According to the Undisputed Source of All Human Knowledge (wikipedia):
"Micro is an English prefix of Greek origin that refers to an object as 
being smaller than an object or scale of focus, in contrast with macro." 

So perhaps "smaller than the regular beam" really is the best definition 
of "microbeam", albeit not very useful as a global standard.


However, since it sounds like you are looking for a criterion for 
admission into the "microbeam club", I imagine you are looking for a 
size of the beam that crosses some threshold where an interesting aspect 
of the experiment changes.  I would propose that this should be about 10 
microns, since this is the size of the first Fresnel zone for 1 A 
radiation with the detector/source at 1 meter distance.  That is, beams 
smaller than this can potentially change the "correlation length".  Ten 
microns is also about the smallest lysozyme crystal that can yield a 
full data set due to radiation damage limits.


Personally, I think the terms "minibeam" and "microbeam" are about as 
useful as the terms "minicomputer" and "microcomputer".  From what I've 
heard, the term "minicomputer" was introduced when computers became 
smaller than a house, and "microcomputer" when they became smaller than 
a desk.  Clearly, this was not three orders of magnitude change in size 
(as one would expect form analogy to milli- vs micro-), but it was three 
orders of magnitude in volume.

I propose that a much more relevant metric for x-ray beams is the volume 
of a crystal than can be probed with it.  After all, scattering power is 
proportional to volume, not linear dimensions.  Yes, crystals can have 
odd shapes, but the part that stays in the beam as the crystal rotates 
is a "round" volume defined by the size of the beam.  The units are 
actually convenient.  For example:
a 1 millimeter cube has a volume of one microliter (uL)
a 100 micron cube is one nanoliter (nL)
a 10 micron cube is a picoliter (pL)
one cubic micron is a femtoliter (fL)

So, why not talk about beams in terms of pL and nL?  In this way, the 
"sphere of confusion" for the goniometer is implicitly incorporated.  
For example, I can say that most beamlines at ALS have 1.0 nL assay 
volumes, and can be apertured down to ~30 pL with a 30 micron pinhole, 
or to 1 pL with a 10 micron pinhole (although the latter is not popular).
 
Since damage is proportional to photons/area, there is no radiation 
damage advantage to shooting a given crystal volume one tiny bit at a 
time.  On the other hand, there is also clearly no point in illuminating 
cryoprotectant, nylon or other non-crystalline crud unnecessarily, so 
using a beam that is bigger than the crystal will only give you 
increased background (unless the crystal is naked).  Therefore, in 
general, it is best practice to match beam size to crystal size.  Yes, 
some crystals are bent, wrinkled or otherwise malformed, and shooting 
just one bit of them can be an advantage (as Tassos pointed out), but in 
my mind I simply view this small "good volume" as "the crystal".  After 
all, anything you don't want to shoot is (by definition) not your 
sample.  But perhaps I am alone in this view.

-James Holton
MAD Scientist


Richard Gillilan wrote:
> Just an interesting question of semantics that annoyingly comes up 
> from time to time when people are comparing x-ray beam diameters.
>
> What counts as "microbeam?"
>
> Of course "micro" has the precise meaning in SI as being a factor of 
> 10^-6.
> The problem is that the prefix "micro" just means "extremely small" in 
> common usage.
>
> The term is used very confusingly everywhere. Take microwaves. 
> Microwaves have wavelengths from 1 millimeter to 1 meter. Go figure. 
> They're just "extremely small" radio waves.
>
> Now I believe that it is more widely accepted that "nanofabrication" 
> is making objects that are measured in nanometers.
>
> So shouldn't microbeams rightly be x-ray beams with diameters measured 
> in microns (i.e. < 1 mm and >= 1 micron). Of course this makes all 
> crystallography beams microbeams and everything smaller than 1 micron 
> a nanobeam. That won't be popular.
>
> I've always called anything smaller than 50 microns microbeam because 
> that's about as small of an aperture-based collimator as we could 
> make. So a user should ask for "microbeam" if regular collimator is 
> too large.
>
> I was always puzzled at the APS habit of calling this "minibeam", but 
> it's starting to sound better all the time.
>
> But in practice, I think "microbeam" sometimes means "smaller beam 
> than yours." So microbeam used to be 30 microns, 10 or 5, now maybe 1 
> micron. Pretty soon no microbeam at all.
>
> I think maybe I'll stick with "small", "smaller than usual", and 
> someday "extremely small."
>
> I'd love to hear people's opinion on the topic.
>
>
> Richard Gillilan
> MacCHESS