tl;dr >1300 combustion scientists logged into the first virtual combustion conference. I was involved in the workshops leading up to the conference and presented our group's work on soot formation. Some main themes included; low T combustion for clean combustion, biofuels for easy fossil fuel replacement, ammonia as a clean fuel and soot formation remaining a mystery.
The combustion symposium happens every two years since 1928 and was planned for 2020 in Adelaide Australia. Due to the pandemic, the conference had to move online and was delayed until January. The talks were all prerecorded and played during the session and then the speaker was live afterward for taking questions. One advantage of this format is that you can access the recorded talks for up to a few months after the conference. A disadvantage was that you are unable to have casual conversations, though they did have a chat feature so you could talk with conference attendees. Below is the interface showing the video box that you play the video within the website.
The lead-up
Before the combustion symposium, there are a variety of workshops that were organised. This year I attended and presented at the 5th International Workshop on Flame Chemistry and the 5th International Sooting Flame Workshop (ISF-5) both done virtually via zoom. I was asked to provide an overview of soot formation at the Flame Chemistry Workshop the slides can be found here. For ISF-5 I worked with Matteo Pelucchi to prepare some summary slides. Below is a nice summary of the talk I gave in a single slide. Credit to Matteo for doing most of the consolidation.
I also presented a poster as a video at ISF-5, which can be viewed online here.
Some of the discussions centred around what is the molecular unit of clustering towards soot formation. Currently, there are two main views that small molecules around the size of pyrene form soot and the other view that molecules around the size of circumpyrene form soot.
It is challenging to know in soot whether the larger molecules seen are formed in the gas phase before they cluster to form soot or if they form after being condensed in soot. We currently think it is the former.
Our work
At the symposium I was involved in three papers and a poster that were accepted and the group contributed in total six talks. Here is a summary slide that we prepared.
The work I was closely involved in focused on localised pi-radicals this was published last year and there is a blog post written on it already. We were able to demonstrate some pi-radicals form from hydrogen addition to pentagonal rings that lead to localised pi-radicals with considerable reactivity. The spin density plotted below shows this localisation.
Another paper Laura Pascazio presented extended this idea to crosslinked molecules. These can form from small PAH crosslinking reactions and lead to a flat molecule due to a double bond.
Laura explored how they physically condense and found they were similar to pericondensed species indicating they are not going to cluster together at flame temperatures. However, in a link with our previous work hydrogen addition to the pentagonal ring gives localised pi-radicals (species 1d in the figure above). Here is the spin density showing localisation for the penta-linked species.
An interesting paper that was presented by KAUST showed that the m/z 154 ion that is usually ascribed to biphenyl is more likely to be acenapthene from fragmentation studies, which is what we predicted!
This raises the question, what is the concentration of the partially saturated species C12H9 that we expect to be a reactive localised pi-radical.
Another interesting paper looked at carbonaceous nanoparticle formation in pyrolysis experiments. It showed firstly that as pyrene is heated in a furnace it breaks up and forms species around the size of 600 Da in size before forming the nanoparticles. It was also shown an increase in nanoparticles with temperature indicating a chemical process and not a purely physical condensation leads to these small particles.
The molecules thought to be involved in premixed flames are a little bit lighter with a peak concentration around 450 Da as was shown in another paper presented at the symposium. There appears to be a convergence in thinking around the size of aromatic molecules that cluster to produce nanoparticles in combustion and pyrolysis systems.
Some insights into what makes these aromatic molecules cluster was made in two papers. The first, already mentioned, showed that there are a reasonable number of aromatic molecules with more hydrogen than would be expected for pericondensed unsaturated aromatic such as pyrene. Second, the concentration of radicals in the flame was shown to be proportional to the amount of hydrogen in the soot. This indicates that hydrogen addition to aromatic can lead to more radicals. These results are all consistent with the aromatic rim-linked hydrogen mechanism we have proposed.
Low temperature combustion
There was a real focus on what role combustion has in a low carbon world. The most interesting technology was the development of low-temperature combustion engines by Mazda. In 2019 they released the first commercial engine that operates in the so-called homogeneous charge compression ignition (HCCI) mode. This practically operates a petrol engine like a diesel engine but at much lower temperatures. The high compression allows the engine to have 20-30% higher efficiency than a normal petrol engine and dramatically reduces the concentration of soot and nitrous oxide (NOx that causes acid rain).
