I apologize for responding late to this question, but there are some issues I would like to discuss. The question was whether one can determine lattice parameters using powder XRD on a thin film sample. The answer is yes, but with many caveats.

First comment is that while Rietveld methods will often work, they may be overkill for simply obtaining lattice parameters. There may be no need to fit a complex structure model, especially if the pattern has reasonably well defined peaks without much overlap. If the crystal is cubic, tetragonal, or hexagonal, it may be simpler to refine lattice parameters using Cohen's method (an old technique, but still very useful). Another option might be to perform a "LeBail fit" (i.e., without a full structure model) to reduce the number of refined variables in your "Rietveld" program.

Second comment is that thin films often exhibit preferred orientation (texture) which will enormously complicate a Rietveld refinement. Powder XRD data, if it is a conventional Bragg-Brentano geometry 2theta scan, will not explore enough of the scattered intensity to allow for any accurate texture determination. So any texture correction (March-Dollase or spherical harmonics) will be somewhat speculative. That said, it probably won't impact the lattice parameter calculation.

Third comment is that thin films may have residual stress due to synthesis, and this stress may impact the lattice parameters. Fortunately the out-of-plane stresses will be small, but they might be sufficient to cause changes in materials that have a stress-induced phase transition.

Fourth comment is to be sure to consider instrumental errors in your analysis. With a thin film sample, the issue of peak shifts due to sample transparency is pretty much avoided, so you need to consider sample height and 2theta zero errors. If you are not sure what I mean, consult any basic text on powder diffraction.

Cheers,

Andrew

Hello Sachin

Thanks for replying

Is fullprof can be applied be applied for thin film ?

Hi.

If you carried out powder XRD in your samples, you can find the lattice parameters and many other physical parameters by using Rietveld refinement analysis of Fullprof software. Also, you can compare the values obtained from the analysis and standard theoretical values.

It depends on your thin film quality. If it is showing peaks clearly u can apply rietveld using fullprof

Generally thinfilms need high resolution XRD. and u need to run for long scan at low angels to get good quality.

See the following files u can have clear idea

Check out this one also..!

Hello Nunna,

Thanks for replying!!

I will try to learn rietveld and then will try to apply for my film

Also in few paper i came to know a software named as "Powdmult computer program but i did not found that on internet, is it also same as reitveld or the different name of rietveld "?

if u have any idea please help me to find the answer

I have software to calculate lattice parameter - both powder and thin films

Rajalingam Sir

If you have then Please help me to find that

Dear Neharana

send to me data my gmail id,

what is yr mailing Id

Rajalingam Sir,

Let me know the software which you have to calculate the lattice parameters for powder as well as thin-films.

You can use PCD (Pearson's crystal Data) if your institute have the access for this. This is a very helpful database which we have here in IITK. Also, I think the XRD machine from where you did your experiment, it has a software too.

I apologize for responding late to this question, but there are some issues I would like to discuss. The question was whether one can determine lattice parameters using powder XRD on a thin film sample. The answer is yes, but with many caveats.

First comment is that while Rietveld methods will often work, they may be overkill for simply obtaining lattice parameters. There may be no need to fit a complex structure model, especially if the pattern has reasonably well defined peaks without much overlap. If the crystal is cubic, tetragonal, or hexagonal, it may be simpler to refine lattice parameters using Cohen's method (an old technique, but still very useful). Another option might be to perform a "LeBail fit" (i.e., without a full structure model) to reduce the number of refined variables in your "Rietveld" program.

Second comment is that thin films often exhibit preferred orientation (texture) which will enormously complicate a Rietveld refinement. Powder XRD data, if it is a conventional Bragg-Brentano geometry 2theta scan, will not explore enough of the scattered intensity to allow for any accurate texture determination. So any texture correction (March-Dollase or spherical harmonics) will be somewhat speculative. That said, it probably won't impact the lattice parameter calculation.

Third comment is that thin films may have residual stress due to synthesis, and this stress may impact the lattice parameters. Fortunately the out-of-plane stresses will be small, but they might be sufficient to cause changes in materials that have a stress-induced phase transition.

Fourth comment is to be sure to consider instrumental errors in your analysis. With a thin film sample, the issue of peak shifts due to sample transparency is pretty much avoided, so you need to consider sample height and 2theta zero errors. If you are not sure what I mean, consult any basic text on powder diffraction.

Cheers,

Andrew

I agree with Edward Payzat

Andrew is absolutely correct. If you are only interested in lattice parameters then the LeBail fit should be quite sufficient. However, you will need to be very carefull in your consideration of in-plane stress and perfered orientation effects.

if you want to get the lattice parameters of your FILM, then a fit (Pawley or Le Bail or traditional Rietveld) is NOT sufficient... this is exactly the point!!! If you do not consider residual stress, then you obtain numbers useful just for playing Bingo... I thought Andrew's answer was clear enough

Matteo-

I am a bit confused by your statment. I am complete agreement with Andrew because he say, and rightfully so, that lattice parametesr can be calculated using a Reitveld, LeBail, or Cohen method. Perhaps the confusion is in my use of the word "sufficient" by which I am seconding that Rietveld is likely overkill for simply extracting lattice parameters. However, as I stated, you will have to take into account film strain for any in-plane lattice components. I do like your imagery of Bingo though.

Kevin, my statement was a bit general... by "insufficient" I mean that you need further models to account for the peak positions shift. So in the end you are not doing a mere Pawley fitting (all parameters free except for peak positions bound to a lattice) and, unless you know the value of the unstrained lattice (which, by the way, is the number you'd like to obtain), you will not be able to obtain an absolute estimate of the cell parameter!

The Rietveld method can be used (a lot of people doing residual stress analysis actually use it), but in that case it MUST be suitably modified to take the peculiar microstructure (texture) and stress into account.

Of course, and I think it was clear even if not explicitly said, a single diffraction pattern cannot give any definitive information on a thin film!