Search

Dopant concentrations higher than 1x1019 cm-3 are required to improve the performances of various GeSn based devices such as photodetectors, electrically pumped lasers and so on. In this study, the in-situ Boron and Phosphorous doping of SiGeSn was investigated, building upon recent studies on in-situ B or P doped GeSn. The surfaces of intrinsic and lowly doped pseudomorphic SiGeSn layers were rough. By contrast, a cross hatch was recovered and surfaces as smooth as the Ge Strain-Relaxed Buffers underneath were obtained for the highest B2H6 or PH3 mass-flows. The surface Root Mean Square roughness and Zrange values were then as low as 0.36 nm and 2.86 nm for SiGeSn:B, and 0.47 nm and 4.60 nm for SiGeSn:P. In addition, Si contents as high as 25% were obtained, notably in SiGeSn:B layers. Dopants were almost fully electrically active in those SiGeSn:B and SiGeSn:P layers, with carrier concentrations as high as 2.0x1020 cm-3 and 2.7x1020 cm-3, respectively. For SiGeSn:P, the shortcoming of in-situ doped GeSn:P was overcome, that is the formation of electrically inactive SnmPnV clusters for high PH3 mass-flows. Such electrically active carrier concentrations will be beneficial for (Si)GeSn based devices, but also for all Group-IV based devices with extremely low thermal budget constraints.

Abstract The continued downscaling of silicon CMOS technology presents challenges for achieving the required low power consumption. While high mobility channel materials hold promise for improved device performance at low power levels, a material system which enables both high mobility n-FETs and p-FETs, that is compatible with Si technology and can be readily integrated into existing fabrication lines is required. Here, we present high performance, vertical nanowire gate-all-around FETs based on the GeSn-material system grown on Si. While the p-FET transconductance is increased to 850 µS/µm by exploiting the small band gap of GeSn as source yielding high injection velocities, the mobility in n-FETs is increased 2.5-fold compared to a Ge reference device, by using GeSn as channel material. The potential of the material system for a future beyond Si CMOS logic and quantum computing applications is demonstrated via a GeSn inverter and steep switching at cryogenic temperatures, respectively.

GeSn alloys are promising materials for CMOS-compatible mid-infrared lasers manufacturing. Indeed, Sn alloying and tensile strain can transform them into direct bandgap semiconductors. This growing laser technology however suffers from a number of limitations, such as poor optical confinement, lack of strain, thermal, and defects management, all of which are poorly discussed in the literature. Herein, a specific GeSn-on-insulator (GeSnOI) stack using stressor layers as dielectric optical claddings is demonstrated to be suitable for a monolithically integration of planar Group-IV semiconductor lasers on a versatile photonic platform for the near- and mid-infrared spectral range. Microdisk-shape resonators on mesa structures were fabricated from GeSnOI, after bonding a Ge0.9Sn0.1 alloy layer grown on a Ge strain-relaxed-buffer, itself on a Si(001) substrate. The GeSnOI microdisk mesas exhibited significantly improved optical gain as compared to that of conventional suspended microdisk resonators formed from the as-grown layer. We further show enhanced vertical out-coupling of the disk whispering gallery mode in-plane radiation, with up to 30% vertical out-coupling efficiency. As a result, the GeSnOI approach can be a valuable asset in the development of silicon-based mid-infrared photonics that combine integrated sources in a photonic platform with complex lightwave engineering.

One of the most important questions concerning the epitaxial growth of Si1-yCy or Si1-x-yGexCy is the ratio of carbon incorporated into substitutional and interstitial sites, which is highly dependent on...

