I'm a graduate student in Bob York's research
group in the Electronics and Photonics group of the Electrical and
Computer Engineering Department at the University of California, Santa
Barbara. I got my bachelor's degree in electrical engineering in 1999
at The Cooper Union for the Advancement of Science and Art in New York.
I grow and characterize BaxSr1-xTiO3
(BST) thin films. BST is a high permittivity dielectric material with a
field-dependent tunability which we use to make tunable capacitors for
RF and microwave applications. Through the application of a DC bias, we
reduce our films' permittivities by more
than 90%. The zero-bias permittivity of our films in thickness
dependent, but typically around 500. For a comparison, commerical Si
varactor diodes exhibit less than 4:1 tuning while our optimized BST
varactors have > 8:1 useful tuning (i.e., tuning without entering
the high leakage current regime preceding device destruction).
BST is a solid solution of BaTiO3 and SrTiO3, which is paraelectric at room temperature and has the following cubic structure:
We deposit our films by RF
magnetron sputtering. At UCSB we have three sputtering chambers
dedicated
to ferroelectrics and related materials. Currently, two of these
machines are in our lab in Engineering 1, while the other one is
located in Engineering 2. Some time in 2005 we will be moving all three
machines in to Susanne
Stemmer's new lab in the Engineering Sciences Building (ESB).
Our test structures,
fabricated in UCSB's research cleanroom,
are simple
parallel plate devices, as depicted below.
 |
A typical parallel plate capacitor structure: a
rectangle of BST (green) on a ground plane (not shown) with metal
contacts (gray) in a ground-signal-ground configuration. Our substrates
are typically c-plane sapphire. |
We use a ground-signal-ground structure because it can be probed
with
coplanar probes for high frequency measurements or needle probes for
low frequency and/or DC measurements. Our measurement lab has four RF
measurement probe stations including a load-pull station for power
measurements and an RF cryo station that uses liquid nitrogen for
measurements down to 80K. We also have three low frequency probe
stations including an MMR probe station that uses a Joule-Thomspon
refrigerator with high pressure nitrogen for measurements between 80K
to 730K. We have a number of network analyzers for high frequency
measurements, and two impedance analyzers for accurate capacitance and
quality factor measurements at lower frequencies.
The focus of my work is the measurement and understanding of loss
mechanisms in thin-film BST. Thus far we have observed "universal"
relaxation in the absence of a DC bias. I have seen similar relaxation
behavior in Jiwei Lu's BZN films but the BZN data is hard to fit
because of the small amount of (total) relaxation over the frequency
range studied. We've found, using measurements on BST films, that we
can take advantage of this observed behavior - we have successfully
used it to predict Q-factor values directly from capacitance data. We
are currently trying to determine if this technique can be used at
higher frequencies where direct Q-factor measurements are difficult.
This past summer I had an intern from CNSI's Apprentice
Researchers
program. Her name is Rachel Stein, and she's currently a senior at
Santa
Barbara High School. Her project was
the measurement of
harmonic generation in thin-film BST capacitors. For our applications,
harmonic generation due to the material's dielectric nonlinearity is a
form of loss - reactive loss. We need to consider the effect of input
amplitude on harmonic generation - most of our test measurements are
performed with AC amplitudes < 500 mV, but Hongtao Xu has integrated
the BST capacitor process with the GaN HEMT process, so we need to
think more about what happens with large AC amplitudes.
The York
Group Projects page has more information our group's work.