We have developed a wavelength-scale embedded active-region photonic-crystal laser using lateral p-i-n structure. Zn diffusion and Si ion implantation are used for p- and n-type doping. Room-temperature continuous-wave lasing behavior is clearly observed from the injection current dependence of the output power, 3dB-bandwidth of the peak, and lasing wavelength. The threshold current is 390 μA and the estimated effective threshold current is 9.4 μA. The output power in output waveguide is 1.82 μW for a 2.0-mA current injection. These results indicate that the embedded active-region structure effectively reduce the thermal resistance. Ultrasmall electrically driven lasers are an important step towards on-chip photonic network applications.
©2012 Optical Society of America
The development of room-temperature (RT) continuous-wave (CW) operated semiconductor lasers is very promising in relation to progress on optical communication systems. In 1970, the first RT CW operated edge-emitting semiconductor lasers were developed by using a double-heterostructure in a GaAs/AlGaAs material system [1, 2]. These lasers are widely used for telecommunication networks ranging from backbone to access networks. In 1988, the CW operation of a vertical-cavity surface-emitting laser (VCSEL) was demonstrated at RT . A reduction of the active volume enables us to reduce the power consumption of the transmitter, which means that an optical transmission system has many advantages compared with its electric counterpart such as data transmission speed and power consumption, even over shorter transmission distances. Therefore, VCSELs are widely used for datacom networks ranging from intra-systems, such as data centers and super-computers, to board-to-board interconnects.
In this context, further reduction of the active volume opens up a novel application area for optical communication systems, namely off-chip and on-chip networks for silicon CMOS. For example, an on-chip photonic network that is integrated with silicon CMOS has attracted a lot of attention because it shows promise as regards overcoming the bandwidth and power consumption limits faced by on-chip electrical interconnects [4–7]. For this photonic integrated circuit (PIC), the density requirement is a critical issue as several thousand lasers, photodetectors, and switches must be integrated on a single chip. Thus the device operating energy should be reduced and the target device energy for the transmitter side should be less than 10 fJ/bit , considering the demands for future interconnects in CMOS based on the International Technology Roadmap for Semiconductors . In addition, it is well known that the use of wavelength division multiplexing (WDM) technology greatly increases the total bandwidth and reduces the total power consumption of a CMOS chip [5, 9]. Thus, the laser requires wavelength controllability because it should oscillate the required wavelength at any position on the chip.
In terms of satisfying the above requirements, lasers incorporating wavelength-scale photonic-crystal (PhC) cavities are ideal . A PhC is an artificial structure having periodic modulation in refractive index which can have a photonic band gap . Because of the strong light confinement achieved through the existence of photonic bandgaps, ultrasmall (comparable to the wavelength) and high-Q (1 × 106) cavities can be realized in PhCs [12, 13]. However, the RT CW operation of an electrically driven wavelength-scale PhC laser has not been achieved. Only RT pulsed operation has been demonstrated in a wavelength-scale PhC laser by employing a vertical p-i-n junction and a current post [14, 15]. However, the current post limits the quality (Q) factor of the cavity (threshold Q-factor (Qth) ~3400), which prevents CW operation. Furthermore, a PhC laser consisting of a vertical p-i-n junction and a current post is unsuitable for constructing a large-scale PIC on a CMOS chip. In this context, PhC lasers with an air-bridge and lateral current injection structure have been demonstrated [16, 17]. However, these devices did not realize laser oscillation even during pulsed operation at RT because their all-active structure, in which a two-dimensional photonic slab consists of a single active region, provides no carrier confinement structure and exhibits high thermal resistance.
