We present an XMM-Newton study of 39 clusters of galaxies (0.4 < z
<1.4, 1.5 < kT < 11 keV) performing a spatially resolved X-ray
spectral analysis, using Cash statistics and modelling the XMM-Newton
background instead of subtracting it.  We did not observe a
statistically significant abundance evolution with redshift.  The most
significant deviation from no evolution (at a 90% confidence level) is
observed considering the emission from the whole cluster (r < 0.6
R500), that could be parametrized as Z~Z_0*(1+z)^{-0.8+/-0.5}.
Dividing the emission in three radial bins, no significant evidence of
abundance evolution could be observed fitting the data with a
power-law.  A substantial agreement with measures presented in
previous works is found.  Computing error-weighted mean of the
spatially resolved abundances in three redshift bins, we found it
consistent to be constant with the redshift.  Although the large error
bars in the measure of the weighted-mean abundance prevented us from
claiming any statistically significant spatially resolved evolution,
the trend with z in the 0.15-0.4 R500 radial bin complements nicely
the measures of Maughan et al. (2008), and broadly agrees with
theoretical predictions.  We also found that the data points derived
from the spatially resolved analysis are well fitted by the relation
Z(r,z) = Z_0*(1+(r/0.15R500)^2)^{-a}*(z/0.6)^{-gamma}, with
Z_0=0.36+/-0.03, a=0.32+/-0.07, and gamma=0.07+/-0.21, which shows a
significant negative trend of Z with the radius and no significant
evolution with the redshift.  The present study is the first attempt
made to spatially resolve the evolution of abundance with redshift.
However, the sample size and the low statistics associated with most
of the clusters in the sample prevents us to draw any statistically
significant conclusion on the different evolutionary path that the
different regions of the clusters may have traversed.