Quantitative Modelling of HIgh Temperature Spectrum of Oxygen (O2)

Oxygen (O2) is one of the most ubiquitous and important molecules on Earth today; both life and combustion rely on its presence. However, while the room temperature spectrum of oxygen is well understood, the spectrum at higher temperatures is not. Extensive high temperature line lists (set of frequencies and intensities of all transitions) produced by high level theoretical techniques utilising all existing experimental data have proved extremely valuable for other molecular species (CO2, H2O, NH3); the production of a similar list for O2 is this project's primary goal. However, O2 raises significant, novel and interesting scientific and technical challenges due to oxygen's complex electronic structure and the fact that all infrared and visible transitions are forbidden by electric dipole, with only weak quadrupole or magnetic dipole transitions allowed. The resulting hot O2 line list will be made widely available to assist in understanding combustion and other high temperature gas environments.

Ultracold Alkali Metal Dimers

Molecules like Rb₂ and LiCs may not seem like reasonable chemical species – they are usually bound more than 10-20 times more weakly than CO. But for ultracold molecular physicists, these alkali metal dimer species are their key weapon of choice in the search for unusual quantum behaviours, unusual chemistry, new physics and a better understanding of the universe. Turns out that these sorts of species can be formed at nano-Kelvin temperatures by using light to form bonds between laser cooled ultracold atomic species such as Li, Na, Rb and Cs. Their ultracold temperature means quantum mechanics is everything and their interactions can be controlled exquisitely to allow discovery of new unusual science. To facilitate these experiments, experimental designers must have access to detailed complete descriptions of their coupled rovibronic structure including their absorption properties. In this project, you will be producing this data for a set of molecules using diverse tools including high-level electronic structure calculations, a newly developed nuclear motion code and literature experimental spectroscopic data.

Intramolecular Energy Flow in Small Molecules after Excitation

Despite their size, at high temperatures, small molecules (with 2-5 atoms) often have thousands to millions of relevant rovibronic or rovibrational energy levels and millions to billions of relevant transitions. The question of what happens to a small hot molecule after excitation to a high level rovibronic or rovibrational energy level thus becomes quite complex. Many undergraduate, post-graduate and post-doctoral projects in the ExoMol group at University College London have been devoted to producing so-called “line-lists” containing data for all the relevant energy levels and transitions for small molecules including water, ammonia, methane, vanadium oxide and more. Part of this data is the intensity of transitions between various energy levels; this data is waiting to be explored in more depth to answer new questions about relaxation processes in isolated molecules.

In this project, we will be using the existing ExoMol line-list data to ask new questions and gain new understanding. Does relaxation from an excited state to the ground state generally only take a few steps, or hundreds? Is the rotational energy generally lost before vibrational energy, is it the reverse or is it a mix? How common is vibrational relaxation, internal conversion and intersystem crossing in an isolated molecule that doesn’t collide with other species? How does this depend on the molecule and on which particular energy level is excited? In this project, you will be given the data, shown how to ask questions of the data and left to explore, asking new questions and finding new answers.

Updating A Highly Cited Reference Database

In 1979, after 10 years of work, Huber & Herzberg released a collation of all existing data on diatomic spectra, including results from about 943 molecules, though for around 300 this only included dissociation energies. And from 1979 - 2018, more than 15,000 scientists have been using their data; on average, the database is still cited about once a day.

And at the same time, a whole army of experimentalists have been producing updating the data, producing new data often on completely new molecules, with theorists working hard to understand the electronic complexity of many of these small but very unusual species (e.g. ArXe!). But this new data is not readily available and thus not cited once a day by scientists worldwide.

My goal is to change this, to get new data of provable usefulness into a modern format, an online queryable database. To achieve this goal, I need to work with a big team with varying expertises and levels of experience.

So far, this has included:

• Systems Managers (trained as computational chemists)

• ORBYTS tutors to lead groups of high school students to find molecular constants

• High school students given a molecule and asked to make it their own, finding out everything that has been found out about their molecule (and, ideally, suggesting what needs to be done next!)

In the future, we need to

• Further develop the online database system, including its website

• Refine the course for delivery in ORBYTS programs

• Import data from existing online databases, e.g. NIST Chemistry Webbook (which has digitisation of original Huber-Herzberg database) and Computational Chemistry database

MARVEL analysis: Obtaining Experimental Energy Levels from Assigned Experiment Transitions

Experimentalists measure transition frequencies, or differences between energy levels, but, as theoreticians and modellers, it is much more useful to work with energy levels when, for example, we compare against theory, refine potential energy curves and produce line lists.

Further, a set of transition frequencies can be internally inconsistent; i.e. there is no set of energy levels that reproduces all the transition frequencies correctly. This issue can be particularly serious if there is a large amount of data from varying sources, but occurs to some extent will all spectroscopic data as the experimental frequencies have non-zero uncertainties.

MARVEL is a software code developed by Tibor Furtenbacher, Attila G. Császár (Hungary) that takes as input assigned transitions with uncertainties and produces energy levels with uncertainties. The program works by building up spectroscopic networks (graphs) by which energy levels (nodes) are connected by transitions (lines); the quality of the energy level determination increases when more transitions are to or from that level. The internal process allows identification of transitions that don't fit into the overall spectroscopic network and recommends increased uncertainties. Using this output as guidance, the user identifies mis-assigned lines, typos and underestimated uncertainties as an iterative process to work toward a self-consistent set of input transition frequencies with output energy levels.

In practice, a MARVEL project proceeds as follows:

1. Find all available spectroscopic data for the molecule via a thorough literature review

2. Digitise available spectroscopic data

3. Convert data to MARVEL format, including consistent quantum numbers

4. Use MARVEL online interface to produce the self consistent list of input transitions and output energy levels

MARVEL projects vary considerably in scope depending on (1) the complexity of the molecule's quantum numbers/ assignments and (2) the quantity and quality of the available spectroscopic data. In some cases, MARVEL may be used as a useful tool to provide energy levels for a line list fit; in other cases, the MARVEL analysis is a significant body of work in itself and is highly publishable (for example, H2O, TiO, C2, NH3, CH4 etc).