A Middle Way: A Review of Physical + Chemical Pathways to Soot Inception

tl;dr Our new preprint "Carbonaceous nanoparticle formation in flames" is out.

The paper is now published in Progress in Energy and Combustion Science

 

A middle way can refer to many things. In common usage it refers to a comprise between two positions. In philosophy or religion, it can refer to a rejection of extremes as exemplified by Aristotle’s golden mean that “every virue is a mean between two extremes, each of which is a vice”. In logic it can refer to a fallacy - halfway between a lie and a truth is still a lie and therefore some care is required in proposing such compromising positions. In science it has been used for a variety of justifiable and unjustifiable positions. One famous example being the middle way between physical scales and another being a position we recently put forward for the formation of the pollutant soot.

In the influential paper “The middle way” published in the Proceedings of the National Academy of Sciences of the USA in 2000, Laughlin et al. discussed the challenge in probing the scale between the atomic and macroscopic dimensions. In this mesoscopic region significant gaps exists in our understanding of how atoms and molecules interact, organise and form complex structures. This intermediate scale is too large to be measured by analytical chemical approaches and too small to be approached from the macroscale. Examples include protein folding, high temperature superconductors and disordered or topologically frustrated materials.

Our recent study on the formation of the pollutant soot illustrates the challenges probing the mesoscopic scale nicely. Figure 1 below shows a schematic of the transformation of fuel molecules into the pollutant soot. Only in the last 5 years have experimental techniques allowed for the aromatic soot precursor molecules as well as the earliest nanoparticles to both be directly imaged. Mass spectrometry has also allowed for the mass of the clustering molecules to be measured during soot formation. However, the mechanism by which these molecules cluster continues to baffle combustion scientists. The prize sought is the ability to understanding and potentially halt the emission of these toxic pollutants from internal combustion engines that damage almost every organ in our bodies as well as contribute to climate change. 

Figure 1 – Schematic for the transformation of fuel into soot inside a flame with insets showing the experimental results from which the schematic is derived. High resolution atomic force microscopy (HRAFM)from Commodo et al. 2019, Helium ion microscopy (HIM) from Schnek et al. 2013, high resolution transmission electron microscopy (HRTEM) from Martin et al. 2018 and scanning electron microscopy (SEM) from Orion carbons. 

Our modelling efforts also struggle to traverse the molecule to nanoparticle transition in soot formation. There are two main classes of models that have been proposed for soot formation. The first is physical nucleation where aromatic molecules grow until the intermolecular interactions between the molecules allows them to stick together and condense. The second is chemical inception where bonds form between the molecular systems. Only recently have accurate computational approaches been developed to explore these suggestions.

Concerning physical nucleation, Prof. Kraft’s group worked with the physical chemist Prof. Alston Misquitta (Queen Mary University) in the 2010s to accurately compute the intermolecular interactions between aromatic species (using a symmetry adapted perturbation with a hybrid density functional approach). From these results it was clear that the clustering species seen in the flame are far too small to possess the significant intermolecular energies required for physical nucleation mechanism. For my PhD, I explored electrical enhancements to physical nucleation that arise from curved aromatic species that possess a strong electric polarisation. While this electrical effect may help explain the electrical control of soot formation it alone cannot justify a nucleation mechanism either.

Concerning chemical inception, we recently undertook a systematic study of the bonds that could form between reactive aromatic soot precursors with Prof. Xiaoqing You’s group at Tsinghua University (made possible by the CARES programme). This was only possible due to the direct imaging of the reactive aromatics in 2019 (see Figure 1) and the recent advances in density functional computational techniques optimised for radicals (the meta hybrid GGA density functional method M06-2X). Figure 2 shows the systematic comparison that was possible with such an approach for small aromatic molecules. The green coloured grid squares correspond with thermally stable species. Mr Angiras Menon was recently able to compute the rate at which each of these crosslinks forms and compared them with the speed of soot formation. We found that for these small species none of the crosslinks formed sufficiently fast enough to explain the rapid clustering of molecules into soot nanoparticles.

