Scientist Recognized for Inventing White LED’s

Ben, possibly a light emitting diode whose silicon substrate is “doped” (a real term used of semiconductors – not a joke on my part) with a particular element that will emit light of the UV -C wavelength range. Power is the key here due to the inverse-square principle of radiating energy which governs light and some other physical phenonmena – probably don’t look for coin cells to provide any meaningful duration (number of treatments); but i’m just guessing with this last assumption on coin cells.

to my limited knowledge, other than white LED’s (which are doped with a substance which, IIRC, produces a bluish light which then in turn causes a phosphor-like coating on the inside of the plastic LED housing to emit “white” light – light having wavelenghts across the visible spectrum and not limited to a narrower range), the doping substance is chosen that will emit light of the desired color/wavelength. “Colored” LEDs don’t use the phosphor since the proper color/wavelength of light is produced by the doping substance.

sorry, my memory is failing me as to the precise doping substances, though i know i knew only a few years ago what several doping substances were and the colors they produced, including the one most commonly used in Nichia White LEDs.

Maybe Rick Drehrer or Dr. Caffin should weigh in here and correct any mistakes i’ve made due to my poor recollection of solid state physics.

Howdy Paul:

Thanks for your insights, as always! :)

Ben, i had a thought on how a small, lightweight Li coin-cell UV-C LED purification unit could be designed:

what might make the Li coin cell powered UV-C LED work, is to have the UV-C light irradiate only a small amount of water passing through a small channel/passage or resevoir. this would reduce the power requirements since a smaller volume of water is being irradiated, so power output can be less. basically turn the unit on and then irradiate “on demand” only the water being drunk. exposure time is still an issue, but with a little more thought perhaps sufficient exposure time can be obtained. the current approx. 45-90 second exposure time for 0.5 to 1.0 liter of water is, i think, based solely on getting all of the water near enough to the UV-C light source, or exposing the water further from the light source a longer period of time to a weakened intensity of UV-C due to the distance the light must travel. irradiating light in one or more small channels or resevoirs would probably greatly reduce the exposure time required to create defects in microbial RNA and/or DNA (depending upon the organism irradiated). basically, a “if Mohammed won’t come to the mountain, then the mountain will come to Mohammed” approach. water swirling of some type is still a good idea to insure that any microbes attached to and shielded by tiny particulate debris will eventually get exposed to the UV-C light. creating a swirling vortex or some turbulent flow is easy enough though as a side effect of the motion of the water passing through a properly designed orifice and into the irradiating channel. this kind of unit might be contained in a larger screw on cap to a Platy or other water bottle/bladder, thus an in-line UV-C purification unit.

i don’t think that i will be long before someone brings a device similar to this to market.

Paul J,
Good thinking. The rule of squares would support your argument. If someone had the gumption to make a UV-C LED, then: Can UV-C be reflected? If so, by what? If so, then any radiation not absorbed by the smaller volume of water could be ‘recycled’ through it again. What materials are transparent to UV-C, and are they suitable/available for construction of the water jacket or tubes. How can you control the residence time/ flow rate to ensure adequatet disinfection? Is this a kitchen-table project?

Vick,

IIRC, UV LEDs are already made – don’t know if it’s the A, B, or C portion/wavelenghts of the UV band.

In fact, maybe it’s Photon who makes one already, but i’m not sure anymore of where i saw that product. [EDIT: just remembered after posting – it IS Photon (or LRI) who makes one (don’t know if it’s the “C” portion of the UV band though). There is a 6 or 7 LED array version of the Photon Freedom headlamp. Can’t remember if there is a single UV LED version of Photon Freedom Microlight.]

Glass and some plastics can block some amount of UV-A and UV-B and my uneducated guess is UV-C may be blocked also (Dr. Caffin, a little help here – please). From my years of working in microbiology, i know that glass and plastic DON’T block as much UV (-C???) as many people (even some authorities on the subject) claim. This is easily determined empirically by placing closed plastic petri dishes innoculated with a bacterium (E. coli for instance) in front of a closed glass window for 20minutes to two hours and then compare growth (or no-growth!!!) to a control dish not so exposed. i used this experiment with undergrads in their first “micro” class as a “graphic” example of what NOT to do.

