Mechanical Stability Assessment of Silicon Microcapsules
Christopher Ikpefua - Parallel I
Author
09/26/2024
Added
3
Plays
Description
Student’s name: Christopher Ikpefua
Home Institution: University of Maryland College Park
NNCI Site: SENIC @ Georgia Tech
REU Principal Investigator: Dr. Michael Filler-Nanotechnology
REU Mentor: Calib Lanier-Nanotechnology
Abstract: This project, a collaboration between the Filler and Vogel Labs at Georgia Tech, aims to revolutionize microelectronics manufacturing by developing a scalable, bottom-up approach to create high-performance nanoelectronic elements. The method utilizes silicon microcapsules produced through a double emulsification process, which serve as growth substrates for nanowires via a volumetric vapor-liquid-solid (VLS) mechanism. Central to this innovative approach is the use of a Powder Coat 300 system for tumbling the microcapsules and introducing precursor gasses to initiate nanowire growth. The success of this method hinges on the mechanical stability of the silicon microcapsules during the tumbling process. By assessing the microcapsules' ability to withstand these mechanical stresses, this research aims to validate and optimize the scalability and efficiency of this novel nanoelectronics manufacturing approach. The findings will contribute to the broader goal of making high-performance microelectronics production more accessible and cost-effective.
Searchable Transcript
Toggle between list and paragraph view.
- [00:00:00.560]All right, good morning. My name is Christopher Ucapua. I'm coming from the University of Maryland
- [00:00:06.240]College Park as a physics undergraduate, and this summer I worked with my mentor Caleb Lanier
- [00:00:11.920]and my PI Dr. Michael Filler at Georgia Tech on the mechanical stability assessment
- [00:00:17.040]of silicon mine capsules for scalable man-wired synthesis.
- [00:00:21.040]So firstly, a little bit of history on the semiconductor industry.
- [00:00:27.840]In 1947, Bell Labs creates the first transistor. This is going to be able to amplify and switch
- [00:00:36.000]electronic signals and lay the foundation for all electronics to come.
- [00:00:39.520]In 1958, mentor Jack Kilby creates the first integrated circuit. Now this is akin to a
- [00:00:48.400]single note going to a symphony but in electronics terms. Then in 1971,
- [00:00:55.680]Intel introduces the first commercial electronic circuit. This is the first commercial electronic
- [00:00:57.680]circuit. This is the first commercial microprocessor. In the following decades,
- [00:01:00.560]semiconductor advancements drive the rise of PCs, phones, as well as the internet,
- [00:01:05.360]culminating into what we have now, which means chat GTB and huge games like Fortnite.
- [00:01:12.000]The current opportunity cost and the limitations in this technology lies in the cost of miniaturization.
- [00:01:19.200]When being compared to other widely used goods like gasoline and aspirin,
- [00:01:24.400]transistors are extremely expensive
- [00:01:27.520]to fabricate, and then also the highly complex planar processing restricts productivity.
- [00:01:33.920]So before looking at how to combat this issue, we're going to look at a
- [00:01:41.040]method of nanowire synthesis called the vapor liquid solvent mechanism.
- [00:01:45.680]So it involves a catalyst, in this case gold, that forms a liquid alloy with the semiconductor
- [00:01:51.760]material. The semiconductor material is supplied in a vapor form which is dissolved into the alloy,
- [00:01:57.360]and then once this alloy becomes supersaturated, the semiconductor precipitates out a volatile
- [00:02:02.720]nanowire which can then be built in different regions to create transducers and answer this
- [00:02:09.280]problem. Now the way the Filler Labs is going to build upon this is through the geoprocess or
- [00:02:15.920]encapsulated VLS synthesis. So as opposed to planar VLS which is restricted to 2D scaling,
- [00:02:27.200]it's going to utilize volumetric scaling. So as you can see in the second image, the gold catalyst
- [00:02:35.280]particles are scattered inside of these silicon microcapsules which are created through a double
- [00:02:40.240]emulsification process. These microcapsules are then milled or tumbled in a chemical vapor
- [00:02:46.720]deposition chamber at 100 rpm and introduced to different doping gases in which the nanowire
- [00:02:53.600]synthesis begins.
- [00:02:57.040]These are grown nanowires and once these nanowires are actually grown in these microcapsules
- [00:03:02.720]that's when we start calling them microgeodes just for clarity.
- [00:03:08.880]So this is where the microcapsules will be tumbled in our PowerCore 300 system and here
- [00:03:14.700]is a quick video of that in action. It's not working so go ahead.
- [00:03:20.960]So and that's here's where I come in so my job was to construct a miniature tumbler to
- [00:03:26.880]assess the mechanical stability of these microcapsules and emulate the tumbling conditions found
- [00:03:32.880]in the CVD and it was also to develop and automate image analysis techniques to extract
- [00:03:38.220]data such as roundness, a flat elongated ratio, and form index. So on the left side is a picture
- [00:03:44.320]of my miniature tumbler and the right is ImageJ, the application I used for measuring the particles.
- [00:03:53.880]Here's a short clip of that in action.
- [00:03:56.720]So in order to actively compare the motion of the powder in my vial to the motion of
- [00:04:03.200]the powder on the side of the chemical vapor deposition chamber, I'm going to be utilizing
- [00:04:08.240]the Frou number. So this is given by angular speed squared times radius over gravity, and
- [00:04:14.540]this is going to describe the ratio of inertial forces to gravitational forces in a rotating
- [00:04:19.480]system. So for our Cpd chamber, the Frou number falls
- [00:04:26.560]in this middle cascading/tumbling motion section, so all of our measurements will be kind of,
- [00:04:35.500]our most important measurements I should say, will be in this range. So at 50 rpm for the
- [00:04:40.900]vial, we, that video is also not playing, but for 50 rpm in the vial, we observe a cascading
- [00:04:49.920]and tumbling motion. For 100 rpm, we observe a crushing and cataracting motion, and for
- [00:04:56.400]150 rpm, we observe a cataract and crushing motion.