The continued downscaling of silicon CMOS technology faces big challenges in achieving the required low power consumption. While high mobility channel materials hold promise for improved device performance at low power levels, to date a material system which enables both high mobility n-FETs and p-FETs, that is compatible with Si technology and can be readily integrated into existing fabrication lines is still missing. Here, we present high performance, vertical nanowire gate-all-around FETs based on the GeSn-material system grown on Si. While the p-FET performance is improved by exploiting the small band gap of GeSn as source yielding high injection velocities, the mobility in n-FETs is increased 2.5-fold compared to a Ge reference device, by using GeSn as channel material. The very high potential of the material system for future beyond Si CMOS logic and quantum computing applications is demonstrated via a GeSn inverter and steep switching at cryogenic temperatures, respectively.

The holy grail of group-IV light emission would be electrically pumped lasing at Room Temperature (RT). In CEA-LETI, the performance of p-i-n light emitting diodes is limited by the use of in-situ doped Ge layers beneath and above the active layer stack [1]. To overcome Sn segregation during the growth of the top electrode and, in thick GeSn/Ge stacks, avoid plastic relaxation, it would be ideal to switch over to in-situ doped GeSn grown at temperatures similar to that of the active region. We have therefore explored the in-situ boron and phosphorous doping of GeSn on Ge strain relaxed buffers with Ge2H6+SnCl4+ B2H6 and Ge2H6+SnCl4+PH3 chemistries. Growth pressure, temperature, F(Ge2H6)/F(H2) and F(SnCl4)/F(H2) Mass Flow Ratio (MFR) were 100 Torr, 349 °C, 7.92x10-4 and 4.78x10-5. X-Ray Diffraction (XRD) measurements were conducted to gain access to the GeSn composition and layer thickness, as shown in Fig. 1 and Fig. 2. As the GeSn peak position was determined by the Sn content and also the B or P substitutional concentrations, the actual layer composition could not be unambiguously determined. Therefore, the Sn content in the following is an apparent Sn content. Thickness fringes outline the high GeSn crystalline quality. As the F(B2H6)/F(Ge2H6) or the F(PH3)/2 F(Ge2H6) MFRs increased, the apparent Sn content stayed at first roughly around 6.5 %, as shown in Fig. 3. This value was the same for i-GeSn, GeSn:B and GeSn:P, in contrast to previous literature findings [2]. As the MFRs reached the highest values, the apparent Sn content reduced significantly for GeSn:B (from 6.5 % down to 4.6 %) and less significantly GeSn:P (from 6.6 % down to 5.6 %). The significantly