*Review* **Review of Si-Based GeSn CVD Growth and Optoelectronic Applications**

**Yuanhao Miao 1,2,\* , Guilei Wang 1,2,3 , Zhenzhen Kong 1,3, Buqing Xu 1,3, Xuewei Zhao 1,3, Xue Luo <sup>2</sup> , Hongxiao Lin <sup>2</sup> , Yan Dong <sup>1</sup> , Bin Lu 2,4, Linpeng Dong 2,5, Jiuren Zhou <sup>6</sup> , Jinbiao Liu <sup>1</sup> and Henry H. Radamson 1,2,3,\***


**Abstract:** GeSn alloys have already attracted extensive attention due to their excellent properties and wide-ranging electronic and optoelectronic applications. Both theoretical and experimental results have shown that direct bandgap GeSn alloys are preferable for Si-based, high-efficiency light source applications. For the abovementioned purposes, molecular beam epitaxy (MBE), physical vapour deposition (PVD), and chemical vapor deposition (CVD) technologies have been extensively explored to grow high-quality GeSn alloys. However, CVD is the dominant growth method in the industry, and it is therefore more easily transferred. This review is focused on the recent progress in GeSn CVD growth (including ion implantation, in situ doping technology, and ohmic contacts), GeSn detectors, GeSn lasers, and GeSn transistors. These review results will provide huge advancements for the research and development of high-performance electronic and optoelectronic devices.

**Keywords:** GeSn; CVD; lasers; detectors; transistors

### **1. Introduction**

Si-based integrated circuits (ICs), which are dominated by Si CMOS technology, have reached their physics limit. The influences of quantum effects, parasitic parameters, and process parameters on data transmission applications are also reaching their limits, as the rapid development of microelectronics has led to higher requirements for data transmission technology. For these reasons, scientists have proposed schemes to integrate optoelectronic devices with microelectronic devices [1–7]. However, Si-based on-chip integrated light source was lacking, and the light sources for existing optoelectronic integrated circuits (OEICs) were all externally coupled; though the coupling efficiency between the edge of the light source and grating coupler was high enough, the lack of an on-chip light source restricted OEICs' applications [8–10]. As such, many research programs started to pay more attention to Si-based monolithic OEIC technology [11–15], which has the following advantages over the baseline technology: (i) it is compatible with mature Si CMOS technology; (ii) has low costs; (iii) has larger wafer sizes and larger scale production; (iv) its partial electrical interconnection can be replaced by optical interconnection, which

**Citation:** Miao, Y.; Wang, G.; Kong, Z.; Xu, B.; Zhao, X.; Luo, X.; Lin, H.; Dong, Y.; Lu, B.; Dong, L.; et al. Review of Si-Based GeSn CVD Growth and Optoelectronic Applications. *Nanomaterials* **2021**, *11*, 2556. https://doi.org/10.3390/nano 11102556

Academic Editor: Filippo Giubileo

Received: 6 August 2021 Accepted: 22 September 2021 Published: 29 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

can realize high-efficiency, high-speed, and low loss data transmission. Si-based monolithic OEIC technology uses Si-compatible semiconductor technology to integrate optoelectronic devices into Si chips in order to improve chip performance, extend chip function, and reduce costs. Though Si-based photonic devices, such as optical waveguides [16,17], photodetectors [18–20], optical modulators [21–23], and optical switches [24,25], have been successfully developed, it is difficult to achieve high-efficiency emission due to the facts that Si is an indirect bandgap semiconductor and its light emission efficiency is about five orders of magnitude lower than that of direct band gap III–V compound semiconductors. Thus, the need for an Si-based high-efficiency light source represents an important technical bottleneck in the development of Si-based monolithic OEICs. Therefore, looking for a direct bandgap semiconductor material that is compatible with the Si CMOS process is of great significance in the creation of large scale Si-based monolithic OEICs [26–28].

Group IV materials are compatible with the traditional Si CMOS process, and Si, SiGe, and Ge are commonly used as indirect band gap semiconductors despite not being suitable for light emission. Fortunately, tensile strain engineering and Sn-alloying engineering have enabled Ge to become a quasi-direct bandgap or direct bandgap material due to the small bandgap difference between its two minima in conduction bands (only 136 meV). Experimental research has shown an optical gain of 0.24% for tensile-strained n<sup>+</sup> -type Ge (the n-type doping level is 1 <sup>×</sup> <sup>10</sup><sup>19</sup> cm−<sup>3</sup> ), which led to the creation of optically injected and electrically injected Ge lasers [29–32]. However, the threshold for a Ge laser is too high, which means that weak tensile-strained n<sup>+</sup> -type Ge is not able to supply enough optical gain to achieve low-threshold lasing.