- [00:05:03.960]And then for 200 rpm, we observe a crushing and then centrifuging motion.
- [00:05:10.910]So this is going to be probably the least helpful because you're not going to be centrifuging
- [00:05:15.730]stuff in a chemical vapor desiccant chamber.
- [00:05:19.670]And then here's an overview of my experiments.
- [00:05:21.970]I took a look at four different rpms, 200 rpm, I didn't have time to image unfortunately.
- [00:05:28.710]This is circularity over time and circularity is just going to be a measure of how close
- [00:05:35.770]a microcapsule is to a perfect circle, one being a perfect circle and zero being a completely
- [00:05:41.950]deformed fragment.
- [00:05:42.950]It's given by 4 pi times area over per meter squared making it a dimensionless quantity
- [00:05:48.930]and the circularity is going to evaluate the milling process or how the milling process
- [00:05:54.830]affects the shape of the microcapsules and it's also going to indicate whether or not
- [00:05:58.670]the process is causing deformation, breakage, or maintaining the original shape of the capsules.
- [00:06:05.170]So moving through the images, as you can see in the control we can see a lot of intact
- [00:06:10.390]capsules which is what you would expect to see, at 50 rpm more or less the same thing
- [00:06:15.590]but a slight bit of breakage, at 100 rpm we're starting to see a little bit more breakage
- [00:06:22.630]but not enough to call the sample completely broken and then at 150 rpm we start seeing
- [00:06:28.630]some noticeable breakage within the samples.
- [00:06:34.350]So this is the image analysis process, this is how I obtained the data that I just showed
- [00:06:40.550]you for the graphs, so I'm going to start out with an original SEM image and this is
- [00:06:46.190]the image after thresholding, fiberization, using the despeckling tool which is a noise
- [00:06:52.370]tool, erosion and manual filling.
- [00:06:58.590]We draw from ImageJ after using the analyze particles tool, as you can see after we overlap
- [00:07:04.970]it, it does a pretty good job of capturing what we believe are particles and leaves out
- [00:07:09.390]noise, things like that, so it does make some mistakes, like these little red dots here
- [00:07:17.350]are counted particles, but they're super, super small so they're extremely easy to filter
- [00:07:23.610]out once I see them in Excel.
- [00:07:28.550]So, in conclusion, although these circularity values do show a negative trend and slightly
- [00:07:32.750]vary, the microcapsules are largely intact and undamaged after tumbling up to 100 RPM.
- [00:07:38.790]This fact stands at even 150 RPM, indicating these microcapsules are highly mechanically
- [00:07:43.990]stable within our operating range.
- [00:07:47.570]This tells us that the double emulsification process used to create the microcapsules is
- [00:07:52.450]successful in making structurally sound capsules, and then this also indicates that the microcapsules
- [00:07:58.510]remain structurally stable when they are actually undergoing nanowire synthesis.
- [00:08:04.050]So some side projects and future steps, I also ran some ball milling experiments with
- [00:08:09.970]varying sizes of steel ball bearings to reliably break open and harvest these grown nanowires.
- [00:08:16.010]These experiments were successful, but the ball milling was an extremely destructive
- [00:08:20.670]method, so additional experimentation is needed with microdios to assess the behavior of nanowires.
- [00:08:28.470]I also created an alternative bioweft vial with interior obstructions like fins inside
- [00:08:38.290]of it to further assess the stability, and I also began the process on automating image
- [00:08:43.990]analysis techniques.
- [00:08:44.990]I didn't get through finishing that because this is essentially creating a neural network
- [00:08:51.010]with the images, that was pretty difficult.
- [00:08:54.970]And these are our acknowledgements, thank you for your time, are there any questions?
- [00:08:58.430]For your circularity graph, is there a reason that all the trials didn't start at the same
- [00:09:10.050]circularity?
- [00:09:11.050]Okay, yeah, so the question was for the circularity graph, is there a reason the trials did not
- [00:09:18.090]start at the same circularity?
- [00:09:19.970]So for this, for these experiments, my mentor advised me to use the same bio for all the
- [00:09:26.390]experiments and just increase the articulation.
- [00:09:28.390]So I was able to use my 200 RPM data to get a more visible gradient of what's happening.
- [00:09:33.910]So it would have been really interesting to get my 200 RPM data because I feel like we
- [00:09:39.690]probably would have seen complete destruction somewhere through there.
- [00:09:44.170]And even still, there are some inconsistencies like right here, for example, this is kind
- [00:09:50.850]of just a, this is a product of an image analysis techniques I was using is 100% not a perfect
- [00:09:58.350]process.
- [00:09:59.350]So thank you for your question.
- [00:10:03.490]Anything else?
- [00:10:04.490]Well, thank you for your time.
- [00:10:06.530]Thank you.
The screen size you are trying to search captions on is too small!
You can always jump over to MediaHub and check it out there.
Log in to post comments
Embed
Copy the following code into your page
HTML
<div style="padding-top: 56.25%; overflow: hidden; position:relative; -webkit-box-flex: 1; flex-grow: 1;"> <iframe style="bottom: 0; left: 0; position: absolute; right: 0; top: 0; border: 0; height: 100%; width: 100%;" src="https://mediahub.unl.edu/media/23113?format=iframe&autoplay=0" title="Video Player: Mechanical Stability Assessment of Silicon Microcapsules" allowfullscreen ></iframe> </div>
Comments
0 Comments