Two more points:
1. as we’ve both mentioned, i’ve thought about the inverse-square rule. i hope that the power required by the UV-C LED can really be quite small because it may be that the sum total of “on-demand” UV-C purification time might add up and make this approach not practical. a knowledgeable person in the field could probably calculate such and also is, perhaps, already set up to empirically determine the viability of this method (they’d probably use bacteria and/or protozoans, and then have a specialized lab test for its effectiveness against viruses – which should be more effective “in vitro” than “in vivo” – probably due to minute particulate matter being able to shield the smaller virus particles better is my guess).

so, testing is a simple way (once a prototype of the device is built) to determine proper flow rate and exposure time. overall area may need to be increased to get an adequate slow flow and proper exposure time, as well as to reduce the amount of “sucking” required to draw water (i.e., increased area reduces resistance to flow and maintains a reasonable flow volume at a slower flow velocity, but then will the device get too large in size??? don’t know).

2. was thinking that a small pressure switch which responds to the delta-P (difference in pressure) caused by a user drawing water from the bite valve could be used to “auto-magically” turn on power to the LED, and then off again as the user stops sucking water.

so, is this approach totally off-the-wall and impractical? don’t know. i’ll let the young entrepreneurs out there give it go. i’ll be retired in a couple of years and don’t have the “drive” anymore to apply my “micro” and engineering backgrounds to this task. SteriPen and AquaStar ought to be doing some R&D in this area – unless they have already determined that it is NOT a viable approach.

You have TWO distinct topics here.

The first is UV purification of water. The Steripen does this with a UV discharge tube right now, and there is a new smaller model coming from the last OR show. These work. The shorter the UV wavelength used, the more grunt the UV photons have for killing the bugs and wogs. UV-C is shorter wavelength than UV-B, which is shoter wavelength than UV-A (which is shorter wavelength than blue light).

The second is whether we can find a semiconductor which can emit UV photons. Well, the shorter the wavelength, the harder this is to do. Some laws of physics come into play here. We do have some UV-A emitting LEDs today, but I don’t know whether they have enough output power (number of photons) or whether the wavelength is short enough to kill the bugs, just yet. Finding a semiconductor material which can emit UV-B photons … well, that’s hard. The energy of the UV-B photons is high enough that few materials have a hope of doing it, and the photons can cause internal damage to the LED as well. VERY crudely, by way of comparison, you can let fire crackers off in a bucket and have it survive, but don’t try it with Semtex!

But dream on guys: there’s a big market out there, believe me, and the developers are trying damn hard!

there you have it!

Thanks Roger for the insightful, informative input.

Very interesting thread. I have done some work in this area. Roger is correct that deep UV emitting semiconductor devices are more challenging than longer wavelength LED’s. However, these devices do exist. Shuji Nakamura’s (the scientist sited in Benjamin Tang’s link: http://news.bbc.co.uk/2/hi/technology/5328446.stm) is credited with the first UVC LED (280 nanometers) – about 3 years ago. These early devices were very inefficient – very low conversion of electrons to UV photons – however, they did indeed create a small amount of germicidal UV light.

About 2 years ago, with funding from the Navy and DARPA, we used some of these devices to treat a very small volume of e.coli spiked water. The testing was successful, and while the technology was at that time completely impractical for making useful volumes of water, this was, to our knowledge, the world’s first demonstration of UV treatment using a Semiconductor.

Over the past 2 years the UVC LED technology has steadily improved and devices are now getting close to the power levels and efficiencies that will eventually make LED water treatment a very good approach – particularly for portable water treatment.

Anyway, there is lots more on this very interesting topic. I have a patent and some pending patents in this area – a pending one that has many of the features that Paul Johnson suggests in his earlier post. For those into looking at patents – go to:

http://www.uspto.gov/patft/index.html

and look at patent number: 6,579,495 and patent aplication (pending patent) number 20060163126

Best,
Miles Maiden
CEO, CTO
Hydro-Photon, Inc.

> patent number: 6,579,495 and patent aplication (pending patent) number 20060163126

Hi Miles
Not asking for disclosure of any sort, but how would compare your patent concept with the current SteriPen, including the latest model?
Yes, I know it’s a CCFL tube, but otherwise?
Especially eficiency expected per battery life?

Hi Roger,
There are a few concepts – hence more than one patent in the works.

