Dear colleague Saeed Moshfeghyeganeh 

it is better to upload the XRD chart

Dear Saeed Moshfeghyegane,

when measuring a single crystal substrate you shouldn't be surprised to see discrete poles, also for in-plane peaks. You can use a pole figure software such as CaRIne to calculate in advance what to expect. The fact that you observe ony two poles (90° apart) instead of four, indicates, however, that your 2theata value isn't well set on the investigated reflection or that your goniometer has some uncorrected angle offsets.

Thank you Dr. Eldenglawey, 

I attached the graph. the red one is (002), the blue one is (004) and the green one is (006) peak. The intensity is in arbitrary units. 

Thank you Dr. Neu,

Is it possible that we have two domains in the substrate? And two peaks come from different domain?

Dear Saeed Moshfeghyegane,

reading your first question, I was actually thinking you refer to in-plane reflections which come from lattice planes perpendicular to the film plane, but apparently you were measuring a (00l) reflection of the (001) substrate in normal Bragg-Brentano geometry on a non-tilted sample? This makes your outcome indeed more complicated to understand, but I can still think of an unwanted sample tilt (either from sample mounting or from the sample holder). I suggest to perform a rocking curve (omega scan) on the (004) reflection to check, whether the maximum is centered or not. The question of crystallographic variants can probably best be answered by the substrate supplier.

Dear Dr. Neu,

you are right. I am measuring (00h) reflections of the (001) substrate. I did the rocking curve on two different substrates. Although (004) reflections happen at different 2theta values, the omega curve has the similar feature (with a tale on the left side) . I attach the graph. Also it seems that appearance of these two peaks is related to sample tilt. As I mount the sample again the first curve disappears. But I get much broader peak.