10.1117/2.1201206.004285

Germanium as the unifyingmaterial for silicon photonicsJurgen Michel and Marco Romagnoli

Germanium, the key enabling material for high-speed photonic links,can be fully integrated into a silicon CMOS process.

In recent years, high-performance active devices based ongermanium (Ge) have been developed for monolithic integrationinto silicon (Si)-based photonic systems. Most of this progressis based on the development of epitaxial Ge growth directlyon Si.1, 2 The initial Ge-based devices were photodetectors us-ing the high absorption coefficient of Ge up to wavelengths ofabout 1550nm. These devices have shown larger than 90% in-ternal quantum efficiencies at 1550nm and bandwidth above30GHz.3, 4 Today, monolithically integrated Ge photodetectorsare ubiquitous and can be found in active optical cables or fullyintegrated with transimpedance amplifiers for telecoms applica-tions.

The next novel Ge-based device was an electro-absorptionmodulator that could be based either on the capacitive electro-optical quantum-confined Stark effect5 or the Franz-Keldysh ef-fect in Ge.6 The best performance for a Franz-Keldysh-basedGe modulator is 30GHz bandwidth7 without resorting to res-onant enhancement as required in plasma dispersion-based Simicroring modulators. Both Ge detectors and modulators are in-tegrated with a Si waveguide, resulting in very low capacitance(of the order of a few femtofarads). If directly integrated with theanalog circuit (driver or transimpedance amplifier), power con-sumption is reduced to a few tens of femtojoules per bit. Thatshould be compared with consumption of picojoules per bit inconventional components based on wire bonding.

We recently demonstrated a Ge-based device, an electricallypumped laser that can be fully integrated into a Si CMOSprocess.8, 9 The lasing in Ge is enabled by strain and n-typedoping to allow dominant direct bandgap recombination in anindirect semiconductor.

The Ge photodetector is the most advanced Ge-basedphotonic device and can already be found integrated with SiCMOS. Integration of Ge modulators will most likely followsoon. The Ge laser is the least advanced device, and more

Figure 1. Wavelength performance range for modulators and detectorsusing different material compositions. Si: Silicon. Ge: Germanium.

development is necessary to increase its reliability and enablefull CMOS integration.

Integrating all Ge devices into one photonic circuit re-quires designing a process flow for fabricating the system.The biggest challenge is to determine the material composi-tion and doping. Strained Ge detectors (0.25%) work best forwavelengths below 1560nm due to the direct band edge on-set at that range. Ge modulators based on the Franz-Keldysheffect will perform best in the same wavelength range whenadding 0.7% Si in Ge.6 The direct band edge of the materialswould be at about 1480nm. A detector with this particular com-position will not be very efficient at absorbing light at 1550nm.While Ge will do so within 5�m, the absorption length for 0.7%Si in Ge is more than five times longer than for 100% Ge. Ina waveguide configuration, the path length of the detector’slight can be extended without a significant performance penalty,mainly because the capacitance is still very low due the device’ssmall size. The benefit of a detector waveguide configuration hasalready been demonstrated.3

We grew Ge detectors and modulators in silicon oxidetrenches with a width of 500nm and lengths between 5 and30�m using ultra-high vacuum chemical vapor deposition. We

Continued on next page

10.1117/2.1201206.004285 Page 2/2

added 0.7% Si to the Ge during growth to tune the modulatorresponse to 1550nm. The Ge devices were fabricated after thesilicide formation but before the back-end-of-line processing.10

We found that limiting the operation range to about 15nm en-ables use of the same GeSi composition for both modulatorsand detectors with a waveguide configuration. Fabrication isthus achieved in a single Ge process sequence, limiting the ther-mal budget and simplifying process flow. Figure 1 shows thewavelength performance range we obtained for modulators anddetectors using different material compositions.

