00:11
Alright. Hello Masako. Hello Asami. Long time no hear. Yeah.
久しぶり。 お久しぶりです。 I had fun with Yusuke-kun, so that was pretty fun.
Yeah. Yeah. Yeah. You know him too, right? I was thinking about that. Maybe I don't...
Yeah. I don't... Do I know him? Maybe. Maybe you barely missed each other.
Yeah. Yeah. I was thinking about whether I knew him or not and... Yeah. Yeah. I may not know him.
I'm not sure. Yeah. Anyway, you're back and it's this time of the month again where we do
科学系ポッドキャスト. By the way, last month's 科学系ポッドキャスト, we had a lot of
反響 from the listeners. Really? What did we talk about last month? We talked about...
What did we talk about? We talked about... Well, I talked about the book I was reading and
there was a scene where a Japanese political figure came in the scene and he needed...
Oh, yeah. Yeah. Right. ...a translator. Yeah. Yeah. Yeah. That thing...
Yeah. That got us a lot of listens and new followers. Great.
Yeah. I don't know. I don't know. Maybe you spoke to people. I don't know if this topic
will speak to people, but today we will be talking about money. And today, this month's
podcast, 科学系ポッドキャストは、ホステッドバイコペテンナイトさん、宇宙の話とかをされている現役大学院生の方の二人組の、一人は社会人なのかな?と大学院生の天文科学を勉強されている方のポッドキャストなんですけれども、
um, money, you know, that's kind of a taboo, I think, in Japanese culture. Is it a taboo? It's a
kind of a serious topic, though. I mean, it's taboo, but also, like, it's taboo in a sense that
not... that's not the first thing you talk with your はじめまして people, but it is also important
to talk about. Especially, I think, it's very fuzzy and unclear how much money it costs to do
03:00
research. At least, I wasn't sure exactly how much money has been invested in research and why it's
so expensive, you know, and it's all coming from taxpayer money, most of us, right? I mean, some of
us maybe have industry collaborations and stuff, but at least all of the research I've been involved
in coming from government money of some sort. So, yeah, like, it's just, I thought it was interesting
to sort of break down, I mean, not like dollar by dollar, you know, how much it costs, because that
would be too depressing, but I thought it might be interesting to share a couple of things about,
like, lab equipments and things that we have in our labs that are weirdly, unexpectedly expensive.
So, I asked you to think for one or two things, I also thought about a couple of things, so
let me share my first surprisingly expensive things in my lab. So, this comes from my PhD
lab, so physical chemistry, experimental physical chemistry lab, and, I mean, there are more obvious,
more expensive things, like, you know, our lasers are expensive, and we have, you know, for our
entire laser system, comprised of three main boxes, so three main lasers, and they each were, you know,
200k, 300k dollars expensive lasers, and needed a lot of maintenance on top of that, but
I think that's, like, not surprising. The one surprising thing that was expensive, though,
is this crystal that we use to generate, we call it a second harmonic crystal, so you shine, let's say,
an 800 nanometer light to this crystal, and if the crystal phase matches with the laser's phase
condition, you can generate from 800 400 nanometer, so it's like a double power,
not power, sorry, double energy, so, you know, 800 nanometer is longer wavelength than 400
nanometer wavelength, which means 400 nanometer has more energy, right? 400 nanometer photons has
more energy, so that's, so, you know, both 800 and 400 is useful for different things, and sometimes,
and many, oftentimes, we need it 400 nanometer from our laser, so we frequently use this
06:06
doubling crystal, or second harmonic crystal, so 400 nanometer is a second harmonic of 800
nanometer Thai sapphire crystal. It's just, like, it's specific to the material of the gain medium
of the laser, and specific to the crystal, so you can just think of it as, like, this magic crystal
that you shine one light, and some other color light comes out from the other end, because that's
kind of what it does, right? I mean, the actual science is more complicated than that, but I'm not
going to go into it. However, so we use ultra-fast pulsed laser, so our pulse duration is anywhere from
100 to 120 femtoseconds, and that's very, very short pulse, and when a short pulse passes through
a material, you know, the pulse interacts with the lattice of the crystal, and that delays, that
changes the property of the pulse, and more often than not, it stretches the pulse, it, like, it makes
the pulse longer than what it entered, because the crystal, the pulse through the crystal, because
the pulse needs to, you know, interact with the crystal's material, right? So you can think of it as
100 femtoseconds pulse entering would come out after a crystal, 100 and maybe 150 femtoseconds.