smaller size of B compared to P (aB = 3.852 Å aP = 5.014 Å) reduced the compressive strain in GeSn:B 2.8 times more than in GeSn:P for the same dopant concentrations, explaining the more pronounced effect. We quantified the substitutional B and P incorporation by assuming that the Sn content was not impacted by B or P incorporation. The estimated B and P substitutional concentrations then increased from 2.8x1019 cm-3 up to 3.8x1020 cm-3 and from 2.5x1019 cm-3 up to 5.4x1020 cm-3 (highest four MFRs for GeSn:B and GeSn:P), shown in Fig. 4. Those values are close indeed to the highest substitutional B or P concentrations obtained in Ge:B or Ge:P layers grown at 350°C, 100 Torr with Ge2H6: 4.8x1020 cm-3 [3] and 4.1x1020 cm-3 [4], respectively. We determined elemental growth rates by multiplying the GR by the Ge or apparent Sn content to have a more detailed look at the growth mechanics (Fig. 5). The introduction of SnCl4 catalyzed the growth rate of GeSn compared to that of pure Ge, in line with the literature [4]. As the F(B2H6)/F(Ge2H6) or the F(PH3)/2F(Ge2H6) MFR increased, the Sn GR component stayed constant at around 2.5 nm min.-1. Meanwhile, the Ge GR component was constant at around 37 nm min.-1 for lower MFRs. For high dopant flows, the Ge GR component increased significantly for GeSn:B and reached 48 nm min.-1. The increase was less significant for GeSn:P, with at most 40 nm min.-1. This is to our knowledge, the first time an increase of the GeSn:P GR is reported for high dopant flows. Similar increases were found for Ge:P and Ge:B grown at 350°C, however [3-4]. The constant Sn GR component and the increase of the Ge GR component at high MFRs are likely due to an increased incorporation of Ge thanks to B and P opening surface sites (Indeed, the Ge GR component is surface rate limited, while the Sn GR component is reaction rate limited [5]). A cross-hatch along was found by Atomic Force Microscopy for all layers, which were otherwise smooth and featureless, as shown in Fig. 6 top for GeSn:B and Fig. 6 bottom for GeSn:P. The only exceptions were the GeSn:B samples grown at the three highest MFRs. There were then some islands, which were likely due to B and/or Sn surface segregation. Other studies will be performed to gain access to the Sn, P and B contents. [1] M. Bertrand et al., 2018 IEEE International Conference on Group-IV Photonics, pp. 45-46. [2] J. Margetis et al., Mater. Sci. Semicond. Process. 70, 38 (2017). [3] J.M. Hartmann et al., ECS PRIME 2020 Meeting Abstract ID 138196, Symposium G03. [4] J. Aubin et al., ECS J. Solid State Sci. Technol. 6, 1 (2017). [5] J. Margetis et al., J. Vac. Sci. Technol. A 37, 021508 (2019). Figure 1

