To view PDF files

You need Adobe Reader 7.0 or later in order to read PDF files on this site.
If Adobe Reader is not installed on your computer, click the button below and go to the download site.

Feature Articles: Plasmon Control Technology

Vol. 21, No. 5, pp. 25–29, May 2023. https://doi.org/10.53829/ntr202305fa3

Ultrafast Optical-to-electrical Conversion Processes in Graphene

Katsumasa Yoshioka, Taro Wakamura, and Norio Kumada

Abstract

Graphene photodetectors have gained significant interest due to their predicted high sensitivity and rapid electrical response to broadband light, offering the potential for superior performance compared with conventional semiconductor devices. However, the experimental evaluation of the operation speed under a zero-bias voltage has been hindered by challenges associated with device architecture and measurement instrumentation, restricting the characterization of graphene°«s intrinsic dynamics and clarification of its operating mechanism. This article presents an ultrafast graphene photodetector developed at NTT laboratories, detailing the measurement technology and operating mechanism.

Keywords: graphene, photodetector, terahertz electronics

PDF PDF

1. Graphene as an optical-to-electrical conversion device

Photodetectors (PDs) play a crucial role in optical information communications and sensors by converting optical signals into electrical signals. Developing a PD that operates at high speeds for broadband optical signals creates new opportunities for advanced information processing through optical sensors and receivers working across a wide range of frequencies. This requirement can be satisfied with graphene, a promising material that exhibits high efficiency with its ability to absorb 2.3% of incident light with a single atomic layer and operate in an ultrawide frequency band, from terahertz to ultraviolet light [1]. The photothermoelectric (PTE) effect enables graphene PDs to operate at zero bias, which is essential for low power consumption and improved signal-to-noise ratio. However, previous research on graphene PDs has limited the experimental operating speed under zero bias to around 70 GHz [2], far from the theoretical expectation of 200 GHz. The limitations in device structure and measurement systems have hindered the investigation of the intrinsic response of graphene and the clarification of its operating mechanism, leading to a lack of design guidelines for maximizing the performance of graphene PDs. Therefore, the demonstration of 200-GHz operation speed and the elucidation of the physical properties of graphene, including the optical-to-electrical (OE) conversion mechanism, is crucial for further developments in graphene PDs.

2. Ultrafast photocurrent detection based on a novel device architecture and measurement approach

Ultrafast OE conversion in a graphene PD requires a device design that facilitates prompt tracking of incident light signals and a measurement technique capable of high-speed photocurrent detection. To meet these requirements, we used a zinc oxide thin film as a gate material to eliminate the current delay caused by capacitive coupling (Fig. 1(a)) and used on-chip terahertz spectroscopy [3] for photocurrent readout (Fig. 1(b)). The OE conversion efficiency of graphene PDs using the PTE effect varies significantly with the charge density. Thus, a gate structure must be integrated into the device. Because the gate electrode must possess high conductivity, gold is commonly used as the gate material; however, this limits the operating speed to below 100 GHz due to unwanted capacitive coupling. We addressed this issue by using a zinc oxide thin film, which enables control of the high-frequency response by controlling growth conditions. This material possesses a unique combination of being a good conductor for direct-current signals while being transparent at terahertz frequencies [4]. As a result, we successfully eliminated the capacitive coupling-induced current delay. The high-speed oscilloscopes commonly used to read out the photocurrent struggle to measure response faster than 100 GHz because of the bandwidth limitation of the electronics. However, we overcame this challenge by using on-chip terahertz spectroscopy, which enabled us to measure the ultrafast response of the graphene PD with a measurement bandwidth approaching 1 THz through the detection of photocurrent using an on-chip photoconductive (PC) switch. The photocurrent generated by irradiating the graphene PD with a pump pulse (femtosecond laser pulse) propagates through the drain electrode to the PC switch (Fig. 1(b)). Because the PC switch becomes conductive when irradiated by the probe pulse (also femtosecond laser pulse), the signal flows to the ammeter only when the generated photocurrent and probe pulse overlap in time. Therefore, measuring the photocurrent waveform with extremely high time resolution is possible by constantly changing the time difference between the pump and probe pulse. Our innovative device architecture and measurement approach have enabled us to explore the ultrafast OE conversion mechanisms in graphene, which was previously unachievable using conventional methods.


