Ferroelectric domain studies of free-standing PbZr0.2Ti0.8O3 (PZT) membranes

Alyssa Simpson Author

07/27/2021 Added

24 Plays

Description

Free-standing ferroelectric oxide membranes are promising materials for building flexible, wearable electronics. For exploiting ferroelectric domain structures to represent the binary logic for information storage, it requires fundamental understanding of the static configuration and dynamic response of the ferroelectric domain walls (DW).

Searchable Transcript

Search:

[00:00:00.000]Hello, my name is Alyssa Simpson, a junior
[00:00:03.423]Physics and Math major at UNL conducting
[00:00:07.613]research on ferroelectric domain studies
[00:00:10.306]of free-standing PZT membranes. In this
[00:00:13.086]talk, I will give an introduction to the
[00:00:16.006]material that we are using and its
[00:00:19.348]importance. I will also give insight into
[00:00:22.961]how we obtained free-standing membranes
[00:00:26.093]and also how we used piezoresponse force
[00:00:28.957]microscopy (PFM) in order to study the
[00:00:32.181]domain walls of our material. Afterwards,
[00:00:35.904]I will give a general summary of our
[00:00:38.023]results and what it means for future
[00:00:40.007]research. To start here is a crystal model
[00:00:43.572]of our material PZT, otherwise known as
[00:00:46.083]Lead Zirconate Titanate. It has positive
[00:00:49.331]lead ions, negative oxygen ions, and a
[00:00:54.541]positive zirconium slash titanium ion that
[00:00:58.541]would form polarized dipoles with the
[00:01:00.912]negative oxygen ion, depending on the
[00:01:03.527]direction of the electric field. These
[00:01:07.397]polarized dipoles, when oriented in the
[00:01:09.831]same direction would form a ferroelectric
[00:01:12.086]domain. And their net dipole moment would
[00:01:15.214]give us the polarization direction. This
[00:01:18.522]here is a graph of switching hysteresis,
[00:01:21.639]which shows us that polarization graphed
[00:01:23.855]against the electric field. This area is
[00:01:26.028]known as the coercive field of the sample
[00:01:29.141]and if we want to polarize the sample,
[00:01:31.603]we'd have to exceed the coercive voltages
[00:01:34.467]that are on either side. The up and down
[00:01:37.873]directions of the polarization are a
[00:01:39.936]binary state. In other words, we can
[00:01:42.042]define either as 0 or 1, in order to store
[00:01:45.311]bit information, like in your computers
[00:01:47.767]and phones. What are we going to do with
[00:01:50.610]this information? Why is it important?
[00:01:53.295]It's because the flexibility of
[00:01:55.427]free-standing PZT membranes can lead to
[00:01:58.315]wearable electronics, which are the future
[00:02:01.425]of modern technology. And we're
[00:02:04.804]specifically studying the domain wall
[00:02:06.807]roughness as the smoother it is the more
[00:02:09.924]information can be stored in a small area.
[00:02:13.444]And we're studying the creep,as that gives
[00:02:16.748]us insight on how quickly our device can
[00:02:21.208]operate at a given time. To fabricate
[00:02:24.816]these membranes, we used off-axis
[00:02:27.571]radio frequency magnetron sputtering.
[00:02:31.422]Within which you could witness a plasma
[00:02:34.021]ball depositing the PZT on our water-
[00:02:38.021]sacrificial Strontium Aluminate buffered
[00:02:41.323]Lanthanum Aluminate (SAO buffered LAO)
[00:02:44.907]base layer. Looking at our atomic force
[00:02:48.811]microscopy image here (AFM), we can see
[00:02:52.518]that our sample was extremely smooth as it
[00:02:55.870]has a roughness value (RMS) of three
[00:02:58.607]armstrong. This is good. We can also see
[00:03:02.330]from this scan of the x-ray diffraction
[00:03:05.662]that our sample was single-crystalline
[00:03:07.574]with no impurity phase, all the peaks are
[00:03:10.608]in the same direction and the only peaks
[00:03:13.360]are at the crystalline structures. In
[00:03:17.170]order to obtain free-standing membranes
[00:03:20.004]we soaked our entire sample in de-ionized
[00:03:22.641]water, and after two hours the water-
[00:03:25.451]sacrificial SAO had dissolved, and we were
[00:03:29.065]able to use a gel-film to pick up the PZT
[00:03:31.608]and deposit it on other conductive
[00:03:33.987]substrates. In this case we used a gold
