completeness of in-situ techniques

[kara-mela]

The ontology includes a (low) amount of in-situ techniques, namely ‘in-situ diffraction’ and ‘in-situ surface diffraction’. I have several questions/issues on that:

1. This list is currently not complete as 'in-situ spectroscopy' is missing, for instance.
2. Although ‘high pressure single crystal diffraction’ is an in-situ technique, it is not related to any in-situ technique through a subclass relation.
3. 'versus sample state' is a superclass of
   (a) ‘versus sample temperature’ (no subclasses)
   (b) ‘versus sample pressure’ (‘high pressure single crystal diffraction’)
   (c) ‘versus sample magnetic field’ (no subclass)
   (d) ‘versus sample electric field’ (no subclass)
4. There does not seem to be a relation to time dependent techniques, although this is often interrelated (change of T over time and thus change in material).
5. The only example for an ex-situ technique is ‘ambient pressure x-ray photoelectron spectroscopy’.

My point is that the PaNET ontology is currently 'imbalanced' and I see three options to continue here

• add terms when requested or 'discovered' at a specific beamline
• create closed terminology, which could result in an explosion of terms
• exclude terms describing physical/chemical experiments

I see similar issues for nanofocus vs microfocus as well as hard and soft x-rays,

I am curious about your opinions and

[spc93]

These are important points.
Terms have been added mainly as required (or taken from facility lists). Some extra terms have been added as placeholders and illustrations.
I would probably not add terms until they are needed.
There is a general point about the separation of technique from sample/subject. The latter is not within the scope of PaNET unless it represents a significant modification of the technique. The separation is not exact!

[paulmillar]

Just to add my 2c-worth :-) ...

I would consider the distinction between in-situ, in-operando, ex-situ as a distinction with the sample and not with the experimental technique.

The information that data was obtained under in-operando conditions is (likely) important to researchers, but that shouldn't mean that PaNET needs to capture this information. This information should be stored and queried through some other fields/standard, perhaps one more focused on capturing the sample and its environment.

The aspect I find particularly relevant is that (I believe) the "in-situ" or "in-operado" label may be applied to almost any technique. Although an experiment being in-situ does have an impact on the experimental setup, whether or not data was obtained "in-situ" or "in-operando" is somewhat decoupled from the experimental technique.

The risk we face here (when including these sample-related aspects in PaNET) is that it creates a combinatorial explosion in the terms: if we have technique X then we would need to add "in-situ X", "in-operando X", "ex-situ X" as additional terms. This would be for each term X. If we consider "high pressure" then we should also add "high pressure in-situ X", "high pressure in-operando X", "high pressure ex-situ X". That's just with one pressure term ("high pressure"), we could want other qualifiers. Now also consider terms describing the magnetic field, the electric field and the sample temperature. This clearly doesn't scale.

In concrete terms, here is my suggestion:

We retire the two techniques that have "in-situ": we keep the terms, but mark them deprecated and provide an alternative as a recommend (e.g., "in-situ diffraction" -> "diffraction")

The "versus sample state" and direct subclasses ("versus sample temperature", etc) are simply removed. Datasets should not have been tagged with these terms. (We could also mark them deprecated, if people prefer).

The term "high pressure single crystal diffraction" is marked deprecated, with the recommendation to use "single crystal diffraction" instead.

[spc93]

I think this is tricky and the distinction is not clear. For something like high pressure single crystal diffraction, I think that most scientists would consider this to be a technique, which a beamline either supports (e.g. it has a high-pressure cell) or doesn't. I think we need to discuss.

[paulmillar]

I think this comes down to a question of scope: just because something could be described by PaNET doesn't mean that it should be. Similarly, just because a researcher (quite reasonably) would like to know something, it doesn't follow that PaNET should provide that information.

Put another way, there's a (I believe, currently unanswered) question on what is a technique? PaNET just assumes that everyone agrees on the boundaries between experimental technique and an experimental setup (for want of a better phrase). When adding more PaNET subclasses (i.e., more specific terms), do we reach a point where the terms are no longer describing techniques but rather a setup?

... but, I agree: this is something we should discuss. Ideally, we would have some guidelines to allow us to make consistent decisions.