So that depends on how much crystal it needs to go through, how much material it needs to go
through, so that's directly related to the thickness of the crystal, right? And we want
as thin crystal as possible that will still give us 400 nanometers, right? So enough interaction
that the second harmonic generation actually happens, but we want it to be as thin as possible
so that we can keep the pulse as short as possible. Oh, I see. Because we work so hard
in the upstream of the laser to generate this very, very short pulses, ultra-fast,
ultra-short laser pulses, we don't want to ruin all of that effort and elongate the pulse.
So making thin crystal turns out to be very hard and very expensive. So these are, let's, sometimes,
I can't remember exact number, but I think we typically used the crystal, the thickness of
maybe like 0.05 millimeters or like 0.1 millimeters on a thicker side of things,
09:07
so it's very, very thin. It's like super thin, super easy to break. And these are maybe the
sides that open the aperture of the crystal, so the part that's exposed to the laser pulse is
maybe like five millimeters by five millimeters. So you can imagine five millimeters by five
millimeters square, that is 0.05 millimeter thickness, so it's very thin. And so it's a tiny
thing, you know, and this thing would cost like anywhere from, if it's a high quality material,
maybe like two thousand, three thousand dollars. Oh, wow. US dollars, exactly. It's so expensive.
Can you use it many times, multiple times? Yes, so as long as you don't break it physically
or burn it, which is more likely to happen when you're using laser, you could burn the crystal
because our pulses are so, so short and pretty intense in energy, like the intensity is high, so
there's a possibility of burning the crystal. But if it doesn't burn, then, you know,
we can use the crystal for as long as we want to. More often than not, we either burn them or
someone who didn't know how to handle crystal accidentally tries to clean the crystal,
which is never a good idea. If you're trying to clean something so thin, you just cannot do that
manually, you know, you just cannot. Yeah, like usually we would take a lens paper and a forcep
to clean the surface of the mirror, for instance. But if you do that to 0.05 millimeter thickness
crystal, you're going to break it. And that's it. Two thousand dollars out of your way. So
that's our weirdly expensive things, number one. So yeah, I remember thinking like,
what? This tiny little thing costs so much? This is insane. We have so many of those
on laser table, you know, not just second harmonic crystal, we have third harmonic crystal,
we have a bunch of other thin crystals like that. All needed to be really thin because
of the nature of what we did as an ultra fast molecular dynamics people generating,
you know, 100 femtosecond lasers. So yeah, that's a weirdly expensive thing.
12:07
Hmm. But what about what about your expensive thing, Masako?
It's just the running cost to just to do experiments. It's quite expensive. We use MRI,
so it's $500 or 50,000 yen, Japanese yen per hour. Per hour? Per hour, yes. And we normally
use two and a half hours per session. And several sessions, sometimes it's just two sessions,
sometimes up to four sessions per participant. Okay, how many participants do you typically
need? It depends on the project. Right now, it's more than 30 subjects for this particular
experiment. Sometimes it's just about 20 subjects per condition. And we have multiple conditions.
And then we have multiple projects. So that would sum up to one project probably will cost
one postdoc salary. So if a postdoc has one project, it will cost two postdocs salary.
Wow. Wow. So it's quite expensive. And you don't even know if the data that comes from
all of these sessions are useful? No, we don't know. We're not sure if we can even publish those
data. Wow, that's so expensive. What a gamble. It's a gamble. Every day, it's a gamble. Wow.
