#It made for a lot of not particularly happy postdocs.

[averbaldumpingground]

“How come I always end up calling you when I can’t fall asleep?”

Maybe he wouldn't pick up this time, she tries to tell herself. Maybe it would just go to voicemail. And then they could pretend that she never called.

Except he wouldn't, probably. Pretend that is. But she'd hold on to her dignity, maybe enough that the next time she saw him, saw that horrible smile on his face, she could meet his eyes for long enough to get through the lecture. Get through the three hours of supervising the undergraduate lab section. Get through the door.

She doesn't want him to pick up. Not when it's two a.m. Again. Not when she's still half drunk, and she can't see the stars beyond the storm clouds on the roof that Darcy left her on somewhere between her second and fourth wine cooler.

She didn't even want to go with Darcy anyway. She's not sure why she did, but the music that she can still somewhat hear from across the quad is starting to give her a headache.

Or maybe it's the long, drawn out seconds between the first ring and the third, or now the fourth. She could just hang up. She could.

But she's not going to. Not until she gets his voicemail.

Because...

Because it's been a crappy night. Because her piece of junk computer crashed while she was mid-compiling.

Because doing her postdoc when she's barely older than most of the seniors has never endeared her to anyone, her weirdly outgoing roommate excepted.

The call connects.

"It's 2:37, Foster."

So much for the small mercies of his voicemail.

The comically put-upon sigh helps dampen the nauseous feeling in her stomach though. She's mostly pretty sure she shouldn't call him.

"The weather forecast lied."

A click. What sounds like blankets shifting.

Oh. He's never been in bed before.

"I could be on a date," he'd said, the last time that she'd called him.

"At four in the morning?"

"It could be a very good date."

She'd hung up on his obnoxious laughter.

But maybe--

Maybe--

"Foster, it's a Tuesday. Why are you drunk on a Tuesday?"

Is it? She probably knows that, but the frustration and the article edits and the lack of sleep are finally starting to catch up with her. The crappy alcohol still working its way through her system is probably not helping things either.

"Were you asleep?"

More blankets shuffling. An over-dramatic sigh.

"Some of us are covering a 9 a.m. lecture this morning, Foster." A beat. "Do you need me to get you?"

He's probably not on a date then. Not that she cares. Not that--

"--Foster? Jane?"

Oh. Had he been saying something? "Huh? What?"

"You're usually a lot more sober when you call me in the middle of the night."

"I don't call you in the middle of the night."

A laugh. The slide of a drawer being opened.

"Of course not, Foster. Where are you?"

She doesn't though. Not really. He's just an assistant professor whose lab sections she's usually stuck babysitting. And maybe she stops by his lectures sometimes. But only because the theoretical framework--

"--Jane?"

"Why are you calling me that?"

"Because it's your name, Foster. You do still remember your name?"

He sounds less asleep now, less soft. She's not sure that she likes it very much.

"Now, Jane, are you going to tell me where you are?"

Why would he--

"The roof," she tells him, almost despite herself.

"The roof. Alright. Which roof?"

She didn't bother checking when she and Darcy climbed up here. One of the residence halls. She'll figure out which one when she sobers up enough to climb down though.

"Jane?"

“How come I always end up calling you when I can’t fall asleep?”

She isn't sure she really meant to say that.

A pause.

"I do hope you aren't trying to sleep on the roof, Jane."

"You're not very good at conversation and you're pretty much always a jerk."

"...I see." She thinks there's something off about the way he says it. She doesn't think she likes the way it sounds. "Should I call campus security instead?"

"No."

"Will you at least tell me where you are, Foster? I've got 300 freshma--"

"--I like that you pick up."

She really doesn't mean to say that either.

"I--Okay?"

"Um. Yeah."

Maybe the wine coolers were an even worse idea than she'd thought. She--

"Okay."

At least he doesn't sound upset anymore. It shouldn't-- the feeling in her gut is probably from the questionable guacamole she'd had. Or those tasteless little cocktail sausage things.

"Do you want me to come get you, Jane?"

It's 3:14 a.m. He'd said he has a class.

"Why would you?"

She hears his car keys. She knows he lives a decent drive off campus.

"The same reason you called."

Edit: I wrote a follow up.

[random2908]

Do you think academia's terribleness is universal? Like, do you think it would genuinely be a waste of most people's time? I just graduated college, and academia as a career is one path I'm considering, so I'd actually be really interested to hear your thoughts

I don’t think I can give a universal answer. But I can say that for most people, if a career is what you want out of it, academia is a scam.

It’s a scam in countries using the French/American hierarchy model because there are not enough assistant-professorships to go around. Not by a long shot. So the chances of any given newly-minted PhD getting one is vanishingly small. 500-2000 people applying for 20-30 positions worldwide small. If you get the job? Well, then there’s the tenure-track period where they grind you into dust for 4-7 years. But if you survive and (professionally) thrive through that and don’t step on the wrong toes, then you’ve made it for life, and that’s what everyone wants, isn’t it? It’s such an amazing goal. Being a scholar, getting to add to humanity’s body of knowledge, in a guaranteed job with guaranteed intellectual freedom for the rest of your working life. Being a teacher, too, if you want. Of course, even then, funding being what it is in most countries other than Germany and China, you won’t have as much freedom as you like, and you could still languish unfunded, and in fact--especially if you’re a scientist--you’ll be spending most of your time chasing funds rather than engaging in intellectual pursuits. But at least you’ll get a title and a paycheck.

And it’s a scam in countries using the German hierarchy model because, although there are assistant-professorships available, they are--even after a PhD and a postdoc and it some cases a habilitation involving a second dissertation--ultimately subordinate positions, secondary long-term postdocs where you answer to the one full professor who really runs things and makes all the creative and high-intellectual decisions for you. And you may never be promoted in that model, either--when your boss finally retires, after possibly decades, you and all your co-assistant-professors may be passed over for promotion in favor of hiring out to one of those lucky few who made it in the French/American system and therefore has actual leadership and creative experience, unlike you who’s been languishing in a subordinate job your entire career.

Through all of this, no one will ever tell you you’re not good enough or you’re not going to be one of the lucky few--because they still need you, and because the people you’ll meet will have been among the lucky few themselves and often won’t realize just how lucky they were. They’re going to keep telling you you can make it, up until you get rejected, repeatedly, and then they’re going to tell you to try again. Very few will tell you there are even alternative paths available to you, other than trying again and being rejected again, and languishing without a career or a steady paycheck.

That’s not to say you can’t get a career. If you like teaching better than research--or well enough to give up on research--in America, at least, there are a lot of jobs available (although I get the sense that the American system is a bit weird in that respect, so I don’t actually know if you could find something similar in Europe). There are the private “teaching” colleges, and some of the most prestigious even allow/expect a small trickle of research from their faculty (I attended undergrad at one of those), although in some years even jobs at those are extremely competitive beyond all possibility of any given person actually being hired anywhere. And in America there are the junior colleges as well--”community” colleges if they’re public; that’s what my parents opted for, eventually, after (especially in my mother’s case) exhausting other options, and I will say my parents are happy there. I think other countries have those, or something similar--those are the schools where we have our technical/vocational training, but they’re also two cheap years for students who can’t afford four years at a university, and a second chance for students who didn’t do well in high school or who left for the work force after high school and want to return when they’re older. If you’re willing to teach that set of students, that’s a path that’s nearly always open--but I don’t have a good sense of how international it is. So... it’s possible this paragraph with its slight ray of hope is actually useless outside of America, but it’s worth looking into.

This doesn’t mean going to grad school is a worthless endeavor. If you really want to spend your 20s doing research, I’m the last person who would tell you no.

In the humanities and social sciences, it’ll leave you overqualified for most jobs afterward, which is a frustrating position to be in, but there are places--particularly in finance--that will take anyone whose resume says they’re smart regardless of the specifics of their background, so you’ll find something eventually. The PhD won’t have done you any good,though, except what you gained from the process. In the sciences, the situation is a lot less bleak, and in fact approximately half of people getting science PhDs go in without the intention of looking for a career in academia. My current job requires a PhD and I’m even working in my actual subfield. At a lot of companies you don’t need a PhD to be hired--for example, at the aerospace company where I interned in undergrad, or at a pharmaceutical company--but it’ll help you get promoted faster, or there may be higher promotions that do require it. I have a friend whose dad spent a lot of years being serially hired by a startup to be the PhD chemist whose degree lent them legitimacy when they sought investment funding; whether that’s a happy or a terrible life depends on whether you’re in on it or whether you’re trying to earnestly do science. And the likes of Google/Amazon on the one hand and McKinsey on the other will take anyone smart and assume they can be trained in whatever the job requires of them; many such giant companies are global.

In the end, that’s the closest to universal advice I can give. If you want to go to grad school because you want to have done the process of getting a graduate degree--the life of study, the research experience, the (probably tiny) contribution to world knowledge--if that’s really how you want to spend your 20s, then go for it. If, afterwards, you want to put your name into the academic-career lottery--and putting your name in is a lot of work, involving tons of jobs applications and even some grant proposals--if you want to do that, despite what I said above about the jobs not even necessarily being that great, then go for it. But understand it’s a lottery, and if you lose let it go; don’t let yourself be drawn into indefinite postdocs or adjunct positions. Have a backup plan and know that you’ll probably have to use it.

If you can go into academia with all those things in mind, with your eyes open to them, with your self-esteem intact through it all, and with robust backup plans, then go for it. I, for one, don’t regret getting my PhD.

[sciencespies]

Creating new opportunities from nanoscale materials

https://sciencespies.com/biology/creating-new-opportunities-from-nanoscale-materials/

Creating new opportunities from nanoscale materials

MIT Professor Frances Ross has designed several custom sample holders for examining nanoscale materials in gases and liquid media in the electron microscope. For liquid environments, thin windows of silicon nitride surround the liquid but allow the electron beam to pass through. For gas environments, the sample holder (shown here) must heat and tilt the sample without compromising its cleanliness. Credit: Denis Paiste/Materials Research Laboratory

More

A hundred years ago, “2d” meant a two-penny, or 1-inch, nail. Today, “2-D” encompasses a broad range of atomically thin flat materials, many with exotic properties not found in the bulk equivalents of the same materials, with graphene—the single-atom-thick form of carbon—perhaps the most prominent. While many researchers at MIT and elsewhere are exploring two-dimensional materials and their special properties, Frances M. Ross, the Ellen Swallow Richards Professor in Materials Science and Engineering, is interested in what happens when these 2-D materials and ordinary 3-D materials come together.

