Particle Atomic Layer Deposition on Long-Afterglow Phosphors

Extending the Lifetime of a Zero-Energy Lighting Technology for Outdoor Applications

Session Notes:


Mark Losego

Associate Professor Georgia Institute of Technology 





Strontium aluminate (SrAl2O4) phosphors activated with Eu2+ and co-doped with RE3+ elements (RE: Nd, Dy, etc) are long-afterglow phosphors (>12 hrs of persistent luminescence) of significant commercial relevance. These phosphors have been widely used as luminescent additives in many commercial products including plastics and textiles. Today, these phosphors are of interest for zero-energy safety lighting both in residential markets and on roadways. However, these phosphors are prone to degradation in moist or humid conditions. This talk will discuss our work to use ALD to protect these phosphors from degradation in aggressive aqueous environments. Powders that are ALD coated with < 10 nm of Al2O3 or TiO2 show good resistance to degradation after being directly immersed in water for over 2 weeks. Achieving good conformality over all powder surfaces appears to be an important requirement to powder stability.

0:00:03   Hello. My name is Mark Losego. I'm a professor in the school of material science and engineering at
0:00:10   Georgia Tech. I'm very happy and pleased that the organizers of this conference invited me to give a
0:00:17   talk here at the first ever particle ale de Summit. I think the very exciting opportunity appreciate
0:00:25   that opportunity. Our lab at Georgia Tech does a lot of work, actually, and using vapor deposition
0:00:33   for modifying commodity sorts of products and materials on we do work some in modifying powders as
0:00:42   well particles on. That's where I'll talk about a bit today. Eyes Recent work we've done on using
0:00:49   atomic layer deposition to protect loan afterglow. Phosphors powder in powder form from degradation,
0:00:59   particularly from a quiz degradation in inhuman or water environments. This work was actually mostly
0:01:08   carried out by visiting students in our lab. Ercole Caracal glue. Hey, came, uh, from from, uh from
0:01:19   from Kirky as a visiting student PhD student just finished up actually recently his PhD there after
0:01:26   after returning. Uh, and he spent a lot of time looking and brought this technology of these long
0:01:31   Africa low phosphorus break based on strong illumined strong illuminate to our lab and we were
0:01:37   looking at just putting down a few nana meters of different metal oxides by OD and how this could
0:01:42   prevent degradation. I think these, um these long after that fosters are actually very exciting
0:01:50   materials. They've been around for quite a while, but still not in widespread use. Um, there they
0:01:59   haven't afterglow. Where if you excite these materials, they can actually continue to glow for not
0:02:05   just a few minutes or an hour, but up to 12 hours. You know, we're OK, Sarah mess. But if you know
0:02:12   we could make these a little better, you get even stronger Africa low for for longer periods of time.
0:02:20   But these these conglomerate and potentially be used for sort no energy lighting situations in a
0:02:26   variety of ah technologies. Eso, for example phosphor, er's and particularly as long afterglow
0:02:34   phosphors are being looked at, for example, of watches, of course, some some sort of long Africa
0:02:40   imposters. Aaron Aaron many watches already. Maybe not as long as 12 hours, but certainly thio
0:02:47   minutes to hours toys. We probably see safety sign ege again. A lot of times these are our powders
0:02:55   or then embedded into into a plastic to give this this glowing property and this idea of potentially
0:03:02   in, you know, zero energy lighting applications like roadway markings. Could we have markings on
0:03:08   roadways that would would soak up energy throughout the day and then re emit that energy is visible
0:03:15   light throughout the evening and in the night. But, uh, you know, one of the problems here, this is
0:03:25   actually this this was this was actually done in Europe. You're a few years ago where they did a
0:03:34   test case where they painted thes thes roads with this long after the Vosper. But what they found
0:03:40   was, you know, after a few weeks to a month, the Rose road stopped glowing after a while. And the
0:03:48   reason was that the markings faded out because of moisture. Degradation from rain and humidity
0:03:55   degraded those phosphors in that paint and the roads quick, glowing eso we're interested in. Other
0:04:03   people worked in this area as well, of how we can use protective coatings to these powders to
0:04:10   improve their long term stability. I think I just want to bring this up because I think it's
0:04:15   interesting. This is sort of ah, from a nice review article. Uh, where they look at a variety of
0:04:22   different Africa of phosphors and they plot here sort of how long, how long they're Africa is. So