Figure 2 – Bond energy between various reactive aromatic soot precursors. Green indicates bonds that have enough thermal stability to be considered as important in flames.

These detailed studies left us with the uncomfortable conclusion that the two main routes proposed for soot formation were unable to describe it. However, something did catch our attention crosslinks that allowed the molecules to both bond and stack, see Figure 2 B), C) and D) sites. This opened up another possibility that both physical and chemical mechanisms could cooperatively contribute to soot formation. Upon exploring these possibilities, we found that π-radicals on five membered rings, site B), formed highly localised states that did not become deactivated as the molecule grew in size, unlike their hexagonal ring equivalent, thereby remaining highly reactive. This allowed for an additive contribution between the physical interactions and the chemical bond only in these so-called aromatic rim-linked hydrocarbons (ARLH). Figure 3 shows the various mechanisms placed on a C/H versus molecular weight schematic to show the middle way suggested.

Figure 3 – A middle way is schematically shown between physical and chemical mechanisms for soot formation. 

As mentioned at the beginning of this article claims to middle ways are poor arguments unless they can be justified. Currently, we have shown that the addition of physical interactions and chemical bonding considerably increases the thermodynamic stability of aromatic rim-linked hydrocarbons. However, we have yet to show that such species can explain the rapid formation of soot in the flame. This requires the collision efficiency between these species and the concentration of the localised π-radicals on five-membered rings to be determined. Experiments are underway in the community to probe such species and close this missing gap between the micro and mesoscale of soot formation.

Are any reactions fast enough for soot formation?

tl;dr In order to stop soot pollution, we need to know what reactions cause soot to form. We used the computer to work out how fast a variety of different reactions between soot molecules to see what reactions could be forming soot. Most of the reactions are too slow and suggest larger molecules are required. 

Soot continues to be a problem for our climate by warming the atmosphere and melting ice. It also damages our bodies and causes significant health impacts. Some recent studies are coming out showing a strong relationship between polluted areas and places where the COVID-19 virus has taken many lives. For example in 66 administrative regions in Italy, Spain, France and Germany, 78% of COVID-19 deaths occurred in the five most polluted regions. (Ogen 2020). With a recent study based in the USA finding that for every 1 microgram per cubic metre of PM2.5 soot pollution is associated with an 11% increase in COVID-19 death rate? (Wu et al. Sci. Adv. 2020). Frustratingly we are still unable to describe how soot forms at the molecular scale and this is inhibiting our ability to reduce the emission of these toxic pollutants.

In this paper, my coworkers and I were able to run a series of calculations on the computer to systematically compare the speed of many reactions thought to happen in the flame. We refined a table of bond energies that we proposed in a previous paper (see this blog post) and by reordering the grid we found that we could categorise reactions into four main classes depending on the type of reactive site involved. 

Next we computed the reaction rate between each of these bonds using transition state theory. This involved computationally stretching the bonds until they were about to break and then determining the likelihood of a collision between these molecules leading to that transition state and ultimately the product with the bond formed. This allowed for a map of reaction rates versus temperature to be plotted and for the various reactions to be compared. 

Surprisingly we found that for all of the reactions between these small aromatics the reaction rates are too low to explain soot formation. This includes all of the mechanisms proposed to date involving small aromatic molecules found in flames. 

So we looked for various effects that could stabilise and enhance the reactions as the molecules enlarge. We found that for the localised pi-radicals the dispersion forces could enhance the equilibrium constant for dimerisation. It is unknown how this effect will impact the forward and backward rate constants but it is suggestive of an enhancement to the forward rate. 

There is more work to be done to work out whether this stabilisation of the larger localised pi-radical dimers will speed up the reactions to explain soot formation and whether they are in high enough concentration. However, we think the main contribution of this paper is being able to rule out a large number of possible reactions that have previously been proposed for soot formation. This is discussed in more detail in the review article that is currently online as a preprint. 

