I am refining XRD data of an oxide material with Rietveld method using GSAS. Chi2 value is coming 0.86. Is there anything wrong in chi2 lesser than 1 (but nearer to 1)? Why is this coming and how can this be solved?
A goodness-of-fit or chi^2 is calculated from Rwp and Re factors. Sometimes the weighting factor used in Rwp calculations is causing the goodness-of-fit to go below 1 (e.g. single crystal refinement using SHELX). Could you please provide your X-ray data and some basic information for your material. I can try Rietvield refinement in FullProf and we can compare it to see whether there is not other problem.
The chi2 value is a squared ratio between Rwp and Rexp, i.e. between R-factor from your refinement and R-factor expected from counting statistics. Therefore, chi2 below 1 means that either you have introduced too many parameters in your refinement (low Rwp, fitting of statistic noise with excessive parameters) or the XRD pattern too bad (high Rexp). Thus coming the suggestion to exclude some meaningless parameters from the refinement or try to acquire better XRD pattern. The later is always better, while in case it is impossible, simply exclude those parameters which are below 3sigma.
It means you are refining the backgrounds as well, thus overestimating the errors.It can never be lesser than one.
In general, I agree with the above suggestions by Igor Plokhikh. If the XRD pattern cannot be re-acquired with a better statistics, I would suggest excluding highly correlated parameters from refinement: when the correlation degree between two parameters is much higher than (say) 50%, one of them can well be fixed... Good luck!..