Fig. 1. Measurement system for ultrafast OE conversion.

3. Overview of the OE conversion processes in graphene

The photocurrent waveform of the fastest response among measured signals is shown in Fig. 2(a). The various frequency components of the photocurrent obtained by the Fourier transform (Fig. 2(b)) show that the 3-dB bandwidth of the graphene PD reaches 220 GHz, more than three times the 70 GHz reported previously and exceeding the theoretical expectation of 200 GHz. This is the first successful extraction of the intrinsic response of graphene unimpeded by device structure or measurement method. To investigate the physical process that determines the dynamical response of the photocurrent, we first focused on the decay time of the photocurrent by fabricating multiple devices with different graphene mobilities. Figure 3 shows that lower graphene mobility is associated with shorter decay times. This result suggests that the decay of the photocurrent in graphene is caused by a decrease in the electron temperature in graphene, which is increased by light irradiation. Because of collisions between acoustic phonons and defects, a sample with lower mobility and more defects loses more energy in a single collision event, resulting in faster decay times. However, the Seebeck coefficient, which corresponds to PD sensitivity, increases as mobility increases (Fig. 3), so there is a tradeoff between bandwidth and sensitivity in graphene PDs. Therefore, it is necessary to select optimal graphene mobility in accordance with the intended application.


Fig. 2. Verification of operating speed of 220 GHz.


Fig. 3. Tradeoff between detection sensitivity and operating speed.

To identify the mechanism that determines the photocurrent generation and propagation response (Fig. 4(a)), we investigated the dependence of the photocurrent on the charge density. The results indicate that a decrease in charge density produced a delay in the peak position of about 4 ps. Such a significant peak shift is unprecedented and provides important clues for understanding the OE conversion processes. We can consider two possible causes: 1) the propagation time of the photocurrent through graphene changes or 2) the time for photocurrent generation after light irradiation changes. To test the first possibility, we measured the peak shift for various device lengths and the irradiation position of the pump light (Fig. 4(b)). All the results fall on a single curve (Fig. 4(c)), indicating that the photocurrent propagation time in graphene is too short to be measured, so the cause of the peak shift must come from the change in the photocurrent-generation time. The ultrashort propagation time in a graphene PD results from its gapless nature. The absence of a bandgap in graphene allows the electric field created by light irradiation to propagate at the speed of light, instantly displacing the electrons near the electrode and generating an immediate flow of current without the need for propagation. The time variation in photocurrent generation is due to the PTE effect, which transforms the electron temperature into a voltage. This means that the non-equilibrium electron state created by the light irradiation quickly settles into a Fermi-Dirac distribution through intraband electron scattering. Our results also indicate that the time it takes for thermal equilibrium is highly dependent on the charge density of the graphene. These findings provide a comprehensive understanding of the ultrafast OE conversion processes in graphene, including the generation, propagation, and decay of photocurrent upon light irradiation [5].


Fig. 4. Dependence of photocurrent-response time on charge density.

4. Future development

The results presented in this article indicate the high potential of graphene as a material for broadband, high-speed PDs. However, our fabrication process of manually exfoliating graphene from graphite is not suitable for mass production. While the quality of graphene deposited over a large area has been historically inferior to that obtained through exfoliation, the gap is closing as deposition techniques continue to improve. Future studies will focus on evaluating PDs that use large-area graphene to facilitate mass production. Additionally, research in the field of creating novel materials by stacking graphene and other two-dimensional materials could be applied to further enhance the operating speed of PDs. The results of this study could be used to generate, control the propagation of, and detect graphene plasmons in the terahertz range. This opens up the possibility for on-chip handling of plasmon signals, bringing us closer to the goal of creating plasmonic circuits.