Reflections on my PhD and what's next?

My PhD is officially conferred this month as well as my fellowship in Western Australia. I wanted to take the opportunity to reflect on my experiences and talk about what's next. 

Cambridge and the University Library

My PhD journey began in 2016 when I arrived in the UK and settled into Churchill College in Cambridge. The Chemical Engineering building was then on Pembroke Street, in the centre of the city (it's since moved out of the city to West Cambridge). I will always remember walking to the market to have lunch. The walk took us past the old Cavendish Laboratories and The Eagle pub where Watson and Crick first announced their discovery of DNA. I will also remember heading to Queens College over the Mathematical Bridge to have lunch in their dining hall followed by frisbee on the green next door. 

The place that is most memorable for me, however, is the Cambridge University Library. This early 20th-century building was absolutely full to the brim with books. There was barely enough room to move past the shelves. You would often find me in the carbon materials science section this was on the left side of the tower in the photo above. There were narrow desks by the windows and a radiator for heating the building. I will always remember biking to the library in the cold of winter to sit next to the radiator while I pored over books. At the end of my PhD, I felt privileged to able to deposit my thesis in this magnificent library. 

Planning

I had planned out my PhD with great detail at the beginning of my first year, but new experiments performed in my and other groups soon redirected my research. This helped me understand my supervisor's reluctance to plan this kind of work too far in advance. I soon realised that my initial plans, while helpful for setting me off in the right direction, were not predictive of the final thesis I submitted. 

An example of this was a paper presented at the 37th International Symposium on Combustion in 2018 (see my blog post on the event). This revealed clear evidence for crosslinking between soot molecules.  This led me to redirect my efforts into exploring the reactivity of soot molecules using computational approaches (see post on this). I also worked directly with experimentalists in the group to find some of the first evidence of curved molecules in early soot particles and to see the impact of pentagon-containing fuels. 

The plan from my first-year report, which was not followed. 

Travel

I had many opportunities to travel during my PhD for which I have to thank my supervisor, Prof. Kraft. 

In my first year I was able to attend the IChemE conference at Bath University and present work on gas molecules interacting with soot. This was a great opportunity to begin to understand the field of chemical engineering and the tools available. I was also able to attend the NanoTec16 conference at Trinity College Dublin, where I had the pleasure of meeting the late Professor Malcolm Heggie and hear him perform one of his famous songs at the conference dinner.

In my second year, I attended the Carbon 2017 conference in Melbourne on my way back to visit my family in New Zealand. At this conference, I began a collaboration with the Carbon Group at Curtin University that led to a letter being published in Physical Review Letters (and to my next research position). Another memorable event was the memorial for the late Professor Mildred Dresselhaus the "Queen of Carbon Science" who I did not meet but whom I have heard a lot about. Read more in my post on the conference and Prof Dresselhaus. 

In my third year, I attended the 37th International Symposium on Combustion 2018 in Ireland. Some of the most memorable moments from this conference were doing my first oral presentation at a scientific conference, the farewell dinner for my colleague Dr Maria Botero and seeing the first images of soot molecules presented. We managed to find a large Airbnb house very close to the conference where a group of us stayed. One night, we were challenged by a professor about whether curved molecules would invert in a flame and not be persistently polar. We went home early from the conference and a group of us managed to calculate the rate of inversion for a large curved aromatic molecule. This involved staying up late into the night and reprogramming software to include the many internal vibrations in these large molecules. To our delight we found it to be very slow - in fact only one flip every two years. We were then able to show this to the professor the next day and get some further critiques. This suggestion then lead to a nice paper and was a good example of the sort of fun challenges that come from a vibrant community. 

