Effects of annealing on performances of 1.3-μm InAs-InGaAs-GaAs quantum dot electroabsorption modulators
© Lee et al.; licensee Springer. 2013
Received: 11 January 2013
Accepted: 30 January 2013
Published: 6 February 2013
In this work, we investigated the effects of quantum dot (QD) annealing (as-grown, 600°C-annealed, and 750°C-annealed) on the preliminary performances of 1.3-μm InAs-InGaAs-GaAs quantum dot electroabsorption modulators (QD-EAMs). Both extinction ratio and insertion loss were found to vary inversely with the annealing temperature. Most importantly, the 3-dB response of the 750°C-annealed lumped-element QD-EAM was found to be 1.6 GHz at zero reverse bias voltage - the lowest reverse bias voltage reported. We believe that this work will be beneficial to researchers working on on-chip integration of QD-EAMs with other devices since energy consumption will be an important consideration.
Quantum dot (QD) lasers are now extensively investigated for applications in low-cost metropolitan access and local area networks. However, most works on QD devices focus on lasers and detectors. There were only a handful of them that were related to quantum dot electroabsorption modulators (QD-EAMs) [1, 2]. For ease of monolithic integration, it is timely to investigate the use of QDs for electroabsorption modulators (EAMs). As such, one can then utilize QDs for both laser and EAM by the identical active layer approach [3, 4].
Recently, Chu et al. reported a small-signal frequency response of 2 GHz for the 1.3-μm QD-EAM . However, the applied reverse bias was 4 V - which could lead to complications for on-chip integration since energy consumption is an issue. We had previously reported the static performance of 1.3-μm QD-EAM based on as-grown QDs . Due to the defined QD potential barriers, one can observe a suppression of absorption at reverse bias <2 V . This implies that our as-grown QD-EAM will also require a significant reverse bias voltage (≥2 V in this case) for small-signal frequency response. Again, this is undesirable for on-chip integration. On the other hand, annealed QDs are proposed to be a good candidate for energy-efficient QD-EAM. By varying the annealing temperature, we are able to induce different diffusion lengths on the QD layers . There are two mechanisms at work, the first being the exchange of In atoms from the InAs QD intermixing with the Ga atoms in its surrounding InGaAs QW and the second being the In-Ga interdiffusion through the InGaAs/GaAs interface . The uniform interdiffusion would result in a more symmetrical Stark shift and consequently reduces the built-in dipole momentum - both of which are beneficial to electroabsorption modulation. An enhanced Stark shift would theoretically provide a cleaner and larger extinction ratio, whereas the reduction of built-in dipole moment in QD would increase the ground-state electron–hole overlap. Therefore, in this work, we investigated the effect of QD annealing on the static and dynamic performances of 1.3-μm QD-EAM.
Label and description of the QD samples under investigation
Annealed at 600°C for 10 s
Annealed at 750°C for 10 s
After the annealing process, the SiO2 was removed using a HF/H2O rinse with a ratio of 1:10. Subsequently, a photoresist was spun on the wafer surface, and the stripe patterns of the ridge waveguide (RWG) structures were defined after UV exposure and photoresist developing. This was followed by wet chemical etching (H3PO4/H2O2/DI = 1:1:5 with an etch rate of approximately 1.2 μm/min) to define the ridge height. After the wet-etching process, the remaining photoresists that acted as etching masks were removed using acetone. The next step was the planarization step. This step involved spin coating of bisbenzocyclobutene (BCB) monomers. The BCB helped to flatten out the sample surface and acted as a passivation step. The waveguide sides that had been coated with BCB also helped to reduce capacitance in high-speed measurements. The etch-back step was then applied to reduce this BCB layer until the waveguide layer was exposed again. Note that this method was preferred over the alternative method of defining photoresist pattern. This was because the RWG EAM devices had heights of approximately 1.2 μm; hence, higher chances of misalignment and poorer yield would be expected if the latter method (i.e., defining photoresist pattern) was employed. The wafer was then lapped down to approximately 100 μm before electron beam evaporation of both p-type and n-type ohmic contact layers. It is worth highlighting that the metallic p-pad, which was needed for probing or wire bonding, was designed to be as small as possible (80 × 80 μm2 in this case). This is because it contributed to the parasitic capacitance and was thus detrimental to the modulation bandwidth of the EAM devices. Finally, the wafer was cleaved into a ridge length of 1,700 μm (i.e., 1.7 mm) for device characterization. For higher yield and easier coupling purposes, the widths of the waveguides fabricated (WG width) were set at 7 μm. The effective index for a 7-μm-wide rib waveguide with an etch depth of 1.2 μm is approximately 3.325 and is still sufficiently narrow to hold single-mode propagation as shown in the simulation in Figure 1 (bottom left). However, careful alignment and cleaving was still necessary in order to avoid exciting higher order modes . Although in actual fabrication the etch depth is 1.4 μm, 0.2 μm has been omitted in this simulation because that is for the GaAs contact layer of higher refractive index and sufficiently far away from the inserted light source that it need not be included when simulating the mode propagation. The microscopic plan view of the QD-EAM devices that were designed as basic top-bottom p-i-n elements as illustrated in Figure 1 (bottom right). The pad sizes of the devices are approximately 80 × 80 μm2 which is sufficiently large for probing and wire bonding purposes but small enough to avoid inducing additional capacitance to the device.
