GeSn alloys exhibit significant potential for mid-infrared devices. The increase of Sn content and the application of tensile strain can convert such alloys into direct bandgap semiconductors. However, a virtual Ge buffer substrate and a grading of the tin concentration are generally required for the growth of high tin content GeSn layers. This growth approach leads to misfit dislocations in the vicinity of the optically active GeSn layer on top and to a degraded optical confinement. In this work, we demonstrate a CMOS-compatible transfer of 200 mm GeSn wafers with a 14-16 % tin content. A comparative study between germane and digermane precursor gases for GeSn growth is also conducted. The crystallinity and composition of as-grown and transferred GeSn layers are characterized using x-rays, Raman and photoluminescence spectroscopies, Atomic Force Microscopy and Transmission Electron Microscopy.
Germanium-tin (GeSn) alloys grown on silicon wafers are promising for light emission (Light Emitting Diodes and lasers) and detection (photodiodes) in the 2 to 5 mu m wavelength range. Strain is a very efficient lever that has a definite impact on semiconductor properties such as band gap, densities of states, oscillator strength, index of refraction and so on. In this work, we show on demand control in the % range and in a reversible way of uniaxial strains in free standing GeSn/Ge resonators. We modulate the strain level by controlling the thermal expansion of Ge strain reservoirs, acting as reconfigurable stretching arms, that are placed on both sides of the central GeSn resonator. Room temperature Raman analysis shows a monotonous decrease of strain in pre-strained resonators upon arm heating. Lasing is observed in the devices at low temperature. More importantly, we demonstrate a blueshift of the laser gain curve, in line with the strain decrease.
Recently, all-group-IV (Si)GeSn alloys attracted great attention as materials for Infra-Red optoelectronics monolithically integrated on Si substrates. In this work, we present the fabrication and the electro-optical characterization of direct bandgap GeSn photodiodes with 15.4% of Sn grown on Ge Strain-Relaxed Buffers, themselves on 200 mm Si(001) wafers. The Ge0.846Sn0.154 photodetectors have a cutoff wavelength of 3.5 mu m, e.g., they are suitable for methane detection around 3.3 mu m. At this wavelength, their specific detectivity D* at room temperature is 3.76 x 10(7) cm.Hz(1/2).W-1. This detectivity is 60 times better than that of previously reported photodetectors with equivalent Sn content. When such Ge0.846Sn0.154 photodiodes are placed in a gas cell together with a commercial Light Emitting Diode emitting at 3.3 mu m, the system presents a limit of detection for methane of 1 600 parts per million with a noise density of 0.78%.Hz(-1/2).
Tin segregation can hinder the use of Ge1-xSnx in devices. A good understanding of this phenomenon is necessary to avoid this undesired phenomenon. Using Ge0.83Sn0.17 micro-disks as a model system, the origin and dynamics of tin segregation during thermal annealing was studied at the nanometer scale. Temperature-steps annealing showed that the onset of segregation occurred at temperatures typically 90 degrees C-95 degrees C higher than the growth temperatures. Interrupted annealing using short 450 degrees C pulses evidenced that segregation always started at single spots and propagated over the entire micro-disk surface in a few seconds. The spatial correlation between dislocations and segregation suggested that such defects were the initiators of such a process. Reducing the density of dislocations in the Ge1-xSnx layers should limit tin segregation during thermal annealing, facilitating the integration of this alloy into technological processes requiring temperatures well above the layers' growth temperature.
Direct bandgap Ge 0.86 Sn 0.14 photodiodes were fabricated on Si substrates using layer transfer. The layer transfer did not degrade the noise of the initial photodiode stacks, while the underlying metallic mirrors improved the responsivity at 300 K by a factor of 2.9 at 2.5 µm.
The reconstruction of thick GeSn crystalline layers implanted with phosphorus is explored. Our study demonstrates (i) the potential for recrystallizing amorphized GeSn crystal under specific conditions and (ii) functional LEDs incorporating an ex-situ doped GeSn injection layer. Mid-infrared direct band-gap light-emitting diodes are compared with either ex-situ GeSn:P or in-situ Ge:P top contact layers.
