Which condensed matter experiment PhD group is the most fun?

[davidbenari]

TL;DR Summary

What type of experimental work is most fun in condensed matter?

This is going to be controversial and might even be taken down, but I think what I will say is absolutely true, and I'm sorry if it offends people.

I'm applying for the second time to condensed matter PhDs. I was in a group that did a lot of device fabrication as part of their experiments and quite honestly this is hardly a physics PhD; I was doing very tedious routine like steps and then leaving all the interesting physics to the theorists.

My idea is that there are some experiment groups that require more thinking than others, which makes them more fun. It seems like some groups have a huge separation to theory, thus leaving the experimentalists absolutely clueless about what they do (of course they know the really petty details but it's almost as if they're studying something that's not a physics PhD). And then it seems like other groups actually require lots of thinking.

I am asking what type of experimental condensed matter groups you consider the most intellectually satisfying and active? Here is my impression so far:

Anything related to optics, like Ultrafast X ray spectroscopy of solids, will require lots of thinking as you have to know how to set up electronics and optical setups.

ARPES and STM publish papers that seem really sophisticated, so I'm guessing it's intellectually pleasing to do.

Then the absolute worst is transport groups and fabrication groups. All they do is IV plots and spend months fabricating a device which literally anyone can do, even people without a degree.

 

[Keith_McClary]

Even decades ago when I went to school I had no idea what experimentalists did.

 

[davidbenari]

Even decades ago when I went to school I had no idea what experimentalists did.

LOL

 

[fluidistic]

Then the absolute worst is transport groups and fabrication groups. All they do is IV plots and spend months fabricating a device which literally anyone can do, even people without a degree.

By IV you mean current/voltage?
I think that it can be quite challenging to get voltages and current in some small anisotropic monocrystalline samples, when their geometry is such that no analytical solution exists for neither the heat equation nor electrodynamics/statics equations inside of it. You have to use numerical solutions, and even then, setting an experiment that will retrieve the quantities of interest is far from a given.
I am not sure the goal should be to build a custom made device that can retrieve the transport properties. Usually you would rely on custom made apparatus when there is no preexisting apparatus that can retrieve what you want, and if you go the custom made one, try to make it easy to reproduce in other labs, otherwise it's just a one punch knock out that works only for you and nobody else can try to even replicate your work.

 

[f95toli]

And all theorists do is programming and code maintenance...

The reality is that irrespective of what you do you will spend a lot (often most) of you time doing things are are not a "core" activity. A good experimental physicist need understand theory well enough to do their work. Often this is NOT when they are the lab (where a different set of skills become important) but before and after the actual experiment; that is when you are planning an experiment, designing up the apparatus and analysing the data.
That said, you also need to understand what you are doing well enough to make good choices during an experiment (what does that slight kink that suddenly appeared in the IV curve mean?); if you are just measuring at random without doing some analysis on the fly you are doing something wrong.

As an experimental PhD student you will almost inevitably need to spend 1-2 years learning all the practical skills you need to do the measurements before you get to the point where you are taking data that is novel/interesting enough to require some serious thinking.,

Also, if you think optics is different from other fields you will be disappointed. The optics people in my department spend most of their time stabilising lasers, measuring vibrations and adjusting optical elements on their tables.

Device fab IS different and there is a reason for why this is typically done by specialists. Never make the mistake if thinking this is easy, the fab done for experiments is almost never "standard" and requires a lot of skill and experience to get right. Also, the people doing the fab are typically also the ones doing the design meaning they need a deep understanding of the physics. The work itself requires you understand not only which buttons to press but also the physical/chemical principles of how the equipment operates. Good device fab is a craft which combines hand-on skills with a lot of science and a bit of witchcraft.

When I looks for a PhD student where the intention is that they will spend most of their time doing fab I always make point of explaining this; if someone start a fab-based project believing that they will be spending a lot of time solving equations something has gone very wrong. ,

 

[davidbenari]

@f95toli

Well I think programming and code maintenance is a lot more intellectually pleasing than doing etching, for example. I get where you're going that you think I am simplifying a field, but I don't think it's a valid critique. There are aspects of experiment that are less intellectually pleasing than programming or solving an equation. I can think of two ways of proving my point:

(1) One is how hands off experimental professors are. Which is why the US system is basically a lot of foreign grad students working as slaves for 8 years and doing the grunt work the professor doesn't want to do; whereas I think the theorist professor is a lot more involved. (2) Another example is based on the fact that in industry, e.g. the semiconductor industry, fab is relegated to technicians. Moreover, in industry they have machines that do the tedious work the grad student has to do by hand. Why? Because doing this work is a nightmare, not because its hard at an intellectual level, but because it's painfully boring. Sorry, but no matter how much you say "you are doing nanofabrication at an atomic level" diping wafers in a bucket isn't fascinating.

Also I disagree with having to "know the physical/chemical principles of how the equipment operates". In real life they just try a bunch of recipes and vary parameters systematically until they get the right fab out of the process. Sure they need some intuition, but they don't need to know much.

And I also disagree that because you're doing the design you need a deep understanding of the physics. I knew people at a top 10 school with papers in Nature and Science working on superconductivity and didn't even know what second quantization was. How is that deep understanding of the physics?

Besides physics that has been known for a long time that kink in the IV might barely be understood by the experimentalists in some fields. Why? It's just too hard and it takes 5+ years to understand, so instead they're just told to look for that kink and then they say the cause is "electronic correlation" or something like that. I've seen this happen a lot. Of course there is some level of analysis involved but its not as fancy as it sounds by reading paper titles and abstracts.Anyways, I'm pretty sure I hate fab. You might disagree on my reasons. I still love physics and want to find what fields are a right fit for me.

