Non-catalyzed dissociation of Ammonia

[Richard Colwell]

Has anyone ever made/used a mechanism for the non catalyzed disscociation of ammonia? Or will I probably have to make it myself?

[Walter Dal'Maz Silva]

According to thermodynamics ammonia is a quite unstable molecule... but in practice its homogeneous decomposition rate is quite slow at atmospheric pressure. Your question is a bit generic and no mechanism can cover the whole range of conditions with same accuracy. 

In my Phd I studied ammonia decomposition for the setup of nitriding of steel in a pressure range from 10000 Pa - 100000 Pa: ammonia simply will not crack below 1200 K if walls of the reactor are inert (in my case with a silica tube and residence times from 1-100 s no more than 2% conversion was observed, in the order of magnitude of the error of the gas chromatography system and flow meter combine - maybe none was cracked). There is a paper from Dirtu et al, 2006 (DOI: 10.2478/s11532-006-0030-4) and same group in 2011 (DOI: 10.1134/S0023158411040112) that I never managed to reproduce and ended giving up on the subject. After a while I realized some of the equations in their chemical mechanism had the same sources/rates as those from GriMech-3 and therefore adapted for a much higher temperature range and low ammonia concentrations. 

If you are dealing with high pressure there are some papers out there and kinetics.nist page can provide you with more information. Otherwise the literature is very poor and nobody invested on the matter because for practical residence times that is generally irrelevant compared to the effect of walls (currently I work on a classified subject with residence times up to 2400 s and all I can tell you is I still have nice results modeling only the walls).

Kind regards,

Walter

[Richard Colwell]

Thank you for the response. I have also come across the 2006 Dirtu paper, as well as a paper by Davidson et al from 1989 "A Pyrolysis Mechanism for Ammonia" (attached). I attempted to replicate the Davidson mechanism in Cantera, but my results do not match very well. I am wondering if the issue is that the efficiencies for the three body reactions are not defined. Cantera will assume an efficiency of 1 if it is not defined, which I think may very likely skew the results. 

You said you had trouble reproducing the results as well, so I am wondering if this is a fruitless effort. Just in case, I am also attaching the .cti mechanism and python code (both in .txt format), if somebody could take a look and see where I am going wrong. 

Thanks, 

Richard