In recent decades, GeSn alloys have demonstrated novel indirect-to-direct bandgap transition, as well excellent carrier transport. Due to their tunable band structures, GeSn materials have become promising candidates to create Si-based OEICs with higher hole mobility, enhanced light absorption, etc. [33–37]. Growing high-quality GeSn layers with relatively high Sn contents has different challenges, e.g., Sn segregation during growth and the poor thermal stability of SnGe layers [38–41]. These issues root from the low solid solubility of Sn in Ge (<1%) and the large lattice mismatch between Si or Ge and GeSn. As early as 1995, the first growth of a GeSn/Ge superlattice was reported using a very low growth temperature in a molecular beam epitaxy (MBE) chamber. Such GeSn layers had an Sn content of 26% [42,43]. Based on these early pioneer works, other growth techniques, such as chemical vapor deposition (CVD) and magnetron sputtering, have been widely used to grow high-quality direct bandgap GeSn materials with high Sn contents [44–49]. Although MBE can grow GeSn materials well, its growth rate is extremely low, which makes it tough to manufacture on a large scale. To achieve a significant impact within the industry, it is very important to develop a commercially available tool to grow highquality GeSn materials. At a very early development stage of GeSn growth via CVD, SnD<sup>4</sup> and Ge2H<sup>6</sup> were chosen as the Sn precursors and Ge precursors, respectively. Although there were many foundational studies on GeSn growth via CVD, SnD<sup>4</sup> is a high-cost material with a short lifetime, which makes it incompatible with the industry. For this reason, other precursors such as SnCl<sup>4</sup> have been explored. The IMEC and KTH groups pioneered the growing of GeSn layers using commercially available reaction precursors (SnCl4/Ge2H6) [50]. A major breakthrough was later demonstrated using the production of commercially available reaction precursors (SnCl4/GeH4) [51,52]. The limitations of incorporating Sn into Ge have been conquered, and two major breakthroughs for GeSn CVD growth have been reached: (i) a world record high Sn content (22.3%) in bulk GeSn materials with PL emission was observed at room temperature (indicating good material quality), and (ii) SiGeSn/GeSn/SiGeSn multiple quantum well (MQW) structure growth and low-temperature PL intensity were later able to be remarkable enhanced [53–56]. Furthermore, the low costs and widespread availability of these chemicals in large-scale fabrication makes them the best choice for GeSn-based optoelectronic integration into CMOS processing. To make GeSn an efficient N-type or P-type semiconductor material for optoelectronic device application, there is an urgent need to research and develop doping

engineering for GeSn. Currently, doping technologies, such as ion implantation and in situ CVD doping, have been optimized regarding their target doping concentration and doping distributions.

After the successful growth of P<sup>+</sup> -Si/i–GeSn/n–GeSn via CVD, Jay Mathews et al. demonstrated the world's first GeSn photodetector with a 2% Sn content in 2009 [57]. The wavelength cutoff was extended to be at least 1750 nm, which means that the GeSn photodetector with a 2% Sn content can cover the entire telecommunication band. Since then, GeSn photoconductor detectors [58–63], and p–GeSn/i–GeSn/n–GeSn heterostructure detectors [64–68] have been demonstrated. Advances in GeSn CVD growth technology have occurred alongside material quality and detector performance improvements, including: (i) the wavelength cutoff for the GeSn photodetector has been progressively broadened from 1800 nm to 2100, 2400, 2600, 2650, and 3650 nm [63]; (ii) based on wafer-bonding technology, the dark current for GeSn photodetector has been suppressed by more than two orders of magnitude [69]; (iii) peak specific detectivity values are now comparable to those of commercial extended-InGaAs detectors (4 <sup>×</sup> <sup>10</sup><sup>10</sup> cm·Hz1/2 ·W−<sup>1</sup> ) at the same wavelength range; (iv) a passivation technique was developed to enhance responsivity and peak specific detectivity [65]; and (v) mid-IR imaging was demonstrated with GeSn photodetectors, and the image quality of the GeSn photodetectors was found to be superior to that of a commercial PbSe detector [63].

