This document summarizes an experiment that measured the temperature-dependent responsivity of plasmonic terahertz detectors. The experiment measured the response of gallium nitride chips under terahertz pulses in both free space and liquid nitrogen conditions. The results showed lower response amplitudes in liquid nitrogen, consistent with decreased electron excitation and conductivity at lower temperatures according to plasmonics and semiconductor physics models. Understanding these material properties allows for optimizing devices that use plasmonic materials.

1. Temperature-dependent Responsivity of Plasmonic THz Detectors
Emeka V. Ikpeazu, Jr.*; Mustafa Karabiyik†, Nezih Pala†
*Department of Electrical and Computer Engineering, Cornell University, Ithaca, New York, USA. †Department of Electrical and Computer Engineering (Integrated
Nanosystems Laboratory), Florida International University, Miami, Florida, USA
Abstract: The importance of measuring responsivity of chips to terahertz pulses of infrared electromagnetic radiation is evinced in its increasingly widening applications such as photodetection, medical imaging, and
security technologies. Introduction: The experiments conducted specifically measured the responsivity of chips for plasmonic detectors under various conditions. Procedure: In these experiments, the equipment we
used was also important, especially in regards to centering the small chip and making sure that wire-bonds soldered to seven different parts of the chip-holder were functioning properly to accurately relay response data.
This paper will discuss the way in which a motorized stage was used for positioning chips on the lab’s optical table setup then it will discuss the results of the responsivity measurements showing how specifically
temperature can have an effect on electromagnetic absorptivity. Conclusions: The importance of this work is showing how understanding material properties can allow for optimally functioning devices.
Background:
In physics, the field of plasmonics measures the interaction between electromagnetic waves and
the electrons in a material. The material we used in this experiment was gallium nitride (GaN). The way
in which the experiments were conducted was measuring the respective response in standard conditions and
in cryostat conditions. The experimentation clues one into understanding the properties of various
semiconductor materials and what they mean for the devices that use them.
Method:
The project had five stages:
1. Arbitrarily center the stage
2. Moving the stage in the x-direction and the y-direction and storing the x and y values
3. Read the voltage value recorded b the lock-in amplifier
4. Display the values of the absorption responsivity for each x-y coordinate as a heat map
5. Write the values to a file.
Figure. 1. Module for Position Axis
The module takes four variables (the number of steps, the number of steps per second, the delay between steps, and the direction (forwards or backwards) of the steps) in order to allow for the motion of the stage in
both the x and y directions.
Figure. 2. Synthesis diagram of position variables and readings
The Xstep Length and Ystep Length are registered in parallel with the readings that come in from the lock-in amplifier.
Figure. 3. Module for Position Axis
The module takes four variables (the number of steps, the number of steps per second, the delay between steps, and the direction
(forwards or backwards) of the steps) in order to allow for the motion of the stage in both the x and y directions.
Procedure:
The first use of the stage was to measure the
terahertz response of gallium nitride (GaN) chip
wire-bonded to a chip holder in free-space and
liquid nitrogen. The motorized stage would move
the chip across a beam of infrared pulses of 10.03
kHz.
Measurements:
There were a total of six THz measurements done
in free space of varying voltages.
Figure 4.
The gallium nitride chip is centered in the middle of the chip holder. Eight wires are soldered to the edges of the chip holder so that the responsivity of the THz source can be recorded.
Measurements (cont’d):
There were a total of eight THz measurements
done in liquid nitrogen of varying voltages.
Figure 5. The free-space measurement of THz response was roughly identical for each voltage value used. This appeared to be essentially the same with the liquid nitrogen measurements. Also the average amplitude of the response voltage was lower in
the liquid nitrogen than in free-space.
Hypothesis:
My prediction regarding the difference in the cryostat measurement and the STP measurements was that the response to the THz pulses
would undergo attenuation in the liquid nitrogen medium. My reasing behind it was that the super-cooling effects of the liquid nitrogen
increases the thermal deBroglie wavelength of the GaN chip. The thermal deBroglie wavelength is expressed as:
Λ 𝑑𝑑 =
ℎ
3𝑚𝑚𝑘𝑘𝐵𝐵 𝑇𝑇
The above equation would have it that the lower temperature would increase the inter-particle spacing in the GaN chip. The increase in this
spacing would lead to more room for loss in absorptivity.