Silicon photonics is widely used for near InfraRed (IR) applications up to 1.6 mu m. It plays a key role in short-range optical data communications. However, silicon photonics does not really address mid-IR applications, particularly in the 1.6-5 mu m wavelength range. This spectral region is essential for environmental/life sensing and safety applications relying on the optical features of molecular vibrations, the aim being to discern and categorize complex chemical entities. Growing markets for such analysis prioritise sensitivity, specificity, compactness, energy-efficient operation and cost effectiveness. The need for a CMOS-compatible integrated photonic platform for the mid-IR is obvious. Such fully-group-IV semiconductor platform should include low-loss guided interconnects, detectors, modulators and, critically, efficient integrated light sources. This paper provides a comprehensive review of recent advances in GeSn-based mid-IR silicon-compatible devices, including optically and electrically pumped lasers, light-emitting diodes and photodetectors. It also discusses the principles underlying these developments, with focuses on material growth techniques and processing methods.
GeSn-based group-IV alloys are attracting great attention in the Si photonics community, as they are considered to be compatible with Complementary Metal Oxide Semiconductor (CMOS) technology. Alloying germanium with more than 8% of tin (Sn) results in direct bandgap semiconductors, with some optical gain in devices such as lasers. Recently, room temperature optically pumped lasing was achieved thanks to high Sn content stacks. GeSn alloys can be used for near, short and mid wavelength infra-red spectral range operation, notably for gas detection. Current efforts are focused on electro-optical GeSn IR devices such as light-emitting devices (LEDs), electrically pumped lasers and photodetectors. In this paper, we evaluate the impact of various types of in situ n-type doped carrier injection layers on top of GeSn direct bandgap LEDs. More precisely, we compare the performances of GeSn:P and Ge:P –capped mid IR LEDs. Using reduced pressure chemical vapor deposition and metastable growth conditions (e.g. fast growth rates at low temperature), high crystalline quality GeSn layers were grown on 200 mm diameter Ge-buffered Si(001) wafers (Figure a). In situ n-type doped Ge layers (sample A) and GeSn layers (sample B) were used to inject carriers into direct band gap Ge0.87Sn0.13 active layers beneath (Figure b). I(V) curves (Figure c) and Electro-Luminescence spectra (Figure d) were compared for sample B (GeSn:P) and sample A (Ge:P). Very similar I(V) behaviors were observed under forward bias for both samples. However, a higher dark current was measured under reverse bias for sample B than for sample A. This could be due to a higher number of defects in the thicker capping layer of sample B (240 nm) compared to that of sample A (50 nm). However, a 2-fold increase in the EL signal was obtained for sample B than sample A (Figure d). Temperature dependence electroluminescence measurements and atom probe tomography data will be used to explain the electroluminescence behavior differences between Samples A and B. Finally, even higher Sn content LEDs were fabricated to emit light at 3.3 µm (Figure e). A direct band gap Ge0.846Sn0.154LED (Sample C) with an in situ n-type doped Ge cap was placed in a gas cell filled with diluted methane. Its emission overlapped well with the absorption of methane (Figure f). The methane detection limit of our setup was evaluated (data not shown). To further increase the emitted power of GeSn LEDs, current spreading was studied for different top contact geometries. Emission maps collected by an InSb camera at room temperature showed a definite dependence of light emission on electrical contact geometry (Figure g). Investigations are underway to further increase light extraction from LEDs and improve their performances for use in environmental sensors. Acknowledgement: This work was supported by the European Union’s Horizon 2020 LASTSTEP Project under the grant agreement ID: 101070208, the EDEN Carnot project and the CEA DRF-DRT Phare project. We gratefully acknowledge the clean room staff from LETI and IRIG for their technical support. Figure 1
GeSn alloys are the most promising direct band gap semiconductors to demonstrate full CMOS-compatible laser integration with a manufacturing from Group-IV materials. Since the first demonstration of lasing with GeSn alloys up to 100 K, many researches were devoted to increase the laser operation up to room temperature. We will discuss the band sructure requirements and the practical issues that have to be addressed in order to reach robust gain with increasing temperature. We show that misfit defects managment and strain engineering are key ingredients. For that purpose we developped a GeSn-On-Insulator platform, that combine strain engineering , defective interfacial layer removal and laser resonator designs ad fabrication. Here we show that room temperature lasing, up to 300 K, can be obtained in microdisk resonators fabricated on a GeSnOI layer both with using high Sn-content in the gain medium, e. g. 17% or with applying tensile strain to a layer with lower Sn-content of 14%.