“We’re interested in the interface between a 2-D material and a 3-D material because every 2-D material that you want to use in an application, such as an electronic device, still has to talk to the outside world, which is three-dimensional,” Ross says.

“We’re at an interesting time because there are immense developments in instrumentation for electron microscopy, and there is great interest in materials with very precisely controlled structures and properties, and these two things cross in a fascinating way,” says Ross.

“The opportunities are very exciting,” Ross says. “We’re going to be really improving the characterization capabilities here at MIT.” Ross specializes in examining how nanoscale materials grow and react in both gases and liquid media, by recording movies using electron microscopy. Microscopy of reactions in liquids is particularly useful for understanding the mechanisms of electrochemical reactions that govern the performance of catalysts, batteries, fuel cells, and other important technologies. “In the case of liquid phase microscopy, you can also look at corrosion where things dissolve away, while in gases you can look at how individual crystals grow or how materials react with, say, oxygen,” she says.

Ross joined the Department of Materials Science and Engineering (DMSE) faculty last year, moving from the nanoscale materials analysis department at the IBM Thomas J. Watson Research Center. “I learned a tremendous amount from my IBM colleagues and hope to extend our research in material design and growth in new directions,” she says.

When gold is deposited on “dirty” graphene (left), blobs of gold collect around impurities. But when gold grows on graphene that has been heated and cleansed of impurities (right), it forms perfect triangles of gold. Credit: Kate Reidy/MIT

More

Recording movies

During a recent visit to her lab, Ross explained an experimental setup donated to MIT by IBM. An ultra-high vacuum evaporator system arrived first, to be attached later directly onto a specially designed transmission electron microscope. “This gives powerful possibilities,” Ross explains. “We can put a sample in the vacuum, clean it, do all sorts of things to it such as heating and adding other materials, then transfer it under vacuum into the microscope, where we can do more experiments while we record images. So we can, for example, deposit silicon or germanium, or evaporate metals, while the sample is in the microscope and the electron beam is shining through it, and we are recording a movie of the process.”

While waiting this spring for the transmission electron microscope to be set up, members of Ross’ seven-member research group, including materials science and engineering postdoc Shu Fen Tan and graduate student Kate Reidy, made and studied a variety of self-assembled structures. The evaporator system was housed temporarily on the fifth-level prototyping space of MIT.nano while Ross’s lab was being readied in Building 13. “MIT.nano had the resources and space; we were happy to be able to help,” says Anna Osherov, MIT.nano assistant director of user services.

“All of us are interested in this grand challenge of materials science, which is: “How do you make a material with the properties you want and, in particular, how do you use nanoscale dimensions to tweak the properties, and create new properties, that you can’t get from bulk materials?”” Ross says.

Using the ultra-high vacuum system, graduate student Kate Reidy formed structures of gold and niobium on several 2-D materials. “Gold loves to grow into little triangles,” Ross notes. “We’ve been talking to people in physics and materials science about which combinations of materials are the most important to them in terms of controlling the structures and the interfaces between the components in order to give some improvement in the properties of the material,” she notes.

Shu Fen Tan synthesized nickel-platinum nanoparticles and examined them using another technique, liquid cell electron microscopy. She could arrange for only the nickel to dissolve, leaving behind spiky skeletons of platinum. “Inside the liquid cell, we are able to see this whole process at high spatial and temporal resolutions,” Tan says. She explains that platinum is a noble metal and less reactive than nickel, so under the right conditions the nickel participates in an electrochemical dissolution reaction and the platinum is left behind.

Professor Frances Ross (left), graduate student Kate Reidy (center), and postdoc Shu Fen Tan work together at the high vacuum evaporator chamber that is part of an electron microscopy suite donated to MIT by IBM. Credit: Denis Paiste/Materials Research Laboratory

More

Platinum is a well-known catalyst in organic chemistry and fuel cell materials, Tan notes, but it is also expensive, so finding combinations with less-expensive materials such as nickel is desirable.

“This is an example of the range of materials reactions you can image in the electron microscope using the liquid cell technique,” Ross says. “You can grow materials; you can etch them away; you can look at, for example, bubble formation and fluid motion.”

A particularly important application of this technique is to study cycling of battery materials. “Obviously, I can’t put an AA battery in here, but you could set up the important materials inside this very small liquid cell and then you can cycle it back and forth and ask, if I charge and discharge it 10 times, what happens? It does not work just as well as before—how does it fail?” Ross asks. “Some kind of failure analysis and all the intermediate stages of charging and discharging can be observed in the liquid cell.”

“Microscopy experiments where you see every step of a reaction give you a much better chance of understanding what’s going on,” Ross says.

Moiré patterns

Graduate student Reidy is interested in how to control the growth of gold on 2-D materials such as graphene, tungsten diselenide, and molybdenum disulfide. When she deposited gold on “dirty” graphene, blobs of gold collected around the impurities. But when Reidy grew gold on graphene that had been heated and cleaned of impurities, she found perfect triangles of gold. Depositing gold on both the top and bottom sides of clean graphene, Reidy saw in the microscope features known as moiré patterns, which are caused when the overlapping crystal structures are out of alignment.

Niobium deposited on top of graphene produces structures that look like the frost patterns formed on the inside of windows in winter, or the feathery patterns of some ferns. They are called dendritic structures. Credit: Kate Reidy/MIT

More

The gold triangles may be useful as photonic and plasmonic structures. “We think this could be important for a lot of applications, and it is always interesting for us to see what happens,” Reidy says. She is planning to extend her clean growth method to form 3-D metal crystals on stacked 2-D materials with various rotation angles and other mixed-layer structures. Reidy is interested in the properties of graphene and hexagonal boron nitride (hBN), as well as two materials that are semiconducting in their 2-D single-layer form, molybdenum disulfide (MoS2) and tungsten diselenide (WSe2). “One aspect that’s very interesting in the 2-D materials community is the contacts between 2-D materials and 3-D metals,” Reidy says. “If they want to make a semiconducting device or a device with graphene, the contact could be ohmic for the graphene case or a Schottky contact for the semiconducting case, and the interface between these materials is really, really important.”

“You can also imagine devices using the graphene just as a spacer layer between two other materials,” Ross adds.

For device makers, Reidy says it is sometimes important to have a 3-D material grow with its atomic arrangement aligned perfectly with the atomic arrangement in the 2-D layer beneath. This is called epitaxial growth. Describing an image of gold grown together with silver on graphene, Reidy explains, “We found that silver doesn’t grow epitaxially, it doesn’t make those perfect single crystals on graphene that we wanted to make, but by first depositing the gold and then depositing silver around it, we can almost force silver to go into an epitaxial shape because it wants to conform to what its gold neighbors are doing.”

Electron microscope images can also show imperfections in a crystal such as rippling or bending, Reidy notes. “One of the great things about electron microscopy is that it is very sensitive to changes in the arrangement of the atoms,” Ross says. “You could have a perfect crystal and it would all look the same shade of gray, but if you have a local change in the structure, even a subtle change, electron microscopy can pick it up. Even if the change is just within the top few layers of atoms without affecting the rest of the material beneath, the image will show distinctive features that allow us to work out what’s going on.”

Reidy also is exploring the possibilities of combining niobium—a metal that is superconducting at low temperatures—with a 2-D topological insulator, bismuth telluride. Topological insulators have fascinating properties whose discovery resulted in the Nobel Prize in Physics in 2016. “If you deposit niobium on top of bismuth telluride, with a very good interface, you can make superconducting junctions. We’ve been looking into niobium deposition, and rather than triangles we see structures that are more dendritic looking,” Reidy says. Dendritic structures look like the frost patterns formed on the inside of windows in winter, or the feathery patterns of some ferns. Changing the temperature and other conditions during the deposition of niobium can change the patterns that the material takes.

All the researchers are eager for new electron microscopes to arrive at MIT.nano to give further insights into the behavior of these materials. “Many things will happen within the next year, things are ramping up already, and I have great people to work with. One new microscope is being installed now in MIT.nano and another will arrive next year. The whole community will see the benefits of improved microscopy characterization capabilities here,” Ross says.

An electron diffraction image of niobium deposited on top of graphene shows that certain crystal planes of niobium align with the crystal planes of the graphene, which is known as epitaxial growth. When a 3-D material is grown on top of a 2-D layer, this perfectly aligned atomic arrangement is often important for device makers. Credit: Kate Reidy/MIT

More

MIT.nano’s Osherov notes that two cryogenic transmission electron microscopes (cryo-TEM) are installed and running. “Our goal is to establish a unique microscopy-centered community. We encourage and hope to facilitate a cross-pollination between the cryo-EM researchers, primarily focused on biological applications and ‘soft’ material, as well as other research communities across campus,” she says. The latest addition of a scanning transmission electron microscope with enhanced analytical capabilities (ultrahigh energy resolution monochromator, 4-D STEM detector, Super-X EDS detector, tomography, and several in situ holders) brought in by John Chipman Associate Professor of Materials Science and Engineering James M. LeBeau, once installed, will substantially enhance the microscopy capabilities of the MIT campus. “We consider Professor Ross to be an immense resource for advising us in how to shape the in situ approach to measurements using the advanced instrumentation that will be shared and available to all the researchers within the MIT community and beyond,” Osherov says.

Little drinking straws

“Sometimes you know more or less what you are going to see during a growth experiment, but very often there’s something that you don’t expect,” Ross says. She shows an example of zinc oxide nanowires that were grown using a germanium catalyst. Some of the long crystals have a hole through their centers, creating structures which are like little drinking straws, circular outside but with a hexagonally shaped interior. “This is a single crystal of zinc oxide, and the interesting question for us is why do the experimental conditions create these facets inside, while the outside is smooth?” Ross asks. “Metal oxide nanostructures have so many different applications, and each new structure can show different properties. In particular, by going to the nanoscale you get access to a diverse set of properties.”

“Ultimately, we’d like to develop techniques for growing well-defined structures out of metal oxides, especially if we can control the composition at each location on the structure,” Ross says. A key to this approach is self-assembly, where the material builds itself into the structure you want without having to individually tweak each component. “Self-assembly works very well for certain materials but the problem is that there’s always some uncertainty, some randomness or fluctuations. There’s poor control over the exact structures that you get. So the idea is to try to understand self-assembly well enough to be able to control it and get the properties that you want,” Ross says.