0:04:29   you could see there's many here. They're out to, you know, 1000 minutes or so. So this is 10 10 plus
0:04:36   hours out here, and they're also colored here by the color of that Foss for. And what you'll see
0:04:46   here is that many of these that exist out Teoh 16 plus hours farthest out are based on sort of
0:04:55   strong iam. Illuminate, illuminate chemistries, aluminum silicate appear sort of strong, illuminate
0:05:02   base chemistries give you sort of these really long Africa low performance. And they're also very
0:05:08   bright in their mission. A relative that's a lot of phosphors. Uh, they fairly stable light
0:05:15   admission after the initial. So we're really Brian, that a stable medium toe, fairly bright
0:05:21   admission. But the challenge is that, you know, they're all very similar chemistries, and they all
0:05:26   have pretty poor waters to build it. And this is what this is what destroying illuminate lattice
0:05:34   looks like, uh and that's the host lattice for all of these phosphors. Uh, and so if you look at
0:05:41   this sort of a little bit more detail. What you find is aluminum oxides or sub structure with strong
0:05:49   IAM ions sort of in cages, uh, or channels inside of this this last material and so to create the
0:05:59   Foss for what you do is you added these activators code opens it basically substitute for the strong
0:06:04   IAM, so we can kind of substitute out something strong minds for these lantern. I'd species of the
0:06:10   your opium is what really creates the emission and the code open. In this case, we use dispose iam
0:06:17   uh, basically elongates the glow. Um, but really, they're just substitution for striking. And
0:06:27   actually, the reason that we have significant degradation here is because water can also get down
0:06:32   these channels and start to dissolve these ions and start to break apart on this network On break up
0:06:38   thing Illuminate Tetra,
0:06:45   I suddenly go over a little bit the experimental system that we used here It was a strong lumet dope
0:06:51   with their opium dispose iam, which started with use basic materials. We did a typical sort of solve
0:06:57   states ceramic synthesis with weighing and mixing grinding and then some thermal analysis and then
0:07:05   appropriate, he treatments to to both get the crystalline structure we want and then the right
0:07:12   oxidation chemistry to get the proper afterglow performance. And then, after making the powder and
0:07:20   confirming its structure and properties, we went ahead and did a lit decoding where we recycled a
0:07:29   number of times with both. Try meth, aluminum water and titanium tetrachloride. Water chemistry is a
0:07:34   separate. We did one with aluminum oxide coating one with a titanium oxide, a coating. A couple
0:07:41   fitness is here. Eso Here's just some basic characterization of the powder itself. Here we show X
0:07:49   ray diffraction patterns for a variety of different conditions. This is This is the pure strong
0:07:56   illuminate powder with a little bit of boron flux on. Then this is with your opium dope in and with
0:08:02   the European disposing doping. The point here is compared to the bottom. Here is the powder
0:08:07   diffraction foul. We see that all the patterns are pretty much the same. There's a little bit shift
0:08:12   speak when you when you dope these, but not a whole lot changes intensities. And then really, if we
0:08:19   compare this blue curve, here's our final Foss for powder and then with different ale de chemistries
0:08:27   coated on either 10 years aluminum oxide or 12 Amber's team oxide. We don't see with those code. You
0:08:33   don't really see any evidence of additional peaks or changes in the peaks were not really affecting.
0:08:39   These are low temperature. Process is Did about 121 150 degree Celsius
0:08:46   way Don't see those Nulty. It changed too to the since to the crystal structure on. Then you know
0:08:58   the first order We just a qualitative check of glow. Here is the the untried of the untreated
0:09:06   uncoated green powder O. R s a two pounder starting aluminum powder with opens we see has this nice
0:09:14   green phosphorescence on. Then if we add either aluminum oxide detain box I coating to it the glow
0:09:24   really doesn't change significantly is visually to the eye. And if we go a little more quantitative
0:09:32   and we look at the emission Spectra, uh, you know, the black line here is the uncoated blue in the
0:09:38   red or the coded. Not again, not much changed a little shift. But you know, just a few few
0:09:43   nanometers really no significant changes in the performance of these materials. What's code and
0:09:53   gradually the micro structure. You know, we're putting out informal coming down. This is only about
0:09:58   10 nanometers of coding, and we don't really see any change in the general morphology and surface
0:10:05   structure of these powders after that code.