[00:03:37.591]substrate and a conductive oxide (LSMO).
[00:03:42.257]These are the optical images of our PZT on
[00:03:46.637]the two substrates and from their AFM
[00:03:50.596]images you can see that the roughness
[00:03:53.330]values have increased, from three to eight
[00:03:59.170]. This suggests that the free-standing
[00:04:01.295]membranes became rougher after transfer.
[00:04:07.767]In order to begin the PFM characterization
[00:04:11.557]we first started by graphing their
[00:04:13.681]switching hysteresis loops. So, for the
[00:04:16.936]PZT on the gold specifically we found the
[00:04:20.533]coercive values to be positive point eight
[00:04:26.533]volts for the polarization upwards and
[00:04:29.613]negative one point four volts for
[00:04:31.607]polarization downwards. These are the two
[00:04:35.294]voltages that we would need to exceed, in
[00:04:38.697]order to polarize our sample in either
[00:04:42.561]direction. As can be seen from the square
[00:04:45.693]domains here, we were successful in this
[00:04:49.237]polarization. As our polarization up state
[00:04:52.983]p-up is a very clearly contrasting color,
[00:04:59.345]from our polarization down, p-down, state
[00:05:02.304]in the middle. In other words, this is
[00:05:06.496]from the phase change that happened due
[00:05:10.830]to the polarization shift. From these
[00:05:13.857]studies we were able to conclude that our
[00:05:17.287]sample had non-volatile bistable states
[00:05:20.942]and robust switching. Now using the PFM
[00:05:24.943]we were able to scan stripe domains and
[00:05:28.684]use MATLAB in order to single the ones we
[00:05:33.258]wanted to use for this calculation out.
[00:05:36.760]Afterwards we could graph their
[00:05:38.787]correlation functions and use the
[00:05:42.607]linearly fit line to find our domain wall
[00:05:47.055]roughness value, as the slope of that line
[00:05:50.579]was double that of our exponent. And we
[00:05:54.975]ended up with point two nine plus or minus
[00:05:57.826]point zero four, this is in range of our
[00:06:01.228]theoretical value for 2D Random Bond
[00:06:04.456]disorder, as that value was point three
[00:06:07.041]one. This means that our domain wall
[00:06:10.036]roughness was dominated by 2D Random Bond
[00:06:12.983]disorder. In order to study the creep
[00:06:16.075]behavior, we pulsed positive ten volts on
[00:06:19.911]an area scanned with negative five volts
[00:06:23.296]and held these pulses for differing
[00:06:25.033]lengths of time. Then, using MATLAB we
[00:06:27.670]were able to extract the radii of each
[00:06:31.279]sample and graphed that against the time
[00:06:34.215]it was pulsed for. Then using the relation
[00:06:37.207]between change in radius and change in
[00:06:39.231]time, we were able to calculate the
[00:06:41.238]velocity of each dot's domain walls, and
[00:06:44.652]graph that against the inversed electric
[00:06:47.286]field at that point. Now due to the semi-
[00:06:49.715]log scale we ended up with a fit line that
[00:06:52.949]was linear and this gave us a creep
[00:06:56.042]exponent of one. Which would disagree with
[00:06:58.591]2D Random Bond disorder that we concluded
[00:07:02.188]earlier. This may be due to the
[00:07:05.835]flexoelectric effect, however, we would
[00:07:09.174]have more studies to conduct.
[00:07:12.774]In conclusion, we have successfully
[00:07:15.238]fabricated single-crystalline
[00:07:18.123]free-standing PZT membranes and found that
[00:07:20.980]the domain wall roughness exponent agreed
[00:07:23.389]with 2D Random Bond disorder, however, the
[00:07:25.672]creep exponent did not, so we'd have to
[00:07:28.075]study that further. Otherwise, this study
[00:07:30.813]provides important information for
[00:07:33.973]information storage applications and their
[00:07:36.894]relatives. I would like to thank Dr. Xia
[00:07:39.801]Hong, my advisor, Qiuchen Wu and Kun Wang,
[00:07:45.231]my graduate student mentors, the National
[00:07:48.457]Science Foundation, Nebraska MRSEC, and
[00:07:51.767]the UNL McNair Scholars Program for
[00:07:54.186]providing funding for this project.
[00:07:57.263]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.

Comments

0 Comments

Ferroelectric domain studies of free-standing PbZr0.2Ti0.8O3 (PZT) membranes

Description

Searchable Transcript

Comments icon comment

Related Channels

Comments