“We have to understand how the atoms end up where they are, then use that self-assembly ability of atoms to make a structure we want. The way to understand how things self-assemble is to watch them do it, and that requires movies with high spatial resolution and good time resolution,” Ross explains. Electron microscopy can be used to acquire structural and compositional information and can even measure strain fields or electric and magnetic fields. “Imagine recording all of these things, but in a movie where you are also controlling how materials grow within the microscope. Once you have made a movie of something happening, you analyze all the steps of the growth process and use that to understand which physical principles were the key ones that determined how the structure nucleated and evolved and ended up the way it does.”

Clean deposition of gold nanoislands on molybdenum disulfide MoS2 with visible moiré patterns. Credit: Kate Reidy/MIT

More

Future directions

Ross hopes to bring in a unique high-resolution, high vacuum TEM with capabilities to image materials growth and other dynamic processes. She intends to develop new capabilities for both water-based and gas-based environments. This custom microscope is still in the planning stages but will be situated in one of the rooms in the Imaging Suite in MIT.nano.

“Professor Ross is a pioneer in this field,” Osherov says. “The majority of TEM studies to-date have been static, rather than dynamic. With static measurements you are observing a sample at one particular snapshot in time, so you don’t gain any information about how it was formed. Using dynamic measurements, you can look at the atoms hopping from state to state until they find the final position. The ability to observe self-assembling processes and growth in real time provides valuable mechanistic insights. We’re looking forward to bringing these advanced capabilities to MIT.nano.” she says.

“Once a certain technique is disseminated to the public, it brings attention,” Osherov says. “When results are published, researchers expand their vision of experimental design based on available state-of-the-art capabilities, leading to many new experiments that will be focused on dynamic applications.”

Rooms in MIT.nano feature the quietest space on the MIT campus, designed to reduce vibrations and electromagnetic interference to as low a level as possible. “There is space available for Professor Ross to continue her research and to develop it further,” Osherov says. “The ability of in situ monitoring the formation of matter and interfaces will find applications in multiple fields across campus, and lead to a further push of the conventional electron microscopy limits.”

Explore further

Frances Ross discusses witnessing nanostructure formation

Provided by Massachusetts Institute of Technology

This story is republished courtesy of MIT News (web.mit.edu/newsoffice/), a popular site that covers news about MIT research, innovation and teaching.

Citation: Creating new opportunities from nanoscale materials (2019, September 6) retrieved 6 September 2019 from https://phys.org/news/2019-09-opportunities-nanoscale-materials.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

#Biology

[houseofmind]

2 Year Anniversary PD Update

A lot has happened since I last brought you up to speed with my life and academic endeavors. Given that my last update post was over 2 years ago, I wasn’t sure where to start. With that said, this post has been opened and worked on (and opened and worked on, and so forth) so it’s a long one :X If you’re new, welcome! Also, please be sure to check out the list of things I learned during my PhD. 

Spring/Summer 2016

Briefly, I moved to Pittsburgh about a week after that update post. I crammed as much of my stuff as I could along with Fausto (my cat) into a rental car and drove to my new place amidst a snowstorm. I made it there safely (although I had to make an overnight stop that I swear gave my cat anxiety). My new home exceeded my expectations (2 bedroom, living room, kitchen, basement, washer/dryer, short commute < $1,000), but leaving NYC and most of my friends was bittersweet. 

I went into lab the next day after arriving in Pittsburgh. THAT felt really weird because it was when reality hit. New job, new role, new boss (and new style), new techniques, new approaches, new life basically. My first challenge in the lab was learning the “bread and butter” technique: in vivo extracellular recordings of VTA dopamine neurons. Initially, this scared the shit out of me because I had 0 electrophysiology experience so I was worried about how long it would take for me to be decent and start collecting good data. I was trained by the RAP in the lab; I shadowed her for a week and copied down as much as I could. Then, I did an animal every day for a month. That’s about how long it took for me to get comfortable. (Side note: if you move to a lab and need to learn a new technique, my advice is to attack it aggressively until you learn it!). At the beginning I encountered problems such as having my animal bleed too much, giving too much anesthetic, taking too long (a whole day for just one surgery), making bad electrodes, not being able to find dopamine neurons, etc. The really cool part was seeing myself get better: not making the same mistakes, being able to correct mistakes when they happened, becoming faster and more efficient, optimizing my procedure. It was a great reminder of why I love doing science and that I am capable of learning new things :]

Then came the part where my advisor asked me to think of projects related to his prior work and funding. Call this my second challenge in the lab. He suggested a “safe” project and a “pie in the sky” project. I really appreciated his flexibility and willingness to have me choose projects and questions that I cared about, while also offering his guidance and steering me in better directions. After I pored over his grants and (then still in press papers), I came up with 2 ideas and presented my experimental design to the lab. Next, I got to work! During this time I received funding from a T32 program within the Psychiatry at Pitt, which would cover my stipend until I was able to secure my own funding (more on this in Fall 2015). 

Work was interrupted for a couple of days in May that I took off to attend my graduation in NYC. My parents and one of my brothers were able to fly in from Puerto Rico and I was really pleased. I am the only (and first!) PhD in either side of the family so I could tell that this was a big deal to them and that they were really proud of me. Obviously, this made me feel proud of myself :D

Here is a pic of me on graduation day:

After I came back, it was time to get down to business. My first cohort of “real animals” was ready for all the single unit dopamine (DA) neuron recordings I had to do. Even though I had only been doing this technique for ~2 months, I managed to get some interesting preliminary data that my PI was enthusiastic about, he suggested I write an F32 (NIH postdoctoral fellowship grant). For the August deadline, meep! Most of June and July was spent working around the clock to not only gather and analyze the data, but also write the proposal. IT WAS INSANE. During this time, I also traveled to San Sebastián in the Basque Country (within Spain) for the International Society for Developmental Psychobiology (ISDP) meeting. It was my first time out of the country since 2013 and it was quite exciting because I was awarded their dissertation prize! I was only there for 4-5 days but I managed to hang out with my PhD advisor and conference buddies, gave a talk on my dissertation research, meet some pretty cool people from all over,  take in some breathtaking views, eat pintxos, and go to the beach. I guess you could call this my “vacation” over the summer. Finally, the last exciting thing to happen over the summer was that I received an invitation from the American College of Neuropsychopharmacology (ACNP) to attend their annual meeting in December (more on this below). 

Fall/Winter 2015

After submitting my grant, I continued to perform experiments related to my F32 proposal project. I also started generating animals for another project in which I would look at the effects of chronic mild stress (i.e. depression model) in a neurodevelopmental model of schizophrenia-like deficits. I also got the grant score for my initial submission, which I knew was not within funding range so that meant resubmission in the spring. What can you do? During this semester I participated in the weekly career and research development (CARD) seminar, which is oriented towards helping postdocs gain the skills necessary for transitioning into an independent research position. This was required for me to get the T32 $$$, but I would strongly advise anyone who has access to these types of programs to attend! Topics covered included but were not limited to: how to negotiate, conflict management, how to prepare a job talk, authorship, mock K review process, etc... There is also a writing block portion of the seminar in which participants are divided into topic specific mini groups and get feedback from at least 2 other faculty on a weekly basis before grant submission. Besides being able to cultivate a positive and professional relationship with my T32 director, I’d say that the CARD seminar was the best thing about participating in this T32 program. 

In terms of travel and meetings, I attended the SfN 2015 conference in Chicago and I’ve got to say that my experience there was so positive it even made me change my mind about Chicago!!! It was the first time I was attending this meeting without having a poster so I focused on trying to meet the dopamine/electrophysiology folks and attending all the socials I could fit in :D Right after SfN, I went to NYC to visit some friends still in grad school over Thanksgiving. I cannot stress how important it is to keep in touch with your friends and make time for them. Anyhow, I got back and then it was time to fly down to Florida for the ACNP meeting. This was my first official presentation as a PD so I was really nervous (also presenting a new technique, new project, new everything!) but my boss was really supportive and the people that came by my poster were kind, insightful and encouraging. The location is also superb and the science top notch, I think this has dethroned Gordon Amygdala as my favorite meeting (most of those people were here anyway!) and I will definitely keep applying for that travel award for as long as I have to. After ACNP I flew home for Christmas and got an invitation from my advisor to co-author a review paper w/ another PD in the lab. This paper has been published already and can be found here. 

Spring 2016

I basically spent the spring working on additional experiments for my main project that I thought were necessary after presenting at ACNP and getting feedback (i.e. extra controls, increasing n, diff stats, etc...). This was a good thing because I also used some of that data for my F32 resubmission, which was due in April. As I wrote that sentence I remembered that this time was particularly busy for me in the lab as the cohort of animals for my other project (MAM-CMS) was born and I was running 2 behaviors + recording all of these animals (8 group total). It was so crazy busy (yet productive) that I swore to myself never to schedule that many experiments for the same couple of months ever again. 

Summer 2016

I presented a new and improved version of my main project poster at the Society for Biological Psychiatry in Atlanta. This was very cool because I had never been there (omg the aquarium!) and I got to reconnect with a lot of people I had met during my PhD but had not seen since the start of my PD. Apparently they really like this meeting and it is one of their go-tos. One of these people was my PhD advisor, whom I worked with on trying to wrap up and expedite my remaining PhD papers. Another person was another female minority graduate student that I knew during my PhD at NYU. It was great catching up with her and hearing about all the great science she will be doing in her PD lab and how the transition was for her. I remembered thinking that there are many paths that can lead to the same outcome, and feeling at peace. 2 interpersonal crises later, I received the score for my F32 resubmission and it was so good I A) almost fell off my chair and B) started crying (like, really). If you’re anything like me, you probably already know that you cannot have it all. It always feels like either I’m great in my personal life and so so at work or I’m killing it at work and everything else has gone to shit. To me, the score meant funding was pretty much a given, so it tipped the scales back to the killing it at work scenario and I settled for being happy with that and harnessing what was left of my happiness into that new positive direction. You can read about my F32 project here. I spent the rest of the summer celebrating my birthday, good fortune and prepping my first PD paper for submission. Also, I GOT A CAR! For the first time in nearly 7 years since I did my PhD in Manhattan and lol cars in NYC. 

Fall 2016

Remember that slump I mentioned I was in at the beginning of the summer? I decided to get over it by taking a solo trip to Thailand :] That is a post in itself (which will likely not end up here) but it was truly a once in a lifetime trip. I liked that it was hard (+16 hours to get there), new (first time ever in Asia) and was something totally out of my comfort zone. I met people from all over the US, rode elephants, visited UNESCO world heritage sites, learned about Buddhism, bar-hopped in Khao San Road and the red light district, shipped in a professional masseuse from the temples for a full body massage, ate weird food, etc. etc. Some of these pics can be found my scrolling down in my IG.  I feel like everyone should do something like that for themselves, at least every once in a while. I also learned that culture shock is funny because it can happen once you get back from your trip. I had a soft meltdown upon arriving at JFK and spent my first night back talking to a friend/processing everything I had witnessed. Can’t wait to go back and this time make it to Krabi or Phuket! 