0:10:13   Let's look now if we take these powders and we immersed, um, disperse them in water for some period
0:10:18   of time. What happens to them on? I'll start with sort of the biggest result up front, which is, you
0:10:27   know, if we look at these pounders here, this is before any water immersion these air, the actual
0:10:33   powders just in visible light. And then if we excite them and look at them in the dark, we could see
0:10:38   their clothes. We just looked at. But now, if we immerse them for two days in water, just completely
0:10:44   dispersed the pounder and agitated in water for two days, we clearly see the uncoated pounder begins
0:10:51   the glow, Marie, much more. Not a fine looking pounder anymore on. And when we tried to get it to
0:10:59   glow, we really don't see any evidence of phosphorescence anymore. In contrast, these powders that
0:11:05   were coated which 10 nanometers L d coatings Ah, I really don't appear much different days. Keep
0:11:14   their fine. Michael. Structural appears to be a fine micro structure, and they glow with same globe.
0:11:20   They had prior to, uh uh, um prior to immerse it immersion in water. And so this is qualitative
0:11:31   course. But if we look at this little more quantitatively, we could track the phosphorescent
0:11:37   intensity. Uh uh uh, for different amounts of emergent times that we're looking zero hours of a
0:11:46   murder Times country means here three hours, 24 hours in two days, and we see by two days we
0:11:53   basically lost all of our phosphorescence in these uncoated powders. But if we had 10 nanometer 12
0:12:00   million Latino to we see, we only lose maybe 10% or so over phosphorescence on even with a with a
0:12:08   fine and immune aluminum oxide coating, we still maintain, you know, better than 90% of our
0:12:14   phosphorescence way. Possibly don't even need a stick of a coded. And this is just the raw spectra
0:12:22   to show that this is again the uncoated foster here again, Dramatic decrease in the phosphorescence
0:12:31   with time eventually a shift. But by this point, it's so degraded, it's really somewhat meaningless.
0:12:36   What's happening to it. We've seen with these 10 Nana meter ish ale de coatings a slight decrease
0:12:44   with these with each with immersion time. But, you know, it kind of starts to plateau, maybe 90% 85
0:12:53   90% of the of the total phosphorescence on then, uh and I think most importantly, is that if we can
0:13:01   reduce the film thickness, that's that's really important because it reduced the costs for commodity
0:13:08   products like this. And so we showed that even five Nanometers Way didn't goeth into the finding in
0:13:13   years. But did you even find the abuses? Is fairly protective? Not really that much different, we
0:13:18   think. Actually, based on some evidence, we have cursory evidence at this point that we're not using
0:13:28   any real rigorous agitation of these pounders. And so we think of the degradation we're seeing here
0:13:33   is really a result of just not perfect informal coating of powdered because the powders air not
0:13:41   being agitated or coding coding small quantities and a basic flow to reactor on. So we think if we
0:13:48   agitated these mawr, we'd get even even better performance from from these coatings
0:13:56   just a little bit about tracking the degradation of the uncoated powders. These air sort of the
0:14:03   different reaction processes that we expect occur when destroying, illuminate, reacts with with
0:14:11   water. Basically, it takes the strong iam ions out of out of the cages. Andi, potentially, That
0:14:19   leaves aluminum hydroxide. Basically, aluminum oxide gets converted toe aluminum hydroxide that
0:14:24   those may combine to form destroying aluminum hydroxide phase. Um, and there's some other
0:14:30   equilibrium that could occur. The thing I want to highlight for you here is that all of these
0:14:35   reactions are mostly reactions result in potential production of hydroxy all, uh, I owns ah
0:14:44   hydroxide ions in solution, and the hydroxide is, of course, we're gonna raise the pH. And so one
0:14:50   way that you contract the degradation of these of these of these phosphors is to look for a rise in
