If you can prepare thin layers of the surface of the ATR prism (e.g.  from suspension by evaporating the solvent) you can perform the test. You should try to compare samples prepared under different conditions and interpret the changes. Alternatively you may try to use diffuse reflection sample holder and optics. That requires a somewhat thicker powder layer. You may also try to prepare KBr pellets and measure in transmission mode.

If you use solid samples than the material should place onto the small crystal area in the ATR-FTIR.

If you want to study in liquid form, then you need to run background measurement with the solvent in which your nanoparticles are dispersed, before running your sample. Similar like solid, the liquid is simply poured onto the crystal. The whole crystal must be covered if performing a quantitative or qualitative analysis.

For ATR FTIR  solid samples, use the flat crystal. to obtain the results, the  sample dimension is  less than 4mm thick. The sample surface must be flat with good  contact between the sample and the ATR crystal, otherwise  can not be obtained good results..

If your sample is powder, it can be processed into a pellet with a diameter of 13 mm and a thickness of about 1 mm to measure.

For ATR FTIR with powdered sample, you should place the sample on the surface on the crystal. The sample is well packed on the crystal to cover it.

PS: Make sure the crystal is clean before starting any analysis. 

I hope this can help

You should prepare a wafer of your material by KBr and then determine all of functional groups on surface of your materials by FTIR spectroscopy.

If you want to measure something like surface modifications of particles do not use dispersions (especially if you have the intention to use the dispersing media for background measurments). The reason is quite simple: In general the solvents have the same functional groups and nearly no light reaches the detector (leading to huge errors).

Particle dispersions leads to the advantage that you can perform HATR measurements after the volatile media is completely removed (the distance between the prism and the particles .is than very small). You can monitor this evaporation by the IR. Additionally you can check the stability of the modification as performed in the attached publication (this publication also contains details concering the preparation).

Article Synthesis and testing of ZnO nanoparticles for photo-initiat...

As Michael said, its better to remove the solvent. What is the concentration of your particles?

In this paper I measured FTIR-ATR of functionalized particles with different organic shells.

https://www.researchgate.net/publication/261101583_Smoothly_Tunable_Surface_Properties_of_Aluminum_Oxide_Core-Shell_Nanoparticles_By_A_Mixed-Ligand_Approach

Article Smoothly Tunable Surface Properties of Aluminum Oxide Core-S...

Dear Sujoy Sarkar what is the recommended crystal to be used?

Hi @Efi Dwi Indari , widely used ATR crystals are ZnSe , Germanium,Silicon and Diamond. Despite the high cost of Diamond ATR crystal, it can be used for the analysis of apparently all materials including corrosive chemicals and powders. ZnSe is the most widely used ATR crystal with similar penetration depth od Diamond.But ZnSe is soft and not chemically inert as other crystals . Depth of penetration of IR wave in crystals may be useful to select the appropriate crystal for the analysis. Low the depth of penetration makes the crystal suitable for the analysis of highly absorbing or scattering samples. Penetration length of few crystals are

1)Diamond crystal- 2um at 1000cm-1

2)ZnSe- 2um at 1000cm-1

3) Germanium-0.67um at 1000cm-1

4) Silicon -0.84um at 1000cm-1