Ge 0.85 Sn 0.15 -based stacks were grown by reduced pressure chemical vapor deposition on 200 mm Si wafers. The resulting photodiodes showed detectivities above 10 8 cm.H z 1/2 .W −1 at 1.55 µm and a cutoff wavelength of 3.5 µm. This detection range opens up promising perspectives for future all-group-IV gas sensors.
Owing to their direct band gaps, (Si)GeSn all-group-IV alloys are promising candidates for light sources, photodetectors and modulators monolithically integrated onto a CMOS-compatible mid-infrared photonic platform. Several research teams have demonstrated optically pumped GeSn lasers, and, more recently, an electrically pumped GeSn laser at low operating temperature. Here, we studied Ge0.85Sn0.15-based light emitting diodes (LEDs) and photodiodes (PDs) operating at room temperature. The stack was grown on a p-doped Ge strain-relaxed buffer at low growth temperatures (below 350°C) in a 200 mm chemical vapor deposition tool. Fabricated GeSn devices were characterized at room temperature with a Fourier-transform infrared spectrometer (FTIR) and an InSb detector. The spectral response of the FTIR InSb detector was calibrated with respect to a Deuterated Triglycine Sulfate detector (DTGS). This spectral response was then used to correct Ge0.85Sn0.15 LEDs emission spectra with emission maximum at 3.3 μm. The cutoff wavelength at 3.7 μm of the GeSn photodiode was finally obtained (at 0V bias) after correction of the Globar incident light spectrum. Such emission and detection open up promising perspectives for all-group-IV LEDs and PDs in applications such as gas sensing.
The structural properties of CVD-grown (Si)GeSn heterostructures were assessed thanks to scanning transmission electron microscopy at the nanometer scale. Quantitative energy dispersive x-ray (EDX) spectroscopy together with precession electron diffraction and geometrical phase analysis (GPA) were performed to probe the chemical and structural properties of the different layers. Results presented in this paper demonstrated the advantages of a multilayer structure, with successive layers grown at decreasing temperatures in order to gradually accommodate the in-plane lattice parameter and incorporate more and more Sn into the stack. It was shown how the GeSn emissive layer could be manufactured with low plastic deformation and a high relaxation rate, necessary for better light emission performances. SiGeSn alloys used as confinement barriers around the emissive layer were also investigated. For such thin layers, we showed the importance of the starting lattice parameter (SLP) prior to the growth on their composition. Indeed, higher SLPs resulted, for the very same process conditions, into higher Sn contents and lower Si contents. The interest in combining EDX, which was accurate enough to detect slight chemical concentration variations, and GPA, for local strain analyses, was clearly demonstrated. Present results will be very useful to predict and control the bandgap and structural quality of (Si)GeSn materials and, in turn, device properties.
Efficient and cost-effective Si-compatible lasers are a longstanding wish of the optoelectronic industry. In principle, there are two options. For many applications, lasers based on III-V compounds provide compelling solutions, even if the integration is complex and therefore costly. However, where low costs and also high integration density are crucial, group-IV-based lasers-made of Ge and GeSn, for example-could be an alternative, provided their performance can be improved. Such progress will come with better materials but also with the development of a more profound understanding of their optical properties. In this work, we demonstrate, using Ge microbridges with strain up to 6.6%, a powerful method for determining the population inversion gain and the material and optical losses of group IV lasers. This is done by deriving the values for the injection carrier densities and the cavity losses from the measurement of the change of the refractive index and the mode linewidth, respectively. We observe a laser threshold consistent with optical gain. Material loss values are obtained from a tight-binding calculation. Lasing in Ge-at steady-state-is found to be limited to low temperatures in a narrow regime of tensile strain at the crossover to the direct-band-gap band structure. We explain this observation by parasitic inter-valence-band absorption that increases rapidly with higher injection densities and temperature. N-doping seems to reduce the material loss at low excitation, but it does not extend the lasing regime. We also discuss the impact of the optically inactive carriers in the L-valley on the linewidth of group IV lasers.