Alongside the significant breakthroughs in GeSn growth and detectors, GeSn lasing had also developed to an advanced stage. Recently reported GeSn laser structures have all been grown via the CVD technique. Following the observation of a PL peak with narrowed line widths, a true direct bandgap GeSn material with an Sn content of up to 10% was experimentally demonstrated in 2014 [33]. Encouraged by this major technical breakthrough, researchers used the injection methods such as optical injection with a Ge laser to check the GeSn waveguide, and lasing behavior was clearly observed at a low temperature in 2015 [70]. Following this breakthrough, several types of GeSn lasers [71–82] were demonstrated, though they still suffer from the problems of lowtemperature operation and high lasing thresholds. To overcome these difficulties, several methods have been proposed to improve performance, such as greater Sn incorporation into Ge [73,75,76], the use of SiGeSn/GeSn/SiGeSn heterostructures or SiGeSn/GeSn/SiGeSn MQWs as the gain medium [83–86], a modulation doping scheme in SiGeSn/GeSn/SiGeSn MQWs [87], defect management [80], and thermal management [81,82]. Considerable efforts in GeSn lasing research have led to an increased maximum lasing temperature of 270 K [76] due to the amazing discovery of strain relaxation growth mechanism [88]. Near-room-temperature lasing was also observed for a GeSn active medium with a 16% Sn content and high uniaxial tensile strain [77]. A breakthrough regarding the optical pumping threshold was reported in 2020, when a low-Sn-content GeSn material with a high uniaxial tensile strain was utilized as an active medium; continuous wave (CW) lasing was also achieved. However, the lasing temperature only reached 100 K due to the low directness of the active medium [80]. In the same year, electrically pumped GeSn/SiGeSn heterostructure lasers with operation temperatures of up to 100 K were demonstrated [89,90]; this was an essential achievement for Si-based electrically pumped group IV interband lasing.

As a group IV material, GeSn is compatible with Si and can realize the transition from indirect band gap to direct band gap by adjusting its Sn content, which makes it the best substitute for group IV materials in Si-based optoelectronic integration applications. GeSn has an extremely high carrier mobility, so it may also be an ideal materials for transistor applications. Due to the significant development of GeSn CVD growth technology, vertically stacked 3-GeSn-nanosheet pGAAFETs (gate-all-around FETs) [91], GeSn p-FinFETs [92,93], GeSn n-channel MOSFETs [94,95], GeSn/Ge vertical nanowire pFETs [96], GeSn GAA nanowire pFETs [97], and GeSn n-FinFETs [98] have been successfully demonstrated. Additionally, GeSn's direct band gap property was found to effectively improve the tunneling probability of electrons, making an excellent material for TFET preparation [99,100], this opening a new development direction for the integrated circuit after Moore's era. The

discovery of this property has attracted considerable research interest in recent years. Since Sn naturally has low solid solubility in Ge (smaller than 1%), growth of high Sn composition single crystal GeSn is difficult. At present, devices prepared with GeSn materials are still in the research and development stage, so they have not been widely used in production.

To the best of our knowledge, there has yet to be a review article that systematically reported on GeSn material growth and counterpart optoelectronic devices using the CVD technique. UHVCVD [101–104], RPCVD [105–110], PECVD [111–113], LPCVD [114–117], and APCVD [118,119] are discussed in this review, with a focus on identifying processes that can be transferred for the commercial production of GeSn. The objective of this comprehensive review article is to provide readers with a full understanding of the recent experimental advancements in GeSn material growth using CVD, as well as their optoelectronic applications. However, due to the large numbers of publications in this area, the authors of this work only selected articles with significant scientific impacts.

### **2. Research Progress for GeSn CVD Growth and Its Potential Applications**

So far, several types of growth techniques, such as MBE, magnetron sputtering, and CVD have been used to grow GeSn materials. CVD is the dominant growth method in the industry, so more easily transferable. Therefore, we decided to review GeSn CVD growth and its potential applications.

### *2.1. Potential Applications*

A literature survey revealed that GeSn materials have numerous potential applications, including Si-based, integrated, high-efficiency light sources [120–122]; high-mobility electronic devices [92–100]; low-cost, Si-based, high-performance shortwave infrared (SWIR) imaging sensors [63–65]; Si-based photovoltaics [123]; optical signal encoding in the midinfrared range [124,125]; high-performance logic applications [126,127]; Si-based integrated thermoelectrics as wearable devices [128,129]; Si-based spintronics [130,131]; Si-based integrated reconfigurable dipoles [132,133]; and Si-based quantum computing [134,135] (Figure 1). GeSn-related fundamental research and development applications have also been extensively investigated (Figure 2).

**Figure 1.** Potential applications of GeSn materials in different research areas.

**Figure 2.** Optoelectronic applications of GeSn as a function of technology readiness level (GeSn transistors, which are still in the technical development stage, are not shown here due to space limitations).