Ge detectors and modulators rely on undoped Ge, but Gelasers need high n-type doping for gain. As long as high n-type Ge can only be grown epitaxially (i.e., in layers), simul-taneous Ge growth for detectors, modulators, and lasers is notpossible. However, the thermal budget for Ge lasers is signif-icantly lower than that for Ge detectors and modulators dueto the high diffusivity of n-type dopants in Ge, requiring lowprocess temperatures to prevent out-diffusion.11 Consequently,it is possible to fabricate Ge lasers after the Ge detectors andmodulators without decreasing their performance. Figure 1 alsoshows the measured gain spectrum for n-type Ge with a dopinglevel of 4 � 1019cm�3. The spectrum spreads over nearly 200nmand covers the performance range of all GeSi compositions fordetectors and modulators.

In summary, using Ge as a material for the most essentialactive photonic components enables development of an inte-grated, CMOS-compatible process flow without adding newmaterials or processes that reduce the yield and reliability ofthe final photonic and electronic systems. We have shown howto fully integrate source, modulation, detection, and electricalnetworks in a CMOS-compatible process in which all packag-ing costs (typically laser, electrical components, and wiring) areeliminated. The reduced cost, together with energy-efficient de-sign, shows that photonic interconnection is a possible solutionto poor scaling of on-chip wires and I/O bandwidth density forfuture technology nodes and promises manufacturability of verylarge volume applications.

Our initial feasibility demonstration is a photonic link be-tween a Ge laser and a Ge photodetector. Moving forward, wewill add active and passive photonic devices to show a fullymonolithically integrated link with increasing complexity.

Author Information

Jurgen Michel and Marco RomagnoliMassachusetts Institute of Technology (MIT)Cambridge, MA

Jurgen Michel is a senior research scientist in MIT’s Micropho-tonics Center. He was formerly at AT&T Bell Laboratories. Heearned his diploma at the University of Cologne and his doc-torate and habilitation at the University of Paderborn. He hasco-authored more than 200 refereed scientific papers.

Marco Romagnoli is a visiting scientist at MIT, developing anoptically integrated multiprocessor chip, and director of theBoston operations of PhotonIC Corp. He has more than 25 yearsof experience in the research field, especially in the area of pho-tonic technologies. He holds a physics degree from the Univer-sity of Rome (La Sapienza). He is the author of more than 160journal papers and conference contributions, and is an inventoron more than 40 patents. He is on the technical committee ofmajor photonics conferences and has served as expert evaluatorfor the European Commission’s 6th Framework Programme. Hewon the Philip Morris prize for optical innovation in 1994.

References

1. L. Colace et al., Ge/Si (001) photodetector for IR light, Solid State Phenom. 54, p. 55,1997.2. H. C. Luan et al., High-quality Ge epilayers on Si with low threading-dislocation den-sities, Appl. Phys. Lett. 75, p. 2909, 1999.3. J. F. Liu et al., Waveguide-integrated Ge p-i-n photodetectors on a silicon-on-insulatorplatform, Opt. Valley China Int’l Symp.Optoelectron., pp. 1–4, 2006.4. D. Feng et al., High-speed Ge photodetector monolithically integrated with large cross-section silicon-on-insulator waveguide, Appl. Phys. Lett. 95, p. 261105, 2009.5. Y.-H. Kuo et al., Strong quantum-confined Stark effect in germanium quantum-wellstructures on silicon, Nature 437, pp. 1334–1336, 2005.6. J. F. Liu et al., Waveguide-integrated, ultra-low energy GeSi electro-absorption modu-lators, Nat. Photon. 2, p. 433, 2008.7. N.-N. Feng et al., 30GHz Ge electro-absorption modulator integrated with 3�m silicon-on-insulator waveguide, Opt. Express 19, p. 7062, 2011.8. J. Michel et al., An electrically pumped Ge-on-Si laser, Opt. Fiber Commun. Conf.Exp/Nat’l Fiber Opt. Eng. Conf., 2012.9. R. Camacho-Aguilera et al., An electrically pumped germanium laser, Opt. Express20 (10), p. 11316, 2012.10. M. A. Beals et al., Process flow innovations for active photonic device integration inCMOS, Proc. SPIE 6898, p. 689804, 2008. doi:10.1117/12.77457611. Y. Cai et al., High n-type doped germanium: dopant diffusion and modeling, J. Appl.Phys. (Accepted for publication.)

c 2012 SPIE