In my fourth year, I attended the Carbon 2019 conference in Kentucky, US with my colleague Angiras. On the way there, I had an opportunity to visit Prof. Green's group at MIT and Dr West's group at Northeastern. I also presented work at Penn State University, meeting many of the top carbon scientists in the field, which was kindly organised by Prof. Terrones' group. The kindness and generosity of these groups were overwhelming and I am sure much will come out of these meetings. At the Carbon conference, I was able to present four different talks - three of my own and one on behalf of a colleague. This was quite a lot to prepare for but I wanted to make sure I justified the travel. I was able to present the 3D graphene work for the first time and managed to 3D print some of the models of disordered carbons to pass around and discuss with people at the conference. There were some mixed reactions and some good discussions that came from the talks. I also very much appreciated the memorial for the pioneering carbon scientist Madame Oberlin (read more about her and the conference in this post).

After my first year in Cambridge UK, I spent the rest of my PhD programme at the University of Cambridge's Singapore centre (Cambridge CARES). This was an amazing opportunity to work with researchers at Asia's top universities, NUS and NTU. The CREATE programme, set up by Singapore's National Research Foundation, also invites other universities to set up centres and encouraged collaborations between these universities. This allowed me to participate in the Commonwealth Science Conference (see blog post), conferences with Shanghai Jiao Tong on biochar for carbon capture and a policy workshop with ETH Zurich.

The Singapore experience also helped me to see the challenges facing the world. In Cambridge, it is easy to think that many of the problems will be solved with some new technology and everyone getting on a bike or using more public transport. Being in Singapore where rising water, increasing temperatures and more erratic weather pose an existential crisis, this is not as easy to brush aside. Singapore is the busiest port in the world and there are endless ships in the harbour, belching out soot. Singapore also hosts the largest oil refinery in Southeast Asia. It brought home to me the significant challenges we face in supplying clean energy to a world reluctant to wean itself off fossil fuels and has lead me to the next project I am undertaking. 

@Chung Kevin https://pxhere.com/en/photo/1609054

Travelling also meant my wife and I were present for many significant events. We were in the UK for the vote concerning EU membership (and were even able to vote due to being Commonwealth citizens). We were also able to take a ferry over the channel to help refugees in Calais. In Singapore, we were on the street by chance when Kim Jong Un drove past to his meeting with President Trump. It was hard to know how to respond to being so close to a dictator and everyone in the crowd was eerily silent. We were also in New Zealand and Singapore during the coronavirus pandemic. With such restrictions on travel at the moment, we feel fortunate to have been able to travel as much as we did.

Collaborations

At the beginning of my time in Cambridge I mentioned in a blog post that there was less collaboration than I thought there was going to be (link to post). I think I was wrong about that. What I realised is that to complete a project required extreme focus and during the generation of results there was little need for collaboration. However, when it came to interpreting the results, extending them and deciding on the next steps I saw the intense collaboration and guidance that took place. 

Working closely with postdocs was highly valuable. My first 1st author paper from my PhD would not have been half as good if not for the careful guidance of Dr Radomir Slavchov. In subsequent work, I saw the importance of this guidance at critical points in my PhD from postdocs and my supervisor. Guidance from others in Cambridge was also invaluable, for example Dr Alan Hayhurst, Dr David Farrien-Jimenez, Dr David Wales, Dr Caterina Ducati, Dr Alston Misquitta and Dr Clive Wells.

Collaborations with international groups also provided new directions and considerable energy. Dr Pascazio visited for a month during her PhD in Italy and over that time produced the main results for the hardness studies that were published recently. She is now a postdoc in the group and we are working closely together on a number of projects. Dr Dingyu Hou also visited for a year during her PhD at Tsinghua University. Her recent computational calibrations of various methods were critical for the paper on reactive crosslinks. 

Working with Laura, Dingyu, Angiras, Gustavo, Kimberly, Chung and others in the CoMo group led to some great insights into soot formation and I have to thank them all. 

CoMo Group and collaborators at the combustion symposium 2018

Challenges

Collaborations are challenging in research - on the one hand, groups have a friendly competition that drives the field forward. However, this does not encourage collaborations and significant trust must be established to achieve productive collaborations. Unfortunately, during my PhD there were times when groups broke our trust and made research more difficult. Examples included trying to block papers being published and having work copied. As a PhD student, this was quite disheartening. However, I learned that you have to keep doing good work and you will be rewarded eventually. 