Figure 2 shows the optoelectronic applications of GeSn as a function of technology readiness level. It can be observed that GeSn detectors are getting closer to the low-cost SWIR imaging applications, indicating that GeSn materials have great potential for use in next-generation civilian night-vision and IR cameras [63–65]. However, there are still some technical problems, which are discussed in Section 3. In addition to detectors (which are being rapidly developed), high-quality SiGeSn/GeSn/SiGeSn MQW growth, roomtemperature, CW, and electrically injected SiGeSn/GeSn/SiGeSn MQW lasers; MQW electro-absorption (EA) modulators; and photovoltaic cells are in the research and development stage.

### *2.2. Research Progress for GeSn CVD Growth*

In 2001, Kouvetakis's group from Arizona State University (ASU) first reported a GeSn alloy on oxidized and oxidized-free Si using UHVCVD [136]; since then, extensive GeSn CVD growth-related research works have been carried out. In 2003, SnD<sup>4</sup> and SiH3GeH<sup>3</sup> were used as reaction precursors, and single-phase SiGeSn on a GeSn buffer was first achieved on Si via UHVCVD at 350 ◦C [137]. To create GeSn materials with higher Sn contents, SnD<sup>4</sup> and Ge2H<sup>6</sup> were chosen as Sn and Ge precursors, respectively; the experimental results showed that SnD<sup>4</sup> is helpful for low-temperature growth, and its reaction with Ge2H<sup>6</sup> can create GeSn with an Sn content of up to 25% [138] (Figure 3). The crystallinity, bandgap, lattice constants, optical properties, photoresponses, photocurrents, and Raman scattering results of GeSn materials grown by UHVCVD have been systematically demonstrated [139–144]. In order to grow GeSn at extremely low temperatures, some authors used Ge3H<sup>8</sup> and Ge4H<sup>10</sup> as Ge precursors [145,146]. By using this method, single crystalline GeSn alloys were successfully deposited at temperatures ranging from 300 to 330 ◦C, the growth rate of the allows was able to meet industrial requirements, and the traditional SK growth mode was avoided. Finally, the authors concluded that Ge3H<sup>8</sup> is a superior solution to grow GeSn alloys via UHVCVD [145,146]. Compared with previously reported reaction precursors (SnD4/Ge2H6), the growth rate of the SnD4/Ge3H<sup>8</sup> combination was found to be improved 3–4 times. For this reason, a 1 µm thick GeSn layer with an Sn content of up to 9% was implemented, and room temperature photoluminescence spectra were observed, indicating that GeSn has great potential to be utilized as a gain medium for a Group IV laser. Later, SiGeSn growth at ultralow temperatures (from 290 to 330 ◦C) using Ge4H10, Si4H10, and SnD<sup>4</sup> were reported [147–149].

**Figure 3.** (**a**) Scanning transmission electron microscopy (STEM) image and (**b**) EDX cross-sectional profile of the GeSn with Sn contents of up to 25%. Reproduced with permission from [138], AIP Publishing, 2001.

Although there have been many foundational studies on GeSn growth via CVD investigated, SnD<sup>4</sup> has high costs, is incompatible with the industry, and is unstable at room temperature. For these reasons, other precursors such as SnCl<sup>4</sup> have been explored. IMEC and KTH were the first groups to propose GeSn growth using commercially available reaction precursors (SnCl4/Ge2H6).

Due to the fact that SnCl<sup>4</sup> is liquid at room temperature, these groups evaporated SnCl<sup>4</sup> using a bubbler that was connected to an RPCVD chamber. Experimental results showed that defect-free doped and undoped GeSn layers with Sn contents of up to 8% were created using RPCVD at atmosphere conditions. Thermal stability was further determined by annealing at different conditions (400 ◦C for 10 min, 400 ◦C for 30 min, 500 ◦C for 10 min, and 500 ◦C for 30 min); the (004) omega-2 theta scan of as-grown and annealed Ge0.92Sn0.08 samples were compared (Figure 4a). For the sample annealed at 500 ◦C for 30 min, the diffraction peaks of GeSn and Ge widened and a clear GeSn peak shift was observed, suggesting possible Ge–Sn interdiffusion. To further confirm this assumption, secondary ion mass spectroscopy (SIMS) was conducted. From the SIMS results, the authors concluded that APCVD-grown GeSn with 8% Sn content was stable at the annealing condition of 500 ◦C for 30 min (Figure 4b). This work paved the way for GeSn growth using both commercially available reaction precursors and CVD production equipment.

**Figure 4.** Comparison of (**a**) (004) omega-2 theta scans and (**b**) Sn content profiles of as-grown and annealed Ge0.92Sn0.08 samples under different annealing conditions. Reproduced with permission from [150], AIP Publishing, 2011.