The idea described in the 6,579,495 patent is essentially a SteriPEN that uses LED’s in place of the CCFL lamp. In a few years, once the deep UV LED’s become a more mature technology (higher electro-optical efficiency, greater UV output power and low production cost) the advantages will include:

small overall SteriPEN size and weight – possibly down in the 2-3 oz range with batteries (by comparrison, our new Adventurer model is 3.6 oz with batteries).

extremely long life – UV LED’s could be expected to provide stable output for in the range of 50,000 hours or more. I believe this represents far more treatments than anyone would expect to do in one lifetime – about 500,000 gallons at a 90 second/liter rate. The current SteriPEN will provide up to 1,250 gallons before a needing a new lamp. Actually though, 1,250 gallons seems to be more than adequate for most people – over the past 6 years I think we have only had 2 SteriPEN’s that needed new lamps because the 5,000 treatments had been used up.

Extreme durability – while the current SteriPEN is actually quite rugged, an LED version will be close to indestructable. LED’s are just extremely rugged devices

High efficiency – in theory, LED’s can be by far the most efficient way to convert electricity to light. A SteriPEN Adventurer, which currently uses two CR123 batteries, may need just one CR123 (or possibly less) for the same performance with LED’s. LED technology in visible, IR and UV is improving all the time – visible and IR devices are further in their evolution but over the next decade or so all these devices will likely continue to improve. And as this happens more and more incandescent and fluorescent aplications will shift to LEDs. This trend is underway now – from flashlights to stop lights to tail lights, etc.

Minimal temperature sensitivity. LED output is not significantly impacted by ambient temperature – they reach full power instantly whether outdoor ambient is cold or hot. SteriPEN’s take a few seconds get up to full power in cold temps – so there is a temperature sensor in the SteriPEN that lengthens the dose a bit when ambient is cold (note- SteriPEN’s lamp is insulated by an outer quartz sleeve -so water temp has almost no effect – only ambient air temp).

Lower cost – once UV LED’s are fully mature and in large scale mass production their lower cost may result in a lower cost SteriPEN – current SteriPEN is $99 and the Adventurer is $129. An LED version might someday be retailed somewhere in the $49 range – maybe even less.

Finally, the LED technology opens the door to a variety of related water treatment possibilities – particularly in the area of flow-through treatment for hydration packs like the Cammelbak system. For a full discussion of the advantages of this go to thehttp://www.uspto.gov/patft/index.html

and look at patent aplication (pending patent) number 20060163126

Best,
Miles
CEO, CTO
Hydro-Photon, Inc.

Miles,

Not to hijack the thread too much but is the new 3.6oz adventurer Steri Pen available now? I’ve been wanting to get one since I saw it highlighted on this site’s review.

Thanks!
Chris

Hi Chris,
We expect to have the Adventurer available in stores in October/November.
Best,
Miles

Gentlemen:

Some of this is “over my head” — but it’s exciting all the same to read about the research done on LED-generated UVC — and the progress made to date.

I wouldn’t be surprised at all if Miles and others will have germicidal LED products to sell in the near future.

I remember the LED as well as LCD displays on calculators and watches back in the 70’s. The progress in both — and especially in large-format color LCD displays — is nothing short of amazing.

Thanks for your very informative posts!

I am the Executive Director of the International Ultraviolet Association (IUVA), and I find this discussion very interesting. There are a few concepts that one needs to understand about ultraviolet disinfection if LEDs are to be ultilized for that purpose.

1. The ‘disinfection’ of microorganisms by UVC is actually an ‘inactivation’, in that the UV photons are absprbed by DNA (or RNA in viruses). This causes a disruption of the structure, so that the microorganism can no longer replicate. Microorganisms that cannot replicate cannot cause disease.
2. The ‘log inactivation’ is proportional to the ‘UV dose’ (measured in unit of J per suqrae meter or millijoule per square cm). 1.0 log inactivation is 90% inactivation, 2.0 logs is 99%, etc. Generally, regulators specify a UV dose of 400 J/m2 or 40 mJ/cm2 to achieve at least 4 logs inactivation of virtually all microorganisms.
3. It does not matter what the source of the UV photons (a low pressure UV lamp or UVC LEDs), the total UV dose must match that in item 2 above.
4. Microorganisms have a UV sensitivity that roughly matches the absorption spectrum of DNA, which peaks at about 260 nm. Thus if UV LEDs are to be effective, they must emit in the range 250-270 nm for optimal performance.