During this semester, I also submitted my initial paper for my main project and added experiments to my secondary project. In more exciting news, I published a first author PhD paper on the effects of early life stress on social behavior and neural activity in cortical and limbic areas. You can read it here. I also got a travel award to present this work at the Society for Social Neuroscience in San Diego jus prior to SfN 2016, where I presented my first SfN poster as a PD. I was overwhelmed by the positive response and I got to do a podcast w/ Deb Budding from Neurocurious. Even though I was dead tired and delirious from my poster session being immediately before, it was SO MUCH FUN <3 I just checked and its still not up but will link it here when its available! SfN 2016 was followed by ACNP 2016 (first time presenting my MAM/CMS data, now DEFINITELY my favorite meeting), although I flew down to Florida early so I could make it to Art Basel (basically a conference, but for art!). I had such a great time that I almost forgot that my paper was rejected :( When I got back to Pitt, I focused on collecting data for the last experiments in yet another (small) project (that I have not mentioned previously yet). Then I went off to Puerto Rico to spend NYE w/ my family and friends :D

Winter 2016-2017 (aka where am I now)

I published another PhD paper. This one is a co-first paper and I’m really proud of it because it represents an effort by our group to incorporate techniques used in humans to facilitate translation of findings between animal models and humans. 

I revised and resubmitted my previously rejected paper. System says it’s still under review so we’ll see what happens... 

I submitted a grant based on an idea I had last fall that ties in nicely with my current work (pitched the idea to advisor at SfN 2016, luckily he agrees it is interesting/worth the shot).

Collected prelim data for above-mentioned grant.

Organizing the outline for a review I want to write that’s related to my grant. A colleague/mentor gave me the idea to do this since it’s a good chance to get a publication of all that lit search you did anyway! 

Put in 2 travel award applications (1 no, 1 still waiting) to present my work 

Nearly done collecting data for that random project 3 that slipped my mind for most of this post. I think this is because it’s not my main project and its a follow-up of someone else’s project. Presented this data at lab meeting and got some feedback about graphing/analyzing data. Working on this and aiming to submit by early summer. 

Traveling to Boston in 2 weeks to present in the Poverty and the Brain symposium at the Eastern Psychological Association meeting

Accepted a position as an ad hoc member of ACNP’s minority task force in order to help increase the participation of underrepresented minorities (URMs) in the college and to help retain them

Accepted a position as Review Editor in the Editorial Board of Frontiers in Behavioral Neuroscience 

WHEW. I finally got it all out. There you have it, what I’ve been up to since starting my postdoc and why I haven’t been more active on Twitter/here. Hopefully it goes someways in attenuating my I-should-be-posting-more-guilt :P 

Also, please feel free to write if you have any questions about PD life, the grad to PD transition or anything else you have read about here : ) In case you didn’t know, I started this blog as a grad student who DID NOT think they were going to make it in academia so the fact that this blog is still alive and so many of you still follow it/engage motivates me to keep it going for as long as I can. Maybe one day I’ll have a job and can look back at everything that led to that there. That’s the dream. Until next time! 

xoxo

Dr. M

[douchebagbrainwaves]

WHY TO DO THINGS

This is so foreign to most people's experience that they don't have to give up on the losing side of a car are to speed. But if it does, even where that motive is not something you can leave running as a background process running, looking for something to spark a thought. Much was changed, but there is little correlation between founders' initial attitudes and how well they do are not orthogonal. Programming languages are for. Growing too slowly is particularly dangerous in a business with network effects, which the best startups it produced would be sucked away to existing startup hubs. When it reaches a certain concentration, it kills off the yeast that produced it. At least, that's what you'll naturally tend to build things that get discovered this way incidentalomas, and they then send me a stream of reasonably high quality ones. You won't have to work quite closely with them.

Hmm, I wonder. In a startup, people treat you as if you're unemployed. They delight in breaking rules, but not like it when voters or other countries refuse to bend to their will, but ultimately it's in all our interest that there's not a single point where you get a product visionary as CEO is for him to go to the real Silicon Valley, the single best predictor of how a startup will change you a lot. Competitors punch you in the long run prove to be an increasing number of early failures, the startup never happens. Because it's too easy for clients to fire them later. Among other things, this shift has created the appearance of ease seems to come over people when they start to talk about that for counterintuitive? This keeps out most fluff, but not with the giant leaps you see in the coming years. The program is canceled. It's much like being a postdoc: you have to jump through in school. Our secret weapon was similar. Similarly, in painting, or in a novel?

That kind of title is the same as with money, avoiding pleasure is no longer much left to copy before the language you've made is Lisp. So if you're running a startup can be the kids' own parents. Increasingly, he has to do something have about eight different countries. Almost everyone's initial plan is broken. He wrote about productivity in lines of code, they see not just that Steve Jobs and Co are industrial design wizards, but because of what they planned to do. When you ask what message a city sends? We'll start with the labels. The reason is that to make Leonardo you need more than that. Vertically integrated companies literally dis-integrated because it was more a flaw their eminence had allowed them to sink into. He couldn't just let the site die. If you want to work on your own projects.

So far the complete list of messages I've picked up from cities is: wealth, style, hipness, physical attractiveness wouldn't have been a total immersion. A round needs to be protected from himself. Just that some kinds of work better sources of habits of mind as well, because they're the solution to a gross problem. If everyone's filters have different probabilities, it will be for the better. Even people sophisticated enough to know about it already, if you want to make their initial users happy. Nerds don't realize this. Now it's just one of the first things Jobs did when they got some money was to be driven by ambition: self-consciously cool person wanted to differentiate himself from preceding fashions e. Perhaps a few will have the energy to try to explain in person to collect a check.

Since you can't derive as much benefit yet from a narrow focus, you may be arguing with a straw man. Could a programming language is a medium of exchange, called the dollar, that doesn't mean such a thing as good and bad investors, one of the readiest to say I don't know till I got to college. It wasn't just Yahoo. When you stretch before running, you put your body into positions much more extreme than they were a race apart. Some angels are, or which kinds of discussions to avoid, but how to procrastinate well. Us. It does whatever you tell it. The two words are pulling in opposite directions. If your startup grows big enough, probably, it was the same at the schools I went to art school run smack into a brick wall. It can be traumatic for the ones that wanted Oracle experience.

Notes

Cit. A company will be the more qualifiers there are already names for this point for me do more harm than good.

I don't know enough about big companies can afford that.

And journalists as part of creating an agreement from scratch is not entirely a coincidence, because the ordering system, written in C and Perl. Kant. The reason for the government. Steven Hauser.

Again, hard work. G.

In practice sufficiently expert doesn't require one to be extra skeptical about Viaweb too. That's the trouble with fleas, they mean statistical distribution. You're investing your own mind.

In technology, so that you were doing Viaweb again, I'd say the rate of change in the Sunday paper. When investors can't make up their minds, they are within any given college.

Thanks to several anonymous CS professors, Jessica Livingston, and Ken Anderson for inviting me to speak.

[josephborrello]

Magnitude and Direction, Issue #42 | 20 Sep 2019

Hardware, Prototyping, and Fabrication

‍♂️ These robots learn how to dig holes by watching humans do it. (So does one dig, while 4 others watch?) ✝ Robot priests are giving new meaning to the term deus ex machina. (I hope you thought that was as witty as I did.) It's amazing how much stuff (and people) you can fit into a fire truck (or ambulance, or rescue helicopter...)

Software and Programming

➕ generated.space is a great collection of very aesthetic, generative artwork. ➡ This AI tries to make emojis more lifelike and is neither particularly good, nor necessary. It is, however, very scary. Clippy is back. Ironically, he's back on Mac, not Windows. (He's also available for websites, like here for example.) The Recomendo newsletter did a better job describing this than I did, so I'll just re-print what they've written:

"I’ve waited all my life for a tool that would create art for me. It’s here. Artbreeder is a website that breeds new visual images from existing images. Using deep learning (AI) algorithms it generates multiple photo-realistic “children” mutations of one image. You — the gardener — select one mutant you like and then breed further generations from its descendants. You can also crossbreed two different images. Very quickly, you can create infinite numbers of highly detailed album covers, logos, game characters, exotic landscapes. I find it fiendishly addictive. Wanna see the zoo of unearthly creatures I found/made? (Note: If Artbreeder is not out of beta use Ganbreeder, it’s predecessor.) — [Kevin Kelly]"

Science, Engineering, and Biomedicine

You've heard of "I Can't Believe It's Not Butter!" but have you heard of "I Can't Believe It's Mostly Water!"? Also on the topic of water, pulling moisture out of thin air in arid, desert climates won't only be the stuff of sci-fi for much longer. The website Relativity, showing how recent observations of a star orbiting near the galactic core helps support Einstein's theory of general relativity, is a great example of the kind of science communication we need to be doing more of.

Mapping, History, and Data Science

There's a (near-infinite) world of colors out there for you to name! (I named one 'Jimmy') ⛓ Do you understand how the blockchain works? If not, maybe this data sonification/visualization will help (it probably won't but it sounds and looks really cool). The orientation of airport runways may not sound very interesting, but I was absorbed by this map almost immediately. (For example, what's with the pretty drastic East/West divide among US runways?) What were the most popular videos on Youtube a decade ago?

Events and Opportunities

We're within spitting distance of Fall and there's lots of exciting autumnal activities coming up:

Sunday, 9/22 The World Health Organization is hosting the first-ever edition of the Walk the Talk: The Health for All Challenge in the United States, being held over 4 miles in New York's beautiful Central Park. This fun run/walk will bring together people from all over the world for a celebration of health, on the eve of the United Nations General Assembly

Monday, 9/23 Derek Brand's monthly ECHO biotech gathering is back and, as always, bringing together some of the NYC area's most exciting life science startups and entrepreneurs.

Monday, 9/23 Coming off a great mid-summer meetup with the Health-Tech Connect group, the Mount Sinai Innovators Group is teaming up with them again for another event featuring some great pre-SINAInnovations topics: Artificial Intelligence and Assistive Devices. Both presenters have extensive experience in the healthcare startup space and this promises to be a can't-miss event!

