TGA gives a global information, but XPS gives a local one. Moreover, the depth probed by XPS is very limited, restraining the analysis to the extreme surface. Therefore, I would certainly not rely only on XPS, even if XPS can indeed brings additional information with respect to TGA.

Alain

Just for the surface analysis, XPS will give you the information you need because it only gives the information to a few namometers (~10 nm) so you can see the Pt-decorated surface with XPS which gives much more additional information in your case as compared to TGA.

Good luck!

Thanks Alain and Muhammad,

That's true that the depth of analysis is limited to the few nanometers, but given the assumption of the homogeneity of the catalyst decoration (equal loading every where) and considering the nanometer size of both catalyst and support that almost coincides with the depth of XPS, and finally remembering the wide area of XPS scan (minimum from 10 to 200 micrometres), why shouldn't we use it as a checkpoint if it has been done at least three times in different zones of a given sample?

I should add that the main focus is on the Pt loading here and for the surface chemistry and oxidation levels of support and catalyst, of course XPS gives the unique information.

The risk I see is the representativity of the scanned zones. You may analyse some regions being either poorer or richer than the average loading. If, however, electron miscroscopy shows a very even distribution, therefore why not basing the analyses on XPS only.

Alain

Ok Alain, Thanks.

You ask for the determination of the Pt loading in MEAs but refer to electrodes for RDE experiments.

For MEAs I think that TGA and XPS are not suitable if you want to determine the Pt loading on a GDE.

The oxidation of the carbon support under O2 in a TGA only works with pure catalyst material, but not with coated GDL/MPL material. Do not forget that also some binder material (for example: PTFE) is needed and mixed into catalyst slurries that usually cannot be oxidized.

We apply a lot of spraying and/or doctor blade techniques to manufacture our own MEAs. Of course, these coating processes are always accompanied by losses. Therefore we have a rather simple method: we weigh our samples before and after coating (once ist has been dried) with a very sensitive balance. From the geometry you can easily deduct the amount of Pt in mg/cm² once you now the amount of catalyst in the slurry. Of course, that needs to be calculated following the recipes for slurry mixtures.

And you do not have to assume an even and homogeneous distribution of the material on the surface that is coated. I doubt if any really homogeneous distribution is ever possible on GDL material even if it is covered with a MPL.

Kind regards

Peter

Hi Peter,

Thanks for your nice comment. It actually implies a delicate reference to the literal attribution of the term “loading”, that is right. Loading should finally mean the mass of catalyst present on unit surface, and basically, on MEA. However, since a similar precise term for the catalyst/support ratio is not developed (at least to the best of my knowledge), sometimes people use the same for this latter purpose. Anyways, my question was looking at the determination of the catalyst/support ratio to be put on GC for RDE experiments.

Of course, as you told, for MEAs the issue is more complicated and TGA and XPS do not work the best. I think even in the “balance” method, as you mentioned, the knowledge of the amount of catalyst in the slurry (or at least catalyst/support ratio) is critical. Maybe this the point that my question concerns about.

Thanks,

Mazdak

Hi Mazdak,

as long as you have a really homogeneous solution that you coat with you know the amonut of catalyst in the slurry with an acceptable low error.

Kind regards

Peter

Pt should not be oxidised when burning the carbon, therefore only the metal should be recovered (in the presence of ashes potentially present initially in the carbon). Washing the ashes should lead to rather pure Pt. Note that a volatile but very instable PtO2 oxide exists, formed at high temperature and which decomposes into Pt + O2 at lower temperature. But the yield in forming such oxide is very low, therefore almost 100% of the Pt will be recovered.

Alain

Hi Rhiyaad,

I agree with what Alain said about the TGA. When doing it in air, carbon is burnt and gone, unless some impurities have been associated with that (like catalyst particles used to synthesize and grow CNTs). Pt remains as metal.

In my opinion, to study Pt loading, you should use ICP.

And if you want to study Pt dispersion (more important than Pt loading), you could use H2-chimisorption technique (using TCD detector). It's a very simple and it's not a local examination technique compared with XPS.

Hi Le Phuc,

Thanks for your opinion. It is however a bit difficult to understand what you meant. ICP usually uses liquid samples (solutions of low concentrations of the element of your interest). A Nebulizer is then used to converts the liquids into an aerosol, and that aerosol can then be swept into the plasma to create the ions. And finally, a mass spectrometer detects and identifies different elements.

In our case there are a couple obstacles for this the main important of which is that we don’t have a liquid sample. It is basically carbon and platinum, both of which are absolutely difficult to dissolve.

Less commonly, the laser ablation has been used as a means of sample introduction for ICP-MS. In this method, a laser is focused on the sample and creates a plume of ablated material which can be swept into the plasma. This is particularly useful for solid samples, though can be difficult to create standards for leading the challenges in quantitative analysis.

Mazdak

Tks Mazdak Hashempour

I talk about Pt dispersion. Sorry about that.

In fact, we're working on FCC catalyst and we also want to quantify Pt loading on fcc solid catalyst. We're trying to determine Pt loading by XRF (http://www.petro-online.com/news/analytical-instrumentation/11/thermo_fisher_scientific_uk/fluid_catalytic_cracking_fcc_catalyst_analysis_using_thermo_scientific_x-ray_fluorescence_spectrometers/12573/)

But until now, because the lack of reference sample, we have not validated our measurement yet.

Best regard,

We've used TGA for this purpose during the last couple of years. It works fine. It remains the most efficient way, although the standard error is rather high.

Regards,

Lionel

hi lionel, could you please explain how you use TGA to determine the platinum loading in an electrode of a fuel cell