the XRD pattern originates from the bulk of the material rather than the surface, unless you have a significant surface to volume ratio; moreover the shift in the XRD pattern can have two reasons: 1-sample displacement error or 2-unit cell parameters change. for the former the shift is uniform across the whole pattern while for the latter the shift increases as 2theta increases. to find out which one is present a simple Rietveld refinement would suffice.

Hi Taher,

practically any variation in chemical composition affects the lattice parameters of a growing crystal. This is obvious in melts where residual components even cause a change of phase, cf.  phase diagrams, Vegard's rule etc. How much these effects become detectable is another question. It is perhaps only detectable by more precise and less averaging techniques, e.g. CBED in TEM. I myself wouldn't assume a measurable effect but it is very responsible to keep an eye on these less probable effects.

Another topic is the discussion about surface effects. Again, a free surface is not comparable with the bulk so that you will have there definitely a relaxation. Cubic cells are then perhaps tetragonal or triclinic (as result of the orientation). However, the major question is, does this really matter, or can you measure this? Since we know that the information depth is in the scale of several microns (dependent on the Bragg angle, the radiation, the absorption) a surface effect which is in the scale of nanometers only cannot be visible in XRD patterns. It is part of the signal but the contribution is below the detection limit.  

Nevertheless, it is worth to discuss the opinion of H. Y. Asl that XRD is a signal coming from the bulk. Well, maybe I have another imagination about the term "bulk" but 5...10µm I wouldn't call bulk. This is for many practical techniques "surface" which is in general a problem for XRD measurements so that techniques like neutron diffraction or the application of synchrotron radiation would much more reliable but they are simply too expensive. Therefore we are using standard XRD but we have to take into account that we are observing only the surface of a sheet or a powder sample.

Finally a remark about the peak shift. I am sure if you scan here at Researchgate you will find similar questions where peak shift has been discussed. It is more problematic as claimed above by H.Y. Asl. Only a misalignment of the detector - the so-called zero shift - will cause a uniform shift in Bragg angle. In contrast, a sample misalignment of a sample is not invariant but strongest at small Bragg angles and becomes Zero at 2theta=180°. The interference of lattice parameter effects and sample defocussing makes a careful placement of the sample essential. Especially for low-absorbing phases the effective diffraction volume is shifted below the surface so that high precise measurements need to consider these effects. 

One good example comes from the world of clays. In XRD, the 001 peak of the mineral smectite changes from about 1.4 nm when air dried, to 1.76 when e. glycol saturated and even more if saturated with other organic compounds. In fact, we use the shift of the 00l peaks in the XRD pattern under different treatments to identify smectite. This mineral is a sheet silicate that has a negative surface charge due to partial isomorphic substitutions inside the cell. Water, hydrated ions and organic molecules are attracted to the surface and cause the expansion.

Of course suface absorption affects cell parameters (volume) in a compound analysed by XRPD (P = powder, diffraction). The case of clay minerals is a good example (as reported by Debora above). In a powder the surface represents a large proportion of the volume, hence reactions taking place on the surface are important. Another thing vould be the surface in a bulk single crystal (even if quite small). In this latter case it may even be negligible.

Please post your XRD data with all experimental details along with sample prep procedure when convenient to enhance responses :-)

I was about to respond then noticed not everyone has been careful to distinguish between adsorption and absorption.

Adsorption will just have the solvent sticking to the surface.  This will not affect lattice parameters of anything - even at the surface.

Absorption means the solvent may be incorporated into the crystal lattice, causing it to expand.  (Although if the pure lattice is rather open, and the solvent is small with high affinity for the solid, it could cause a contraction.)

W.S.I. Boyer's observation is certainly a good point.  However, when Taher Rabizadeh posed the question, he articulated it by saying: "there are surface adsorptions and crystal incorporations" therefore he was contemplating both adsorption and absorption (= crystal incorporation). The latter is certainly to be taken into account as the possible cause of the phenomena observed (peak shifts and cell volume changes).

All of us know that the X-ray diffraction peaks result from the diffraction from the planes in the crystal lattice.It is necessary that the long range order exists within the material to build up this  periodicity.In the absence of the  periodicity one would not expect to obtain the diffraction peaks.When  any gaseous or solvent species are adsorbed on the crystal surface, the species are distributed at random and no periodicity would occur  within the adsorbed species to generate a new lattice at lower concentration of the adsorbed species.At higher concentration this may result in the formation of a solvate and this should result in the formation of a new lattice yielding additional peaks.However, if these species enter the bulk of the crystal lattice , it may result in the change in the unit cell dimensions of the original crystal  and subsequently bring about the change in the unit cell volume of the crystal

Dear Taher,

                                       if , during growth, surface adsorption is not followed by incorporation into the crystal lattice, XRD diagrams will not be affected at all, but only  crystal morphology . However, if absorption follows, this can occur either in a disordered or in ordered way. In these cases both position and shape of the X-ray peaks can be affected.

I'm sending you three recent papers, in which we showed the double effect of adsorption/absorption.

All the best

Dino