Yeah. I'm surprised because, you know, fMRI machine is, you know, it's already commercialized.
It's still like a pretty groundbreaking technology, but it's not new anymore. You know,
it's been around for how long now? Like 10, 20 years, maybe? Even the high res ones, right? So
I'm surprised that the operation cost is that expensive.
Yeah. It's to use the system. To use, right? And that doesn't even count the maintenance fee.
No, it's so expensive. I don't know the exact amount, but it's always akaji.
Wow. Now I have a newfound respect for people who publish fMRI data now.
15:04
Wow. It's like all of a sudden, such a risky experiment.
I mean, I, as a participant to your project, it was very easy. I just lay down,
hang out in an fMRI machine for like, you know, half an hour.
We don't, we won't tell participants how expensive those, you know,
experiments would be. Otherwise it would be difficult for them to.
Yeah. Wow. Huh. I never would have thought that fMRI machines are that expensive.
I wonder, I wonder, I mean, thankfully I've never been in a situation where I needed an fMRI
machine at the hospital, but I wonder if the lab tests, you know, if you have injuries or something
to do MRI scans, if that's also expensive or if like even gomaiam per hour is still like at a
discount to research rate. Yeah. In some other places it's like a gomaiam per hour, I guess,
including the, yeah, like a technical assistant fee. So, but in some places it's free if you're
collaborating with the researchers at the institute. Oh, free to you, but not free to
someone else. Right. Yeah. Probably. Yeah. Probably costing someone else's money.
Yeah. So it depends, but it's, it's quite expensive to run the experiment.
I mean, my beam time experiment is also expensive. Weird. I mean, not, maybe not
weirdly, because if you think about running a giant laser that goes on for like two kilometers,
you know, that's a sensitive instrument, you need like 10 people on a team. So just like,
yeah, like you said, just the operation cost is expensive. And yeah, you don't even know
if it's going to be worth it, the data, you know, so I guess it's also expensive in that sense.
Um, but like, at least it's not like a daily thing that we do, you know, like you need MRI
machine every day. And yeah, at least several times a week. Yeah. We use it. Right. Like,
like you need them to be available to you mostly at all times. Whereas like beam time as, you know,
expensive as it is, it only happens like once a year if we're lucky. And I mean, technically
doesn't come from our pocket money, right? It comes from the grant money, which comes from
18:05
someone else. So there's that. But interesting, interesting. Yeah, I'll go back to my expensive
things in a lab. I have one more expensive thing that might be kind of, kind of surprising.
Um, not quite. Let's see how I don't, I don't know how. Um, I don't know if this is particularly
cheap or particularly expensive, because it's a technician cost. So once in a while,
uh, when the laser is out of alignment or out of condition so much to the point that
a grad student cannot fix, we will call a field technician, a field engineer from the,
from the, from the manufacturer. So like the vendor person will come, you know, someone who
knows inside out of the instrument who can like open up the boxes and fix it.
Um, I have had a field engineer visit maybe two times in my entire time in, you know,
entire five years in grad school. And, um, each time it costs us about $5,000
for two days of his working day. What? I know, like, I mean, you know, that includes like
travel costs, whatever, but it's mostly for his skills that we pay for. And of course,
it's from the corporation. So like, we are probably paying a hefty amount of markup.
And, you know, he probably doesn't get all that $5,000 in his pocket, but that's how much
his visit cost to us. Um, so does it include, um, like changing some parts of the device or?
No, no, no. If there, if there's a different, if, if we need a new parts,
if you need a complete part replacement, then it's 5,000 plus the cost of whatever
parts that we need to replace. You know, if it's a diode head or different grating
or some other things. Um, so it like the 5,000 is just purely the human cost.