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(Ge2H6)/F(H2), F(Si2H6)/F(H2) and the F(SnCl4)/F(H2) 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. B2H6 and PH3 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 SnCl4 was mass-transport limited19 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 PH3 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 2x1020 cm-3 were achieved in SiGeSn:B (Figure 1 (e)). These were seven times higher than the 3x1019 cm-3 obtained for GeSn:B. For GeSn:P, c active was likely limited by the formation of SnmPnV nanoclusters. It was at most 7x1019 cm-3 (Figure 1 (f)). Four times higher c active values were obtained for SiGeSn:P with at most 3x1020 cm-3. No decrease of c active was observed for high PH3 flows in SiGeSn:P. This might have been due to the formation of fewer SnmPnV 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

Doped Ge can be employed in the sources and drains of Ge pMOS transistors 1 or as p-type films in Ge p-i-n Photo-Detectors (PDs) 2. If the growth temperature is low enough, they can even be used in GeSn PDs and Light Emitting Devices 3. Those Short and Mid Wavelength Infra-Red devices are rapidly gaining interest especially since the demonstration of electrically pumped lasing in GeSn at temperatures up to 100K by the University of Arkansas 4. The electrically active carrier concentration in binary Si:P was significantly increased thanks to a dissolution with nanosecond laser annealing (NLA) of P clusters 5. Ge:B alloys with metastable, ultra-high substitutional concentrations of B can also be obtained thanks to in-situ doping 6. However, electrical activation in such Ge:B binaries was limited by the formation of electrically inactive clusters 7. In this work, we evaluated whether or not they could be dissolved by NLA. The melt threshold shifted from 0.875 Jcm-2 to 0.85 Jcm-2 as the substitutional B concentration increased (from Time Resolved Reflectivity maps during Ge:B NLA; Figure 1 (a)). This was likely due to differences in terms of crystalline quality, surface roughness and/or B concentration itself. Figure 1 (b) shows normalized ω-2ϴ scans around the (0 0 4) X-Ray Diffraction order. The Ge:B peak did not significantly change below the melt threshold at 0.825 Jcm-2. Above it, the Ge:B peak became less and less intense and vanished at the full melt threshold (1.05 Jcm-2). NLA at even higher Energy Densities (ED, 2.00 Jcm-2) did not result in a recovery of the Ge:B peak. At the melt threshold, localized surface structures melted, as shown by Atomic Force Microscopy in Figure 1 (c) with a 30 nm z-scale. Surface structures were rectangular with sides along the directions, as for the cross-hatch of the as-grown layer. The size of all surface structures was quite constant. Comparable surface structures were previously seen for SiGe 8. A High Resolution Transmission Electron Microscopy image of the Ge:B layer after NLA with an ED of 0.85 Jcm-2 is shown in Figure 1 (d). The Ge:B layer had a thickness of 33 nm, which was slightly lower than that of the as-grown layer, i.e. 39 nm. It might be that material agglomerated in the 12 nm high surface structures 8. Surface structures otherwise had a brighter contrast, which might be due to the accumulation of lighter B atoms in those surface structures. Well defined Fast Fourier Transform features outlined the somewhat good crystalline quality of NLA Ge:B. However, they were less bright than for a superior quality epitaxial layer. Laser annealing resulted in a redistribution of B, with the formation of electrically inactive clusters that did not contribute to strain. Accordingly, the sheet resistance increased by 70%, from 39.82 Ω/□ up to 68.62 Ω/□, when the layer melted. This corresponded to an electrically active carrier loss of around 50%, from 8.1x1020 cm-3 down to 3.8x1020cm-3, shown in Figure 1 (e). This behavior was independent of the as-grown substitutional B concentration. Even multiple shots with various energy densities at the same position, shown in Figure 1 (f), were not able to improve electrical activation. However, there was some slight improvements of the sheet resistance in the sub melt regime, which would need to be confirmed in future experiments. We have seen above that single shot NLA resulted in the formation of (i) surface structures becoming larger for high EDs and (ii) electrically inactive clusters not contributing to strain. With the use of multiple shots, surface structures merged and formed even larger surface structures. This led to more B redistribution and likely to the formation of more electrically inactive boron-interstitial clusters in the melt regime. In the sub melt regime, multi shot NLA might improve contact resistance. NLA in the melt regime therefore seemed not to be able to improve electrical activation in heavily in-situ boron-doped Ge layers as it did for heavily in-situ phosphorous doped Si. Vohra, A. et al. Jpn. J. Appl. Phys. 58, SBBA04 (2019). Osmond, J. et al. Appl. Phys. Lett. 95, 151116 (2009). Casiez, L. et al. IEEE Photonics Conference (IPC) 1–2 (IEEE, 2020). Zhou, Y. et al. Optica 7, 924 (2020). Rosseel, E. et al. ECS Trans. 64, 977–987 (2014). Hartmann, J.-M. et al. ECS Trans. 98, 203–214 (2020). Porret, C. et al. Phys. Status Solidi A 217, 1900628 (2020). Dagault, L. et al. ECS J. Solid State Sci. Technol. 8, P202–P208 (2019). Figure 1