To overcome these problems, we have developed a wavelength (lambda)-scale embedded active-region PhC (LEAP) laser, in which wavelength-scale InGaAsP-based active region is embedded with an InP based line-defect PhC waveguide . This structure allows us to greatly reduce the thermal resistance and realize the strong confinement of both photons and carriers in the active region, because the thermal conductivity and bandgap of the InP layer are larger than those of the InGaAsP layer. Thus, we have demonstrated a low threshold input power of 6.8 μW with a Qth of 14,300, an output power of −10.3 dBm, and 20-Gbit/s direct modulation with 8.76 fJ/bit by optical pumping . Furthermore, we controlled the lasing wavelength over a wavelength range of more than 100 nm by controlling the lattice constant of the PhC .
The purpose of the present study is, therefore, the development of a CW operated LEAP laser using lateral current injection at RT. The modulated line-defect cavity [11, 18] is suitable for making a lateral p-i-n junction. We have reported a successful fabrication of ultrahigh-Q nanocavities based on this design with lateral p-i-n junctions in Si photonic crystals by employing ion implantation . To fabricate a lateral p-i-n current junction in InP photonic crystals, here we employ Zn diffusion and Si ion implantation respectively for p- and n-type doping into an i-InP layer. Room-temperature continuous-wave lasing behavior is clearly observed from the injection current dependence of the output power, 3dB-bandwidth of the peak, and lasing wavelength. The threshold current is 390 μA and the estimated effective threshold current is 9.4 μA. The output power in output waveguide is 1.82 μW for a 2.0-mA current injection. These results indicate that the embedded active-region structure enables us to reduce the thermal resistance of the device and confine both the carrier and photon within wavelength-scale cavity.
2. Device structure and fabrication
Figure 1(a) shows the schematic diagram of LEAP laser and Fig. 1(b) shows cross-sectional view of the device. We place an extremely small embedded active region in a straight-line defect waveguide in an InP-PhC slab. This structure allows us to greatly reduce the thermal resistance and realize the strong confinement of both photons and carriers in the active region, because the thermal conductivity and bandgap of the InP layer are larger than those of the InGaAsP layer . Furthermore, a cavity based on a line-defect waveguide is expected to provide an extraordinarily high-Q (Q>1 million) nanocavity with a mode volume (Veff) of ~(λ/n)3 by implementing the slight local structural or reflective index modulation of a line-defect waveguide [11, 21]. The output waveguide is placed in an offset position with respectto the line-defect waveguide including the active region to obtain effective coupling between the cavity and output waveguide .
To fabricate a lateral p-i-n current injection structure, we employ Zn diffusion and Si ion implantation respectively for p- and n-type doping into an i-InP layer. Silicon ions were implanted for n-type doping with an energy of 240 keV and a dose of 1.2 × 1014 cm−2 and photoresist was used to pattern the n-type doping region. After removing the photoresist, we deposited a SiO2 layer. The sample was then annealed at 700°C for 3 min to activate the dopants. Then, a p-type InP layer was formed by Zn diffusion with 550°C for 1 min. A patterned SiO2 layer was used to define the p-type region. A fabrication process that employs impurity implantation and diffusion will result in greater performance repeatability and increased yield because it is the same fabrication technique that is used for silicon CMOS.
Figure 2(a) shows a scanning electron micrograph (SEM) image of our fabricated LEAP laser with an air-bridge structure. We use a 3-quantum well structure for the active region with 2.6 × 0.3 × 0.15 μm3 (0.12 μm3). The active region was embedded within line-defect PhC waveguide with flat surface, which is important to obtain high-Q cavity. The output waveguide is placed in an offset position with respect to the line-defect including the active region. The cavity design is basically same as previous developed optically pumped laser in ref. 7, in which we achieved a high external differential quantum efficiency of 53% and ultra-low energy cost of 8.76 fJ/bit with 20-Gbit/s NRZ signal. However, main purpose of this manuscript is to achieve the electrically driven laser operated at RT CW condition. In this context, the reduction of the threshold current is the most promising way because RT CW operation of the laser is suppressed by the increase of the active region temperature due to the current injection. Thus, compared with the device in ref. 7, we decrease the active region length to reduce the threshold current and increase the Q-factor to reduce the threshold current density. Therefore, the output waveguide is positioned further away from the cavity than with the previous device to increase the Q-factor. The output waveguide consists of a 3-μm wide waveguide, a taper waveguide and the line-defect of the PhC. The field profile of the cavity mode obtained by using a finite-difference time-domain (FDTD) calculation is shown in Fig. 2(b). Q-factor was ~4200 and Veff was ~0.15 μm3. This clearly shows that the embedded active region largely overlaps the cavity mode profile.