I hope this helps. You can obtain information about the International Ultraviolet Association (a non-profit association) at http://www.iuva.org, or send me an email at jim.bolton@iuva.org

Hi Miles

> The idea described in the 6,579,495 patent is essentially a SteriPEN that uses LED’s in place of the CCFL lamp. In a few years, once the deep UV LED’s become a more mature technology (higher electro-optical efficiency, greater UV output power and low production cost) the advantages will include…

From Laser Focus World, May 2006:
“In the last few years, a group led by Asif Khan at the University of South Carolina (USC; Columbia, SC) has worked at pushing light-emitting-diode (LED) wavelengths well down into the 200 nm range. Now, a nearby company has commercialized the technology after making its own improvements, including advances in materials and device technology. While not a spinoff of the group at USC, Sensor Electronic Technology (SETI; also in Columbia, SC) does have Khan as one of its principals. The company’s deep-UV III-nitride-based LEDs are the only ones on the market, and range from 365 down to 250 nm in wavelength. The emitters were exhibited at this year’s Pittcon conference (March 12-17; Orlando, FL).
One intended application is UV water purification, which is currently done using the 254 nm emission line from expensive and power-hungry mercury-arc lamps. Banks of deep-UV LEDs would serve the same purpose; the SETI LED produces greater than 0.6 mW of light at an optical power density of 10 mW/cm2 from about 120 mW of electrical power. Unlike mercury-arc lamps, which must be optically filtered, the entire output of the LED is directly used for sterilization. “The wavelengths that can be provided by UV LEDs in the range of 260 to 268 nm are much more effective than the 254 nm wavelength produced by mercury lamps, because they target the peaks in DNA absorption spectra,” says Bilenko. “

Reckon your time-line of ‘a few years’ might even get accelerated a bit, maybe?

In the meantime, the new Steripen looks rather cute!

Cheers

Roger, Miles, and James, excellent info.

“10mW/cm2” – based upon James Bolton’s post 40 mJ/cm2 is required. Still some questions in my mind on this matter: for how long a period of time, or is this total exposure within some reasonable time frame; how long would that time frame be? So, at 10mW/cm2 are we looking at longer exposure times here for purification, are we talking just four seconds of exposure to get the 40mJ/cm2, or are we below some practical minimum threshold – if such a min. thres. even exists?

Can Bolton, Caffin, or Maiden (or anyone else for that matter) please clarify?

Many thanks.

> “10mW/cm2” – based upon James Bolton’s post 40 mJ/cm2 is required.

I confess I don’t fully understand this one, and I hope that JB or MM can explain further. 40 mJ per square centimetre is fine, but this is for an area. How deep is the water behind this area? One millimetre, or one metre?

It is of course possible that water is fairly transparent at this wavelength so that it is only the bugs which stop the UV. If this is the case, how nice! Please advise.

Perhaps JB or MM can also advise what plastics will stop UV in this region. We have Lexan water bottles, and I personally use PET botttles left over from fizzy mineral water. Will these plastics stop the UV from hitting me? Will these plastics degrade under the UV? All very important things.

Please educate us.

FWIW many PET bottles, especially hot fill like gatoraid, juice stuff, have a UV blocker in the resin to extend shelf lifes. I don’t know the wavelengths it blocks off the top of my head but I can find out. I presently work in a PET division of my company. I’ve been trying to find a product that lends itself to collapsing with minor modification for “cheap” platypus’s.

Roger, thanks for replying. I see that you have some of the same questions that i do, e.g., the time aspect of the proper level of UV-C dosing for a given area – of course Watts easily convert to Joules per second, hence my comment about 4seconds, but i think you understood that given your Physics background. i’m guessing area is used since any volume involves a third dimension, viz. distance, which now has effects on the intensity of the radiated light, so at the remotest area the proper dosing level must be achieved.

Hope JB or MM are able to respond soon with answers to the questions in both of our posts.

[BTW, great article you wrote on CO and combustion. Thanks for taking the time to research it and write it.]

Hi Roger, sorry for not responding sooner – I had thought this discussion had sort of ended – so I didn’t check back.

A couple of points: 1 joule = 1 watt second. So, for example, if an area of say, 1 square inch is illuminated by 1 watt of sunlight then for every second there is 1 joule of energy. And after 60 seconds the sunlight has provided 60 joules to that square inch…. and so on. Of course, if the sunlight within the square inch was – say 5 watts, then for every second there are 5 joules of energy incident on that area… and 300 joules in a minute and so on… One point here is that the same amount of joules (the same dose) can be applied in longer or shorter time periods, depending on UV power. Also, the same UV power applied over the same time period, but to – say half the volume of water (agitated water) will theoretically provide twice the dose.