In recent years, Ge-based group-IV alloys (GeSn, SiGeSn) have received a significant amount of attention as candidates to replace Silicon for future low power and high performance nanoelectronics [1]. The interest in these materials stems primarily from the fact that, by varying the Sn-content of the alloy, it is possible to precisely tune its bandgap from indirect to direct [2], which even opens up the possibility to switch the carrier transport from larger mass low mobility L-valley electrons to the lower mass and high mobility Γ-valley electrons. Adding Si atoms into GeSn alloys enables additional strain engineering by decoupling the lattice constant from the band gap and enables the fabrication of devices to target specific applications. Ge exhibits superior hole mobility over Si and GeSn is predicted to further improve carrier mobilities for both electrons and holes, while still retaining Si CMOS compatibility [3]. In this Abstract, we present the fabrication and characterization of Ge- and GeSn-based vertical gate-all-around (GAA) nanowire (NW) p-MOSFETs. Multilayer stacks of Ge and GeSn were grown on a Ge virtual substrate (Ge-VS) using industrial CVD reactors and subsequently characterized, confirming the high quality of the alloys. On these GeSn/Ge heterostructures, vertical GAA nanowire FETs were fabricated using a top-down approach. First, nanowires were defined by electron-beam lithography and subsequently etched anisotropically using reactive ion etching (RIE). The diameter of the nanowires was reduced by digital etching, consisting of repeated combined GeOx layer formation by plasma oxidation and removal in diluted HF solution. This way nanowires with a diameter down to 20 nm and a height of 210 nm were fabricated. A two-step process was employed for gate dielectric formation to ensure a low interface trap density: (i), deposition of a thin layer of Al2O3, followed by an O2-plasma post-oxidation step; (ii) deposition of a HfO2 dielectric layer to reach the required EOT (equivalent oxide thickness). TiN deposited by sputtering forms the gate metal. Planarization and isotropic dry etching were performed to remove the TiN on the top of the nanowire. After a second planarization step, NiGe-contacts were formed on the exposed top nanowire by Ni-deposition followed by a forming-gas annealing step. Finally, metal contacts for gate and source/drain were added. The resulting Ge-NW-pMOSFETs exhibit high electrical performances. A low subthreshold slope (SS) of 66 mV/dec, a low drain-induced barrier lowering (DIBL) of 35 mV/V and an I on/I off-ratio of 2.1×106 were measured for nanowires with a diameter of 20 nm. For 65 nm NWs, the I on/I off-ratio improves, which is attributed to the decreased contact resistance on top of the NWs, leading to larger on-currents. The peak transconductance for the Ge NWs reached ~190 µS/µm (V DS=-0.5 V). Adopting a GeSn/Ge-heterostructure, with GeSn on top of the nanowire used as source the device performances are strongly enhanced. The on-current I on was increased by ~32%, mostly due to the reduced contact resistivity of the smaller bandgap of GeSn compared to Ge. It was also observed that adopting GeSn alloys leads to an increase in transconductance, G max, to a respectable value of ~870 µS/µm, almost 3 times larger as reported to date for Ge NWs. Moreover, both SS and DIBL are improved by decreasing the NW diameter as a consequence of improved electrostatic gate control over the channel. These results demonstrate that the incorporation of GeSn into Ge-MOSFET technology yields a significant advantage and confirm its high potential for low-power-high-performance nanoelectronics. Fig. 1: (a) Schematic of the GAA nanowire FET based on a GeSn/Ge-heterostructure. (b) Optical image on the metallic contacts (c) Transfer curve of a Ge nanowire pFET with a diameter of 20 nm. The SS is 68 mV/dec and the DIBL is 35 mV/V. (d) Transfer curves of Ge0.92Sn0.08/Ge nanowire pFETs with a diameter of 65 nm and different EOTs. Acknowledgments The authors acknowledge support from the German BMBF project “SiGeSn NanoFETs”. References: [1] M. Liu et al. ACS Appl. Nano Mater. 4, 94-101 (2021) [2] S. Wirths et al. Nature Photonics 9, 88-92 (2015) [3] J. Kouvetakis, J. Menendez, A. V. G. Chizmeshya: Annu. Rev. Mater. Res. 36:497-554 (2006) Figure 1

International audience; Tensile strained GeSn lasers could be used as tunable infrared light sources in silicon photonics. We explore here the relationship between the uniaxial tensile strain applied to suspended microbridges and the maximum lasing temperature reached in GeSn 13% cavities.

Direct-band gap GeSn alloys have strong potentials as MIR light emitters. We investigate LEDs with 13% and 16% of tin. Light emission is observed up to room temperature for both types of LEDs. The Ge 0.84 Sn 0.16 LED is used in a gas sensor to detect CH4.