Tuesday, 9/24 Join NYDesigns for a tour of their 5,000 square foot fabrication facility and learn about how you can make use of all the impressive equipment there at their upcoming open house.

Tuesday, 9/24 Join GeoNYC and Doctors Without Borders for a special map-a-thon to fill in missing geospatial data for underserved regions in order to provide international and local NGOs and individuals with the data they need to better respond to crises.

Wednesday, 9/25 The NYC JLABS crew is back for their next Innovators and Entrepreneurs mixer, which promises, as always, to be a great event to meet local life science startups and biotech enthusiasts.

Wednesday, 9/25 The RobotLab meetup's September event focuses on the good, the bad, and the ugly of Industry 4.0 and autonomous manufacturing.

Thursday, 9/26 Join GRO-Biotech, MSIG, and the Petri biotech accelerator at Mount Sinai for an info session on opportunities to get involved in their new startup program focused on helping formation-stage innovators realize the next frontier of biology and engineering

Thursday, 9/26 It's been touched on in previous Existential Medicine events, but the next science seminar collab between New Lab and JLABS dives deep into the revolutionary, and sometimes controversial technology of CRISPR. Use code "NewLab2019" to unlock the event registration.

Saturday, 9/28 Admission is just the swipe of a metro card for the Parade of Trains at the Brighton Beach station. Vintage train cars from all periods of the subway's history will be on display, as well as taking passengers on short trips around south Brooklyn.

Monday, 9/30 The NY Hardware Startup meetup is back with their regularly scheduled programming, which, in this case, means presentations from the likes of Lime, Light Phone, and more, down at General Assembly in the Flatiron.

Tuesday, 10/1 The next stop on Ogilvy's healthcare innovation pop-up series takes them to Hudson Yards, where they're teaming up with the HITLAB and SAP.iO Foundry for an event that will focus primarily on the female and underserved health innovators who are disrupting healthcare today.

Tuesday, 10/1 NYDesigns' next Women in Tech happy hour is back at Bierocracy in Long Island City. As always, individuals who identify as female and men are welcome to attend, too!

Wednesday, 10/2 The next edition of Larger Than Life Science at New York BioLabs is back and focusing on a crucial aspect of bringing healthcare innovations to market: preparing for and conducting clinical trials.

Some other upcoming events to keep on your radar...

Monday, 10/7 Join GRO-Biotech and Insight Data Sciences for a talk and Q&A about different careers in data, the most suitable backgrounds for each of them, and how Insight Data Sciences can help you make the transition.

Saturday, 10/12 The next edition of Hot Glass Cold Beer returns to the Brooklyn Glass studios in Gowanus, featuring live glass blowing, open studios, and effectively endless amounts of beer. As always, getting a ticker in advance (versus at the door) means you'll be guaranteed to get one of their hand-made glasses (which you can subsequently drink out of for the rest of the night).

October 11-16 Innovation Week at Mount Sinai. What started as just the SINAInnovations conference is now a week's worth of activities dedicated to bringing New York's biomedical innovation communities together. Here's the full lineup:

Friday-Sunday, 10/11-13 Mount Sinai Health Hackathon. The 4th annual Mount Sinai Health Hackathon will be an exciting 48-hour transdisciplinary competition focused on creating novel technology solutions for problems in healthcare. This year’s theme is Artificial Intelligence – Expanding the Limits of Human Performance.

Tuesday, 10/15 Careers & Connections 2019. October may feel far away, but I promise you it's not and you'll want to be sure to mark your calendars for GRO-Biotech's next big event, the Careers & Connections mini-conference and networking event, held concurrently with emerging healthcare technologies conference, SINAInnovations.

Tuesday & Wednesday, 10/15-16 SINAInnovations Conference. The Icahn School of Medicine at Mount Sinai is hosting its eighth annual SINAInnovations conference around the theme of Artificial Intelligence. A range of talks and panels will focus on the explosive growth of AI in our society and in particular in medicine, featuring international thought leaders across the range of relevant domains.

Wednesday, 10/16 Right after Careers & Connections, GRO-Biotech is hosting a fireside chat at BioLabs with Adam Wollowick from Stryker and Jack Wu from Adlai Nortye on what a career in business development looks like and how you can start a career in bizdev.

Saturday, 10/19 New Lab's annual open house birthday celebration is back, with a theme this year of Light+Motion. As always, you can expect pretty much everyone affiliated with technology, design, science, and/or entrepreneurship to turn up for what's one of the bigger bashes of the year.

Saturday & Sunday 10/19-20 The biggest bi-annual graduate career symposium in the country is back at NYU Med showcasing all the career trajectories you can pursue post-PhD. This is one of the best opportunities for graduate students and postdocs to learn about the breadth of career paths for doctorates and an amazing place to network with the next generation of scientists. More info on the two-day conference can be found here, and the registration link is here.

Saturday, 10/26 The Future of Care conference is back at Rockefeller University featuring some of the latest breakthroughs in clinical care and the innovators helping shepherd them from bench to bedside. Apply to attend the conference by September 6th.

Tuesday, 10/29 Join Columbia Nano Labs for their annual Industry Day conference. Learn how you can use and leverage the Nano Labs facilities, hear from a panel of entrepreneurs who have done just that, and listen to faculty and technical experts discuss the way these sophisticated tools contribute to cutting-edge research. (Yes, this was rescheduled from the originally planed date of 9/5.)

Thursday, 10/31 Pitching your startup in front of investors doesn't have to be spooky. The Mid Atlantic Bio Angels 1st Pitch events offer NYC's biotech entrepreneurs the chance to pitch their innovations in front of a panel of real investors and receive critical feedback on their pitches and business plans. The 1st Pitch events are also a great place to learn about the latest innovations in the NYC biotech ecosystem and connect with some of its major players.

Friday-Sunday, 11/8-10 For 36 hours on November 8-10, HackPrinceton will bring together 600 developers and designers from across the country to create incredible software and hardware projects. They'll have swag, workshops, mentors, prizes, games, free food, and more.

Map of the Month

When we hear about the 2-3 Celsius increase in temperature that's going to set us on path to irreversible environmental changes, it often sounds like it's still a ways off. As this map from the Washington Post shows, that future is already becoming a reality in some parts of the US.

Odds & Ends

Bovine Obstruction. I can't explain this.

[dorcasrempel]

Creating new opportunities from nanoscale materials

A hundred years ago, “2d” meant a two-penny, or 1-inch, nail. Today, “2-D” encompasses a broad range of atomically thin flat materials, many with exotic properties not found in the bulk equivalents of the same materials, with graphene — the single-atom-thick form of carbon — perhaps the most prominent. While many researchers at MIT and elsewhere are exploring two-dimensional materials and their special properties, Frances M. Ross, the Ellen Swallow Richards Professor in Materials Science and Engineering, is interested in what happens when these 2-D materials and ordinary 3-D materials come together.

“We’re interested in the interface between a 2-D material and a 3-D material because every 2-D material that you want to use in an application, such as an electronic device, still has to talk to the outside world, which is three-dimensional,” Ross says.

“We’re at an interesting time because there are immense developments in instrumentation for electron microscopy, and there is great interest in materials with very precisely controlled structures and properties, and these two things cross in a fascinating way,” says Ross. 

“The opportunities are very exciting,” Ross says. “We’re going to be really improving the characterization capabilities here at MIT.” Ross specializes in examining how nanoscale materials grow and react in both gases and liquid media, by recording movies using electron microscopy. Microscopy of reactions in liquids is particularly useful for understanding the mechanisms of electrochemical reactions that govern the performance of catalysts, batteries, fuel cells, and other important technologies. “In the case of liquid phase microscopy, you can also look at corrosion where things dissolve away, while in gases you can look at how individual crystals grow or how materials react with, say, oxygen,” she says.

Ross joined the Department of Materials Science and Engineering (DMSE) faculty last year, moving from the Nanoscale Materials Analysis department at the IBM Thomas J. Watson Research Center. “I learned a tremendous amount from my IBM colleagues and hope to extend our research in material design and growth in new directions,” she says.

Recording movies

During a recent visit to her lab, Ross explained an experimental setup donated to MIT by IBM. An ultra-high vacuum evaporator system arrived first, to be attached later directly onto a specially designed transmission electron microscope. “This gives powerful possibilities,” Ross explains. “We can put a sample in the vacuum, clean it, do all sorts of things to it such as heating and adding other materials, then transfer it under vacuum into the microscope, where we can do more experiments while we record images. So we can, for example, deposit silicon or germanium, or evaporate metals, while the sample is in the microscope and the electron beam is shining through it, and we are recording a movie of the process.”

While waiting this spring for the transmission electron microscope to be set up, members of Ross’ seven-member research group, including materials science and engineering postdoc Shu Fen Tan and graduate student Kate Reidy, made and studied a variety of self-assembled structures. The evaporator system was housed temporarily on the fifth-level prototyping space of MIT.nano while Ross’s lab was being readied in Building 13. “MIT.nano had the resources and space; we were happy to be able to help,” says Anna Osherov, MIT.nano assistant director of user services.

“All of us are interested in this grand challenge of materials science, which is: ‘How do you make a material with the properties you want and, in particular, how do you use nanoscale dimensions to tweak the properties, and create new properties, that you can’t get from bulk materials?’” Ross says.

Using the ultra-high vacuum system, graduate student Kate Reidy formed structures of gold and niobium on several 2-D materials. “Gold loves to grow into little triangles,” Ross notes. “We’ve been talking to people in physics and materials science about which combinations of materials are the most important to them in terms of controlling the structures and the interfaces between the components in order to give some improvement in the properties of the material,” she notes.

Shu Fen Tan synthesized nickel-platinum nanoparticles and examined them using another technique, liquid cell electron microscopy. She could arrange for only the nickel to dissolve, leaving behind spiky skeletons of platinum. “Inside the liquid cell, we are able to see this whole process at high spatial and temporal resolutions,” Tan says. She explains that platinum is a noble metal and less reactive than nickel, so under the right conditions the nickel participates in an electrochemical dissolution reaction and the platinum is left behind.

Platinum is a well-known catalyst in organic chemistry and fuel cell materials, Tan notes, but it is also expensive, so finding combinations with less-expensive materials such as nickel is desirable.

“This is an example of the range of materials reactions you can image in the electron microscope using the liquid cell technique,” Ross says. “You can grow materials; you can etch them away; you can look at, for example, bubble formation and fluid motion.”