Since then, there has been a sharp increase in the scientific knowledge of GeSn CVD growth, as shown by a number of publications (Figure 5a). The number of publications on GeSn CVD growth grew dramatically in 2013 and reaches 19 in 2018 (Figure 5a). The rapid development of GeSn CVD growth techniques has meant that the number of GeSn optoelectronic device publications followed the similar tendency (Figure 5b): (i) following the world's first demonstration of a GeSn detector, GeSn detector-related publications grew from 1 in 2008 to 30 in 2019; (ii) since the world's first demonstration of an optically pumped GeSn laser, publications related to GeSn lasers continually increased from 10 in 2015 to 25 in 2019, and the majority of these laser publications reported experimental results; (iii) there are still few publications regarding GeSn modulators, and a CVD-grown modulator has not been achieved (the majority of the modulator publications have been theoretical investigations).

**Figure 5.** (**a**) Number of publications/year on GeSn materials grown by the CVD technique; (**b**) number of the publications/year on GeSn optoelectronic devices (theoretical calculations and conference proceedings are included).

To help readers to understand the research status of CVD growth techniques, Figure 6 summarizes research on GeSn CVD growth since the introduction of CVD in 2001 in terms of the research institution, growth chamber, year of deposition, and corresponding reference. Figure 6 shows several types of growth chambers, such as UHVCVD (baby color dot), RPCVD (dark color dot), APCVD (red color dot), PECVD (orange color dot), LPCVD (coffee color dot), and RTCVD (green color dot), that have been used to grow GeSn materials. Following pioneer works from ASU and IMEC, research groups from KTH Royal Institute of Technology (KTH), Applied Materials Inc (AM), PGI (Peter Grünberg Institute), and UA (University of Arkansas) started researching GeSn growth using CVD technology in 2013. Since then, research groups from ASM, University of Warwick (UW), National Taiwan University (NTWU), and Université de Montréal (EPM) have also researched GeSn CVD growth. Among all CVD growth technologies, RPCVD growth chamber is most widely accepted due to its commercial availability and more easily transferability (six research groups have used RPCVD chambers to grow GeSn). After the successful demonstration of the low-temperature growth of high-quality Ge on Si using PECVD, plasma-enhanced techniques came to be regarded as promising methods to grow GeSn materials. Thus, plasma-enhanced GeSn growth techniques aroused researchers' attentions from UA and ASU.

*Review* 

**Applications** 

**Review of Si-Based GeSn CVD Growth and Optoelectronic** 

**Yan Dong 1, Bin Lu 2,4, Linpeng Dong 2,5, Jiuren Zhou 6, Jinbiao Liu 1 and Henry H. Radamson 1,2,3,\*** 

researchers' attentions from UA and ASU.

**Yuanhao Miao 1,2,\*, Guilei Wang 1,2,3, Zhenzhen Kong 1,3, Buqing Xu 1,3, Xuewei Zhao 1,3, Xue Luo 2, Hongxiao Lin 2,** 

To help readers to understand the research status of CVD growth techniques, Figure 6 summarizes research on GeSn CVD growth since the introduction of CVD in 2001 in terms of the research institution, growth chamber, year of deposition, and corresponding reference. Figure 6 shows several types of growth chambers, such as UHVCVD(baby color dot), RPCVD (dark color dot), APCVD (red color dot), PECVD(orange color dot), LPCVD(coffee color dot), and RTCVD (green color dot), that have been used to grow GeSn materials. Following pioneer works from ASU and IMEC, research groups from KTH Royal Institute of Technology (KTH), Applied Materials Inc (AM), PGI (Peter Grünberg Institute), and UA (University of Arkansas) started researching GeSn growth using CVD technology in 2013. Since then, research groups from ASM, University of Warwick (UW), National Taiwan University (NTWU), and Université de Montréal (EPM) have also researched GeSn CVD growth. Among all CVD growth technologies, RPCVD growth chamber is most widely accepted due to its commercial availability and more easily transferability (five research groups have used RPCVD chambers to grow GeSn). After the successful demonstration of the low-temperature growth of high-quality Ge on Si using PECVD, plasma-enhanced techniques came to be regarded as promising methods to grow GeSn materials. Thus, plasma-enhanced GeSn growth techniques aroused

**Figure 6.** List of GeSn CVD growth papers by different groups. **Figure 6.** List of GeSn CVD growth papers by different groups.