Tin segregation in Ge1-xSnx alloys is one of the major problems potentially hindering the use of this material in devices. Ge1-xSnx microdisks fabricated from layers with Sn concentrations up to 16.9% underwent here annealing at temperatures as high as 400 °C for 20 min without Sn segregation, in contrast with the full segregation observed in the corresponding blanket layers annealed simultaneously. After annealing, no changes in the elemental composition of the microdisks were evidenced. An enhancement of the total integrated photoluminescence, with no modifications of the emission energy, was also observed. These findings show that microstructuring offers a completely new path in maintaining the stability of high Sn concentration Ge1-xSnx layers at temperatures much higher than those used for growth. This approach enables the use of thermal annealing processes to improve the properties of this alloy in optoelectronic devices (such as light emitting diodes, lasers, photodetectors, or modulators). It should also facilitate the integration of Ge1-xSnx into well-established technologies requiring medium temperature processes. The same strategy may help to prevent Sn segregation during high temperature processes in similar metastable alloys.
GeSn alloys are promising materials for light emitters monolithically grown on silicon. In this work, we demonstrate room temperature (RT) lasing in a GeSn hetero-structure with 17.2% of Sn. We report a threshold of 3.27 MW cm−2 at 305 K with peak emission at 353 meV. We ascribe these improvements to a higher tin concentration in the GeSn active layer with lower Sn content barriers on each side and to a better thermal dissipation provided by an adapted pedestal architecture beneath the GeSn micro-disk. This outcome is a major milestone for a fully integrated group-IV semiconductor laser on Si.
We report on lasing in two types of Germanium-Tin (GeSn) photonic crystal lasers, with band-edge and H4 hexagonal cavities. A GeSn 16% step-graded structure was used as the optical gain material. The strong out-of-plane emission observed in the band-edge cavity was attributed to the thick optical gain layer, in addition with the presence of leaky band-edge modes above the light cones and a high photonic density of states. The maximum lasing temperature in both types of photonic crystal lasers, 180 K, was lower than in a micro-disk laser fabricated from the same stack, i.e., 230 K. Meanwhile, the lasing thresholds at low temperature (15 K) were very similar (120 kW/cm2 for a band-edge cavity and 140 kW/cm2 for a H4 cavity, to be compared with 134 kW/cm2 for a micro-disk cavity). Since the injected carrier density is governed by the pump power, we suggest the strong variation of optical gain at low temperature, for a small variation of injected carrier density, as a possible cause of such a phenomenon.
GeSn alloys are the most promising direct band gap semiconductors to demonstrate full CMOS-compatible laser integration with a manufacturing from Group-IV materials. Here, we show that room temperature lasing, up to 300 K, can be obtained with GeSn. This is achieved in microdisk resonators fabricated on a GeSn-On-Insulator platform by combining strain engineering with a thick layer of high Sn content GeSn.
A CMOS compatible, direct bandgap material for optical interconnects can be obtained by alloying Ge with Sn 1 , applying tensile stress to Ge 2 or both 3 . Lasing in GeSn was demonstrated in 2015 4 by Wirths et al., followed in 2020 by electrically pumped lasing up to 100K 5 and, in 2022, optically pumped lasing at room temperature 6,7 . In-situ doped SiGeSn might offer high dopant incorporation, while delivering good electronic confinement, improving thereby the performances of devices. Such doped layers can be used in photodetectors 8–10 , light-emitting diodes 11–13 and modulators 14,15 operating at wavelengths higher than 1.55 µm, enabling their use in future CMOS compatible lab-on-a-chip devices with integrated light sources 5,16,17 . The in-situ doping of SiGeSn was compared to that of GeSn. All layers were grown at 349 °C, 100 Torr in a 200 mm Epi Centura 5200 RP-CVD tool from Applied Materials. Ge strain relaxed buffers were used to accommodate the lattice mismatch between (Si)GeSn and the Si substrates 18 . The F(Ge 2 H 6 )/F(H 2 ), F(Si 2 H 6 )/F(H 2 ) and the F(SnCl 4 )/F(H 2 ) Mass-Flow Ratios (MFRs) were constant at 7.92x10 -4 , 1.25x10 -3 , and 4.69x10 -5 , respectively. In-situ doped SiGeSn Growth Rates (GRs), shown in Figure 1 (a) , were around 30 nm min. -1 . They were below that of GeSn:B and GeSn:P (40 nm min. -1 ). The latter significantly increased for high dopant flows. Meanwhile, SiGeSn:B GR slightly increased and SiGeSn:P GR decreased as the dopant flow increased. B 2 H 6 and PH 3 might have opened surface sites for GeSn and SiGeSn:B, while the formation of gas phase intermediates might have reduced the SiGeSn:P GR. Interestingly, the surface quality improved significantly for in-situ doped SiGeSn, reaching the same quality as that of GeSn for high dopant flows, as shown in Figure 1 (b) . Surfaces had RMS roughness values below 0.40 nm, close to that of GeSn, with a full surface cross-hatch recovery. There was, for in-situ doped GeSn, a Sn content reduction for high dopant flows ( Figure 1 (c) + (d) ) most likely because SnCl 4 was mass-transport limited 19 and not impacted by a larger amount of open surface sites. The influence of dopants on the layer composition was even more pronounced in SiGeSn:B. Si/Sn ratios of 3.5, with Si contents of up to 25%, were obtained, which should result in improved electrical confinement. The formation of Si and Ge gas phase intermediates might explain why Sn contents were higher, in SiGeSn:P, for high PH 3 flows. Such insights should yield better control of Si and Sn contents in stacks for optical or electronic purposes. Electrically active carrier concentrations c active of the order of 2x10 20 cm -3 were achieved in SiGeSn:B ( Figure 1 (e) ). These were seven times higher than the 3x10 19 cm -3 obtained for GeSn:B. For GeSn:P, c active was likely limited by the formation of Sn m P n V nanoclusters. It was at most 7x10 19 cm -3 ( Figure 1 (f) ). Four times higher c active values were obtained for SiGeSn:P with at most 3x10 20 cm -3 . No decrease of c active was observed for high PH 3 flows in SiGeSn:P. This might have been due to the formation of fewer Sn m P n V nanoclusters. Such c active values and better electrical confinement were used to fabricate (Si)GeSn based photodiodes with improved electroluminescent integrated intensity compared to photodiodes with doped Ge contact layers. Gassenq, A. et al. Appl. Phys. Lett. 109 , 242107 (2016). Elbaz, A. et al. Nat. Photonics 14 , 375–382 (2020). Chrétien, J. et al. ACS Photonics 6 , 2462–2469 (2019). Wirths, S. et al. Nat. Photonics 9 , 88–92 (2015). Zhou, Y. et al. Optica 7 , 924 (2020). Chrétien, J. et al. Appl. Phys. Lett. 120 , 051107 (2022). Bjelajac, A. et al. Opt. Express 30 , 3954 (2022). Li, X. et al. Photonics Res. 9 , 494 (2021). Zhou, H. et al. Opt. Express 28 , 10280 (2020). Wu, S. et al. IEEE J. Sel. Top. Quantum Electron. 28 , 1–9 (2022). Stange, D. et al. Optica 4 , 185 (2017). Oehme, M. et al. IEEE Photonics Technol. Lett. 26 , 187–189 (2014). Schwartz, B. et al. Opt. Lett. 40 , 3209 (2015). Bertrand, M. et al. 2020 IEEE Photonics Conference (IPC) 1–2 (IEEE, 2020). Zhou, H. et al. Opt. Express 28 , 34772 (2020). Casiez, L. et al. 2020 IEEE Photonics Conference (IPC) 1–2 (IEEE, 2020). Soref, R. Nat. Photonics 4 , 495–497 (2010). Hartmann, J. M. & Aubin, J. J. Cryst. Growth 488 , 43–50 (2018). Margetis, J. et al. Vac. Sci. Technol. A 37 , 021508 (2019). Figure 1
We investigate the recrystallization of thick phosphorus-implanted GeSn layers using 308 nm Ultraviolet Nanosecond Laser Annealing (UV-NLA). We identify the optimal annealing conditions leading to the reconstruction of Ge0.92Sn0.08 crystal amorphized by dopant implantation. The fully recrystallized GeSn layers present specific structures with localized tin and strain variations. Above the non-amorphized and unmelted Ge0.92Sn0.08 seed layer, a first highly tensile strained GeSn sublayer is formed, with a tin gradient from 2.5% up to 10.5%. Closer to the surface, a second sublayer consists of tin-enriched vertical structures in a Ge0.93Sn0.07 matrix. Laser annealing enables us to reverse the strain of the GeSn layer. The initial GeSn presents a compressive strain of −0.10%, while the recrystallized Ge0.93Sn0.07 matrix is tensile strained at 0.39%. UV-NLA presents the advantages of (i) local annealing that recrystallizes amorphized GeSn layers after implantation without excessive tin segregation and (ii) reversing the strain of epitaxial GeSn layers from compressive to tensile. Our results open up promising perspectives for the integration of GeSn mid-IR photonic devices.