On Fri, 14 May 1999, Frank May wrote:
I decided to answer to the list, since this may be of interest not only to
Frank
> Q1. Using powder XRD, can meaningful information be obtained about the
> populations of sizes in a specimen comprised of "crystallites" ranging from
> 20-200 Angstroms? Note that in addition to the distribution of broadening
> due to crystallite size variations, there is also the probability of
> variations in unit cell size caused by the presence of a second (or more)
> alloy component(s). The latter condition manifests itself by a peak
> centroid shift.
A: Yes, but it ma take a lot of effort, see below
>
> Q2. Is the measurement of a 30 Angstrom "crystallite size" meaningful? It
> is accepted by practitioners of surface science techniques (ESCA, etc.)
> that a "surface" is about 30 Angstroms deep.
A: We have been hunting for the evidence of the surface modification
in nano-powders for quite some time and has not yet found any indication
of such an effect. Which means that if it is there it doesn't penetrate
deep into the bulk. I have seen myself SiC nanopowders with 30A
crystallites and literature lattice constants. I know of the work on
platinum catalyst with 40A crystallites - again a perfect Pt structure.
HRTEM pictures published in the literature show the same uniform structure
over the whole volume of the nanoparticles.
>
> Q3. If practical, what would be required to obtain both the data and the
> software to process the data? Obtaining synchotron data is not an option
> in this case. Software may run on either PC or Macintosh.
> Please respond both to this group and directly to me at the address below.
A: We have a set of programs to calculate powder diffraction patterns of
nanoparticles from first principles i.e. for particles defined as sets of
atoms. It works in such a way that one defines a unit cell, grows it into
a larger crystallite and cuts out of such an assembly a particle of a
desired shape. Atomic coordinates are written into a disk file and
another two programs calculate the diffraction pattern of an ensemble of
identical crystallites. Having a set of such patterns calculated for
various sizes one can combine them to get a size-distribution. Practical
limits: two atomic species and 150A maximum size (the number of
operations is proportional to the 6-th power of linear dimension). We also
have a program that can go up to 1000A, but the particles must be
spherical.
The software IS NOT user friendly but it works and I can provide it to
those interested. It is available as a C+Fortran code (compiles
smoothly on many systems) or IBM-PC executable.
The really hard part is to find the right model of the size and shape
distribution to fit your data.
For the measurements: if the stuff doesn't contain Fe or Ni and you can
provide 1 cm3 there should be no problem taking some patterns.
Stan
--------------------------+---------------------------------------
Stan(islaw) Gierlotka | High Pressure Research Centre
[EMAIL PROTECTED] | Polish Academy of Sciences
X-Ray Diffraction Lab. | Sokolowska 29, 01-142 Warszawa, Poland
phone +(48 22) 364433 | phone (international) +(48) 3912 3276
| | phone (local) +(48 22) 6325010
| | | fax (international) +(48) 3912 0331
| | | | | fax (local) +(48 22) 6324218
_!_!____!___!_!__!__!!____!________________________________________