Wow. Thankfully, um, the guy who came like both of the times are very, very competent person. So,
um, yeah, he usually fixed the problem in two days. And sometimes, uh, because he's worked
21:03
with our lab, my boss's lab, uh, for a long time before. So he, like, if he couldn't fix it in two
days, he would like come for the third day and try and help us, you know, still at $5,000. Um,
but yeah, like, so that's, that's really like crazy. Right. To think that $2,500 per day.
Yeah. Yeah. For this person. Um, so it was a big deal to have him in the lab. Either we have major
issues that we cannot solve or, um, yeah, like, like we've tested everything else that we can,
and we realized, okay, this is we, we need professional help and, you know, for him to
come and fix it. And it's not like we can always get ahold of him either. You know,
he's not like a plumber, right. He has his schedule. He has the entire, uh, you know,
America pretty much because he covered the California region and Northeast region. So
he goes around different university fixing lasers or different, you know, companies fixing lasers.
So he's not always available. And, um, yeah. So, so there was that. Yeah.
Wow. So several, another weirdly expensive thing. $100 per hour. Yeah. So my boss had a system.
Um, he had, so in our university, a grad student cost it roughly about like 80 to 90 K
a year, right. Cause, um, that's, that includes tuition fee, our salary. Um, you know, our salary
is not 80, 90 K, right. Like that's not the money that I get paid as a grad student. That's the
money he needs to pay to the university to hire us. But only some, like, you know, I would say
maybe like 35, 38 K is like our salary roughly. And, um, and so, so, so he had to pay almost the
double of what we costed because, you know, tuition fee and overhead costs and other things.
So he had an idea of how much our daily labor costs him. Right. And he also had some spreadsheet
of, uh, like items we need to purchase or like sometimes, you know, is this the level that we
need the $5,000 field engineers and whatnot. So he had all these like variables input and he would
calculate for a specific given problem, the total cost of how much it would potentially take to
24:05
solve this one problem. You know, let's say fixing this particular section of the laser,
you know, costs X many dollars and he would divide it by our daily wages and he would know,
okay, this cost us two weeks of Asami salary. Is this something she can fix in two weeks?
Does it look like she needs more than two weeks? If so, when is a good time to request a $5,000
help? You know, or can we solve the problem by buying things? Like maybe the current part that
she's troubleshooting might be fixable, but if it goes beyond two weeks of her salary,
then it breaks, like, it doesn't make sense to keep investing her as the way to solve the problem
when it's easier to buy the item and just completely replace the part. Like we're not
going to know what was the root cause, but we will replace and that will save her time.
Right. So he had this like multivariable calculus in his spreadsheet of how many, how much cost it
should, like monetary cost, how much financial cost it should take to solve this problem and
divided by the grad students' daily wage and decided that this is the time span it should take.
And that's how he decided actually to make a decision on whether to purchase item or request
professional help or whatever, or if it's something that I should be fixing. Because like maybe
the answer comes out to be, oh, this is like two days of Asami's work. Then maybe it's okay if I
take four days to fix it. It's still doable. Like maybe it's still cheaper than hiring
field engineers. But if it's going to cost me one month to fix this problem and if it's,
if it looks like it might not fix the problem, then might as well just get the field engineer
from the beginning. Right. Like that sort of was the idea that he was thinking. And I really liked
that he had a practical approach to this, like very German of him to like think about this in a,
yeah, very clear cut way of deciding how much financial cost it should cost to solve a problem.
And rather than wasting grad students time and energy, he would decide based on his little
spreadsheet, if it's worth investing the money to buy new parts and get professional help, or if it's
27:01
worth my issue to fix. Because there are parts that we cannot outsource a problem, right? Like if it's
like our, like lasers, we bought our lasers commercially, but they were like 50% of our system
was not commercialable. Like we were making a homemade instrument. So for the homemade instrument,
there's no one who can fix this. We need to work at it. So there's only so many things he can do.