A particularly important application of this technique is to study cycling of battery materials. “Obviously, I can’t put an AA battery in here, but you could set up the important materials inside this very small liquid cell and then you can cycle it back and forth and ask, if I charge and discharge it 10 times, what happens? It does not work just as well as before — how does it fail?” Ross asks. “Some kind of failure analysis and all the intermediate stages of charging and discharging can be observed in the liquid cell.”

“Microscopy experiments where you see every step of a reaction give you a much better chance of understanding what’s going on,” Ross says.

Moiré patterns

Graduate student Reidy is interested in how to control the growth of gold on 2-D materials such as graphene, tungsten diselenide, and molybdenum disulfide. When she deposited gold on “dirty” graphene, blobs of gold collected around the impurities. But when Reidy grew gold on graphene that had been heated and cleaned of impurities, she found perfect triangles of gold. Depositing gold on both the top and bottom sides of clean graphene, Reidy saw in the microscope features known as moiré patterns, which are caused when the overlapping crystal structures are out of alignment.

The gold triangles may be useful as photonic and plasmonic structures. “We think this could be important for a lot of applications, and it is always interesting for us to see what happens,” Reidy says. She is planning to extend her clean growth method to form 3-D metal crystals on stacked 2-D materials with various rotation angles and other mixed-layer structures. Reidy is interested in the properties of graphene and hexagonal boron nitride (hBN), as well as two materials that are semiconducting in their 2-D single-layer form, molybdenum disulfide (MoS2) and tungsten diselenide (WSe2). “One aspect that’s very interesting in the 2-D materials community is the contacts between 2-D materials and 3-D metals,” Reidy says. “If they want to make a semiconducting device or a device with graphene, the contact could be ohmic for the graphene case or a Schottky contact for the semiconducting case, and the interface between these materials is really, really important.”

“You can also imagine devices using the graphene just as a spacer layer between two other materials,” Ross adds.

For device makers, Reidy says it is sometimes important to have a 3-D material grow with its atomic arrangement aligned perfectly with the atomic arrangement in the 2-D layer beneath. This is called epitaxial growth. Describing an image of gold grown together with silver on graphene, Reidy explains, “We found that silver doesn’t grow epitaxially, it doesn’t make those perfect single crystals on graphene that we wanted to make, but by first depositing the gold and then depositing silver around it, we can almost force silver to go into an epitaxial shape because it wants to conform to what its gold neighbors are doing.”

Electron microscope images can also show imperfections in a crystal such as rippling or bending, Reidy notes. “One of the great things about electron microscopy is that it is very sensitive to changes in the arrangement of the atoms,” Ross says. “You could have a perfect crystal and it would all look the same shade of gray, but if you have a local change in the structure, even a subtle change, electron microscopy can pick it up. Even if the change is just within the top few layers of atoms without affecting the rest of the material beneath, the image will show distinctive features that allow us to work out what’s going on.”

Reidy also is exploring the possibilities of combining niobium — a metal that is superconducting at low temperatures — with a 2-D topological insulator, bismuth telluride. Topological insulators have fascinating properties whose discovery resulted in the Nobel Prize in Physics in 2016. “If you deposit niobium on top of bismuth telluride, with a very good interface, you can make superconducting junctions. We’ve been looking into niobium deposition, and rather than triangles we see structures that are more dendritic looking,” Reidy says. Dendritic structures look like the frost patterns formed on the inside of windows in winter, or the feathery patterns of some ferns. Changing the temperature and other conditions during the deposition of niobium can change the patterns that the material takes.

All the researchers are eager for new electron microscopes to arrive at MIT.nano to give further insights into the behavior of these materials. “Many things will happen within the next year, things are ramping up already, and I have great people to work with. One new microscope is being installed now in MIT.nano and another will arrive next year. The whole community will see the benefits of improved microscopy characterization capabilities here,” Ross says.

MIT.nano’s Osherov notes that two cryogenic transmission electron microscopes (cryo-TEM) are installed and running. “Our goal is to establish a unique microscopy-centered community. We encourage and hope to facilitate a cross-pollination between the cryo-EM researchers, primarily focused on biological applications and ‘soft’ material, as well as other research communities across campus,” she says. The latest addition of a scanning transmission electron microscope with enhanced analytical capabilities (ultrahigh energy resolution monochromator, 4-D STEM detector, Super-X EDS detector, tomography, and several in situ holders) brought in by John Chipman Associate Professor of Materials Science and Engineering James M. LeBeau, once installed, will substantially enhance the microscopy capabilities of the MIT campus. “We consider Professor Ross to be an immense resource for advising us in how to shape the in situ approach to measurements using the advanced instrumentation that will be shared and available to all the researchers within the MIT community and beyond,” Osherov says.

Little drinking straws

“Sometimes you know more or less what you are going to see during a growth experiment, but very often there’s something that you don’t expect,” Ross says. She shows an example of zinc oxide nanowires that were grown using a germanium catalyst. Some of the long crystals have a hole through their centers, creating structures which are like little drinking straws, circular outside but with a hexagonally shaped interior. “This is a single crystal of zinc oxide, and the interesting question for us is why do the experimental conditions create these facets inside, while the outside is smooth?” Ross asks. “Metal oxide nanostructures have so many different applications, and each new structure can show different properties. In particular, by going to the nanoscale you get access to a diverse set of properties.”

“Ultimately, we’d like to develop techniques for growing well-defined structures out of metal oxides, especially if we can control the composition at each location on the structure,” Ross says. A key to this approach is self-assembly, where the material builds itself into the structure you want without having to individually tweak each component. “Self-assembly works very well for certain materials but the problem is that there’s always some uncertainty, some randomness or fluctuations. There’s poor control over the exact structures that you get. So the idea is to try to understand self-assembly well enough to be able to control it and get the properties that you want,” Ross says.

“We have to understand how the atoms end up where they are, then use that self-assembly ability of atoms to make a structure we want. The way to understand how things self-assemble is to watch them do it, and that requires movies with high spatial resolution and good time resolution,” Ross explains. Electron microscopy can be used to acquire structural and compositional information and can even measure strain fields or electric and magnetic fields. “Imagine recording all of these things, but in a movie where you are also controlling how materials grow within the microscope. Once you have made a movie of something happening, you analyze all the steps of the growth process and use that to understand which physical principles were the key ones that determined how the structure nucleated and evolved and ended up the way it does.”

Future directions

Ross hopes to bring in a unique high-resolution, high vacuum TEM with capabilities to image materials growth and other dynamic processes. She intends to develop new capabilities for both water-based and gas-based environments. This custom microscope is still in the planning stages but will be situated in one of the rooms in the Imaging Suite in MIT.nano.

“Professor Ross is a pioneer in this field,” Osherov says. “The majority of TEM studies to-date have been static, rather than dynamic. With static measurements you are observing a sample at one particular snapshot in time, so you don’t gain any information about how it was formed. Using dynamic measurements, you can look at the atoms hopping from state to state until they find the final position. The ability to observe self-assembling processes and growth in real time provides valuable mechanistic insights. We’re looking forward to bringing these advanced capabilities to MIT.nano.” she says.

“Once a certain technique is disseminated to the public, it brings attention,” Osherov says. “When results are published, researchers expand their vision of experimental design based on available state-of-the-art capabilities, leading to many new experiments that will be focused on dynamic applications.”

Rooms in MIT.nano feature the quietest space on the MIT campus, designed to reduce vibrations and electromagnetic interference to as low a level as possible. “There is space available for Professor Ross to continue her research and to develop it further,” Osherov says. “The ability of in situ monitoring the formation of matter and interfaces will find applications in multiple fields across campus, and lead to a further push of the conventional electron microscopy limits.”

Creating new opportunities from nanoscale materials syndicated from https://osmowaterfilters.blogspot.com/

[djgblogger-blog]

Personalized “deep learning” equips robots for autism therapy

http://bit.ly/2tQX3XA

An example of a therapy session augmented with humanoid robot NAO [SoftBank Robotics], which was used in the EngageMe study. Tracking of limbs/faces was performed using the CMU Perceptual Lab’s OpenPose utility. Image: MIT Media Lab

By Becky Ham

Children with autism spectrum conditions often have trouble recognizing the emotional states of people around them — distinguishing a happy face from a fearful face, for instance. To remedy this, some therapists use a kid-friendly robot to demonstrate those emotions and to engage the children in imitating the emotions and responding to them in appropriate ways.

This type of therapy works best, however, if the robot can smoothly interpret the child’s own behavior — whether he or she is interested and excited or paying attention — during the therapy. Researchers at the MIT Media Lab have now developed a type of personalized machine learning that helps robots estimate the engagement and interest of each child during these interactions, using data that are unique to that child.

Armed with this personalized “deep learning” network, the robots’ perception of the children’s responses agreed with assessments by human experts, with a correlation score of 60 percent, the scientists report June 27 in Science Robotics.

It can be challenging for human observers to reach high levels of agreement about a child’s engagement and behavior. Their correlation scores are usually between 50 and 55 percent. Rudovic and his colleagues suggest that robots that are trained on human observations, as in this study, could someday provide more consistent estimates of these behaviors.

“The long-term goal is not to create robots that will replace human therapists, but to augment them with key information that the therapists can use to personalize the therapy content and also make more engaging and naturalistic interactions between the robots and children with autism,” explains Oggi Rudovic, a postdoc at the Media Lab and first author of the study.

Rosalind Picard, a co-author on the paper and professor at MIT who leads research in affective computing, says that personalization is especially important in autism therapy: A famous adage is, “If you have met one person, with autism, you have met one person with autism.”

“The challenge of creating machine learning and AI [artificial intelligence] that works in autism is particularly vexing, because the usual AI methods require a lot of data that are similar for each category that is learned. In autism where heterogeneity reigns, the normal AI approaches fail,” says Picard. Rudovic, Picard, and their teammates have also been using personalized deep learning in other areas, finding that it improves results for pain monitoring and for forecasting Alzheimer’s disease progression.  

Meeting NAO

Robot-assisted therapy for autism often works something like this: A human therapist shows a child photos or flash cards of different faces meant to represent different emotions, to teach them how to recognize expressions of fear, sadness, or joy. The therapist then programs the robot to show these same emotions to the child, and observes the child as she or he engages with the robot. The child’s behavior provides valuable feedback that the robot and therapist need to go forward with the lesson.

The researchers used SoftBank Robotics NAO humanoid robots in this study. Almost 2 feet tall and resembling an armored superhero or a droid, NAO conveys different emotions by changing the color of its eyes, the motion of its limbs, and the tone of its voice.