 

[f95toli]

@f95toli
(1) One is how hands off experimental professors are. Which is why the US system is basically a lot of foreign grad students working as slaves for 8 years and doing the grunt work the professor doesn't want to do; whereas I think the theorist professor is a lot more involved.

This happens everywhere. I am nominally an experimental physicist (in the UK) but I have been doing this for a long time now (I finished my PhD 15 years ago) and my time is -literally- too valuable for me to spend a lot of time in the lab. I have a research group to run, meetings to participate in etc. Hence, whereas I would love to to spend more time in the lab, I literally do not have time. This happens in all professions, once you get really good at something you are expected to do less hands-on work and spend more time teaching and managing others.

(2) Another example is based on the fact that in industry, e.g. the semiconductor industry, fab is relegated to technicians. Moreover, in industry they have machines that do the tedious work the grad student has to do by hand.

This is simply wrong. Or at least mostly wrong. It is true that much of the "routine fab" in industry is done by technicians but process development is still mostly done by people with PhDs. They might still have titles as "technician" (but most often it is something with "engineer" in it) but that does not mean that is is not a highly skilled profession.
Now, it is quite rare for the work done in academic groups to involve "standard" fab simply because the research for these have already been done. Mostly, we are using strange/unusual materials, have very strict demands when it comes to imperfections etc. Moreover. most fab is only done for handful of samples before you move on to the next process which means that you are always in state of developing something new. This means that nothing is ever routine. When we DO need something that is "standard" we typically just order that from a foundry; it usually doesn't make sense to develop a whole process for this.

Why? Because doing this work is a nightmare, not because its hard at an intellectual level, but because it's painfully boring. Sorry, but no matter how much you say "you are doing nanofabrication at an atomic level" diping wafers in a bucket isn't fascinating.

I can't help you there. Some people really like fab, some people hate it. If you don't like doing things with your hands it is certainly not for you. I did a bit of fab as a PhD student but it was not for me.

Also I disagree with having to "know the physical/chemical principles of how the equipment operates". In real life they just try a bunch of recipes and vary parameters systematically until they get the right fab out of the process. Sure they need some intuition, but they don't need to know much.

There is some truth in that, but in order to work efficiently you also need to be able to use and interpret data from diagnostic tools such as XRD, AFM, SEM and TEM (just to mention a few). Just varying parameters at random and until you get one good sample CAN work, but it rarely works if you need your results to be reproducible. You also need to understand enough about materials physics so that you can optimise your films (grain boundaries etc)
I am fortunate enough to collaborate with some very good fab people and they certainly understand the physics (although most of them also do some measurements).

And I also disagree that because you're doing the design you need a deep understanding of the physics. I knew people at a top 10 school with papers in Nature and Science working on superconductivity and didn't even know what second quantization was. How is that deep understanding of the physics?

That depends on what you are doing. Someone working on superconducting devices does not need to know everything there is to know about BCS theory, but they do need to understand the physics of the device, which if you are working with superconducting devices can often be adequately modeled as a lossless metal with some kinetic inductance (and an annoying habit of trapping flux!).
As it happens I've worked on SC devices for about 20 years and my most used Hamiltonian is probably the Jaynes-Cummings Hamiltonian (I've even published a couple of theory/modelling papers) which is what I use to model and design devices, this is probably as "advanced" as BCS theory but the difference is that I've never actually had to use anything but the results of the latter. As with most experimental physicists I've dabbled in many different fields but I've only got "deep" understand in a few very specific areas. I've e.g. also worked on electron spin resonance and know quite a lot about decoherence of electron spins in solids, but I don't claim to be an expert in atomic physics.

Besides physics that has been known for a long time that kink in the IV might barely be understood by the experimentalists in some fields. Why? It's just too hard and it takes 5+ years to understand, so instead they're just told to look for that kink and then they say the cause is "electronic correlation" or something like that. I've seen this happen a lot. Of course there is some level of analysis involved but its not as fancy as it sounds by reading paper titles and abstracts.

Then you've been unlucky. I've spent many, many hours of my life discussing kinks in IVs and so far we've always been able to eventually identify the cause for all of them;although some causes require some quite complicated physics to explain (especially some many-body effects and more recently topological effects).

Anyways, I'm pretty sure I hate fab. You might disagree on my reasons. I still love physics and want to find what fields are a right fit for me.

That is fair enough. However, it is important that you have realistic expectations. Real life physics is very different from what many people imagine it to be and mostly you are NOT "doing physics" when working as a a physicist; if you really don't like the "hand-on" routine stuff in the lab you should probably look for something else.
This is all down to personal preference and I've had seen promising students leave physics altogether after their MSc/MRes because they decided that lab work was not for them.

 

[davidbenari]

@f95toli

Out of humility and not wanting to undermine people that spent 8 years doing this I won't argue anymore, although I don't agree with what you said haha.

I just genuinely want to find a good fit. I haven't given up on physics. I think experiment comes in many different flavors, for example, particle physics experimental people can be essentially coders. I think academic counseling has to do a better job at describing all these different flavors of experimental work.

 

[hutchphd]

There is one more angle to this discussion that has not been broached. There is a certain "zen" to doing routine work that I have always found immensely useful. As my value increased there was more and more pressure to do only high-value-added activities: this I resisted vehemently.
I have had perhaps 10 really good scientific ideas in my nearly 3 score and ten years on the planet. Every one of them came to me while I was engaged in some routine task that let my brain off-leash, or by my noting unexpected (perhaps subtly so) behavior in what should have been a routine process and adding that to my pile of useful information. A good physicist never stops doing physics IMHO.

Be careful what you wish for.