TGA analysis can only show weight loss by heating and not identify the type of coke deposited on your sample. I think most of coke species deposit on the metallic sites of catalyst.

Yes i agree with Mohammad, same think happened with me also but 2-3% coke is not that much it might be deposited on metallic site and coke itself has acidic nature(pH near 2) better you repeat the analysis for confirmation or do SEM or TEM of used HZSM-5 so you get to know the morphology to coke deposition

Mohammad Reza Toosi Thank you for the quick reply. Can you tell me how to identify the type of coke? Does it have relationship to the reaction time? This used HZSM-5 catalyst only run for 4 h. The TGA has also been tested on the metallic catalyst, it showed very little coking on it.

Thanks.

Amrin Punekar Thank you for your reply. The metallic catalyst and acid catalyst have been sieved to 40-60 mesh, therefore, they were separated after reaction. The TGA ananlysis and NH3-TPD were only for the used HZSM-5.

Thanks.

It is difficult to identify and it may be characterized by TEM or XPS. An easy method is TPO analysis and refer to the literatures.

4 hours for a catalytic reaction is not long. All catalysts show coke deposition in the initial period. You must analyze coke deposition in a long period more than 24 h or after several runs.

Mohammad Reza Toosi Actually, I have done the TGA for another HZSM-5 after 100h run of reaction. I am wondering the TGA showed there are coking on the surface or inside of this used HZSM-5 after 4 h run, is it possible that the coking was blowed away by He in the pretreatment of NH3-TPD? We do that under 350 C, that temperature is not enough to remove all the coking, right?

But yesterday I tried to do the pretreatment without He, still under 350 C, the result is the same as previous one, also close to the fresh HZSM-5 catalyst. I am stuck.

As you mentioned, a part of coke will be removed by pretreatment below 350 C. NH3-TPD is known as a method for determining acidic sites of fresh catalyst not spent one. It seems that some NH3 molecules could adsorb on coke species. Moreover, NH3 TPD generally show Bronsted sites not Lewis sites. B/L ratio is an important factor in your work that must be determined. I think it's not important to know variation of acidity after and before reaction. Mechanism of DME from formaldehyde is known for all chemists . However, if you want to show any change in spent catalyst I suggest use TPO analysis.

What type of detector your NH3-TPD instrument have? Is it a TCD detector? If yes, make sure the temperature range does not include the range in which carbon is decomposing. This will affect your data. Other than that, the acid sites in zeolites are mostly internal (in the microporous structure), if deactivating occurs at the external surface you may observe these results too.

Alexandre A. S. Gonçalves Yes, it's TCD detector actually. Also, I tried to reduce the pretreatment temperature from 350 C to 150 C, it showed the similar results. So I guess that is not the problem of carbon, because the TGA showed the carbon was fully combusted until 900 C.

Alexandre A. S. Gonçalves Another question is if the deactivation only occurs at the external surface of zeolite, however, the DME yield was reduced to almost zero, is it because of the very high concentration of MeOH leading to the temporary deactivation or I mean some hydrocarbon cover the surface of zeolite?

As I mentioned, by determination of types of coke you can detect location of coke (internal or external surface) on your catalyst. In TPO analysis, first peak as a large signal shows coke on external surface and next small peaks indicate internal coke samples. In addition, dehydration occurs inside of the channels in zeolite and the conversion will reduce after fouling the channels resulting blockage of them.

Mohammad Reza Toosi Thanks! I will probably try TPO first.

Hi Xiao Fan,

I did not understand what you mean with pretreatment. I call as pretreatment the 2 initial steps of NH3-TPD; (1) surface cleaning and (2) NH3 sorption at given temperature. The carbon decomposing will affect the results during the desorption step which is performed by increasing temperature, because the strong acidity of zeolites may catalyze the carbon decomposition even when oxygen is not present; the TCD detector will not distinguish between CO, CO2, NH3, or other hydrocarbons. Note that you need to use high temperatures to completely desorb NH3 from zeolites. However, if you use an MS detector instead of a TCD detector, this problem can be solved. In this paper (open access) I described how I used both HR-TG and NH3-TPD, but I did not have the carbon problem during my NH3-TPD measurements: Article One-Pot Synthesis of Mesoporous Ni–Ti–Al Ternary Oxides: Hig...

The TPO will give you similar results than that of TG, both will burn the deposited carbon. Particularly, I find the HR-TGA better due to resolution. You can distinguish better the events and can play with kinetic studies as well (Article Evaporation-induced self-assembly synthesis of nanostructure...

The deactivation may not occur only at the surface. Carbon is deposited by polymerization followed by carbonization. The polymerization if facilitated if precursors do not diffuse out of the pores quickly, which is quite common for zeolites. That's one of the reasons many researches work on the development of mesoporous zeolites to increase catalyst stability. Being said that, you need to identify the nature of the deactivation. If that is by carbon deposition, you need to characterize if carbon is being deposited into the micropores or external surface. If it is externally deposited, you reaction is likely occurring only at the external surface rather than internal (diffusional constrain); considering your catalyst results.

Better than any other techniques mentioned above, I would recommend low-temperature nitrogen sorption of fresh and spent catalysts. By looking at the micropore volume, you will know if carbon is being deposited internally. The same apply if you are depositing metals in the zeolite. Ideally, you can built your whole discussion by connecting all these techniques.

Best Regards,

Alexandre Gonçalves

Alexandre A. S. Gonçalves Your answers are really helpful! Thank you so much, Alex.