Another challenge that is not unique to my PhD was arranging visas for me and my wife. I was surprised to learn that spouses of PhD students in Singapore are not eligible to apply for a dependent's visa. After a few months of unsuccessful job searching, my wife travelled back to New Zealand for three months while we sorted out a different kind of visa. My research organisation was very helpful and in the end my wife was able to find a job and live in Singapore with me. However, it was frustrating that it's assumed that anyone doing a PhD does not have a family. 

The final challenge surprised me and that was mental health. Throughout the PhD there were some times when I did not cope as well as I would have hoped. Through support from my friends, family, my very supportive wife and visiting a counselor I managed through the stress. This taught me the struggles that many PhD students face and that I am not invincible. 

Finally, coronavirus was the final challenge in my PhD. Though I am particularly fortunate in having done most of my work computationally there were some aspects I missed out on due to the coronavirus. I was unable to have an in-person oral examination for my PhD defense. This was a shame as I wanted to be physically present to defend my thesis and then to celebrate with my colleagues afterward. It would have also been nice to hand in my final thesis and to actually see it. However, there were some positives to having a virtual defense. We could go through the literature online and look through the thesis on my computer in a way I do not think we could have done in person. However, if I could do it again it would be nice to have done it in person and to have shaken my examiners' hands afterward. Finally, I graduated in absentia and do not know when I will be able to travel to the UK to attend a formal ceremony. This is something to look forward to as many of my friends will be graduating with me and will be a signal that the world has moved back to some sort of normalcy.

Defending my PhD in lockdown

What's next? - Hydrogen in Perth

I have accepted a Forrest Fellowship in Perth to undertake research in the area of hydrogen storage. This fellowship is based at Forrest Hall, which brings together PhD students and Fellows from around the world to work in Western Australia. The fellowship funds my research for three years and allows me to mentor the PhD students at Forrest Hall. 

The research I will be undertaking will bring together two areas of impact that I care deeply about: decarbonisation and the science of carbon nanomaterials. The goal is to make a 'sponge' for hydrogen out of a novel material called 3D graphene. This will allow for the storage and transport of green hydrogen produced from solar or wind power. 

What makes the project so exciting is the groups I will be working with in the Physics department at Curtin University. The Carbon Group, headed by Dr Nigel Marks and Dr. Irene Suarez-Martinez, has developed some of the most accurate descriptions of carbon materials to date and I have collaborated with them previously (read more in this blog post). The second group is the Hydrogen Storage Research Group headed by Prof. Craig Buckley, a world leader in hydrogen storage materials. Bringing these two world-leading groups together, the aim is to push carbon materials to hydrogen storage capacities not yet achieved.

Combustion webinar - Carbonaceous nanoparticle formation in flames

At the end of last month Prof. Kraft, my PhD supervisor, presented the recent work on soot formation from the Computational Modelling group. I helped plan out the talk with colleagues and it is a very nice overview of the research we have been doing recently on the formation of soot. It is definitely for a more technical audience so be warned. For a less technical description, I recommend the webinar I recently gave at Churchill College. 

Are new reactive molecules present in flames?

tl;dr

Which reactive molecules lead to the formation of soot? Recently we showed that adding hydrogen to the edges of soot molecules makes the edges reactive. In this paper, we showed that these sites are very common in flames, making them likely to be important for soot formation.

In my previous blog post, I talked about our systematic comparison of possible crosslinks between reactive aromatic molecules that had recently been detected using atomic force microscopy. This gave the bond energies and allowed us to work out which bonds could be stable at flame temperature. It showed that adding hydrogen to a pentagonal ring gives a reactive localised π-radical that allows crosslinking and stacking.  

In our recent paper that just got accepted in the Proceedings of the Combustion Institute, "Reactive localised π-radicals on rim-based pentagonal rings: properties and concentration in flames" (see preprint here), we showed that these reactive sites are present in the flame in significant concentrations. 