Results and discussion
A good indication of the single-mode propagation obtained was by observing the single-mode Fabry-Perot (FP) spectrum as shown in the inset of Figure 3. The calculated propagation loss based on the respective FP spectra was 4.0 dB/cm for AG, 3.7 dB/cm for 600A, and 3.0 dB/cm for 750A at the wavelength range of 1,308 to 1,315 nm. The improvement in the propagation loss indicated the diffusion of the QD layers and an unintentional passivation of the device. When measuring the propagation loss, shorter waveguides from the batch of devices were cleaved instead of using actual DUTs. This was because longer devices will give much finer mode spacing, and this would result in less accurate data.
Note that although the DC extinction ratio of 600A (750A) was reduced to less than 70% (30%) of its original modulation ability, RF measurement on the devices was still possible due to lower propagation loss after annealing. The 3-dB bandwidth of both 600A and 750A is approximately 1.6 GHz. Noting that these are preliminary RF results, similar frequency responses of approximately 1.6 GHz for both 600A and 750A might be due to the non-optimized WG structures and RF matching. That is, the obtained RF performance was limited by the device design and not by the QD materials. Therefore, we believe that an improvement in the high-speed performance will be expected following the optimization of QD waveguide design and improved RF matching.
The realization of RF measurement on the processed (annealed) lumped-element QD-EAM confirms the prospect of QD epiwafer in monolithic integration for future references. By applying low-cost intermixing, such integration will have low insertion loss and polarization-independent properties . This is because the integrated devices would actually be made from the same epilayers unlike other types of integration. Therefore, the EAMs would naturally be tuned to the same polarization as that of the emitted radiation from the corresponding QD lasers, and improved extinction ratio may even be observed due to the improved absorption strength of the same platform that integrated devices share.
In this work, we investigated the effects of annealing on the static and dynamic performances of lumped-element QD-EAM operating at the wavelength of 1.3 μm. The extinction ratio at −8 V (propagation loss) for the as-grown, 600°C, and 750°C DUTs was found to be 10 dB (4.0 dB/cm), 7 dB (3.7 dB/cm), and <3 dB (3.0 dB/cm), respectively. Hence, both the extinction ratio and the insertion loss decrease upon increase in annealing temperature. Most significantly, the 3-dB response of the 750°C-annealed lumped-element QD-EAM was found to be 1.6 GHz at zero reverse bias voltage. This suggests a cost- and design-effective solution to enhance transmission and will be beneficial for researchers working on the implementation of QD-EAMs in monolithic integration through the intermixing process method.
This work was supported in part by the DSTA Defense Innovative Research Project (POD0613635).
