World’s Shortest Wavelength Laser Diode Emits Deep UV Light at Room Temperature | Hacker News Reader

World’s Shortest Wavelength Laser Diode Emits Deep UV Light at Room Temperature

iopscience.iop.org

101 pointsbookofjoe64 comments

In the atomic research world, diode lasers are great because they’re cheap and fairly tunable.

Many of atomic cooling studies done 20 years ago were done using rubidium because it had a primary absorption wavelength that matched what commercial CD players used. This made the 780nm diodes dirt cheap.

Having new wavelengths of diode lasers makes certain basic research more accessible. I’ve been out of the field too long to know where this one may help but it’s exciting to see progress.

Nowadays in atomic physics frequency-quadrupled diode lasers are commonly used for accessing DUV wavelengths. They produce a lot of power and have narrow linewidths, but are very expensive (>> 100k€). Here is a chart from one manufacturer that shows commonly accessed transitions: https://www.toptica.com/fileadmin/Editors_English/15_downloa...

In the paper the authors only demonstrate pulsed operation of a bare laser diode. To make it usable for atomic physics, they would have to achieve continuous-wave operation with feedback from an external cavity. This might be pretty difficult to do.

This is good, the usage for UV is huge and anything that reduces costs will only help.

Whilst many don't need it, I'm sure having a UV filter of your water would not go amiss.

Or maybe for photolithography hobbyist.

Yeah, especially because that hinges on having coherent light, whereas sterilization does not AFAIK.

As for incoherent UV light, I wonder how this laser compares in efficiency to arc lamps. Is that still how incoherent UV is generated? Deuterium and mercury vapor? Or have LEDs taken over?

A solution that does not use mercury, then that will make for cheaper and equally safer production and dispose.

DUV for lithography is usually produced with excimer lasers at 248 or 193 nm. While they are technically lasers, they are barely coherent and have horrible beam quality.

For other applications such as sterilization or curing adhesives, mercury vapor lamps and LEDs are popular.

Is photolithography at home limited by laser'd wavelength?

DUV LEDs have existed for quite a while and produce a lot more power.

UVB/UVC LEDs barely touch 2% efficiency converting energy into photons. When they can hit 30%, I'll have the perfect mineral hunting lamp designed and built.

Handheld UV water purifiers are already common. A lot of backpackers use them. I'm not sure how they work. Presumably they're not using LEDs.

They use mercury lamps. I took one apart to use the lamp as a wavelength standard. Don't try that at home, the sterilizers have interlocks to prevent accidental eye and skin exposure.

I have been considering doing this too. Do you have any specific advice, you would have loved to give your prior self?

Yes. Don't do it. Buy the nice little mercury-argon lamp from Ocean Optics.

First of all, the mercury lamp is dangerous, for real. Second, the newer sterilizer pens contain complicated interlock circuitry, making it impossible to figure out how to actually get the lamp to turn on by itself. I got one lamp to light up from an external power supply, then set it aside and bought a few of the Ocean Optics boxes.

As a wavelength standard, it was also a poor configuration -- most of the light can't be directed where you need it, whereas the Ocean Optics produces a nice compact beam and has a fiber optic connector.

thank you very much for the swift reply,

I actually don't need it as a wavelength standard, but as a small source for entangled photons.

I'm well aware of the dangers of deep UV and ozone (and of mercury if I would accidentally break the lamp by opening the sterilizer pen).

I'm unable to identify what Ocean Optics box you are referring to?

If I do have resort to the sterilizer pens, any advice regarding opening up and analyzing the interlock circuitry? I have an oscilloscope...

Aha, the company name changed. But I've bought these under the Ocean Optics brand, and they've always worked fine. Note that they have a variety of elements, not just mercury. If you don't need great gobs of photons, a small neon lamp is a lot more friendly to play with, still involves dangerous high voltage, but no UV. The NE-2 lamp is a common one, available from electronic parts suppliers.

I found the circuit of the sterilizer pen to be impenetrable, as it contains a microcontroller, which is a "black box" to me. I think they wanted to make it hard to jimmy.

https://www.oceaninsight.com/products/light-sources/calibrat...

Understand that I have to be conservative about this stuff. I'm not an engineer, so I have to assume the worst about the hazards.

Also, I do know from using any of these lamps, that a timer is a good investment, so you don't burn them out prematurely.

Even more, they are getting pretty close to x-ray now which begins at 10nm.

The laser structure was grown on the (0001) face of a single-crystal aluminum nitride substrate. The measured lasing wavelength was 271.8 nm with a pulsed duration of 50 ns and a repetition frequency of 2 kHz.

I know this is a pipe-dream, but a 2.5Å x-ray laser would be a scientific revolution for protein crystallography. Instead of ~2-10(!) years to find the structure of a protein, we’d be limited by the chemistry.