References

[1] M. Romagnoli, V. Sorianello, M. Midrio, F. H. L. Koppens, C. Huyghebaert, D. Neumaier, P. Galli, W. Templ, A. D’Errico, and A. C. Ferrari, “Graphene-based Integrated Photonics for Next-generation Datacom and Telecom,” Nat. Rev. Mater, Vol. 3, No. 10, pp. 392–414, Oct. 2018.
https://doi.org/10.1038/s41578-018-0040-9
[2] S. Marconi, M. A. Giambra, A. Montanaro, V. Mišeikis, S. Soresi, S. Tirelli, P. Galli, F. Buchali, W. Templ, C. Coletti, V. Sorianello, and M. Romagnoli, “Photo Thermal Effect Graphene Detector Featuring 105 Gbit s−1 NRZ and 120 Gbit s−1 PAM4 Direct Detection,” Nat. Commun., Vol. 12, No. 1, p. 806, Dec. 2021.
https://doi.org/10.1038/s41467-021-21137-z
[3] K. Yoshioka, N. Kumada, K. Muraki, and M. Hashisaka, “On-chip Coherent Frequency-domain THz Spectroscopy for Electrical Transport,” Appl. Phys. Lett., Vol. 117, No. 16, p. 161103, Oct. 2020.
https://doi.org/10.1063/5.0024089
[4] N. H. Tu, K. Yoshioka, S. Sasaki, M. Takamura, K. Muraki, and N. Kumada, “Active Spatial Control of Terahertz Plasmons in Graphene,” Commun. Mater., Vol. 1, No. 1, p. 7, Dec. 2020.
https://doi.org/10.1038/s43246-019-0002-9
[5] K. Yoshioka, T. Wakamura, M. Hashisaka, K. Watanabe, T. Taniguchi, and N. Kumada, “Ultrafast Intrinsic Optical-to-electrical Conversion Dynamics in a Graphene Photodetector,” Nat. Photon., Vol. 16, No. 10, pp. 718–723, Oct. 2022.
https://doi.org/10.1038/s41566-022-01058-z
Katsumasa Yoshioka
Researcher, Quantum Solid-State Physics Research Group, Quantum Science and Technology Laboratory, NTT Basic Research Laboratories.
He received a B.S., M.S., and Ph.D. in physics from Yokohama National University, Kanagawa, in 2014, 2016, and 2019. He joined NTT Basic Research Laboratories in 2019. Since then, he has been engaged in the study of on-chip THz spectroscopy of two-dimensional electronic systems. He is a member of the Physical Society of Japan and the Japan Society of Applied Physics.
Taro Wakamura
Research Scientist, Quantum Solid-State Physics Research Group, Quantum Science and Technology Laboratory, NTT Basic Research Laboratories.
He received a B.E., M. Sc., and Ph.D. in science from the University of Tokyo in 2010, 2012, and 2015. He joined NTT Basic Research Laboratories in 2019 and constructed a semi-automatic heterostructure fabrication system with novel two-dimensional materials including graphene and transition-metal dichalcogenides (TMDs). His current interests are broad, and he has recently been working on nonreciprocal transport in noncentrosymmetric TMDs. In addition to his own projects, he also contributes to collaborative works for developing novel PDs based on two-dimensional materials such as graphene. He is a member of the Physical Society of Japan.
Norio Kumada
Distinguished Researcher and Group Leader, Quantum Solid-State Physics Research Group, Quantum Science and Technology Laboratory, NTT Basic Research Laboratories.
He received a B.S., M.S., and Ph.D. in physics from Tohoku University, Miyagi, in 1998, 2000, and 2003. He joined NTT Basic Research Laboratories in 2003. Since then, he has been engaged in the study of highly correlated electronic states confined in two dimensions. From 2013 to 2014, he was a visiting scientist at CEA Saclay, France. He received the Young Scientist Award of the Physical Society of Japan in 2008 and the Young Scientists’ Prize from the Minister of Education, Culture, Sports, Science and Technology in 2012. He is a member of the Physical Society of Japan.

↑ TOP