- Chu Y, Thompson MG, Penty RV, White IH, Kovsh AR: 1.3 μm quantum-dot electro-absorption modulator. In CLEO’07: Conference on Lasers and Electro-Optics: May 6–11 2007; Baltimore. Piscataway: IEEE; 2007:1–2.Google Scholar
- Ngo CY, Yoon SF, Loke WK, Cao Q, Lim DR, Wong V, Sim YK, Chua SJ: Investigation of semiconductor quantum dots for waveguide electroabsorption modulator. Nanoscale Res Lett 2008, 3: 486–490. 10.1007/s11671-008-9184-7View ArticleGoogle Scholar
- Ramdane A, Devaux F, Souli N, Delprat D, Ougazzaden A: Monolithic integration of multiple-quantum-well lasers and modulators for high-speed transmission. IEEE J Sel Top Quantum Electron 1996, 2: 326–335. 10.1109/2944.577388View ArticleGoogle Scholar
- Chacinski M, Westergren U, Stoltz B, Thylen L: Monolithically integrated DFB-EA for 100 Gb/s Ethernet. IEEE Electron Device Lett 2008, 29: 1312–1314.View ArticleGoogle Scholar
- Ngo CY, Yoon SF, Lee SY, Zhao HX, Wang R, Lim DR, Wong V, Chua SJ: Electroabsorption characteristics of single-mode 1.3-μm InAs-InGaAs-GaAs ten-layer quantum-dot waveguide. Photonics Technology Letters IEEE 2010, 22: 1717–1719.View ArticleGoogle Scholar
- Ngo CY, Yoon SF, Loke WK, Cao Q, Lim DR, Wong V, Sim YK, Chua SJ: Characteristics of 1.3 μm InAs/InGaAs/GaAs quantum dot electroabsorption modulator. Appl Phys Lett 2009, 94: 143108. 10.1063/1.3119186View ArticleGoogle Scholar
- McKerracher I, Wong-Leung J, Jolley G, Fu L, Tan HH, Jagadish C: Selective intermixing of InGaAs/GaAs quantum dot infrared photodetectors. J Quantum Electronics 2011, 47: 577–590.View ArticleGoogle Scholar
- Cao Q, Yoon SF, Liu CY, Tong CZ: Effects of rapid thermal annealing on optical properties of p-doped and undoped InAs/InGaAs dots-in-a-well structures. J Appl Phys 2008, 104: 033522–033526. 10.1063/1.2963689View ArticleGoogle Scholar
- Song P, Lian J, Gao S, Li P, Wang X, Wu S, Ma Z: PECVD grown SiO2 film process optimization. In SOPO’11: Symposium on Photonics and Optoelectronics: May 16–18 2011; Wuhan. Piscataway: IEEE; 2011:1–4.Google Scholar
- Lee SY, Yang H, Li YC, Mei T: Integration of multimode interference device with electroabsorption modulators as simple switches. In AOM’10: OSA-IEEE-COS Advances in Optoelectronics and Micro/Nano-Optics: December 3–6 2010; Guangzhou. Piscataway: IEEE; 2010:1–4.Google Scholar
- Wang Y, Djie HS, Ooi BS: Quantum-confined Stark effect in interdiffused quantum dots. Appl Phys Lett 2006, 89: 151104. 10.1063/1.2358296View ArticleGoogle Scholar
- Wang Y, Negro D, Sjie HD, Ooi BS: Quantum-confined Stark effects in interdiffused semiconductor quantum dots. In Proc of SPIE. Volume 6468. Bellingham: SPIE; 2007:64681C.Google Scholar
- Vazquez C, Aramburu C, Galarza M, Lopez-Amo M: Experimental assessment of access guide first-order mode effect on multimode interference couplers. Optical Engineering 2001, 40: 1160–1162. 10.1117/1.1385338View ArticleGoogle Scholar
- Yang T, Tatebayashi J, Aoki K, Nishioka M, Arakawa Y: Effects of rapid thermal annealing on the emission properties of highly uniform self-assembled InAs/GaAs quantum dots emitting at 1.3 μm. Appl Phys Lett 2007, 90: 111912–111913. 10.1063/1.2713135View ArticleGoogle Scholar
- Shin DS, Yu PKL, Pappert SA: High-power electroabsorption modulator using intra-step-barrier quantum wells. J Appl Phys 2001, 89: 1515. 10.1063/1.1333719View ArticleGoogle Scholar
- Wood TH, Pastalan JZ, Burrus CA Jr, Johnson BC, Miller BI, de Miguel JL, Koren U, Young MG: Electric field screening by photogenerated holes in multiple quantum wells: a new mechanism for absorption saturation. Appl Phys Lett 1990, 57: 1081. 10.1063/1.103539View ArticleGoogle Scholar
- Sonnet AM, Khayer MA, Haque A: Analysis of compressively strained GaInAsP-InP quantum-wire electro-absorption modulators. Quantum Electronics IEEE J 2007, 43: 1198–1203.View ArticleGoogle Scholar
- Welstand RB, Pappert SA, Nichols DT, Lembo LJ, Liu YZ, Yu PKL: Enhancement in electroabsorption waveguide modulator slope efficiency at high optical power. Photonics Technology Letters IEEE 1998, 10: 961–963.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.