Now, your observation about the dose in mJ (millijoules) being in a 2 dimensional area – sq.cm. rather than a 3 dimensional space is astute. Indeed, the joules per sq.cm. do vary a great deal depending on distance from the light source. The reason for using a 2 dimensional space when talking about UV doses in mJ has to do with the way that the microbiological testing for determining a given microbe’s UV sensitivity is done. This testing involves placing a small sample – say e.coli bacteria – in a specific area -normally a square centimeter. Then a collumated beam (light through a tube) of UV light with a crossectional area of 1 sq.cm. and a known intensity (say 1mW) is directed over the 1 sq.cm. e.coli sample, for a specific number of seconds – say 10 seconds. So, in this case the sq.cm. of e.coli has been exposed to 10mJ/sq.cm. of 254nm UV energy. Next, the sample is placed on a petri dish and grown overnight. E.coli colonies are counted and then compared to an unexposed control sample. From this comparrison the reduction of e.coli by exposure to 10mJ/sq.cm. can be determined – the convention is to express the microbial destruction as logs of reduction. So – if the control sample shows 1,000 colonies of e.coli and the treated sample shows 10 colonies then there has been a 99% reduction or a 2 log reduction.

Now, obviously this test method can be used on virtually any microbe and the dose time and energy can be varied – so, a researcher could do several test runs for a given microbe at a variety of doses – mJ/sq.cm. – in order to learn what dose can be expected to result in what log reduction.

Now, while you can do calculations and tests with radiometers to derive theoretical average energy density within a 3 dimensional volume, the way it is often done is with a microbe. If you take a specific volume of water (or rate of flow)spiked with a microbe with a known UV sensitivity and thoroughly illuminate the volume with a germicidal UV source for a specific amount of time – and then grow the control and treated samples and compare – you will know how many mJ/sq.cm. were applied to the volume treated within the dose period.

Hope this explanation makes some sense.

Finally, actually, Asif Khan’s group and the company you refer to – SET were our major LED suppliers during our Navy/DARPA work. The first deep UV LED’s we used, however, were from Shuji Nakamura and Steve denBaars lab at U.C. Santa Barbara.

You are correct that the UV LED technology is moving fast. HUGE progress has been made in efficiency and output over the past 3 years. The devices do however, need to improve a bit more before they will be practical.

Best regards,
Miles

Miles,

Good explanation on the 2D vs. 3D aspect. Your entire explanation is quite understandable.

Sounds like agitation is the key to making this linear dosing scheme work.

Many thanks for posting back with the info.

Hi Roger,
Sorry to not respond to your question about UV transmission through plastics sooner. As it turns out, very few plastics are even moderately transparent to deep UV light. Polycarbonates like Lexan are virtually opaque to UV. The only plastics I know of that can be fairly UV transmissive are fluoropolymers (Teflons) – but only if they are fairly thin – say .04″ or less.

In addition to poor transmission through most plastics, deep UV does not penetrate well through most other materials. Standard glass has very poor transmission. And of course, any material that is opaque to visible wavelengths will be opaque to UV as well.

Of the few materials that are transparent to deep UV, the most commonly used is quartz. Other materials that are quite UV transparent but not generally practical for common use are saphire and diamond.

Finally, another question that often is asked in conjunction with the plastic question is, “am I exposed to UV coming through the surface of the water?” Actually, the underside of the air/water interface is a very good reflector – next time you are in a swimming pool with a mask on swim down 10 feet or so and then look straight up at the surface of the water. You may be surprised to see a very nice circular mirrored surface centered over your head. This is the phenomenon behind what is known as “total internal reflectance.” So, anyway, due to reflectance, very little UV escapes. We have confirmed this many times with radiometer testing.

Best regards,
Miles

On a tanget, here’s a decent article on WalMart adapting LED technologies, among other things, in an effort to bolster environmentalism:

http://www.usatoday.com/money/industries/retail/2006-09-24-wal-mart-cover-usat_x.htm

They’ve been using LED’s to light store sinage for some time. They’re developing a LED lighting system for refrigerator case displays, at a cost of $30 million. It will supposedly cut WalMart’s lighting costs by 50%, which accounts for 30% of retailer’s energy costs.