### *2.3. GeSn CVD Growth Strategy*

*Nanomaterials* **2021**, *11*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/nanomaterials Later, high-quality GeSn with a world-record high Sn content of 22.3% was crafted after the discovery of strain-relaxation-enhanced (SRE) GeSn CVD growth mechanism To have a full understanding of the GeSn CVD growth strategy, it is necessary to calibrate the Ge growth at low temperatures (below 450 ◦C). After calibration, the flow rate between Ge precursor and Sn precursor needs to be taken into consideration due to the possible etching effect of the generated Cl\* species on the GeSn surface. Therefore, there is a critical flow rate, and the growth rate for GeSn growth has to be high enough to overcome the etching rate. More importantly, the effects of temperature, pressure, carrier gas, and strain relaxation on material growth must be canvassed.

### 2.3.1. Temperature and Pressure Effect on GeSn Growth

Previous GeSn CVD growth work has demonstrated that Sn content is closely related to growth temperature because the decreasing temperature moves the growth conditions further from equilibrium, thus increasing Sn content. Therefore, we summarize most GeSn CVD growth results in Figure 7. In GeSn growth using the SnCl4/GeH<sup>4</sup> reaction precursor combination, SnCl<sup>4</sup> and GeH<sup>4</sup> lose their reactivity at a temperature of 280 ◦C and growth is totally ceased. Below 285 ◦C, GeH<sup>4</sup> is not well-adsorbed, which may suggest the generation of GeH<sup>2</sup> and/or 2H. Therefore, the growth temperature for GeSn RPCVD growth with the SnCl4/GeH<sup>4</sup> reaction precursor combination is usually higher than 280 ◦C. For the Ge2H<sup>6</sup> and SnCl<sup>4</sup> precursor combination, GeSn growth temperature could be as low as 275 ◦C.

Significantly, UA demonstrated GeSn growth using PECVD with the commercially available GeH<sup>4</sup> and SnCl4; low-temperature growth at 350 ◦C for GeSn epitaxy on an Si substrate was achieved with an Sn content of up to 6% [113]. By using a 1064 pulsed laser as the light source, a PL signal was also observed at the peak wavelength of 2000 nm, as shown in Figure 8 (Spot III).

**Figure 7.** Temperature effect on the Sn content from different research groups (UW, UA, and EPM denote the University of Warwick, University of Arkansas, and Université de Montréal, respectively) [113,115–117,151–156].

**Figure 8.** Room-temperature PL spectra for GeSn grown by PECVD technology (the Sn content is 6%). Reproduced from [114], open access by OSA Library, 2018.

Their follow-up work verified that the PECVD system was able to grow a high-Sncontent (>10%, with an PL emission peak at approximately 2100 nm) GeSn layer at ultralow temperatures (250, 260, and 270 ◦C) [157] (Figure 9). The realization of GeSn PECVD growth at such low temperatures using a SnCl4/GeH<sup>4</sup> precursor combination mainly benefits from plasma-assisted reactivity improvements [157]. With proper growth optimization, the Sn content of the GeSn grown by PECVD should be higher than that of other CVD chambers. Compared to GeH4, Ge2H<sup>6</sup> is more reactive and possesses lower growth temperature capabilities, indicating that the reactivity of the Ge-hydride is the only limiting factor for low-temperature GeSn growth. For GeSn RPCVD growth using GeH4, Sn incorporation was found to drastically decrease at ~285 ◦C, whereas the growth temperature limit for using Ge2H<sup>6</sup> was found to be 270 ◦C [153].

**Figure 9.** PL spectra for the GeSn grown at temperatures of 250, 260, and 270 ◦C. Reproduced from [157], open access by ScholarWorks@UARK.

For GeSn growth in a UHVCVD chamber [158–161], growth pressure is usually kept in the range of 1 <sup>×</sup> <sup>10</sup>−4–2.5 <sup>×</sup> <sup>10</sup>−<sup>4</sup> Torr, and Sn content rises with decreasing growth temperatures. Even when different combinations of precursors (SnD4/Ge2H6, SnD4/Ge3H8, and SnD4/Ge4H10) are chosen, similar Sn content variation trends are observed (Figure 10). However, growth temperatures with different precursor combinations are varied; the lowest reported growth temperatures for SnD4/Ge2H6, SnD4/Ge3H8, and SnD4/Ge4H<sup>10</sup> are 250, 350, and 150 ◦C, respectively [145,146,161]. Different from UHVCVD, pressures for GeSn growth in LPCVD and APCVD chambers have been found to range from 10 to 760 Torr [115–117,150]. The surface morphology of a layer GeSn grown by APCVD is shown in Figure 11, where surfaces are milky and pyramidical defects are observed at pressures of 10 and 100 Torr; this issue can solved by further increasing the growth rate (keep the SnCl<sup>4</sup> constant and increase the Ge2H<sup>6</sup> gas flow).