Result:
My prediction was correct but the reasoning was not. Plasmonics deals with electronic excitations and it turns out that temperature has a
significant effect in this domain. Decreasing the temperature pushes the material closer to thermal equilibrium (T = 0 K) and thus reduces
electron excitation. The conductivity of the material is expressed as:
𝜎𝜎 =
𝑛𝑛𝑒𝑒2
𝜏𝜏
𝑚𝑚
1 − 𝑗𝑗 𝑗𝑗𝑗𝑗
The most important variable here is 𝜏𝜏, the average scattering time; the average scattering rate, 𝑄𝑄 = 1/𝜏𝜏. The conductivity as a function of
the scattering rate is expressed:
𝜎𝜎 =
𝑛𝑛𝑒𝑒2
𝑚𝑚𝑚𝑚
1 − 𝑗𝑗 𝑗𝑗/𝑄𝑄
=
𝑛𝑛𝑒𝑒2
/𝑚𝑚
𝑄𝑄 − 𝑗𝑗 𝑗𝑗
Conclusions:
The measurements of the absorption response are clearly indicative of how a change in
conditions can lead to an a drastic effect on the performance of the device chips.
The LabVIEW program was effective in registering certain values for the response of
the chip as it was guided across the stage. The well-soldered edges of the chip were
instrumental in relaying the THz signal form the chip to the individual nodes of the
detector (Figure 6). The results were in line with what one would expect having an
adequate background in solid-state and semiconductor physics.
The results also evince the efficacy of the Drude model in understanding the behavior
and properties of these sorts of materials. As we were dealing with materials with high
electron density rather than individual particles. In this way, one need not appeal to
quantum mechanics in order to explain the results acquired.
For semiconductor materials in general, understanding the plasmonics is a good tool for
being able to optimize engineering applications through their use.
The increased conductivity at lower temperatures evinces a robust potential for
applications in optical systems, particularly waveguides. The low absorption rate means
less loss and increased transmission for EM waves traveling through GaN waveguides,
More research should be directed towards the study of plasmonic nanostructures and
materials in order to optimize the performance of the devices which are to make use of
them.
Figure 6. The nodes of the detector correspond to the
appropriate parts of the holder for the GaN chip.
References:
1. Popov, Vyacheslav V., D. M. Ermolaev, and Kirill V. Maremyanin. "High-responsivity Terahertz Detection by On-chip InGaAs/GaAsField-effect-transistor Array." High-responsivity TerahertzDetection by On-chip InGaAs/GaAs Field-effect-transistorArray. Applied Physics Letters,11 Aug. 2011. Web. 19 Aug. 2014.
2. Wood, Christopher D., John E. Cunningham, and Prashanth C. Upadhya. "On-chip Photoconductive Excitation and Detection of Pulsed Terahertz Radiation at Cryogenic Temperatures." ResearchGate.Applied Physics Letters, 3 Apr. 2006. Web. 19 Aug. 2014.
3. Sun, Guan, Guibao Xu, Yujie J. Ding, Hongping Zhao, Guangyu Liu, Jing Zhang, and Nelson Tansu. "Efficient Terahertz Generation Within InGaN/GaNMultiple Quantum Wells." IEEE Journal of SelectedTopics in QuantumElectronics 17.1 (2011): 48-53. IEEE Xplore.IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, 4 Feb. 2011. Web. 19 Aug. 2014.
4. Bouillard, Jean-Sebastien G., Wayne Dickson, Daniel P. O'Connor, Gregory A. Wurtz, and Anatoly V. Zayats."Low-Temperature Plasmonics of Metallic Nanostructures." NanoLetters(ACS Publications).American Chemical Society, 16 Feb. 2012. Web. 09 Feb. 2015.
Figure 7. A sample diagram showing conductivity (σ) as a function of scattering rate (Q)
This sample diagram shows the magnitude of the conductivity of the material as a function of scattering rate. As one can see the conductivity decreases as the scattering rate increases.
The relation to temperature is that decreasing the temperature—and thus the thermal energy—decreases the scattering rate and increases the conductivity. This increase in conductivity
leads to ohmic losses in the material and thus poorer absorption.
Figure 8. Nanocharacterzation apparatus
With this device we would not only set the chip in the chip-holder, but we would also check the chip for
impurities before proceeding to wash it with alcohol (propanol) and water and then dry it.