> I'm sure having a UV filter of your water would not go amiss

You don't need lased light for that (and using a low efficiency, high cost device also not ideal?!).

UV LEDs seem to be in development. The problem with UV fluorescent tubes is that they have horrible cycle life.

I guess everyone means UVC here. A UVA flashlight is cheaper than lunch ;) edit: my facts were scrambled. C is the nasty one... maybe my favorite (unsafe) language will help me remember that from now on.

Would be very useful for vitiligo treatment.

> I'm sure having a UV filter of your water would not go amiss.

UV "filter"? I thought the UV thing in water treatment was merely to kill most (if not all) of what micro-life would otherwise still live in yr water.

Filters are needed to actually get some (dissolved) solids --possibly UV-zapped micro org corpses-- out of yr water.

Most places add chlorine/chloramine to the tap water to prevent micro orgs becoming so prevalent "at the faucet" that you can get "cannot go to work tomorrow" sick.

Yes, if you live in an area with municipal water treatment, but if you rely on a private well, accessible and inexpensive UV treatment would be great. Of course, you also (generally) need a water softener and perhaps more expensive filtering equipment depending on the what metals and minerals might be present.

Why would one need a UV if there's also going to me actual filters after that? (honest question) Only thing I can think of is to prevent contamination of the filters.

Because if you have an actual filter that is a small enough micron to capture pathogens, it's also going to capture all of the really fine sediment and plug up quickly.

That why most water filter systems have "stages", every stage "covers" for the next.

It is entirely site specific. I don't have any filters, just a water conditioner (since I don't want to deal with a softener). Adding UV would probably be unnecessary but would provide some nice peace of mind.

Bit of a tangent, but do the blue LEDs that hand driers often have now do anything? Or is it just marketing buzz?

Normally, in LEDs and solid state lasers, there is a direct relation between wavelength(energy quantum) (electron-Volts) and operating voltage (Volts).

For example Red = 660 nanometers = 1.8 electron-Volts = 1.8 Volts, which is the typical operating voltage for a Red LED.

Blue = 470 nm = 3.3 eV = 3.3 Volts operating voltage.

So: Deep UV (270 nm) is supposed to operate on ~4.6 Volts.

But article mentions 13.8 Volts. I wonder why there is such a huge gap?

Down conversion perhaps?

I'm not sure if that has any meaning, as I know nothing about lasers, but 4.6 * 3 = 13.8 , could be pure coincidence.

The bands of laser diodes are not necessarily sharp, so there can be a significant difference between the injection voltage and the photon energy. The paper mentions that their particular doping method produces a step-less valence band profile.

Furthermore, they need a fairly large current (0.4 A) to get this thing to lase. The "knee" in the U-I curve is at around 9 V and 0.05 A, so there seems to be a fair bit of voltage drop due to internal resistance of the device. This is probably also why they have to pulse the laser with a 0.01 % duty cycle in order not to produce too much heat.

By sharp, do you mean a direct band-gap?

I recall that you sometimes want to pump your atoms in a higher energy state than the one that corresponds to the output energy, but I forgot its usefulness. Could someone enlighten me on this? Or is it just for optically pumped mediums (thus pumped with higher energy photons than the emitted ones)?

It is impossible to achieve inversion (and therefore gain) by optically pumping a medium exactly at the emission wavelength. Therefore, optically pumped laser gain media are at least three (usually four) level systems.

Based on the abstract, it is a pulsed system (in contrast to continuous wave LEDs) so I think there is a bias voltage of 14V.

> But article mentions 13.8 Volts. I wonder why there is such a huge gap?

I'm spitballing here, but I'd guess that the difference is due to a resistive drop across one or both of the contact layers.

In the article, Fig 2 says that they have 0.5A flowing with a pulse width of 50 ns and a pulse rate of 2 kHz. Their duty cycle is then 1/10000, so that's a lot of current flowing during the active part of the cycle. I haven't found all the details about the dimensions and doping, but any small amount of resistance might cause enough potential drop between the contact and the lasing region to account for the difference (especially that side contact to the bottom n-type layer.)

Three laser diodes in series, electrically?

I'd wager that they're just using a pulsed array of three LEDs in series to get their required photon output. Typical for solid state lasing.

Does this have any implication for chip fabs? Right now EUV requires crazy Rube Goldberg machines using tin plasma or something like that, which is insanely expensive.

EUV is 20x smaller wavelength, 10000x higher power, and far larger area

it's noteworthy that even when it's on, this laser is mostly off. its duty cycle is 0.00001% (50 ns at 2khz).

Who is the lead writer of this paper? According to google Ziyi Zhang is a famous Chinese actress.