0:14:58   ph of the dispersion. Uh, while while the powders are immersed in water, that's in fact, what we see
0:15:06   this is immersion time in water here, the pH that we're tracking. And you can see the black data
0:15:14   points here for the uncoated powder. And ph rises quite rapidly within a day, two days where we've
0:15:22   saturated uh, no ph of 13 13. So versus thes Cody and powder samples show a much more modest rise in
0:15:34   pH. Probably some surface chemistry that's occurring. We're getting some some hydroxide creation,
0:15:42   but it's plateaus and appears to be pretty stable. You know, at this point,
0:15:51   eso we contract this a number of other ways as well. We could see in the X ray diffraction patterns
0:15:57   that you know, here, here, here's the this is again the uncoated pounders. Initially, this is the
0:16:02   pattern saw, of course, on destroying illuminate phase is labeled with these these purple pee
0:16:10   circles. But with time in water, we start to see these other reflections appear. That's trying
0:16:17   aluminum hydroxide, which is what we noted that last slide. This is one of the possible reaction
0:16:23   products, and we start to see these these peaks starting to appear, in fact, by 48 hours. That's
0:16:29   pretty much all that's in in the material to balance out the slick geometry because this is a bit
0:16:35   deficient aluminum. Well, we think the aluminum eyes is left. Is this aluminum hydroxide phase
0:16:42   primarily on then sort of the uninteresting case, but the But the important case is that when we
0:16:50   able to coat these powders, here's the original pattern and we don't see change many hours of
0:16:58   immersion in water. No change to what we could detect by x ray diffraction on. Then you just to show.
0:17:10   Obviously, these reactions were basically hydrating. The water's getting in there and hydrating the
0:17:14   ceramic so we can track how much water is in there by basically heating taking. You use immersed
0:17:20   samples and heating them up to 900 seeing driving off all the water, even the a draw Oxley thing. Go
0:17:26   back to oxides and we could figure out how much water really reacted with the powers. And when we do
0:17:33   that, you know the uncoated we see after 48 hours of immersion, 25 weight percent, almost of the
0:17:40   powder is really water reacting with the powder. Whereas in these coated films were we've only, you
0:17:47   know, reactive, maybe three here with the two were below 1.5 way percent water. And in fact, you
0:17:57   know there's some water on a surface to begin with, So even before emerging, you're seeing half away
0:18:03   percent. So we're really talking about maybe one way percent of water, uh, reacting with the city's
0:18:08   with these powders that are coded warning 123 kind of weight percent on we see in the micro
0:18:16   structures. Well, of course, in the powders, you could see that photographs that showed Start, But
0:18:22   you can see it here. And this is optical my cross to be and then electron my cross to be on bright.
0:18:28   This is for again the uncoated sample. This is before immersion in water is after three hours of
0:18:36   emerging, you start to already see a fairly significant difference in what the powder looks like.
0:18:43   And C powder here. What's things were happening here degrading on. And then after two days, you're
0:18:52   seeing significant changes in the powder and the morphology of the powder particles on. In fact,
0:18:59   when we looked at electron, my cross be here, we could start to see these cubic species which relate
0:19:06   to the strong IAM aluminum hydroxide actually is a cubic morphology. And if you look at it, you can
0:19:12   actually see these cubes of more strongly, um, than aluminum in them, which is which is what slick
0:19:19   geometry is for that phase and then these other sort of less, more amorphous, sort of looking
0:19:25   structures. Sometimes they have someone of excitable shape. This aluminum hydroxide phase much or
0:19:32   there may be some other, but they're more rich. An aluminum hydroxide basically so we could track
0:19:38   the story geometry, actually, in the manuscript, which was listed the beginning of the stock.