In CEA-LETI, GeSn/SiGeSn stacks are typically grown at 100 Torr with Ge2H6, Si2H6 and SnCl4 and temperatures in the 301°C-363°C range. Ge2H6 has however a reputation of being unstable for high concentrations (10% in H2). The short shelf time can be adressed by switching over to lower concentrations and regularly changing bottles, which is complex as Ge2H6 is costly and complicated to order. One solution could be to switch from digermane to germane, which was shown by several R&D institutes to yield high quality GeSn layers. The epitaxy tools used, then, were capable of Atmospheric Pressure CVD (i. e. @ 760 Torr), while we have Reduced Pressure CVD chambers that can operate only up to a few hundreds of Torr. We have thus investigated the growth of tens of nm thick GeSn layers on 2.5 µm thick Ge Strain-Relaxed Buffers, themselves on Si(001). First GeSn growth trials at 349°C, 100 Torr with the same GeH4 flow than Ge2H6 did not yield any peaks in X-Ray Diffraction (XRD). Changing the Mass-Flow Controller and using a GeH4 flow 4 times higher resulted in modest Sn content layers, with some Sn surface segregation, however. Increasing, for fixed GeH4 and SnCl4 flows, the precursor partial pressures by (i) reducing the H2 carrier flow or (ii) increasing the chamber pressure resulted in linear GeSn Growth Rate (GR) and Sn content increases. Sn droplets were present on the surface up to 300 Torr, however. By contrast, GeSn layers grown at 400 Torr were high quality in XRD (well defined, intense peaks with thickness fringes, as shown in Fig. 1), with cross-hatched surfaces as the Ge SRBs underneath (Fig. 2). For a fixed GeH4 partial pressure (1.28 Torr), we have then evidenced a linear increase of the GeSn GR with the SnCl4 flow at 349°C. This increase was sub-linear at 325°C, with a slope change for P(SnCl4) = 0.008 Torr (Fig. 3). By contrast, the Sn content x was almost steady with the SnCl4 flow at 349°C. At 325°C, we had first of all an increase of the Sn content up to P(SnCl4) = 0.008 Torr, then a decrease, as shown in Fig. 4. Two SnCl4 partial pressures were therefore selected for the investigation of the impact of temperature on GeSn growth kinetics: 0.008 and 0.016 Torr (i.e. F(SnCl4)/F(H2) = 2E-05 or 4E-05). We had exponential increases, over the 301°C - 349°C range, of the GeSn GR with the temperature, with similar activation energies for both chemistries: 8.6 – 11.8 kcal mol.-1 for GeH4 ó 10.4 kcal. mol.-1 for Ge2H6 (Fig. 5). Meanwhile, the Sn content linearly decreased as the growth temperature increased, with a slope which was less for GeH4 than for Ge2H6 (- 1.0 or -1.1 % / 10°C, to be compared with - 1.7% / 10°C) (Fig. 6). Although the GeH4 flow was 4 times the Ge2H6 one and the chamber pressure 4 times higher (400 Torr ó 100 Torr), the GeSn GR was almost the same for a given SnCl4 flow. Sn contents were by contrast significantly lower with GeH4 than with Ge2H6. Halving, for the same GeH4 flow, the SnCl4 flow resulted in GeSn GR 40% lower but slighly higher Sn contents, which was once again counterintuitive! The Ge GR component (= (1-x)*GeSn GR) increased exponentially with the temperature, with similar values and activation energies for both precursors. Meanwhile, we had, over the 301°C-349°C range, a slight decrease of the Sn GR component (= x*GeSn GR) as the temperature increased. This was likely due to Sn sublimation. Sn GR components were 50% higher, for the same SnCl4 flow, with Ge2H6 than with GeH4. Complex interplays between precursors thus govern the GeSn growh kinetics with both precursors. Higher Sn contents and GeSn GR are nevertheless feasible with Ge2H6 + SnCl4 instead of GeH4 + SnCl4. Finally, while the GeSn GR and Sn contents were rather uniform over the 200mm wafer surface with Ge2H6 + SnCl4, the situation was definitely worse with GeH4 + SnCl4. GR were then significantly higher at the wafer edges than close to the center (although the Sn content was fairly uniform). Figure 1

