Genetic engineering to produce oil from waste from Lanzatech.

Last week when I was researching sustainable biofuels, my wife who works in L&T Hydrocarbon, suggested I look up Lanzatech’s Bio-Ethanol process which has been commissioned at IOCL Panipat, last year. What makes this technology unique is that instead of conventional raw materials, it uses waste gases – in this case from the Hydrogen Generation Unit - to produce ethanol. With sustainable aviation fuel soon expected to become a mandate in many countries, this is clearly a strategic move by India’s largest oil marketing company.

Alcohol made from fermenting food crops like sugarcane, maize, soya or molasses is referred to as 1G Ethanol. The need to look beyond may have been influenced by the inevitable debate over prioritising fuel over food, but I suspect the main motivation was to look for cheaper feedstock - plant residues, grass, algae & the like. The problem with the 2G feedstock is that it is naturally difficult to process. For example, in case of plant residues, only the cellulose component is of use, the lignin is not – so straightaway we lose about 40% yield. It is precisely this limitation that the 3G technology looks to address. Lanzatech’s process is based on fermentation of gases & it uses a type of bacteria called acetogens. The biochemical process ensures that all the carbon available participates in the reaction – even if the feedstock is of low carbon content like municipal waste or is difficult to extract like lignin.

Generally, the starting point for sustainable fuel manufacture is syngas, a mixture of carbon monoxide & hydrogen. Once you get there, usually Fischer-Tropsch synthesis takes over, converting carbon monoxide and hydrogen to a mixture of hydrocarbons. But while the FT route is well established, Lanzatech just might have a better solution to offer. ??

Waste comes in all forms – if it is off-gases like what is being used at IOCL, it can directly go the fermenter; solid feed like municipal waste or crop residue first needs to undergo pyrolysis. This involves heating the material with using superheated steam in an airtight vessel – where the steam is both a heat source as well as a reactant. The output is syngas with some amount of methane and carbon dioxide. ?Although the syngas, technically, can directly go into the fermenter, it usually undergoes a refining process to remove impurities like sulphur or nitrogen compounds which could reduce the potency of the bacteria used for fermentation. During the fermentation step, a lot of processes happen simultaneously. The syngas is bubbled through bacterial culture where the acetogens ferment it and produce ethanol & while the reaction is in progress, the substrate is continuously drawn out & distilled. Having harvested the alcohol, the rest of the substrate is sent back to the reactor to continue the fermentation process.

Inherently biochemical & chemical processes have complementary features. Chemical reactions are obviously faster, and often reversible whereas biochemical reactions are slower but unidirectional (like fermenting milk to make curd). Operating conditions often differ widely. With syngas as starting point FTP operates at 300C & 30 bar pressure whereas Lanzatech’s fermentation takes place at 37C under atmospheric pressure. Biochemical catalysis reactions are rather complicated, but as the catalyst regenerates as the reaction progresses instead of getting depleted, it allows for continuous processing – which helps increase throughput & reduce plant costs. But the biggest advantage of the Lanzatech process is its flexibility to handle any H2/CO ratio unlike the FTP which needs it to be 2:1.

Despite all these positives associated with the biochemical route, there are massive challenges to surmount to make it a commercially viable alternative. For one, since the feed is low in carbon content, the reaction needs to be fast and selective so that the production rates are reasonable and the yield of the preferred grade is high enough to keep separation costs under control. Of course that means choosing the right kind of strain but more importantly it is about ensuring, through understanding of the metabolic pathway, that the reaction is coaxed to move in the right pathway. This is easier said than done!

To understand Lanzatech’s approach, we will use the analogy of a car design studio – the kind that Dilip Chhabaria pioneered in the early 2000s where you could give him any car and, for a lot of money, he would convert it to the car of your dreams. Now imagine a car garage like that getting a customer who wants to transform his sedan into a race car. Now, the car has two components, broadly: the hardware – which is the engine, transmission, braking system etc & the software that controls the fuel injection & other critical functions. If you want to reprogram the software, you either need the source code or the ability to hack it. The hardware changes may seem simpler because it generally involves altering or replacing some parts but unless you test it, you can’t be absolutely sure that the changes will be integrated seamlessly. Plus, there could be parts that you would not want in the race car like the energy recovery system that you would rather repurpose to give a faster pickup. So essentially you may need to replace, repurpose or just remove a few parts from the current car and put in new features instead.

Now let us get back to Lanzatech. The feedstock composition is a critical input in the screening of likely bacterial strains candidates & more often than not, there is no unique solution. But over time, Lanzatech has developed predictive models of microbial metabolism which helps them identify the high potential strains as well as the enhancements required to drive the reactions towards high selectivity and yield. Augmentation of properties, in this context, means genetic modification and the two types of bacteria – gram positive & gram negative present vastly different challenges. In case of gram negative bacteria, breaching the cell wall, is like trying to entering our office through the lobby. The security will not let a visitor in unless his host physically comes and accompanies him inside. Gram negative bacteria has cell membrane instead & trying to gain access is like entering the office through the car park in the basement which has access control with face detection facility. It is effective but not foolproof and typically if there is a power failure, I have seen it switch off for a minute or two till the DG backup kicks in - not a huge security risk but it is enough to let someone slip through unnoticed.

DNA transfer can be done through either electroportation or conjugation. Congujation, which is used for gram negative bacteria, is like entering through the lobby – it needs cell to cell contact for DNA transfer. So, while incorporation is difficult, once it gets in, the DNA gets assimilated into the host cell much more easily. Electroporation on the other hand is used for gram positive bacteria & it uses a short burst of electricity to momentarily depolarise the cell membrane which allows the foreign DNA to infiltrate. However, since the cell views it as an intrusion, it activates its defence mechanism against it. So although electroportation is easier, cheaper and less selective, the inserted DNA needs careful management to ensure that it survives after it enters the bacterial cell. The objective of gene editing is usually to either suppress or express a particular functionality. For a particular strain, if ethanol and butanol could be equally likely pathways - the objective could be to stop the generation of butanol so that the entire yield shifts to ethanol – just like the trade-off we made of preferring pickup over fuel economy in the car studio. In many cases it may be used just to knock off some genes that are potentially troublesome, especially when they pose danger to the inserted DNA but that is a little more complicated. ?

Starting from their first site in Shougang in China, Lanzatech has come a long way. It has tied up with Boeing and Virgin Atlantic for aviation fuel and is partnering with Dow, Evonik & Invista among others for green chemicals. The 3G ethanol plant at Panipat has already commenced production & is expected to reduce 11,600 tons of CO2 emissions from the refinery annually – and this is just the beginning! Lanzatech’s cutting edge capabilities in modelling of metabolic pathways, automated genome sequencing & editing tools using tools like CRISPR backed by extensive IP helps them design processes that are robust & scalable for processes for a variety of industries.

At one level this is classic industrial symbiosis where the waste gases from burning fuel in one process becomes the raw material for the Lanzatech plant which converts it back to fuel - which is used in original process again, thus creating a closed loop carbon cycle – but the technology has the potential to do far more. Using this technology, it is possible to have a plant with massive flexibility where just by changing operating parameters and microbial load, a wide range of chemicals could be manufactured from the same facility with the mix keeping in step with market demand – promising to offer a kind of flexibility that chemical plants would find tough to replicate! Ethanol is just one of the many products possible. ?

Honestly, when I started researching for this article, I had hardly bargained to find cutting edge genetic engineering and predictive modelling of metabolic pathways to be the enabler. But that is the beauty of science. At a time, when we are bombarded by doomsday news of the earth heating up and teetering on the brink of a precipice, stuff like this helps restore faith. It is a message worth amplifying.