The 35 children with autism who participated in this study, 17 from Japan and 18 from Serbia, ranged in age from 3 to 13. They reacted in various ways to the robots during their 35-minute sessions, from looking bored and sleepy in some cases to jumping around the room with excitement, clapping their hands, and laughing or touching the robot.

Most of the children in the study reacted to the robot “not just as a toy but related to NAO respectfully as it if was a real person,” especially during storytelling, where the therapists asked how NAO would feel if the children took the robot for an ice cream treat, according to Rudovic.

One 4-year-old girl hid behind her mother while participating in the session but became much more open to the robot and ended up laughing by the end of the therapy. The sister of one of the Serbian children gave NAO a hug and said “Robot, I love you!” at the end of a session, saying she was happy to see how much her brother liked playing with the robot.

“Therapists say that engaging the child for even a few seconds can be a big challenge for them, and robots attract the attention of the child,” says Rudovic, explaining why robots have been useful in this type of therapy. “Also, humans change their expressions in many different ways, but the robots always do it in the same way, and this is less frustrating for the child because the child learns in a very structured way how the expressions will be shown.”

Personalized machine learning

The MIT research team realized that a kind of machine learning called deep learning would be useful for the therapy robots to have, to perceive the children’s behavior more naturally. A deep-learning system uses hierarchical, multiple layers of data processing to improve its tasks, with each successive layer amounting to a slightly more abstract representation of the original raw data.

Although the concept of deep learning has been around since the 1980s, says Rudovic, it’s only recently that there has been enough computing power to implement this kind of artificial intelligence. Deep learning has been used in automatic speech and object-recognition programs, making it well-suited for a problem such as making sense of the multiple features of the face, body, and voice that go into understanding a more abstract concept such as a child’s engagement.

“In the case of facial expressions, for instance, what parts of the face are the most important for estimation of engagement?” Rudovic says. “Deep learning allows the robot to directly extract the most important information from that data without the need for humans to manually craft those features.”

For the therapy robots, Rudovic and his colleagues took the idea of deep learning one step further and built a personalized framework that could learn from data collected on each individual child. The researchers captured video of each child’s facial expressions, head and body movements, poses and gestures, audio recordings and data on heart rate, body temperature, and skin sweat response from a monitor on the child’s wrist.

The robots’ personalized deep learning networks were built from layers of these video, audio, and physiological data, information about the child’s autism diagnosis and abilities, their culture and their gender. The researchers then compared their estimates of the children’s behavior with estimates from five human experts, who coded the children’s video and audio recordings on a continuous scale to determine how pleased or upset, how interested, and how engaged the child seemed during the session.

Trained on these personalized data coded by the humans, and tested on data not used in training or tuning the models, the networks significantly improved the robot’s automatic estimation of the child’s behavior for most of the children in the study, beyond what would be estimated if the network combined all the children’s data in a “one-size-fits-all” approach, the researchers found.

Rudovic and colleagues were also able to probe how the deep learning network made its estimations, which uncovered some interesting cultural differences between the children. “For instance, children from Japan showed more body movements during episodes of high engagement, while in Serbs large body movements were associated with disengagement episodes,” Rudovic says.

The study was funded by grants from the Japanese Ministry of Education, Culture, Sports, Science and Technology; Chubu University; and the European Union’s HORIZON 2020 grant (EngageME).

[dorcasrempel]

Creating new opportunities from nanoscale materials

A hundred years ago, “2d” meant a two-penny, or 1-inch, nail. Today, “2-D” encompasses a broad range of atomically thin flat materials, many with exotic properties not found in the bulk equivalents of the same materials, with graphene — the single-atom-thick form of carbon — perhaps the most prominent. While many researchers at MIT and elsewhere are exploring two-dimensional materials and their special properties, Frances M. Ross, the Ellen Swallow Richards Professor in Materials Science and Engineering, is interested in what happens when these 2-D materials and ordinary 3-D materials come together.

“We’re interested in the interface between a 2-D material and a 3-D material because every 2-D material that you want to use in an application, such as an electronic device, still has to talk to the outside world, which is three-dimensional,” Ross says.

“We’re at an interesting time because there are immense developments in instrumentation for electron microscopy, and there is great interest in materials with very precisely controlled structures and properties, and these two things cross in a fascinating way,” says Ross. 

“The opportunities are very exciting,” Ross says. “We’re going to be really improving the characterization capabilities here at MIT.” Ross specializes in examining how nanoscale materials grow and react in both gases and liquid media, by recording movies using electron microscopy. Microscopy of reactions in liquids is particularly useful for understanding the mechanisms of electrochemical reactions that govern the performance of catalysts, batteries, fuel cells, and other important technologies. “In the case of liquid phase microscopy, you can also look at corrosion where things dissolve away, while in gases you can look at how individual crystals grow or how materials react with, say, oxygen,” she says.

Ross joined the Department of Materials Science and Engineering (DMSE) faculty last year, moving from the nanoscale materials analysis department at the IBM Thomas J. Watson Research Center. “I learned a tremendous amount from my IBM colleagues and hope to extend our research in material design and growth in new directions,” she says.

Recording movies

During a recent visit to her lab, Ross explained an experimental setup donated to MIT by IBM. An ultra-high vacuum evaporator system arrived first, to be attached later directly onto a specially designed transmission electron microscope. “This gives powerful possibilities,” Ross explains. “We can put a sample in the vacuum, clean it, do all sorts of things to it such as heating and adding other materials, then transfer it under vacuum into the microscope, where we can do more experiments while we record images. So we can, for example, deposit silicon or germanium, or evaporate metals, while the sample is in the microscope and the electron beam is shining through it, and we are recording a movie of the process.”

While waiting this spring for the transmission electron microscope to be set up, members of Ross’ seven-member research group, including materials science and engineering postdoc Shu Fen Tan and graduate student Kate Reidy, made and studied a variety of self-assembled structures. The evaporator system was housed temporarily on the fifth-level prototyping space of MIT.nano while Ross’s lab was being readied in Building 13. “MIT.nano had the resources and space; we were happy to be able to help,” says Anna Osherov, MIT.nano assistant director of user services.

“All of us are interested in this grand challenge of materials science, which is: ‘How do you make a material with the properties you want and, in particular, how do you use nanoscale dimensions to tweak the properties, and create new properties, that you can’t get from bulk materials?’” Ross says.

Using the ultra-high vacuum system, graduate student Kate Reidy formed structures of gold and niobium on several 2-D materials. “Gold loves to grow into little triangles,” Ross notes. “We’ve been talking to people in physics and materials science about which combinations of materials are the most important to them in terms of controlling the structures and the interfaces between the components in order to give some improvement in the properties of the material,” she notes.

Shu Fen Tan synthesized nickel-platinum nanoparticles and examined them using another technique, liquid cell electron microscopy. She could arrange for only the nickel to dissolve, leaving behind spiky skeletons of platinum. “Inside the liquid cell, we are able to see this whole process at high spatial and temporal resolutions,” Tan says. She explains that platinum is a noble metal and less reactive than nickel, so under the right conditions the nickel participates in an electrochemical dissolution reaction and the platinum is left behind.

Platinum is a well-known catalyst in organic chemistry and fuel cell materials, Tan notes, but it is also expensive, so finding combinations with less-expensive materials such as nickel is desirable.

“This is an example of the range of materials reactions you can image in the electron microscope using the liquid cell technique,” Ross says. “You can grow materials; you can etch them away; you can look at, for example, bubble formation and fluid motion.”

A particularly important application of this technique is to study cycling of battery materials. “Obviously, I can’t put an AA battery in here, but you could set up the important materials inside this very small liquid cell and then you can cycle it back and forth and ask, if I charge and discharge it 10 times, what happens? It does not work just as well as before — how does it fail?” Ross asks. “Some kind of failure analysis and all the intermediate stages of charging and discharging can be observed in the liquid cell.”

“Microscopy experiments where you see every step of a reaction give you a much better chance of understanding what’s going on,” Ross says.

Moiré patterns

Graduate student Reidy is interested in how to control the growth of gold on 2-D materials such as graphene, tungsten diselenide, and molybdenum disulfide. When she deposited gold on “dirty” graphene, blobs of gold collected around the impurities. But when Reidy grew gold on graphene that had been heated and cleaned of impurities, she found perfect triangles of gold. Depositing gold on both the top and bottom sides of clean graphene, Reidy saw in the microscope features known as moiré patterns, which are caused when the overlapping crystal structures are out of alignment.

The gold triangles may be useful as photonic and plasmonic structures. “We think this could be important for a lot of applications, and it is always interesting for us to see what happens,” Reidy says. She is planning to extend her clean growth method to form 3-D metal crystals on stacked 2-D materials with various rotation angles and other mixed-layer structures. Reidy is interested in the properties of graphene and hexagonal boron nitride (hBN), as well as two materials that are semiconducting in their 2-D single-layer form, molybdenum disulfide (MoS2) and tungsten diselenide (WSe2). “One aspect that’s very interesting in the 2-D materials community is the contacts between 2-D materials and 3-D metals,” Reidy says. “If they want to make a semiconducting device or a device with graphene, the contact could be ohmic for the graphene case or a Schottky contact for the semiconducting case, and the interface between these materials is really, really important.”

“You can also imagine devices using the graphene just as a spacer layer between two other materials,” Ross adds.

For device makers, Reidy says it is sometimes important to have a 3-D material grow with its atomic arrangement aligned perfectly with the atomic arrangement in the 2-D layer beneath. This is called epitaxial growth. Describing an image of gold grown together with silver on graphene, Reidy explains, “We found that silver doesn’t grow epitaxially, it doesn’t make those perfect single crystals on graphene that we wanted to make, but by first depositing the gold and then depositing silver around it, we can almost force silver to go into an epitaxial shape because it wants to conform to what its gold neighbors are doing.”

Electron microscope images can also show imperfections in a crystal such as rippling or bending, Reidy notes. “One of the great things about electron microscopy is that it is very sensitive to changes in the arrangement of the atoms,” Ross says. “You could have a perfect crystal and it would all look the same shade of gray, but if you have a local change in the structure, even a subtle change, electron microscopy can pick it up. Even if the change is just within the top few layers of atoms without affecting the rest of the material beneath, the image will show distinctive features that allow us to work out what’s going on.”