0:19:43   Actually, it's in its published in the journal American Ceramic Society earlier this year. You can
0:19:49   fly. It's in the details here, but we were able to kind of pull off the street geometries and match
0:19:53   them to what we're seeing on the X ray diffraction patterns on. And then again, sort of the the UN's
0:20:01   exciting result. But really, truly, the exciting result is that when we coat thes the powder really
0:20:08   don't see any change as prepared toe after 48 hours and emergent, there's really no change in the
0:20:14   more falling to hear that powders or state system, which is what we've seen with all the other
0:20:18   characterization techniques that we've shown eso. You know, Uh, the take away here is that, you know,
0:20:27   a few nanometers, 10 even five centimeters of oven oxide coating on these powders that could be a
0:20:36   really, really great barrier to water water degradation of these of these ceramic powders, even
0:20:46   though they're not, you know, not smooth or not. They're not regular shape. They're just powders
0:20:52   that we milled with 30 sorts of ceramic processing tools. We can still apply thes vapor phase ale de
0:21:01   techniques to create protection the layers to these to these commodity. We might consider mawr,
0:21:08   commodity materials and traditional microelectronics so that I'll end with this one slide. Talk a
0:21:15   little about my perspective on commoditize ing Atomic layer. Deposition may be particle ale de but
0:21:23   commoditize ING LD In general, I think it's something of interest to me. The research that we do. Um,
0:21:30   I think you know, I my students think about this quite a bit. How it's great if weaken find some
0:21:37   interesting signs. But can we make this a manufacturer herbal process? And I think is speaking to
0:21:43   the academics in the audience. I think we have an opportunity to develop fundamental scientific
0:21:49   principles and approaches that help to support me faction, and they're different maybe, than some of
0:21:53   the some of the fundamental signs. So we often try to do, I think is really interesting science
0:21:59   behind ah manufacturing, applying vapor phase processing tool to sort of ah high throughput. High
0:22:10   volume. Uh, how do we make inexpensive? How do you make it safe? All these things require way could
0:22:17   certainly do from a from a trial and error kind of approach. You could. You could try it that way.
0:22:22   But there think there's scientific questions and and fundamental science that can be applied and
0:22:28   developed and researched to support these manufacturing things. Manufacturing efforts. Uh, I list a
0:22:37   few here. I'm sure there's others but us, particularly particles of powder. I mean, how do we
0:22:43   promote con formality in the science of how do we agitate thes pounders and even other sorts of
0:22:50   products? You can think about fabrics and films and other things. But how do we How do we How do we
0:22:57   agitate them so we could get some formal coatings? How do we flow the gases? And there's a lot of
0:23:02   science. They're not just not just not just engineering optimization, but there's potential signs
0:23:08   there that we could develop. And then and then you lend to our our manufacturing colleagues or
0:23:15   industry colleagues to use Teoh help to help propel the commoditization of ale on. Then you think
0:23:24   about how can we develop simple processes or or different processes. Maybe we think about a lady to
0:23:30   play o. D from microelectronics in one way. But are there other ways to design or dozing schemes and
0:23:36   things to make them simpler and make them more impactful? Can we get more deposition, her cycle or
0:23:43   other ways designed processes that give us the same sort of end result but accelerated, simplify the
0:23:52   overall process or make it more inexpensive? We talk a lot about reducing the number of cycles, but
0:23:59   I think there's also a lot of science around precursor designer selection We could think about doing
0:24:06   obviously consumption of precursor in process. How do we minimize active use of excess? Precursor?
0:24:14   All these things have, I think, a rich opportunity for developing scientific principles, asking
0:24:21   fundamental scientific questions to support the manufacture, sale de manufacturing and then finally,
0:24:30   uh, safety on. And there's a lot of things and safety, but you can think of just even simple simply
0:24:36   heat heat management, um, especially powders. There's a lot of pence potential. There's lost service
0:24:45   areas. There's a lot of potential reactions happening. You're doing a lot more heat than you
0:24:48   otherwise might be generating for a typical microelectronic ale due process. So how do we manage
0:24:54   heat flow as well as mass flow in all of these things? And so there's again rich scientific
0:25:00   questions to be explored there in academia that help support our industry colleagues. With that,
0:25:09   I'll summarize with with this sort of slide here, showing the country pounder than our ale de coded
0:25:18   powders and how they survive many hours of immersion versus versus no treatment on. I'll thank the
0:25:29   lab A to George Attack. 1,000,000 students again. Erico here in the center was visiting student. It
0:25:37   really did most of this work that shared with you today and finally acknowledged funding, especially
0:25:44   from a tuba talk from Turkey Research Institute for allowing their goal to come and spend about 11
0:25:54   months with us to do this. Research s so with that, I'll conclude, and I'd be happy to take any
0:26:03   questions. Thanks