**Figure 10.** Temperature effect on Sn content from UHVCVD, LPCVD, and APCVD growth.

**Figure 11.** GeSn surface morphology vs. growth pressure (growth temperature: 320 ◦C; growth pressure: 10, 100, and 760 Torr; precursors: Ge2H<sup>6</sup> and SnCl<sup>4</sup> ). Reproduced with permission from [119], IOP Publishing, 2018.

For GeSn APCVD growth at a temperature of 320 ◦C, pressure was found to be a main factor in the growth of high-Sn-content GeSn materials (the achieved Sn contents at 10 and 760 Torr were 16.7% and 6.6%, respectively) [119]. For LPCVD growth at 120 Torr and 320 ◦C, the Sn content for GeSn was almost the same as that of APCVD.

### 2.3.2. Carrier Gas Effect on GeSn Growth

The effect of carrier gas on GeSn CVD growth is important and of great significance for the good mixing of precursor gases in a CVD chamber [105,152,153]. In contrast to pure Ge growth, the GeSn CVD growth mechanism has changed due to the introduction of Sn precursors, which have made GeSn CVD growth more complex. In several instances, a thickness reduction or an absence of GeSn occurs when choosing N<sup>2</sup> as the carrier gas; this indicates that the growth rate has already changed and is below the etching rate from HCl. Furthermore, the Sn content of GeSn grown with an N<sup>2</sup> carrier gas is different from that grown with an H<sup>2</sup> carrier gas (Sn% difference is usually approximately 1%; see Figure 12). This Sn content reduction may be mainly attributed to the lower growth rate found when using N<sup>2</sup> as the carrier gas.

**Figure 12.** Sn content vs. SnCl4/GeH<sup>4</sup> ratio (growth temperatures: 320 and 350 ◦C; carrier gases: N<sup>2</sup> and H<sup>2</sup> ; precursors: GeH<sup>4</sup> and SnCl<sup>4</sup> ). Reproduced from [153], open access by ASU library.

### 2.3.3. Strain Relaxation Effect on GeSn CVD Growth

F. Gencarelli et al. discovered a composition-dependent strain relaxation mechanism, and they found that high-Sn-content materials show a classical strain relaxation behavior [162]. Their AFM results showed that the island size and density of their low-Sn-content GeSn layers increased with strain relaxation degree (Figure 13) for the following reasons: higher amounts of Sn precursors were needed for high-Sn-content GeSn growth, extra Cl doses were exposed to the surface of GeSn and thus likely avoided Ge–Sn diffusion, Cl atoms could be regarded as the surfactants to mediate the enhancement of island size and density.

**Figure 13.** AFM images of (**a**) GeSn with an Sn content of 6.4%; the strain relaxations for (**a1**), (**a2**), and (**a3**) are 8%, 33%, and 75%, respectively. (**b**) GeSn with a strain relaxation of 75%; the Sn contents of (**b1**), (**b2**), and (**b3**) are 12.6%, 8.1%, and 6.4%, respectively. Reproduced with permission from [162], IOP Publishing, 2012.

Later, high-quality GeSn with a world-record high Sn content of 22.3% was crafted after the discovery of strain-relaxation-enhanced (SRE) GeSn CVD growth mechanism [113], thus showing that compressive strain is the primary limiting factor for achieving greater Sn incorporation under an Sn oversaturation condition (Figure 14). In this research, the following growth strategy was proposed: (i) for first GeSn layer growth, they used a growth recipe of 9–12% Sn (the Sn content ranged from 8.8 to 11.9%); (ii) for second GeSn layer growth, they used the same growth recipe, and the SnCl<sup>4</sup> flow fraction increased by ~8% compared to the first GeSn layer (the Sn content ranged from 12.5 to 16.5%); and (iii) for third GeSn layer growth, they used the same growth recipe, and the SnCl<sup>4</sup> flow fraction increased by ~8% compared to the second GeSn layer. It should be noted that the grading rate of Sn incorporation was well-designed to suppress the growth breakdown. Inspired by the discovery of the SRE GeSn CVD growth mechanism, S. Assali et al. grew a high-quality GeSn layer with 15% Sn using low pressure chemical vapor deposition (LPCVD) in 2018 [115,116].

deposition (LPCVD) in 2018 [115,116].