3. Device characteristics
The fabricated electrically driven LEAP laser is CW operated at RT. The light and voltage versus current characteristic is shown in Fig. 3(a) . The laser output light from an output line-defect waveguide was collected into a single-mode fiber. We used a tapered waveguide from the output line-defect waveguide to a 3-μm-wide waveguide. We estimated a coupling loss of 10 dB between the optical fiber and the line-defect waveguide . The device exhibits a clear kink at a threshold of 390 μA. The output power coupled to the output line-defect waveguide was 1.82 μW when the injection current was 2.0 mA. The images captured using an IR camera are shown in Fig. 3(b) and (c) for injection current of 0.4 and 2.0 mA, respectively. The output lights are observed from the cavity to the surface normal direction and the facet of the output waveguide because the output waveguide has no anti-reflection coating. As described in section 2, the coupling efficiency with the output waveguide was decreased and therefore the output from the surface normal direction was increased. This could be improved by optimizing the cavity design. In addition, as shown in Fig. 3(b) and 3(c), electroluminescent light was observed under the p-contact electrode (lower-side of electrode). Since the image was monitored through the 1450-nm optical long-pass filter, the light was emitted from the InGaAs sacrificial layer due to the current leakage. This means that undoped InP buffer layer shown in Fig. 1(b) cannot suppress the current through the InGaAs sacrificial layer. As shown in Fig. 3(a), the injection current was increased from bias voltage of 0.4 V, which agree with the built-in potential of the InGaAs layer.
The injection current dependence of the 3-dB bandwidth of the peak, and the lasing wavelength are shown in Fig. 4(a) . Below the threshold region (less than 390 μA), the 3-dB bandwidth decreased as the injection current increased because the loss in the cavity decreased. The lasing wavelength also decreased as the injection current increased because the generated carriers induced band filling and plasma effects, which led to a reduction in the refractive index of the cavity and shortened the cavity resonance wavelength. In the threshold region, where the laser field underwent a phase transition, the linewidth broadened with increasing output power as the gain-refractive index coupling rapidly started to influence the linewidth . The 3-dB bandwidth at the threshold was 0.035 nm, corresponding to a Q-factor of 44,600. Above the threshold region, the lasing wavelength was constant because the carrier density was clamped and the 3-dB bandwidth decreased with increasing input power. Our experimental results agreed with the calculated results , and we confirmed that our laser clearly showed lasing oscillation with RT CW operation.
As shown in Fig. 3(b) and 3(c), the present device has large leakage current. To estimate the effective current input into the laser cavity, we measured the device characteristics by using optical pumping. Optical pumping measurements were performed using a laser diode (λ = 980 nm) as the excitation source. The pump laser beam was incident from the surface normal direction and focused to a 3-μm diameter spot on the sample surface with a microscope objective (50x, numerical aperture = 0.42). Since the absorption coefficient of the InP layer for 980 nm-pump light is negligible, we correctly estimated the absorbed optical power by using an absorption coefficient, α, of 14850 cm−1 and an absorbed region ( = active region) of 0.25 x 0.3 x 0.15 μm3. The absorption coefficient α was calculated using the 6-band k·p analysis and the Fermi’s golden rule . The effective absorbed power, Peff, is given by,Fig. 4(b), where the red and blue circles represent the experimental results for electrical and optical pumping, respectively. The output light was detected from the output waveguide in both measurements. Thus, the threshold output powers are the same because the optical property of the cavity is not related to the pumping methods and the required threshold gain must be the same for both optical and electrical pumping. Thus, we shifted the absorbed optical power (upper X-axis in Fig. 4(b)) to agree with the output power dependence of the injection current. This indicates that the effective absorbed optical power at the threshold was 11.8 μW, corresponding to an effective threshold current of 9.4 μA. The present device has a large leakage current through the InGaAs sacrificial layer. Thus, we can reduce the operating current by modifying the epitaxial structure and mask design. The threshold optical input power was larger than that of previously developed optical pumped lasers reported in ref. 7 (Pth = 3.2 μW). This is because the carrier absorption such as the intervalence band absorption and the free carrier absorption in the p-InP region increases the loss in the cavity. Thus, we need to optimize the device fabrication process.
