Fracture faces are often difficult to study by XRD as they are uneven and are diffcult to place in the XRD so that they are properly aligned, what you obtain is a lot of noise. We made some good experience with a micro XRD (most institutions haven't got one) but the easiest way to obtain a straight fracture face is to put a large vickers indent (10 kg) with the wing cracks normal to the sides of a bending bar (~3x4 mm2 diameter) and break the bar in a 3 or 4pt setup, you end up with a perfectly straight fracture face which is easy to align in the XRD. You should obtain a spectrum with a good signal/noise ratio. If the signal is still too poor use 4-5 bars aligned (to fix them we use modelling clay, but be careful this clay usually contains crystalline fillers, so the bars have to stand off by some millimeters)

Fracture faces are often difficult to study by XRD as they are uneven and are diffcult to place in the XRD so that they are properly aligned, what you obtain is a lot of noise. We made some good experience with a micro XRD (most institutions haven't got one) but the easiest way to obtain a straight fracture face is to put a large vickers indent (10 kg) with the wing cracks normal to the sides of a bending bar (~3x4 mm2 diameter) and break the bar in a 3 or 4pt setup, you end up with a perfectly straight fracture face which is easy to align in the XRD. You should obtain a spectrum with a good signal/noise ratio. If the signal is still too poor use 4-5 bars aligned (to fix them we use modelling clay, but be careful this clay usually contains crystalline fillers, so the bars have to stand off by some millimeters)

Two suggestions. One. If you have uneven surfaces you need to use parallel beam optics rather than divergent beam optics. This will compensate for any parts of the surface that are not at the focus position. Two. Your diffractogram shows very few counts, yet there is some evidence of crystallinity together with the amorphous signal. You just need to gather data for much longer counting times, get bettter peak shapes, better signal to noise.

Yes, I would experiment with longer counting times, since you do see some small intensity peaks in the pattern.

Did you use a Mo radiation for this sample? In this case you have another problem with flourescence!?

I think measurement through micro-focus XRD system will work for this problem. few days back we did it for one of my colleague in our laboratory. Peak profile was quite okk in his case.

You do have crystalline peaks in your pattern. You just need to measure again with long counting time in order to increase the signal to noise sufficiently to fit the pattern. You also should make some effort to decrease the background (using a flight tube, tighter slits, shielding for air scatter, etc...

The rising background at higher angles is an indication that you are not blocking air scatter at higher angles. Also the "amorphous peaks you see at low angle could in fact be a nano-crystalline phase or simply background from perhaps the sample holder? (if it is amorphous).

Dear Nasser

Can you inform me why you want to measure the XRD to the fracture surface of ZrB2 ceramic? I think there are other techniques for the analysis of the results of this ceramic fracture such as EDX and EDS with FESEM technique. Also what is the software that was used in the analysis of the XRD results?

Dear Uday 

I have done the EDS analysis with FESEM, I just want to know that the new phase is crystalline or not. Also I used Xpert High Score Plus to analysis the XRD pattern.

I recommend to spin the sample as fast as your device can and don't use a parallel beam. Besides, increase the step duration time and resolution. Therefore your intensities will be higher. You can do overnight scanning. Because of the signal coming from the crystal peaks will be sharper than the other part it will be easier to analyse your spectrum. Finally, normalize your spectrum and use savitzky golay filter for a better view.

Good luck,

Mehmet.

Mehmet, I disagree. I think if the sample is a fracture surface then it will be uneven and subject to errors from variable sample displacement. This can be overcome by parallel beam. Why do you suggest not to use parallel beam? I agree with spining and longer counting.

http://www.xos.com/techniques/xrd/parallel-beam-x-ray-diffraction/

http://www.rigaku.com/products/xrd/ultima/app016

Ian,

Errors coming from uneven form can be estimated and explained (such as peak broadening). It is a trade off : If you use parallel beam, you need to use a parallel beam collimator which blocks the some part of the beam and limits the beam.

http://en.wikipedia.org/wiki/Collimator

That is, some of the X-ray will be reflected from metal collimator and only parallel reflected beams are transmitted to surface. As you can see from his spectrum, he is frustrated because of low intensity of crystal peaks. I suggest to use largest slit with a very long parallel length as much as possible for this case. However, if he doesn't have a slit like that or if peaks are not weak,  as you have mentioned, i recommend to use parallel beam as well. 

The fracture surface should be polished.