[113], thus showing that compressive strain is the primary limiting factor for achieving greater Sn incorporation under an Sn oversaturation condition (Figure 14). In this research, the following growth strategy was proposed: (i) for first GeSn layer growth, they used a growth recipe of 9–12% Sn (the Sn content ranged from 8.8 to 11.9%); (ii) for second GeSn layer growth, they used the same growth recipe, and the SnCl4 flow fraction increased by ~8% compared to the first GeSn layer (the Sn content ranged from 12.5 to 16.5%); and (iii) for third GeSn layer growth, they used the same growth recipe, and the SnCl4 flow fraction increased by ~8% compared to the second GeSn layer. It should be noted that the grading rate of Sn incorporation was well-designed to suppress the growth breakdown. Inspired by the discovery of the SRE GeSn CVD growth mechanism, S. Assali et al. grew a high-quality GeSn layer with 15% Sn using low pressure chemical vapor

### *2.4. Doping for GeSn*

Mainstream GeSn doping technologies, such as ion implantation and in situ CVD doping, have been intensively studied for future electronics and photonics applications. Low contact resistivity plays a vital role in the creation of high-performance devices. Table 1 presents a summary of reported B, BF<sup>2</sup> + , and P-doped GeSn via ion implantation in terms of year, institution, Sn content, doping type, doping concentration, activation temperature, and contact metal. Additionally, Tables 2 and 3 present summaries of B-doped GeSn, P-doped GeSn, and As-doped GeSn in terms of year, institution, Sn content, doping type, doping concentration, contact metal, and contact resistivity.


**Table 1.** Summary of reported B, BF<sup>2</sup> + , and P-doped GeSn via ion implantation in terms of year, institution, Sn content, doping type, doping concentration, activation temperature, and contact metal.


**Table 1.** *Cont.*

### 2.4.1. Ion Implantation for GeSn

Ion implantation is a widely used technique for doping semiconductor materials, and its advantages include low-temperature operation, precise dose control, good uniformity, and extremely small lateral diffusion. The research and development of GeSn's ion implantation technology is also of great significance for future device application. So far, researchers have carried out extensive research into GeSn ion implantation technology (although most GeSn has been grown in MBE chambers, which are also significant).

Phosphorus has been widely adopted for ion implantation to achieve efficient N-type doping in GeSn layers because its doping concentrations usually ranges from 2.1 <sup>×</sup> <sup>10</sup><sup>19</sup> to 2.1 <sup>×</sup> <sup>10</sup><sup>21</sup> cm−<sup>3</sup> . For the P-type doping, there are two options: boron and BF<sup>2</sup> + . The highest P-type doping concentration can reach up to 1 <sup>×</sup> <sup>10</sup><sup>20</sup> cm−<sup>3</sup> .

### 2.4.2. In Situ GeSn CVD Doping

Optoelectronic devices, such as GeSn LEDs, GeSn lasers, and GeSn detectors, generally need highly doped GeSn for efficient carrier recombination and low contact resistance. Electronic devices, such as GeSn MOSFETs, GeSn TFETs, GeSn FinFETs, and GeSn GAAFETs (gate-all-around), require lower ohmic contacts, higher dopant concentrations, and selective doping. The use of in situ doping technology for GeSn is an attractive route for improving the performance of optoelectronic and electronic devices because it enables the doping of GeSn at low temperatures with a high doping efficiency and selective doping. Indeed, GeSn transitions from an indirect to direct bandgap material with an Sn content as high as 10%, and this property has led to research interest in Si-based, high-efficiency light sources. The first electrically injected GeSn lasers were recently demonstrated with Sn contents of 11% and 15%. It is definitely true that we require better solutions to create direct bandgap, high-quality doped GeSn, and the selection of an appropriate reaction doping gas and the optimization of epitaxial process are vital for this purpose. To this end, the growth of B-doped GeSn, P-doped GeSn, and As-doped GeSn using CVD has been

reported by several institutions, as summarized in Table 2. However, there are several key points to consider: (I) Sn loss occurs for B-doped GeSn CVD growth, indicating that there is a competition between Sn and B atoms [150,185]; (II) excess partial pressure for PH<sup>3</sup> contributes to poor material quality due to P segregation; (III) B2H<sup>6</sup> partial pressure has no degradation effect on material quality, though it increases the activation doping concentration; (IV) more P could be incorporated into Ge and GeSn by using high order precursors; (V) boron δ-doping layers are helpful for highly doped GeSn growth, and the maximum B concentration can reach up to 1 <sup>×</sup> <sup>10</sup><sup>20</sup> cm−<sup>3</sup> ; and (VI) the doping efficiency of As-doped GeSn is better than that of P-doped GeSn [110].


**Table 2.** Summary of reported B-doped GeSn, P-doped GeSn, and As-doped GeSn in terms of year, institution, Sn content, doping type, doping concentration, and contact metal.


**Table 3.** Summary of reported B-doped GeSn, P-doped GeSn, and As-doped GeSn in terms of institution, Sn content, doping type, doping concentration, and contact metal.
