Transforming the optical properties of silicon using femtosecond laser pulses

Presentation Date: 

Monday, December 20, 2010

Location: 

Horizons of Nanophotonics and Nanoelectronics, a Keio-Harvard Workshop, Harvard University (Cambridge, MA)

Presentation Slides: 

We developed a technique, optical hyperdoping (OHD), for doping semiconductors to unusually high levels and endowing them with remarkable optoelectronic properties. By irradiating Si with a train of femtosecond laser pulses in the presence of chalcogen (column VI) compounds, a 300-nm thin layer of Si is doped to previously unreported, non-equilibrium levels (about 1%). When the dopant is chosen from the heavy chalcogens (sulfur, selenium, tellurium), the doped silicon exhibits remarkable optoelectronic properties: near-unity absorptance from the ultraviolet ( λ < 250 nm) to the near-infrared ( λ > 2500 nm) [1, 2], even though crystalline silicon is transparent to wavelengths λ > 1100 nm due to its band gap at 1.1 eV. Additionally, the chalcogen dopants act as donor states: after doping, we observe the formation of a junction between the laser modified region and the substrate [3].

Our current work focuses on understanding what contributes to the remarkable properties of this material and applying the OHD technique to a range of other semiconductors and on understanding the physics of the doping process. It has become clear that two processes are involved: melting and ablation. The rapid melting and resolidification are responsible for the inclusion of dopants in the irradiated material, while ablation alters the surface morphology. We have recently developed a technique for decoupling the doping resulting from the melting and resolidification from changes in surface morphology, making it possible to obtain specularly flat surfaces of OHD materials. This makes it possible to study the properties of OHD materials without contributions from an altered morphology.

To better understand the properties of OHD silicon we carried out a large number of optical and electronic measurements. Optical measurements indicate that the OHD process leads to the formation of an impurity donor band in the optical gap between the conduction and valence band of silicon, extending from about 400 meV to 200 meV below the bottom of the conduction band. This band extends absorptance of the material well into the infrared until past 8 µm. Hall measurements show that the band is centered on a previously identified substitutional sulfur site at 318 meV below the conduction band and that the negative carrier density is on the order of 1017 cm–3 at room temperature. This carrier density indicates that about one out of every thousand implanted sulfur atoms is an ionized donor at room temperature. The Hall measurements also show that the mobility of the majority carriers is excellent and comparable of that silicon doped to much lower levels by traditional techniques.

Using this technique we fabricated silicon-based photodiodes that have a spectral responsivity five orders of magnitude larger than commercially available silicon photodiodes at wavelengths as long as 1.7 μm [3]. Devices made from OHD silicon show excellent rectification and a remarkable gain, even at a bias as low as 3 V. The low bias excludes impact excitation and avalanching as a possible mechanism for the observed gain. Instead, measurements suggest the high responsivity (up to 500 A/W at 1 µm) is a result of the carrier lifetime greatly exceeding to the carrier transit time through the device, a phenomenon called photoconductive gain. We attribute the large carrier lifetime to trapping of holes in the partially depleted OHD layer. This trapping of holes makes it possible for the photoelectron to cycle through the device many times before recombining.

We are currently broadening the scope of our research. In the area of photovoltaic research, we have built prototype solar cells exhibiting promising efficiencies. In addition, we are applying the OHD technique to other materials and have observed similar results in a broad class of semiconductors, including wide band gap oxides such as TiO2.

REFERENCES

[1] C. Wu, C. H. Crouch, L. Zhao, J. E. Carey, R. Younkin, J. A. Levinson, E. Mazur, R. M. Farrell, P. Gothoskar, and A. Karger, Near-Unity Below-Band-Gap Absorption by Microstructured Silicon. Applied Physics Letters, 78, 1850-1852 (2001).
[2] B. R. Tull, M. T. Winkler, and E. Mazur, The Role of Diffusion in Broadband Infrared Absorption in Chalcogen-Doped Silicon. Applied Physics a-Materials Science & Processing, 96, 327-334 (2009)
[3] J. E. Carey, C. H. Crouch, M. Y. Shen, and E. Mazur, Visible and near-Infrared Responsivity of Femtosecond-Laser Microstructured Silicon Photodiodes. Optics Letters, 30, 1773-1775 (2005)

BIOGRAPHY

Eric Mazur is Balkanski Professor of Physics and Applied Physics and Dean of Applied Physics at Harvard University. He obtained a PhD degree in experimental physics at the University of Leiden in the Netherlands in 1981 and joined the Harvard faculty in 1984. His work includes spectroscopy, light scattering, and studies of electronic and structural events in solids that occur on the femtosecond time scale. He is also interested in education, science policy, outreach, and the public perception of science.