Likely not the same person. From the article:

Author e-mails zhang.zc@om.asahi-kasei.co.jp Author affiliations 1 Innovative Devices R&D Center, Corporate Research & Development, Asahi Kasei Corporation, Fuji, Shizuoka 416-8501, Japan

2 Center for Integrated Research of Future Electronics, Institute of Materials Research and System for Sustainability, Nagoya University, Furo-cho, Chikusa-ku, Aichi 464-8601, Japan

I'm sure some of the people named Will Smith are scientists.

Zhang is an extremely common Chinese family name, so I wouldn't be surprised to see a lot of people with the same name.

But can I use it to burn things?

What does “Burn” mean?

Break the bonds of just about anything except fluorides and metals? Probably.

Deposit enough energy I locally heat the material above it’s auto ignition T? Well that’s a flux question

Yes, to raise the temperature of a material in which ignition can occur.

I know the cheap hobby laser machines use a 405-450nm module, which works decent on stuff like wood. Now since it's about half the wavelength, is it more affective in cutting / etching these materials with a lower optical output power.

That's just straight up energy transfer. Doesn't matter exactly what wavelength you use provided you dump enough energy in, and this one is so inefficient it doesn't make sense.

If you look at CO2 lasers, at 40W ($350 on eBay) you can cut plywood, at 150W ($5000 on eBay) you can cut metal. Both of those are 1040 nm.

The 450 nm modules you can get up to 5W ($100 on eBay), but those can hardly cut anything, they're more like engravers.

CO2 lasers are typically at 10.6 microns. They can cut metal but you need at least 100W and RF pumping. Metal-cutting lasers are more commonly fiber lasers with wavelengths of ~1050nm.

“ is it more affective in cutting / etching these materials with a lower optical output power.”

Probably because it’s more likely to be absorbed.

On the other hand, absorption can’t be > 100%. So a longer wavelength laser an cut just as well for the same power as long as the mtls absorbs all the photons

This looks like a great HN comment, but could you please stop creating accounts for every few comments you post? We ban accounts that do that. This is in the site guidelines: https://news.ycombinator.com/newsguidelines.html.

HN is a community and we want it to remain one. For that, users need some identity for others to relate to. Otherwise we may as well have no usernames and no community, and that would be a different kind of forum. https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme...

You needn't use your real name, of course.

“ This looks like a great HN comment, ”

Thank you. I thought it was rushed w/out the explanation that it deserves (it has fascinating implications about the lack of the color blue in nature). Also, I felt it came out as snarky or arrogant, which I didn’t intend. But I was juggling other things :S

Unfortunately, I’m afraid you’ll have to ban me. I’ve abused the site far too long but I also have no intention to expose myself w/ a profile (any user with a karma > 1000 is trivial to dox).

> any user with a karma > 1000 is trivial to dox

How so?

Can you elaborate? Detach to another thread maybe?

It's result of frequent shadowban without reason.

We don't shadowban accounts that have significant history on HN. We tell them why we've banned them: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...

We use shadowbanning when accounts don't have much history and there's evidence of prior abuse. There's more explanation at https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que... if anyone wants it.

It's not true.

If it weren't true, people would be jumping in with counterexamples all over the place and we'd never hear the end of it.

If you know of a counterexample, you should state what it is so readers can make up their own minds. Also, it's possible that we made a mistake, in which case a link would be helpful so we can correct it.

It's also possible that there were complicating factors. We've dealt with thousands of these cases. There's no easy formula for summing them all up. But the principles I just described here are certainly the ones we apply and the ones that we teach to moderators.

I think you are fighting an uphill battle there. HN is a community, yes, but one with a heavy pro-privacy slant. There's always been a disconnect between that and the admins' values who have always opposed adding the ability to delete anything, not even usernames on comments. Rotating usernames is a predictable consequence of not being able to delete anything, even superficially (and much more effective).

It's okay to have an alt account, you just shouldn't create a new alt account for every new comment. This allows for their moderation systems to be effective.

My account is a great example. This account is separate from my real identity, and I use it to post about things I don't want associated with my real identity. I went through a period of poor-quality commenting that led to well-justified moderator action against it(I'm not exactly sure what, but I'm consistently brought to the bottom of threads). This is important for the site because it helps maintain a good signal-to-noise ratio and keep the site interesting and productive. If I were to always create a new account, it would bypass this important moderation tool. Reputation is important in any community, including this one.

We delete comments and rename accounts for people every day. We just don't offer it wholesale. No one asks for help with that gets turned away, nor, I believe, has anyone gone away unsatisfied. Maybe I'm forgetting one or two.

We care a lot about individuals' need for privacy. We just need to balance it with the community's need to preserve its archive, and the right of repliers not to have the context of their comments pulled out from under them. We bend over backwards to get that right. For example, I have written custom code to help individual users with complex requests on several occasions.

I should probably add this to the FAQ.

Reply on news.ycombinator.com