The lasing spectrum was measured during RT pulsed operation to determine the increase in the active region temperature. Figure 5(a) shows the lasing spectra for pulse widths of 20 ns and 180 μs with a 200-μs repetition. The peak wavelength of the device operated with a 180-μs pulse is the same as the lasing wavelength operated under a 1.2-mA CW condition. Using the temperature dependence of the lasing wavelength for our laser (0.095 nm/K) , we estimated the increase in the active region temperature caused by CW operation to be 7.1 K for an injection current of 1.2 mA. The lasing spectra and increase in the active region temperature for various injection currents ranging from 0.5 to 2.0 mA are shown in Fig. 5(b) and (c), respectively. Although the active region temperature increases with the injection current, it was less than 15 K when the injection current was 2.0 mA.
Lateral current injected LEAP laser realizes CW operation at RT, and exhibits a threshold current of 390 μA. If the leakage current flowing into the substrate and the InGaAs sacrificial layer is subtracted, the threshold current is estimated to be ~9.5 μA. We believe that a further reduction in the threshold current could be achieved by optimizing the fabrication process and epitaxial layer design. The output power in output waveguide is 1.82 μW for a 2.0-mA current injection. These results indicate that the combination of the embedded active-region structure and photonic crystal cavity is essential to obtain electrical driven ultrasmall cavity laser. Hence, the LEAP laser constitutes a new and exciting class of lasers, with the potential to employ data transmission in a silicon CMOS chip.
We thank K. Kato for fruitful discussions. We also thank K. Ishibashi and Y. Shouji for fabricating the device. Part of this work was supported by the New Energy and Industrial Technology Development Organization (NEDO).
References and links
1. I. Hayashi, M. B. Panish, P. W. Foy, and S. Sumski, “Junction lasers which operate continuously at room temperature,” Appl. Phys. Lett. 17(3), 109–111 (1970). [CrossRef]
2. Zh. I. Alferov, V. M. Andreev, D. Z. Garbuzov, Yu. V. Zhilyaev, E. P. Morozov, E. L. Portnoi, and V. G. Trofim, “Investigation of the influence of the AlAs–GaAs heterostructure parameters on the laser threshold current and the realization of continuous emission at room temperature,” Fiz. Tekh. Poluprovodn. 4, 1826 (1970).