Reidy also is exploring the possibilities of combining niobium — a metal that is superconducting at low temperatures — with a 2-D topological insulator, bismuth telluride. Topological insulators have fascinating properties whose discovery resulted in the Nobel Prize in Physics in 2016. “If you deposit niobium on top of bismuth telluride, with a very good interface, you can make superconducting junctions. We’ve been looking into niobium deposition, and rather than triangles we see structures that are more dendritic looking,” Reidy says. Dendritic structures look like the frost patterns formed on the inside of windows in winter, or the feathery patterns of some ferns. Changing the temperature and other conditions during the deposition of niobium can change the patterns that the material takes.

All the researchers are eager for new electron microscopes to arrive at MIT.nano to give further insights into the behavior of these materials. “Many things will happen within the next year, things are ramping up already, and I have great people to work with. One new microscope is being installed now in MIT.nano and another will arrive next year. The whole community will see the benefits of improved microscopy characterization capabilities here,” Ross says.

MIT.nano’s Osherov notes that two cryogenic transmission electron microscopes (cryo-TEM) are installed and running. “Our goal is to establish a unique microscopy-centered community. We encourage and hope to facilitate a cross-pollination between the cryo-EM researchers, primarily focused on biological applications and ‘soft’ material, as well as other research communities across campus,” she says. The latest addition of a scanning transmission electron microscope with enhanced analytical capabilities (ultrahigh energy resolution monochromator, 4-D STEM detector, Super-X EDS detector, tomography, and several in situ holders) brought in by John Chipman Associate Professor of Materials Science and Engineering James M. LeBeau, once installed, will substantially enhance the microscopy capabilities of the MIT campus. “We consider Professor Ross to be an immense resource for advising us in how to shape the in situ approach to measurements using the advanced instrumentation that will be shared and available to all the researchers within the MIT community and beyond,” Osherov says.

Little drinking straws

“Sometimes you know more or less what you are going to see during a growth experiment, but very often there’s something that you don’t expect,” Ross says. She shows an example of zinc oxide nanowires that were grown using a germanium catalyst. Some of the long crystals have a hole through their centers, creating structures which are like little drinking straws, circular outside but with a hexagonally shaped interior. “This is a single crystal of zinc oxide, and the interesting question for us is why do the experimental conditions create these facets inside, while the outside is smooth?” Ross asks. “Metal oxide nanostructures have so many different applications, and each new structure can show different properties. In particular, by going to the nanoscale you get access to a diverse set of properties.”

“Ultimately, we’d like to develop techniques for growing well-defined structures out of metal oxides, especially if we can control the composition at each location on the structure,” Ross says. A key to this approach is self-assembly, where the material builds itself into the structure you want without having to individually tweak each component. “Self-assembly works very well for certain materials but the problem is that there’s always some uncertainty, some randomness or fluctuations. There’s poor control over the exact structures that you get. So the idea is to try to understand self-assembly well enough to be able to control it and get the properties that you want,” Ross says.

“We have to understand how the atoms end up where they are, then use that self-assembly ability of atoms to make a structure we want. The way to understand how things self-assemble is to watch them do it, and that requires movies with high spatial resolution and good time resolution,” Ross explains. Electron microscopy can be used to acquire structural and compositional information and can even measure strain fields or electric and magnetic fields. “Imagine recording all of these things, but in a movie where you are also controlling how materials grow within the microscope. Once you have made a movie of something happening, you analyze all the steps of the growth process and use that to understand which physical principles were the key ones that determined how the structure nucleated and evolved and ended up the way it does.”

Future directions

Ross hopes to bring in a unique high-resolution, high vacuum TEM with capabilities to image materials growth and other dynamic processes. She intends to develop new capabilities for both water-based and gas-based environments. This custom microscope is still in the planning stages but will be situated in one of the rooms in the Imaging Suite in MIT.nano.

“Professor Ross is a pioneer in this field,” Osherov says. “The majority of TEM studies to-date have been static, rather than dynamic. With static measurements you are observing a sample at one particular snapshot in time, so you don’t gain any information about how it was formed. Using dynamic measurements, you can look at the atoms hopping from state to state until they find the final position. The ability to observe self-assembling processes and growth in real time provides valuable mechanistic insights. We’re looking forward to bringing these advanced capabilities to MIT.nano.” she says.

“Once a certain technique is disseminated to the public, it brings attention,” Osherov says. “When results are published, researchers expand their vision of experimental design based on available state-of-the-art capabilities, leading to many new experiments that will be focused on dynamic applications.”

Rooms in MIT.nano feature the quietest space on the MIT campus, designed to reduce vibrations and electromagnetic interference to as low a level as possible. “There is space available for Professor Ross to continue her research and to develop it further,” Osherov says. “The ability of in situ monitoring the formation of matter and interfaces will find applications in multiple fields across campus, and lead to a further push of the conventional electron microscopy limits.”

Creating new opportunities from nanoscale materials syndicated from https://osmowaterfilters.blogspot.com/

[dorcasrempel]

Personalized “deep learning” equips robots for autism therapy

Children with autism spectrum conditions often have trouble recognizing the emotional states of people around them — distinguishing a happy face from a fearful face, for instance. To remedy this, some therapists use a kid-friendly robot to demonstrate those emotions and to engage the children in imitating the emotions and responding to them in appropriate ways.

This type of therapy works best, however, if the robot can smoothly interpret the child’s own behavior — whether he or she is interested and excited or paying attention — during the therapy. Researchers at the MIT Media Lab have now developed a type of personalized machine learning that helps robots estimate the engagement and interest of each child during these interactions, using data that are unique to that child.

Armed with this personalized “deep learning” network, the robots’ perception of the children’s responses agreed with assessments by human experts, with a correlation score of 60 percent, the scientists report June 27 in Science Robotics.

It can be challenging for human observers to reach high levels of agreement about a child’s engagement and behavior. Their correlation scores are usually between 50 and 55 percent. Rudovic and his colleagues suggest that robots that are trained on human observations, as in this study, could someday provide more consistent estimates of these behaviors.

“The long-term goal is not to create robots that will replace human therapists, but to augment them with key information that the therapists can use to personalize the therapy content and also make more engaging and naturalistic interactions between the robots and children with autism,” explains Oggi Rudovic, a postdoc at the Media Lab and first author of the study.

Rosalind Picard, a co-author on the paper and professor at MIT who leads research in affective computing, says that personalization is especially important in autism therapy: A famous adage is, “If you have met one person, with autism, you have met one person with autism.”

“The challenge of creating machine learning and AI [artificial intelligence] that works in autism is particularly vexing, because the usual AI methods require a lot of data that are similar for each category that is learned. In autism where heterogeneity reigns, the normal AI approaches fail,” says Picard. Rudovic, Picard, and their teammates have also been using personalized deep learning in other areas, finding that it improves results for pain monitoring and for forecasting Alzheimer’s disease progression.  

Meeting NAO

Robot-assisted therapy for autism often works something like this: A human therapist shows a child photos or flash cards of different faces meant to represent different emotions, to teach them how to recognize expressions of fear, sadness, or joy. The therapist then programs the robot to show these same emotions to the child, and observes the child as she or he engages with the robot. The child’s behavior provides valuable feedback that the robot and therapist need to go forward with the lesson.

The researchers used SoftBank Robotics NAO humanoid robots in this study. Almost 2 feet tall and resembling an armored superhero or a droid, NAO conveys different emotions by changing the color of its eyes, the motion of its limbs, and the tone of its voice.

The 35 children with autism who participated in this study, 17 from Japan and 18 from Serbia, ranged in age from 3 to 13. They reacted in various ways to the robots during their 35-minute sessions, from looking bored and sleepy in some cases to jumping around the room with excitement, clapping their hands, and laughing or touching the robot.

Most of the children in the study reacted to the robot “not just as a toy but related to NAO respectfully as it if was a real person,” especially during storytelling, where the therapists asked how NAO would feel if the children took the robot for an ice cream treat, according to Rudovic.

One 4-year-old girl hid behind her mother while participating in the session but became much more open to the robot and ended up laughing by the end of the therapy. The sister of one of the Serbian children gave NAO a hug and said “Robot, I love you!” at the end of a session, saying she was happy to see how much her brother liked playing with the robot.

“Therapists say that engaging the child for even a few seconds can be a big challenge for them, and robots attract the attention of the child,” says Rudovic, explaining why robots have been useful in this type of therapy. “Also, humans change their expressions in many different ways, but the robots always do it in the same way, and this is less frustrating for the child because the child learns in a very structured way how the expressions will be shown.”

Personalized machine learning

The MIT research team realized that a kind of machine learning called deep learning would be useful for the therapy robots to have, to perceive the children’s behavior more naturally. A deep-learning system uses hierarchical, multiple layers of data processing to improve its tasks, with each successive layer amounting to a slightly more abstract representation of the original raw data.

Although the concept of deep learning has been around since the 1980s, says Rudovic, it’s only recently that there has been enough computing power to implement this kind of artificial intelligence. Deep learning has been used in automatic speech and object-recognition programs, making it well-suited for a problem such as making sense of the multiple features of the face, body, and voice that go into understanding a more abstract concept such as a child’s engagement.

“In the case of facial expressions, for instance, what parts of the face are the most important for estimation of engagement?” Rudovic says. “Deep learning allows the robot to directly extract the most important information from that data without the need for humans to manually craft those features.”

For the therapy robots, Rudovic and his colleagues took the idea of deep learning one step further and built a personalized framework that could learn from data collected on each individual child. The researchers captured video of each child’s facial expressions, head and body movements, poses and gestures, audio recordings and data on heart rate, body temperature, and skin sweat response from a monitor on the child’s wrist.

The robots’ personalized deep learning networks were built from layers of these video, audio, and physiological data, information about the child’s autism diagnosis and abilities, their culture and their gender. The researchers then compared their estimates of the children’s behavior with estimates from five human experts, who coded the children’s video and audio recordings on a continuous scale to determine how pleased or upset, how interested, and how engaged the child seemed during the session.

Trained on these personalized data coded by the humans, and tested on data not used in training or tuning the models, the networks significantly improved the robot’s automatic estimation of the child’s behavior for most of the children in the study, beyond what would be estimated if the network combined all the children’s data in a “one-size-fits-all” approach, the researchers found.

Rudovic and colleagues were also able to probe how the deep learning network made its estimations, which uncovered some interesting cultural differences between the children. “For instance, children from Japan showed more body movements during episodes of high engagement, while in Serbs large body movements were associated with disengagement episodes,” Rudovic says.

The study was funded by grants from the Japanese Ministry of Education, Culture, Sports, Science and Technology; Chubu University; and the European Union’s HORIZON 2020 grant (EngageME).

Personalized “deep learning” equips robots for autism therapy syndicated from https://osmowaterfilters.blogspot.com/