 345 (1989). 2. Š. Urban, N. Tu, K. Narahari Rao, and G. Guelachvili, J. Mol. Spectrosc. 133, 312(1989). 3. I. Kleiner, G. Tarrago, and L. R. Brown J. Mol. Spectrosc. 173, 120 (1995). 4. L. R. Brown and J. S. Margolis, JQSRT 56, 283 -294 (1996). 5. C. Cottaz, I. Kleiner, G. Tarrago, L. R. Brown, J. S. Margolis, P. L. Poynter, H. M. Pickett, T. Fouchet, and P. Drossart, J. Mol. Spectrosc. 203, 285 -309 (2000). 6. C. Cottaz, G. Tarrago, I. Kleiner, and L. R. Brown, J. Mol. Spectrosc. 209, 30 (2001). 7. C. Cottaz, 4 -μm hot bands (thesis, U. Paris-Sud). 8. J. Pearson Rotational, ν 2 - ν 2, and ν 2 -ground state prediction (private communication, 2000). 9. I. Kleiner, L. R. Brown, G. Tarrago, Q. -L. Kou, N. Picque, G. Guelachvili, V. Dana, and J. -Y. Mandin, J. Mol. Spectrosc. 196, 46 (1999). Contains errors Large portions of HITRAN are unassigned. ● ● ● Some regions of the spectrum are missing (and limited to < 5300 cm-1).

• Initial Analysis Step 1: Remove 1027 15 NH 3 transitions Step 2: 1723 unassigned transitions were identified and separated. ● Step 3: 933 lines with conflicting a/s symmetries identified and corrected Step 4: Selection rule checking showed 24 further strictly forbidden transitions. ● 6 corrected by swapping the sign of l ● ● Result: 26 255 “allowed” and assigned 14 NH 3 transitions ● to be checked via combination differences and other measures ● Still gave many errors and inconsistencies. Eg Many upper level errors found to come in bunches corresponding to data from different sources. ●

An Improved Energy Level list. . . Many inconsistencies in lower state energies: Replace by: P. Chen et al. J. Mol. Spectrosc. 236, 116 (2006) For g/s and v 2 levels • Cottaz et al. J. Mol. Spectrosc. 203, 285 (2000) Chosen as most reliable source of 2 v 2 levels. 2 v 4 transition data replaced with original source data. Combination differences now consistent to 0. 01 cm-1 or better. Used to give a new list of empirical energy levels. Some levels more reliable than others.

Corrections prior to search for new assignments

Trivial Assignments Wavenumber is difference between two levels within experimental error, Transition obeys the selection rules. 12 new assignments

Combination Difference Assignments One or more potential upper energies coincide to within the experimental error Transition obeys the selection rules. 116 new assignments Many re-assigments/re-labels

NH 3: Comparison of our predictions with HITRAN ).

Fine tuning potential Before Refinement of the PES: Very elaborate After

Assignment using variational linelist BYTe linelist ●S. N. Yurchenko, R. J. Barber & J. Tennyson , MNRAS, 413, 1828 (2011) ● Compare unassigned lines with BYTe Look for both frequency and intensity matches Initially all the assigned HITRAN lines removed from BYTe list. ● HITRAN

Removal of verified line assignments from BYTe -Match BYTe and HITRAN energy levels Remove assigned HITRAN lines from BYTe All strong [>104 cm-1/mol] BYTe lines found in the HITRAN list. ● Many weaker lines missing from HITRAN BYTe - HITRAN B Unassigned HITRAN

Example: Assignment of 104 cm-1/mol intensity band Problem tackled in decreasing orders of magnitude. 4 -1 ●Intensity above 10 cm /mol are accounted for. 3 -1 4 ● 10 - 10 cm /mol lines -90% matched automatically ●

Assignments ● Surplus BYTe lines suggest missing experimental data from HITRAN.

Trivial using lower energies: 1, 186 Trivial: 12 CD using lower energies: 985 5025 lines I have tried to assign/reassign Progress so far 5025 problem lines 1154 need assigning after initial analysis 82 lines assigned for 104 > I > 103 cm-1/mol 281 lines assigned for 103 > I > 102 cm-1/mol 7 lines missing with I > 103 cm-1/mol 13 lines with I > 103 cm-1/mol outside the HITRAN frequency range Work still in progress

Method of Branches (Polyansky et al, Ap J Lett, 1997) ● Obs-Calc residuals between vary systematically with J and K within a vibrational band. ● Vibrational band errors used to give an idea accurate prediction ● Dependence on J and K within given branch also aids assignment.

www. worldscibooks. com/physics/7574. html About the first edition “The best book for anyone who is embarking on research in astronomical spectroscopy” Contemporary Physics (2006) Just published