3. F. Koyama, S. Kinoshita, and K. Iga, “Room-temperature continuous wave lasing characteristics of a GaAs vertical cavity surface-emitting laser,” Appl. Phys. Lett. 55(3), 221–222 (1989). [CrossRef]
4. Y. Vlasov, W. M. J. Green, and F. Xia, “High-throughput silicon nanophotonic wavelength-insensitive switch for on-chip optical networks,” Nat. Photonics 2(4), 242–246 (2008). [CrossRef]
5. A. Shacham, K. Bergman, and L. P. Carloni, “Photonic networks-on-chip for future generations of chip multiprocessors,” IEEE Trans. Comput. 57(9), 1246–1260 (2008). [CrossRef]
6. D. A. B. Miller, “Device requirements for optical Interconnects to silicon chips,” Proc. IEEE 97(7), 1166–1185 (2009). [CrossRef]
7. S. Matsuo, A. Shinya, C.-H. Chen, K. Nozaki, T. Sato, Y. Kawaguchi, H. Taniyama, and M. Notomi, “20-Gbit/s directly modulated photonic crystal nanocavity laser with ultra-low power consumption,” Opt. Express 19(3), 2242–2250 (2011). [CrossRef] [PubMed]
9. M. Haurylau, G. Chen, H. Chen, J. Zhang, N. A. Nelson, D. H. Albonesi, E. G. Friedman, and P. M. Fauchet, “On-chip optical interconnect roadmap: challenges and critical directions,” IEEE J. Sel. Top. Quantum Electron. 12(6), 1699–1705 (2006). [CrossRef]
10. M. Notomi, A. Shinya, K. Nozaki, T. Tanabe, S. Matsuo, E. Kuramochi, T. Sato, H. Taniyama, and H. Sumikura, “Low power nanophotonic devices based on photonic crystals towards dense photonic network on chip,” IET Circuits, Devices Syst. 5(2), 84–93 (2011). [CrossRef]
11. M. Notomi, “Strong light confinement with periodicity,” Proc. IEEE 99, 1768–1779 (2011).
12. T. Tanabe, M. Notomi, E. Kuramochi, A. Shinya, and H. Taniyama, “Trapping and delaying photons for one nanosecond in an ultrasmall high-Q photonic-crystal nanocavity,” Nat. Photonics 1(1), 49–52 (2007). [CrossRef]
14. H. G. Park, S. H. Kim, S. H. Kwon, Y. G. Ju, J. K. Yang, J. H. Baek, S. B. Kim, and Y. H. Lee, “Electrically driven single-cell photonic crystal laser,” Science 305(5689), 1444–1447 (2004). [CrossRef] [PubMed]
15. M.-K. Seo, K.-Y. Jeong, J.-K. Yang, Y.-H. Lee, H.-G. Park, and S.-B. Kim, “Low threshold current single-cell hexapole mode photonic crystal laser,” Appl. Phys. Lett. 90(17), 171122 (2007). [CrossRef]
16. C. M. Long, A. V. Giannopoulos, and K. D. Choquette, “Modified spontaneous emission from laterally injected photonic crystal emitter,” Electron. Lett. 45(4), 227–228 (2009). [CrossRef]
17. B. Ellis, T. Sarmiento, M. Mayer, B. Zhang, J. Harris, E. E. Haller, and J. Vuckovic, “Electrically pumped photonic crystal nanocavity light sources using a laterally doped p-i-n junction,” Appl. Phys. Lett. 96(18), 181103 (2010). [CrossRef]
18. S. Matsuo, A. Shinya, T. Kakitsuka, K. Nozaki, T. Segawa, T. Sato, Y. Kawaguchi, and M. Notomi, “High-speed ultracompact buried heterostructure photonic-crystal laser with 13 fJ of energy consumed per bit transmitted,” Nat. Photonics 4(9), 648–654 (2010). [CrossRef]
19. K. Takeda, T. Sato, A. Shinya, K. Nozaki, C.-H. Chen, Y. Kawaguchi, H. Taniyama, M. Notomi, and S. Matsuo, “80°C continuous wave operation of photonic-crystal nanocavity lasers,” 23rd International Conference on Indium Phosphide and Related Materials, Berlin, May 2011.
20. T. Tanabe, K. Nishiguchi, E. Kuramochi, and M. Notomi, “Low power and fast electro-optic silicon modulator with lateral p-i-n embedded photonic crystal nanocavity,” Opt. Express 17(25), 22505–22513 (2009). [CrossRef] [PubMed]
22. G. Björk, A. Karlsson, and Y. Yamamoto, “On the linewidth of microcavity laser,” Appl. Phys. Lett. 60(3), 304–306 (1992). [CrossRef]
23. S. L. Chuang, Physics of Optoelectronic Devices (John Willey & Sons, 1995).