Toward Computational Electron Ghost Imaging

Lucas Clark Burnette Author

07/27/2021 Added

18 Plays

Description

Working towards developing a method to effectively construct an array of nanotip fibers used for computational electron ghost imaging.

Searchable Transcript

Search:

[00:00:01.032]Hello, my name is Lucas Clark Burnette.
[00:00:03.490]I am from Minnesota State University in Mooorhead,
[00:00:06.778]and I've spent my summer research working
[00:00:08.360]toward computational electron ghost imaging.
[00:00:12.230]I figured it would be best to start with a more simple
[00:00:14.330]explanation of what ghost imaging is,
[00:00:16.230]and then work my way towards what my research was on.
[00:00:20.290]So very briefly, we have to discuss classical ghost imaging.
[00:00:24.340]First, a laser passes through a beam splitter.
[00:00:26.970]This results in two highly correlated beams of light.
[00:00:30.320]We have two detectors here,
[00:00:31.700]a bucket detector and a pixelated detector.
[00:00:34.460]The bucket detector can only tell when photons are fitted,
[00:00:37.460]so there is no spatial information being recorded.
[00:00:40.870]On the other hand,
[00:00:41.910]the pixelated detector gives us the spatial information
[00:00:44.710]since it can record where the photons sit.
[00:00:47.810]The sample objects sits in front of the bucket detector
[00:00:50.480]and the pixelated detector is programmed to only record
[00:00:53.230]a pixel when the light strikes both the bucket detector
[00:00:56.040]and itself.
[00:00:57.230]Since both the beams are correlated in position
[00:00:59.890]by sweeping across the whole sample,
[00:01:01.770]a silhouette of the sample object is created.
[00:01:06.440]Next, we have computational ghost imaging.
[00:01:09.250]In this set up, the beam splitter is replaced by a mirror.
[00:01:12.640]Instead of splitting into two separate beams,
[00:01:14.840]the laser shines on a spatial light modulator
[00:01:17.220]or SLM for short.
[00:01:19.220]The SLM encodes the beam with spatial information,
[00:01:21.950]so we don't have to rely on the pixilated detector
[00:01:24.390]to give us the spatial data.
[00:01:27.580]For computational electron ghost imaging,
[00:01:30.090]instead of a laser beam,
[00:01:31.370]we wanna try to use an electron beam.
[00:01:33.800]So naturally the first question that arises is,
[00:01:36.270]how do we get an electron source?
[00:01:38.760]Unfortunately, for me, my mentor, Dr. Botolan
[00:01:41.800]had already conducted research in the past,
[00:01:43.700]to manufacturer these nanotip fibers that you see here
[00:01:46.690]in the gray with the lasers pointing at them.
[00:01:51.260]In the past, Dr. Botolan and his colleagues
[00:01:53.470]use these fibers to observe photo field electron emission.
[00:01:57.350]They have a whole paper on the topic,
[00:01:58.730]so I'm not gonna get into the physical principles,
[00:02:01.330]but basically you can shine a laser into these fibers
[00:02:03.870]and they emit electrons from their tips.
[00:02:08.620]So,
[00:02:09.740]we know that we have a source of electrons
[00:02:11.380]to use for ghost imaging.
[00:02:13.000]Now, why would we want to do it?
[00:02:17.900]First off, the spatial resolution of classical ghost imaging
[00:02:21.040]is on the order of millimeters.
[00:02:22.890]By focusing an electron beam,
[00:02:24.370]the spatial resolution of about one nanometer
[00:02:26.890]can be achieved.
[00:02:28.400]This will allow us to image things that are many times
[00:02:30.660]smaller than what could normally be done
[00:02:32.350]with regular ghosts imaging.
[00:02:34.870]The second advantage of electron ghost imaging,
[00:02:37.630]which can actually be applied to other forms
[00:02:39.820]of ghost imaging too,
[00:02:41.080]is the option to replace imaging cameras with a simple
[00:02:43.700]bucket detector.
[00:02:46.350]Another huge advantage of electron ghost imaging,
[00:02:48.890]is the reduced exposure time.
[00:02:51.100]This could have a pretty big impact in the field
[00:02:53.070]of nanomedicine imaging.
[00:02:55.230]Right now, it is hard to image certain types
[00:02:57.270]of biological tissues of traditional electron microscopy,
[00:03:00.870]because the contrast is too low.
[00:03:03.560]So let's say we wanna improve the contrast,
[00:03:05.850]so we turn up the power on our electron microscope,
[00:03:08.160]a little bit.
[00:03:09.830]From experience,
[00:03:12.300]we found out that doing this will actually destroy
[00:03:14.950]any living cells that we were trying to look at.
[00:03:17.370]So if we had a way to take an image of the cell
[00:03:20.010]with shorter exposure time and still get the high contrast
[00:03:23.270]that we need, we would have opportunities
[00:03:26.000]to make advancements in our nanomedicine.
[00:03:30.560]So, that's our background on ghost imaging.
[00:03:33.360]Now let's get to where my research fits into the puzzle.
[00:03:37.180]Earlier, I mentioned the nanotip fibers
[00:03:39.250]were going to be used as the source of electrons.
[00:03:42.090]My job was to figure out how to create
[00:03:43.740]a nice array of these fibers.