HCCI engines had been notoriously difficult to build requiring recirculation of the exhaust gas into the intake and superchargers to allow for the high compression of the fuel without autoigniting. The trick that Mazda developed was to make use of the spark plug to start things off. The video from Mazda shows the details of how this works.
What is not discussed in the video is how Mazda developed a swirling fuel mixture that provides a slightly higher fuel mixture right where the spark plug is to allow it to ignite and push the charge above the critical pressure to auto-ignite in a homogeneous way. It can switch dynamically from spark ignition to the low-temperature HCCI combustion mode completely dynamically - some truly remarkable engineering.
So how does it reduce the soot and NOx pollutants? It comes down to the unique combustion mode that is able to operate in a sweet spot. The most helpful way to explain this is by using the equivalence vs temperature plot below.
 |
| Reference |
Diesel engines work by injecting fuel into a high-pressure chamber upon which it autoignites. A diesel engine will therefore have an equivalence ratio almost always >1 leading to soot, carbon monoxide and nitrous oxide.
In a spark-ignition engine (like most petrol engines), the equivalence ratio is always close to one but because the flame front is quite concentrated and high temperature you get nitrous oxide formation.
The clever thing about HCCI engines is they can operate in the 1400-2000 K by spreading out the flame so it homogenously burns and does not form NOx and by keeping the equivalence ratio <1 soot is not formed.
Prof. Yiguang Ju gave a very nice talk on the chemistry of this low-temperature combustion and how multiple oxygen molecules attacking fuel molecules allowing for different types of flames if you are interested in the details.
Ammonia
There was a lot of work on ammonia combustion because it is being seriously considered as a low carbon fuel for ship and energy storage. Combustion of ammonia leads to water and nitrogen - so no greenhouse gases that accumulate.
4 NH3 + 3 O2 → 2 N2 + 6 H2O (g)
Ammonia is also considered an chemical storage method for hydrogen. Hydrogen can be made from solar or wind powered splitting of water. Combining hydrogen with nitrogen in the air gives ammonia. Here is the usual sales pitch.
The reason you would want to convert hydrogen into ammonia is that the latter is much easier to store and transport. Many of the challenges of using this fuel were addressed at the combustion symposium including, the low flame speed and the formation of pollutants nitrous oxides (NOx) or worse cyanides were all discussed.
The first issue the low flame speed means that ammonia by itself makes for a very poor fuel it can easily be blown off a burner and be extinguished. Here is what blow off looks like from a burner. This is a significant industrial problem for turbines i.e. you don't want the flame to blow off in your planes engine...
To maintain stability you must run ammonia in a fuel rich condition that leads to slip of the toxic gas ammonia through the combustor which is not ideal (called ammonia slip).
There are a couple of solutions to the low flame speed problem. The most widely studied at the symposium was the addition of hydrogen gives a much nicer fuel with flame speeds approaching hydrocarbons. The other option is to add some methane to improve the burning conditions. This could potentially form soot so a nice study was conducted to look into the production of soot in an ammonia/methane flame. With low mixing of ammonia into a methane flame much of the soot was removed which is a good sign. Mixtures of ammonia, methane and hydrogen were also considered. As was the addition of biofuels. However, this mixed fuel with a hydrocarbon still produces CO2 and therefore the hydrogen addition is preferable.
The second issue is the pollutants produced. Basically at the high temperatures in the flame oxygen combines with nitrogen in the air (N2) or in with ammonia related species producing nitrous oxide (NOx). This is a toxic gas that can turn into acid rain in the environment or react with hydrocarbons in the air forming smog. It was clear that ammonia combustion produces a lot of nitrous oxide more than a hydrocarbon flame. One study demonstrated the highly toxic compound hydrogen cyanide could also be formed in parts per million concentration, which needs to be avoided at all cost.
There was quite a lot of work on the chemistry of ammonia combustion highlighting in particular the importance of the OH chemistry. As well as considering the impact of pressure and flame structure.
The most interesting idea was the use of MILD combustion methods to reduce the emission of NOx. MILD combustion is a highly efficient flame-less form of combustion that preheats the inlet air and fuel so they they homogeneously combust in the reactor. It can increase efficiencies of furnaces by up to 30%. It is a form of low-temperature combustion that was mentioned in the previous section. A video of a furnace switch to the MILD combustion mode can be seen here.
For a real in-depth look at how MILD combustion works here is a section from a very informative lecture.
The conference was excellently put together, however, the online format did not capture the buzz that you have with an international conference with everyone in the same place and timezone. I also missed catching up with friends after the conference. Hopefully, the conference in two years will be in person.
No comments:
Post a Comment