In-situ boron doping of Si and Ge can be of use in various types of devices. Poly-Si:B gates can be integrated in p-type metal oxide semiconductor (pMOS) transistors. Monocrystalline Si:B layers can be used in p-i-n junctions for RF switches, attenuators, photo-detectors (PDs) or phase shifters. Ge:B layers can be integrated in the sources and drains of Ge pMOS transistors or as the p-type films in Ge p-i-n PDs. If the growth temperature is low enough, they can even be used in GeSn PDs or light emitting devices. We have explored here the boron-doping of Si and Ge with Si2H6 + B2H6 and Ge2H6 + B2H6. Such chemistries enable Si (Ge) epitaxy at temperatures lower than with conventional SiH4 (GeH4) precursors. Growth temperatures and pressures in our 200mm RP-CVD tool were 550°C, 20 Torr (Si) and 350°C, 100 Torr (Ge). X-Ray Diffraction (XRD) was used to convert the tensile strain in Si:B or Ge:B into substitutional B concentrations, as shown in Fig. 1 and Fig. 2, with layer peaks moving away from the Si(001) substrate or the Ge Strain-Relaxed Buffers (SRB) as the diborane flow increased. The presence of marked thickness fringes testified to the high crystalline quality of the films. Secondary Ions Mass Spectrometry (SIMS) gave us the atomic B concentration. X-Ray Reflectivity on Si0.5Ge0.5 / Si:B stacks yielded Si:B Growth Rates (GR). Finally, differential weighting gave us the thickness of Ge:B layers grown on Ge SRBs and other GR values. A huge Si:B GR increase, from 9.5 up to 40 nm min.-1, was evidenced when adding relatively small amounts of B2H6 to Si2H6, as shown in Fig. 3. This was due to the presence of B atoms on the surface which catalyzed H desorption, freeing sites for growth. Ultra-high substitutional and atomic boron contents were obtained in those Si:B layers (at most: 2.7x1020 and 1.1x1021 atoms cm-3), as shown in Fig. 4. Films became poly-cristalline for the highest F(B2H6)/F(Si2H6) Mass-Flow Ratios (MFRs) probed, without XRD peaks anymore and a boron ion concentration which was reduced compared to its peak value of 2x1020 cm-3. The good agreement between substitutional, atomic and ionic B concentrations for low to medium MFRs is to be underlined. By contrast, the growing discrepancy between XRD and SIMS for the highest diborane flows probed might partly be due to a SIMS artefact. Indeed, the Relative Sensistivity Factor used to convert the raw SIMS signals into B concentrations was for doped layers and not higher boron content binary alloys. The presence of B atoms in nanoclusters that did not contribute to XRD cannot be excluded, however. A definite growth rate increase (from 6 up to 21 nm min.-1) and really high substitutional B concentrations (at most: 4.8x1020 atoms cm-3) were also obtained for Ge:B, as shown in Fig. 5 and Fig. 6. The good agreement between XRD and SIMS for medium F(B2H6)/F(Ge2H6) MFR is to be underlined. The growing discrepancy for high diborane flows might be due to the same SIMS artifact as before and to the presence of B atoms in nanoclusters. x/(1-x) = m*F(B2H6)/F(Si2H6 or Ge2H6) relationships accounted for the semi-linear increase of the substitutional or atomic B concentration x with the diborane flow, with (i) msubst. = 1.85 (matom. = 7.45) in Si:B and (ii) msubst. = 0.15 (matom. = 0.68) in Ge:B. m values above 1 meant that B incorporation was easier than Si incorporation in Si:B. In contrast, B incorporation was more difficult than Ge incorporation in Ge:B alloys, as m was less than 1, then. Current growth rate increases and B concentrations are definitely higher than with mainstream chemistries. For Si:B grown at 650°C, 20 Torr with SiH4 + B2H6, we had for instance GR increases from 9 up to 23 nm min.-1, with at most 9x1019 cm-3 B ions concentrations before layers became poly-crystalline (ECS Trans. 16 (10), 485 (2008)). For Ge:B grown at 400°C, 100 Torr with SiH4 + B2H6, GR increased from 10 up to 18 nm min.-1, with at most 1020 cm-3 B ions concentrations (Thin Solid Films 557, 4 (2014)). The advantages of switching over to higher order Si and Ge precursors and adopting lower growth temperatures are obvious. Figure 1