To show what we mean by localisation the electron's spin density is shown below. The spin density shows where the reactive electron is likely to be able to form a bond with higher values indicating higher reactivity. We find two classes of π-radicals, those that 1) delocalise in 6-membered ring aromatic molecules and 2) reactive localised π-radical for pentagonal rings or methylene (–CH2) that does not delocalise across the molecule. 

While these reactive sites have been seen in molecules sampled from the flame it was not clear whether they also exist in the flame. For example, sampling these molecules from the flame could lead to hydrogen being added while the molecules in the flame could actually be lacking this hydrogen i.e. it could be an artifact of sampling. 

In order to determine if these species are present in the flame, we calculated all of the reactions that could allow hydrogen to be added or removed (thanks to Angiras and Dingyu for this). We could then consider, given the concentrations of hydrogen species in the flame, what sort of concentration we would expect. The reactions are shown below (those barrierless reactions do not make computing the rates very easy but it can be done with sufficient approximations).

We found that between 1-10% of the molecules contained a localised π-radical on their rim-based pentagon (for between 1400-1500K which are temperatures within a flame where soot begins to form). Comparing these results with the HR-AFM structures recently imaged we found a consistent frequency of rim-based pentagonal sites with a ratio of 27:12:4 for the unsaturated, saturated and partially saturated rim-based pentagonal rings, showing that these species are present in the flame in significant concentrations.

We explored another exciting possibility - that multiple localised π-radicals are present on a single molecule. These species are also likely to be present in reasonable fractions (thanks to Gustavo and Angiras for developing the KMC simulations). Of the molecules that were recently imaged using HR-AFM, over half contained one rim-based pentagon and roughly a quarter had two rim-based pentagons suggesting that the formation of multiradicals in flames is likely.

These results suggest a new mechanism for soot formation where molecules with two or more localised π-radicals can polymerise (a rapid chain reaction) - what we called the aromatic rim-linked hydrocarbon mechanism (ARLH).

There is a nice historical connection with New Zealander Prof. John Abrahamson, Canterbury University. During a sabbatical in the '70s at the Chemical Engineering department at the University of Cambridge (where I did my PhD) he wrote a paper proposing that the partial saturation of aromatic platelets forms soot (see structure below). I spoke with him recently during the lockdown in New Zealand about the HR-AFM results that show partial saturation of aromatic platelets and our results showing the localised π-radicals and he was happy to hear about the recent insights and how close he got in 1977.

Abrahamson, John. "Saturated platelets are new intermediates in hydrocarbon pyrolysis and carbon formation." Nature 266.24 (1977): 323-327.

The Chemical History of a Candle and Structure of an Ember - Webinar

A webinar presented on the 6th of May 2020 for the Churchill College MCR, University of Cambridge, by Jacob W. Martin.

Abstract: Grab a hot chocolate and get cosy for a fireside chat about a candle's flame and the embers left behind. Following the lead of Michael Faraday in his 1848 Royal Institution Christmas lectures, I will provide an updated chemical history of a candle - including some experiments for you to try at home. The fun doesn't end there - after a flame is extinguished, there's a wealth of discoveries to be made in the embers. Understanding the structure of these carbon materials begins with Rosalind Franklin, best-known for her work on DNA, and continues with simulations on a supercomputer. While there are still many unanswered questions around flames and carbon materials, these recent insights are enlightening and important for cleaning up our planet.

Defending my PhD thesis in lockdown

Just passed the oral examination for my PhD! A bit strange doing it via video but thankfully no technical difficulties. Thanks to everyone for your support. In particular, to my supervisor Professor Markus Kraft, members of the CoMo Group. To my parents/inlaws Debbie Martin and Brent Martin as well as Perine and Laurie Renwick for their moral and financial support. Finally, of course, to Louise Martin, you really made this happen.

If you are interested in what my PhD was about I recently gave a webinar that goes through the main findings.