So he will prioritize us spending more time and energy on the parts that other people cannot help
us. And for the parts that other people can help us or other like purchasing item can help us,
he prioritized that. But every time a field engineer comes, like I need to be physically
there for all of the time that he's working and like just learn, absorb everything from him.
Like just follow everything he's doing, observe everything he's doing, take notes.
And I would try and replicate that when I see similar problems, right?
Yeah.
It was, it was really helpful. And he always has like, you know, secret tricks up in his sleeves.
So something that the manual doesn't necessarily say how to do, he'd be like, Oh, this is how,
this is a little trick that I do to check or something. And it's funny because he was kind
of like a grumpy grandpa type character. He did not like that I hang out with him to observe his,
I mean, he didn't sound like at the first session, he seemed like he completely ignored me. He just
like did his work. And I asked him questions. And he said yes, or no, or like, just like flat out
ignored. But second time I saw him, a few years later, he was a lot more friendly. I think he
got used to me. Yeah, he like, Oh, he's this is the girl who just wouldn't leave me alone. So I
might as well enjoy our chat. So he shared more tips with me, which was great. But yeah, like,
I don't know how many PIs have this very practical approach of listing how much it
costs versus and weighing that out with a grad student cost. So, I mean, I don't know if I'll
ever become a PI. But if I do, and then or if I start hiring people, you know, even if I don't
become a PI, right? I want to keep that in mind as an approach, because like, there are definitely
things that I've seen other labs operate and they some labs, you know, cannot afford a $5,000 field
engineers. So they make their grad student slave away at this laser for, I don't know, like two
30:07
months. And maybe the problem was a lot easier to fix if you just buy the new crystal or buy a new
mirror or something. But like, they wouldn't let them do that first, like they need and that's two
months of precious time as a grad student, just gone for fixing problems. Not even doing experiments,
like trying to fix the problem to do the experiments. You know, so like, I think that's
a source of resentment and this discouragement. Yeah, you know, so if I do become a PI, I want to
be mindful of that. Yeah, yeah. I'll keep that in my mind too. And usually when something breaks down,
that's something we can't really fix. So we have to ask. Yeah, I think it's a very
physical chemistry thing to try and fix our own instruments. Like maybe physics also do that,
or like some engineering labs. I know that, for instance, a biochemistry friend of mine,
they fix something very simple, like chiller system or something. But
if there's something wrong with their minus 70 fridge or their GCMS,
they're not going to fix it themselves. They're going to have the engineers come right away. So
this approach of fixing up lasers, like we definitely touched and opened the boxes that
we were not supposed to be opening and closing as well. Like, you know, those stuff that would
get us in trouble with warranty, we definitely have done that. And maybe that's not usually
the case for your type of research. Because I mean, I don't know, I'm sure you've used MRI machine
for a long time, but you probably don't know how to fix the MRI. We were not supposed to.
Yeah, like that's not your training. So it's not surprising. But like, it's the stuff that I think,
fairly common in experimental physical chemistry to build our instrument and also
fix our instruments. Interesting. Yeah. So it's a our conclusions would be it's I mean,
I don't know. Money. I mean, research is expensive. Yeah. For a good reason. Yeah.
33:03
For a good reason. Yeah. I want to close this episode, though, by asking you this question.
What's the first song that comes to your mind when money is mentioned?
Song? Yeah. I have no idea. No, it doesn't. It doesn't really connect to a particular music.
Okay, because when I heard this topic, that we're going to talk about money.
What I thought about was like Pink Floyd's money song. Yeah. I'll send you the the
the episode, the link to YouTube or something. But he basically like the intro to the song
is a cashier sound. And then he sings about money and how we need money.
So I wanted to ask, I thought you might say money, money, money,
money. Yeah, that's, that's one thing.
That's it. Okay, thanks for listening. That's it for the show today. Thanks for listening and find
us at Eigo de Science on Twitter. That is E-I-G-O-D-E S-C-I-E-N-C-E. See you next time.
