Dear Bikash,

as Alan pointed out, the peaks (if you will get any)  will be very broad  due to the small size of your particles ( 3-4nm is 30 -40A and that is about only 15 to 20 atomic layers !!!.) and thus they will be very small in height. Furthermore these tiny (broad) structures are on top of a huge background of your support (bear in mind that you have only 0,5% of your material in the x-rays beam which will contribute to your 'Au-peaks').

In order to resolve your 'peaks' from the statistical noise of the background  you have to wait a very long, long time;  A day or even much more.......

If your nanoparticle  like material is only on top of your support you may use grazing incidence XRD, so you will be more surface sensitive. Thats the only way to increase your signal to background ratio and thus reducing measurement time.

Taking advantage of the total reflection properties of your substrate you may be successfull with GIXRD in your case.

https://en.wikipedia.org/wiki/Grazing_incidence_diffraction

https://www.rigaku.com/en/products/xrd/ultima/app029

Possibly  with 24 hours irradiation under the x-ray beam....

The peak broadening will be large and peaks will be diffuse, though.

Thanks Dr Alan. Can you please add some more explanation to it,

Dear Bikash,

as Alan pointed out, the peaks (if you will get any)  will be very broad  due to the small size of your particles ( 3-4nm is 30 -40A and that is about only 15 to 20 atomic layers !!!.) and thus they will be very small in height. Furthermore these tiny (broad) structures are on top of a huge background of your support (bear in mind that you have only 0,5% of your material in the x-rays beam which will contribute to your 'Au-peaks').

In order to resolve your 'peaks' from the statistical noise of the background  you have to wait a very long, long time;  A day or even much more.......

If your nanoparticle  like material is only on top of your support you may use grazing incidence XRD, so you will be more surface sensitive. Thats the only way to increase your signal to background ratio and thus reducing measurement time.

Taking advantage of the total reflection properties of your substrate you may be successfull with GIXRD in your case.

https://en.wikipedia.org/wiki/Grazing_incidence_diffraction

https://www.rigaku.com/en/products/xrd/ultima/app029

Thanks Dr. Martens for the valuable answer and reference given is also very helpful.

@ Bikash Sharma   In the 70's (not my 70's - which are to come, hopefully - but the 1970's) in my Ph.D work I successfully got XRD Debye-Scherrer films from nominally 5% Au on SiO2  (Davison 70 if I recall) - made by hydrazine hydrate reduction if you want to know.....  Caveats: 24 hours irradiation; sizes (number mean)  were typically 10 nm (we always used Ångstroms then  - 100 Å) by TEM and Scherrer equation.  I calculated line broadening by hand.... See attached for an old picture from my thesis (developed and printed all my pictures too; those were the days...). For some work from this see: https://doi.org/10.1016/1381-1169(96)00027-1 

Nominally 5% (found ~ 4.5%) is 10 times your quoted metal loading.  Sufficient to get peaks good enough for lattice parameter calculations and looking at deviations from Vegard's Rule for the alloy systems we studied (Pd-Au, Pd-Ag).  However, in the nearly 50 years that have elapsed since then XRD equipment (especially) detectors and processing has advanced markedly and I would expect this to be feasible.

Thanks Dr. Alan for the valuable answer. The reference mentioned discuss the catalytic reduction of isoprene with Pd and bimetallic Pd with group 11 metal like Au and Ag supported on SiO2 and their different selectivity to the possible isomers.

It is an interesting paper to read how half hydrogenated allylic radicals get further hydrogenated and it depends on the charge on the incoming Hydrogen which is effected by the concentration of the coinage metal present in the bimetallic Pd particles.

But the article don't have any discussion regarding the XRD and the crystallite size.

@ Bikash  The paper you mention did not deal with the XRD side.  I've summarized that in my answers above.  We also used chemisorption, XPS (ESCA), plus a bunch of other techniques (e.g. TPR, TEM) for characterization.  XRD measurements for the powder were conducted in a Lindemann tube with a really old Philips Analytical instrument on a Debye Scherrer camera (I think 114.6 mm diameter) with the film removed 24 hours later.

Thanks  Dr. Alan for the valuable answer.

 Hi Bikash

I have attached a paper my group published several years ago that compared the performance of 0-10 wt.% Au/TiO2 photocatalysts for H2 production in ethanol-water mixtures. The Au nanoparticle size in our system was 4-6 nm at all Au loadings.

The supplementary information shows XRD patterns for the 0-10 wt.% Au/TiO2 photocatalysts. Due to severe line broadening caused by the small Au nanoparticles, the Au reflections were difficult to discern at low loadings (e.g. 2 wt.%), though the analysis was complicated here by overlap of the main Au(111) reflection with that of the TiO2 support. XPS, XRF, EDAX and UV-Vis were much better techniques for detecting and quantifying Au at low loadings.

However, if your support is amorphous (e.g. SiO2), then it is easier to detect Au nanoparticles at low loadings by XRD since the Au peaks will not be swamped by intense diffraction peaks from the support. We have made a number of Au/SiO2 catalysts and detection of 5 nm Au nanoparticles at loadings as low as 0.5 wt.% is readily achievable by XRD. Ideally, the Au(111) reflection should not overlap with peaks from your support.

Best regards

Geoff