[00:03:46.270]For this, a larger array of fibers would mean
[00:03:48.790]the resolution of our images would also be higher.
[00:03:51.950]I found the easiest way to stack these fibers
[00:03:53.970]was in this triangular pattern that you can see here.
[00:04:03.640]All right,
[00:04:04.500]so now I have my list of issues that I ran into
[00:04:07.070]and the solutions I came up with.
[00:04:09.410]And this list is not comprehensive,
[00:04:11.790]but I thought it would be
[00:04:13.664]good to hit some of the main points.
[00:04:17.460]So right off the bat,
[00:04:18.370]the first issue is that the fibers are very small
[00:04:21.090]and very fragile.
[00:04:22.430]The whole fiber is about 20 centimeters long
[00:04:25.130]with a diameter of 125 microns.
[00:04:28.470]And then the tips are about two millimeters long
[00:04:31.350]and tapered to a point of about 50 nanometers.
[00:04:35.840]So if you touch the tip, it's almost guaranteed to break.
[00:04:40.160]To lower the risk of breakage,
[00:04:41.470]I designed a fiber clamp that mounts onto a 3D micrometer.
[00:04:45.640]This allowed me to grab the fiber where it's less fragile,
[00:04:49.740]meaning back from the tip and make very fine adjustments
[00:04:52.980]that would have been impossible to do by hand.
[00:04:57.730]Issue number two, is the fact that each fiber
[00:04:59.890]needed a separate voltage in order to emit electrons.
[00:05:04.125]To avoid messing up the symmetry of the array
[00:05:07.020]a smaller 25 micron tungsten wire was adhered to the fiber
[00:05:11.250]with a conductive silver paste.
[00:05:14.320]Issue number three is the fact that the fibers need to be
[00:05:16.840]in a uniform array, ideally as close as possible.
[00:05:20.560]Since each fiber is coated in a thin film of gold
[00:05:23.070]at the tip, this means the fibers
[00:05:24.970]must be electrically insulated from each other.
[00:05:28.260]To maintain uniformity and adhesive with low expansion
[00:05:31.410]needed to be used.
[00:05:33.230]I found that coating the wire in thin film of nail polish
[00:05:36.280]worked as a good insulator.
[00:05:38.750]However, this leads me into my fourth issue,
[00:05:41.550]which is the fact that this array of fibers,
[00:05:43.330]needs to be put in a vacuum.
[00:05:45.700]Unfortunately, I never had time to test the nail polish
[00:05:49.020]in a vacuum, so this issue still remains unresolved.
[00:05:52.980]From what I've read, nail polish is an uncommon material
[00:05:56.610]to put a vacuum system.
[00:05:58.100]So I think this would be something to experimentally test
[00:06:01.580]in the future.
[00:06:03.910]And then this brings us to the final issue that I had,
[00:06:06.360]which was with the adhesive,
[00:06:08.160]possibly reacting with the cladding of the fiber,
[00:06:10.730]and then just messing up the electron emission.
[00:06:15.350]To test this, I basically did trial and error
[00:06:17.900]with a bunch of different adhesives,
[00:06:19.970]including the silver paste, tire seal,
[00:06:22.270]liquid electrical tape, and nail polish.
[00:06:24.930]To test if the power transmission was effective,
[00:06:27.090]I measured the power through a fiber
[00:06:28.860]before anything was changed.
[00:06:31.120]Then I put on whatever adhesive I was testing
[00:06:33.890]and gave it 24 hours for any possible reactions
[00:06:37.020]to take place and then measured power again, the next day.
[00:06:41.450]I kind of gave it away, I gave away the answer already,
[00:06:43.960]but the nail polish ended up being the best material,
[00:06:47.050]since it was low reactive and it had the right viscosity
[00:06:49.830]that made it easy to apply to the fibers.
[00:06:55.920]To conclude my research,
[00:06:57.130]I found that the best method to create an array
[00:06:59.930]was to assemble each fiber with its supply voltage
[00:07:02.570]and insulate each fiber before putting them together.
[00:07:06.490]On average, the diameter of the component increased from 125
[00:07:10.115]to 350 microns.
[00:07:12.300]This value is largely dependent on the amount of insulation
[00:07:15.110]each fiber had,
[00:07:16.480]which was pretty inconsistent due to the polish
[00:07:18.520]being applied by hand.
[00:07:20.600]Moving forward, I think it would be best to investigate
[00:07:23.070]how the array of fibers will hold up in a vacuum
[00:07:26.390]as well as how the nail polish out gases in the vacuum.
[00:07:29.860]Additionally, I think it could be beneficial
[00:07:32.090]to explore ways to revise the current method of insulating
[00:07:35.040]the fibers or exploring alternative methods altogether.
[00:07:42.340]Well, that about concludes my research for the summer,
[00:07:45.180]for my acknowledgements.
[00:07:46.320]I would like to give a special thanks to Sam Keramati,
[00:07:49.000]Rachel Corzine, Will Sippel
[00:07:51.030]and the rest of the physics department at UNL
[00:07:53.070]for their support and assistance this summer.

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.

Comments

0 Comments

Toward Computational Electron Ghost Imaging

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments