Mid-infrared resonant light emission from GeSn resonant-cavity surface-emitting LEDs with a lateral p-i-n structure

Chen-Yang Chang; Po-Lun Yeh; Yue-Tong Jheng; Lung-Yi Hsu; Kuo-Chih Lee; Hui Li; H. H. Cheng; Guo-En Chang

doi:10.1364/PRJ.457193

1. INTRODUCTION

Electronic–photonic integrated circuits (EPICs) have received increasing attention for a wide range of emerging applications, including telecommunications, on-chip optical interconnection, sensing, and quantum computing, owing to their unique advantages of high data rate, large bandwidth, and low energy consumption [1–4]. In contrast to III-V semiconductors, group IV materials are fully compatible with standard CMOS processing, making them more suitable for high-volume, scalable manufacturing of monolithic EPICs [1–4]. Although several key building blocks for Si-based EPICs, including photodetectors [5–12], modulators [13–16], and waveguides [17] have been demonstrated, an efficient electrically injected group IV light emitter remains challenging to complete EPICs because of the indirectness of the band structure of conventional group IV alloys [1,2].

Fortunately, pure Ge is a promising material for efficient light emitters because of its quasi-direct bandgap, in which the direct conduction lies just 136.5 meV above the lowest direct conduction band at the $L$ -point in the momentum space [18,19]. A viable solution to reduce this energy difference is to alloy Ge with Sn, another group IV element [20]. When the Sn content exceeds $\sim 6 %$ , bulk GeSn alloys can be transformed from an indirect bandgap material into a direct bandgap material [21], making them suitable for use as efficient light emitters. Despite the challenges of the limited equilibrium solid solubility of Sn in Ge of $\sim 1 %$ [21,22], the recent advances in low-temperature growth techniques using molecular beam epitaxy (MBE) [21–23] and chemical vapor deposition (CVD) [24–28] have enabled the growth of high-quality GeSn alloys with a Sn content of up to 36% [28]. This material development has encouraged the development of efficient GeSn light emitters for EPICs. Recently, optically pumped GeSn lasers have been demonstrated on Si and insulator platforms at cryogenic temperatures [29–34], and even up to room temperature ( $T = 300 K$ ) [35,36], showing great promise for efficient group IV light emitters.

However, practical EPICs crave efficient electrically injected GeSn light emitters operating at room temperature. Recently, electrically injected GeSn LEDs at room temperature also have been demonstrated [37–41]. The first electrically injected GeSn laser operated under pulsed current injection was recently demonstrated with a strong light output power of a few milliwatts at cryogenic temperatures up to $T = 100 K$ [42,43], demonstrating the feasibility of electrically injected GeSn lasers. However, most of the demonstrated group IV GeSn LEDs and lasers are vertical p-i-n diode structures, which suffer from a strong optical loss induced by the metal pads and heavily doped regions that could significantly increase the threshold and/or decrease the lasing temperature [41,43]. To overcome this problem, lateral p-i-n structures have been proposed by keeping the lossy metal pads and heavily doped regions far away from the active region [41,44]. In addition, planar laser structures are highly desired to meet the requirements of EPICs. Thus, lateral p-i-n structures integrated with optical resonators are considered a promising approach for efficient group IV lasers.

In this study, we demonstrate room-temperature resonant electroluminescence (EL) from electrically injected GeSn resonant-cavity LEDs (RCLEDs) with a lateral p-i-n homojunction on a silicon-on-insulator (SOI) substrate. A vertical cavity structure was introduced for the GeSn active layer to enhance the light-emitting efficiency. A lateral p-i-n junction was designed for the current injection to reduce the optical loss from the metal contacts. By increasing the Sn content to 4.3%, the energy difference between the direct and indirect conduction bands was reduced from 136.5 to 89 meV, thereby enhancing direct-gap interband transitions. The device features a planar structure, which is highly desirable for monolithic integration in EPICs. The room-temperature EL experiments clearly demonstrated resonant cavity modes and enhanced light emission, which our simulations supported. The effects of the $Q$ factor of the cavity and defects were theoretically analyzed to evaluate the threshold to achieve lasing action. These results pave the way for efficient electrically injected GeSn lasers at room temperature for EPICs.

2. DEVICE DESIGN AND SIMULATION

Figure 1(a) shows a schematic diagram of the designed surface-emitting GeSn RCLEDs with a lateral p-i-n homojunction. The designed layer structure is shown in the inset of Fig. 1(a). The GeSn active layer was grown on an SOI substrate with a Ge virtual substrate (VS). The device consists of a lateral GeSn p-i-n homojunction with a hexagonal GeSn mesa with a side length of $L$ and a height of $t$ for electrical injection, and a ${SiO}_{2}$ layer on the GeSn active layer as the passivation layer and top reflector. The wavelength-dependent refractive indices (RIs) of the materials were obtained from Refs. [45,46], and Fig. 1(b) shows the RI profile along the $z$ direction of the structure at $λ = 1980 nm$ . The low-RI buried oxide (BOX) and deposited ${SiO}_{2}$ layers serve, respectively, as the bottom and top reflectors, creating a vertical cavity to enhance light emission. To confirm the vertical cavity effect, finite element method (FEM) simulations were performed for the designed structure using a plane wave as the light source excited at the center of the GeSn active layer. Figure 1(c) shows the simulated optical field at $λ = 1980 nm$ for the structure ( $L = 20 μm$ ). Proper reflection at the top ${SiO}_{2} / GeSn$ and bottom BOX/Si-substrate interfaces was observed due to the large RI difference, highlighting standing wave patterns in the cavity and confirming good optical confinement by the vertical cavity to enhance the light emission. From the FEM analysis, it also is found that the cavity supports multiple transverse modes because of the relatively wide mesa. Figure 1(d) shows the simulated energy distribution of selected transverse modes supported in the cavity at $λ = 1980 nm$ . It is clear that the deposited ${SiO}_{2}$ on the sidewall provides good optical confinement for the GeSn active region because of the large RI difference.

Fig. 1. Device design and simulation. (a) Schematic diagram of the designed lateral GeSn p-i-n diodes on an SOI substrate. The inset shows the layer structure of the grown sample (not to scale). (b) Refractive index profile along the $z$ direction. The Si cap layer is not presented because of its small thickness. (c) Simulated energy distribution along the $z$ direction at $λ = 1980 nm$ using the finite element method, showing clear standing wave patterns by cavity effects. (d) Energy distribution of selected transverse modes supported in the cavity at $λ = 1980 nm$ , showing good transverse optical confinement.

Download Full Size | PDF

3. MATERIAL GROWTH AND CHARACTERIZATION

The material growth was conducted on a (001)-oriented SOI substrate (with a 2.5 μm thick top Si layer and a 1 μm thick BOX layer) using MBE with a unique low-temperature growth technique [37] at a base pressure less than $2 \times 10^{- 10}$ Torr $(1 Torr = 133.32 Pa)$ . The epitaxy started with the growth of a high-quality Ge VS consisting of: (i) a 100 nm thick Si buffer layer grown at 600°C, (ii) a 100 nm thick Si buffer layer grown at 350°C, (iii) a 117.5 nm thick Ge seed layer grown at 350°C followed by an in-situ annealing at 800°C for 5 min, and (iv) a 117.5 nm thick Ge buffer layer grown at 550°C. The growth temperature then was decreased to 150°C to suppress unwanted Sn segregation for the growth of a 540 nm thick intrinsic layer of GeSn, followed by a 3 nm thick Si cap layer.

The structural properties of the grown sample were systematically investigated using cross-sectional transmission electron microscopy (XTEM), secondary-ion mass spectrometry (SIMS), X-ray diffraction (XRD), and photoluminescence (PL) techniques. The PL experiments were conducted at room temperature using a 532 nm continuous-wave (CW) laser with an optical power of 1 W, which was optically chopped at 200 Hz and then focused onto the sample with a spot size of $\sim 300 μm$ . The PL emission from the sample was filtered through a 1200 nm longpass filter, collected by lenses, and then read using a spectrometer equipped with a monochromator with a 600 lines/mm grating blazed at $λ = 1600 nm$ and an ${LN}_{2}$ -cooled extended InGaAs photodetector operated in the spectral range of 1.3–2.2 μm. A lock-in technique was adopted to enhance the signal-to-noise ratio (SNR) of the PL emission signals.

Figure 2(a) shows an XTEM image of the grown sample, with corresponding energy-dispersive spectroscopy (EDS) mappings shown in Figs. 2(b)–2(d). Significant dislocations were found at the interface between the Ge VS and the top Si layers owing to the lattice mismatch, suggesting the strain relaxation in the Ge VS. In addition, the Ge VS was shown to efficiently confine misfit dislocations at the interface between the top Si layer and Ge VS to reduce misfit dislocations in the GeSn active layer. No obvious defects were found at the GeSn/Ge interface, indicating that the GeSn layer and Ge VS were coherent. The EDS mapping also revealed a uniform distribution of Sn atoms in the GeSn layer and a sharp interface between the Ge/GeSn interface. Figure 2(e) shows the secondary ion mass spectrometry (SIMS) atomic distributions of Ge and Sn atoms along the growth direction. A uniform distribution of Sn atoms in the GeSn layer and a sharp interface between the GeSn/Ge interface can be observed in the SIMS depth profile. Figure 2(f) shows the X-ray reciprocal space mapping (RSM) of the (224) plane, with two distinct peaks associated with the Ge VS and GeSn layer. Most importantly, the two peaks were aligned vertically, indicating that the GeSn layer was pseudomorphic to the underlying Ge VS. Figure 2(g) shows the XRD (004) $ω - 2 θ$ scan of the grown sample. Three distinct peaks were observed, associated with the Ge VS, the GeSn active layer, and the Si substrate. The peak of the Ge VS is much weaker with a much broader FWHM owing to the confinement of defects in the Ge VS and the small thickness of the Ge VS. In contrast, the peak of the GeSn active layer is much stronger and narrower than that of the Ge VS, indicating good material quality. The Bragg angle of the Ge VS is smaller than the position of the bulk Ge, suggesting the existence of a biaxial strain of $ε = 0.12 %$ in the Ge VS, which was induced by the mismatch of thermal expansion coefficient between Si and Ge [5]. From the RSM data, the in-plane and out-of-plane lattice constants of the GeSn layer were extracted. Results of an analysis of the lattice constants demonstrate that the GeSn active layer has a Sn content of 4.3% and a biaxial strain of $ε = - 0.52 %$ . Figure 2(h) shows the PL spectrum of the grown sample measured at room temperature, as well as the mode envelope. Unlike the single emission peak in conventional GeSn thin films on Si [37], the observed PL spectra exhibit several emission peaks with a free spectrum range (FSR) of $\sim 110 nm$ , providing clear evidence for Fabry–Perot resonant cavity modes in the vertical cavity. From the envelope of the emission peaks depicted in Fig. 2(h), we obtained an envelope maximum of 1980 nm, corresponding to a photon energy of 626 meV, which is assigned to the lowest direct bandgap of the GeSn active layer.

Fig. 2. Material characterization results of the grown sample. (a) Cross-sectional transmission electron microscopic (XTEM) image and corresponding energy-dispersive spectroscopy (EDS) mappings of (b) Si, (c) Ge, and (d) Sn atoms. (e) Secondary ion mass spectrometry (SIMS) atomic distribution of Ge and Sn atoms. (f) (224) reciprocal space mapping (RSM) of the grown sample, revealing pseudomorphic GeSn/Ge structure. (g) X-ray $ω - 2 θ$ scan. The vertical dashed line presents the Bragg angle of bulk Ge. (h) Room-temperature photoluminescence spectrum with mode envelope (dashed line).

Download Full Size | PDF

4. DEVICE FABRICATION AND CHARACTERIZATION

A. Device Fabrication

Figure 3(a) shows the key fabrication steps for the GeSn RCLEDs. The fabrication process started with the definition of hexagonal mesas with different side lengths using standard optical lithography and reactive ion etching (RIE) techniques down to the BOX layer. Then, $n$ and $p$ regions were introduced by ion implantation with boron and phosphorous ions, respectively, with an energy of 20 keV and a dose of $10^{15} {cm}^{- 2}$ . To activate the dopants, microwave thermal annealing was performed with a power of 1650 W for 100 s, which yields a high dopant activation rate without significant Sn segregation [47]. Subsequently, a ${SiO}_{2}$ passive layer was deposited using plasma-enhanced CVD. Windows were then introduced using wet etching with a buffered oxide etchant. Finally, contact electrodes were introduced by depositing an Au/Cr bilayer with a thickness of 200/20 nm using an e-beam evaporator and patterning using the lift-off technique. An SEM image of the fabricated device is shown in the inset of Fig. 3(b).

Fig. 3. (a) Key fabrication steps for the GeSn RCLEDs. (b) Scanning electron microscopy image of the fabricated device. The inset shows the zoom-in image of the hexagonal GeSn active region.

Download Full Size | PDF

B. Electrical Characteristics

The current–voltage (I–V) characteristics of the fabricated devices were measured at room temperature using a sourcemeter (Keithley 2400). The measured I–V characteristics of the fabricated devices ( $L = 20 μm$ ) are shown in Fig. 4. The fabricated GeSn diodes exhibited a clear rectifying behavior. In the forward bias region, the injected current increased rapidly at a small bias voltage, followed by a linear increase as the bias voltage increased further because of the series resistance effect. The I–V characteristics can be fitted using the standard diode model [48]:

I = I_{0} {\exp [\frac{q (V - I R_{s})}{η k_{B} T}] - 1},

where

I_{0}

is the saturation current in the reverse bias region,

R_{s}

is the series resistance,

η

is the diode ideal factor,

k_{B}

is the Boltzmann factor, and

T

is the temperature. From Eq. (1), the series resistance can be obtained from the derived equation [48]

\frac{d V}{d I} = \frac{η k_{B} T}{q} I^{- 1} + R_{s} .

Fig. 4. (a) Current–voltage (I–V) characteristics of the fabricated GeSn horizontal p-i-n RCLEDs at room temperature. The dashed line represents a linear fit to the data in the linear region for the determination of turn-on voltage. (b) Plot of $d V / d I$ versus $I^{- 1}$ . The dashed line represents a linear fit to the data for the determination of series resistance.

Download Full Size | PDF

Figure 4(b) shows a plot of $d V / d I$ versus $I^{- 1}$ and the fitting curve using Eq. (2). The series resistance obtained from the fitting was $1.84 k Ω$ . This relatively high series resistance is likely attributable to the metal–semiconductor contact and the large quasi-neutral region, which may induce significant Joule heating during CW current injection. Further optimization of the doping and annealing processes could lower the series resistance to suppress Joule heating and enhance the current injection efficiency. By fitting the linear region of the I-V curve with a fixed slope of $1 / R_{s}$ [48], as shown in Fig. 4(a), the turn-on voltage of the diode was determined to be 0.51 V.

C. Electroluminescence Characteristics

The EL from the devices was measured at room temperature by injecting CW currents into the devices using the Keithley 2400 sourcemeter. The emission signals from each device were optically chopped at 200 Hz, and then sent to a spectrometer consisting of a monochromator with a grating blazed at 1600 nm and a liquid $N_{2}$ -cooled extended InGaAs photodetector with a cutoff wavelength of 2.2 μm. The signal was then read using a locking amplifier to enhance the SNR.

Figure 5(a) shows the measured electroluminescence spectra of the device ( $L = 20 μm$ ) with different CW injected currents ( $I_{inj}$ ). Several emission peaks were observed with an FSR of $\sim 110 nm$ and an FWHM of $Δ λ \sim 70 nm$ . The strongest emission peak is located at $\sim 1960 nm$ , which agrees with the PL results. As the injected current increased, the EL intensity significantly increased owing to the increased injected carrier density. In addition, the peak position of the emission peak slightly red shifted, mainly because of the Joule heating effect during the CW current injection. The FWHM and peak position ( $λ_{0}$ ) of the emission peaks were then obtained by fitting the EL spectra with Gaussian functions for the determination of the $Q$ factor ( $Q = λ_{0} / Δ λ$ ). Figure 5(b) shows the resonance wavelength and the $Q$ factor of the emission peak at $\sim 1980 nm$ as a function of the injected current. The corresponding injected current density ( $J_{inj}$ ) can be converted from the expression $J_{inj} = I_{inj} / A_{c} = I_{inj} / (L \times t)$ , where $A_{c}$ is the cross-sectional area the current injects through the GeSn active region. As the injected current density increased, the resonance peak was slightly red-shifted, which is mainly attributable to the Joule heating effect during CW current injection. In contrast, the $Q$ factor decreased when the injected current increased from 20 to 25 mA, due to the Joule heating effect that increases the free-carrier absorption [49]. When the injected current increased further, the $Q$ factor also increased, indicating absorption bleaching in the cavity under electrical injection caused by the electron-hole recombination via direct-bandgap transitions. With the $Q$ factor, the spectral emission enhancement factor ( $G_{e}$ ) of the device can then be further evaluated using [48]

G_{e} = \frac{2 \sqrt{r_{1} r_{2}}}{1 - r_{1} r_{2}},

where

r_{1}

and

r_{2}

are the reflection coefficient, respectively, of the top and bottom reflectors. The product

r_{1} r_{2}

can be obtained from the

Q

factor using the expression

Q = \frac{2 \bar{n} L_{eff}}{λ} \times \frac{π \sqrt{r_{1} r_{2}}}{1 - r_{1} r_{2}},

where

\bar{n}

and

L_{eff}

are the effective index and length of the cavity. The quantity

\bar{n} L_{eff}

can be obtained from the FSR via the relation

\bar{n} L_{eff} = λ^{2} / Δ λ

[48]. Figure 4(c) shows the spectral enhancement factor with different injected currents. A good spectral enhancement factor of

\sim 2

was obtained, demonstrating enhanced light emission at room temperature under CW current injection. To further understand the light emission properties, from the driven current-dependent EL spectra, the integrated EL intensity (

L

) as a function of injected current was extracted, as shown in Fig. 5(d). Their dependence can then be characterized using

L = A I^{m}

, where

A

is a constant and

m

is an indicator of the emission efficiency of the device. Fitting the data yields

m = 1.48

. Compared to the

m

value of

\sim 1

for Ge LEDs [44], the observed higher

m

value of our devices highlights the higher radiative efficiency due to the enhanced direct bandgap in the GeSn active layer.

Fig. 5. Electroluminescence characteristics. (a) Room-temperature electroluminescence (EL) spectra of the fabricated GeSn devices ( $L = 20 μm$ ) with various injected currents. (b) Extracted $Q$ factor and resonant emission wavelength. The inset shows a schematic diagram of current injection into the GeSn active region. (c) Spectral enhancement factor as a function of injected current. (d) Integrated EL intensity as a function of injected current.

Download Full Size | PDF

5. THEORETICAL ANALYSIS

To further understand the resonant EL behavior of the GeSn RCLEDs, theoretical simulations were performed. We started with the determination of the band structure of the GeSn active layer using a multiband $k \cdot p$ model considering the strain effect [19,50,51] and the parameters used for the simulation were taken from Ref. [19]. The bandgap energies of bulk GeSn alloys can be determined using [19]

E_{g}^{η} ({Ge}_{1 - x} {Sn}_{x}) = x E_{g}^{η} (Ge) + (1 - x) E_{g}^{η} (Sn) - b_{η} x (1 - x),

where

E_{g}^{η} (Ge)

and

E_{g}^{η} (Sn)

are the bandgap energies of Ge and

α - Sn

, respectively,

b_{η}

is the bowing parameter, and

η

represents the different valleys of the energy band. The bandgap energy parameters used in these calculations are given in Table 1.

Table 1. Bandgap Energies of GeSn Alloys at $T = 300$ K [19,37]

View Table

Figure 6(a) shows the calculated band structure of the GeSn active layer at $T = 300 K$ . The introduction of 4.3% Sn into the active layer lowered both the direct and indirect conduction band edges and reduced their energy difference from 136.5 to 89 meV. Thus, the electron density populating the direct conduction band can be significantly enhanced under current injection, thereby enhancing the direct-gap light emission efficiency. The 0.52% compressive strain in the GeSn active layer shifted the heavy-hole (HH) band above the light-hole (LH) band. As a result, the direct band is defined as the lowest direct-gap transition from the $Γ$ -conduction band to the HH band. The calculated $c Γ \to HH$ transition energy is 653 meV, which is in good agreement with the PL and EL experimental results, confirming that the observed emission originates from the $c Γ \to HH$ optical transition in the GeSn active layer.

Fig. 6. Theoretical analysis. (a) Calculated band structure of the GeSn active layer at $T = 300 K$ . (b) Calculated optical gain (solid lines) and loss (dashed lines) spectra of the GeSn active layer at room temperature with various injected current densities. The shaded areas represent net gain. (c) Calculated net gain (solid line) as a function of injected carrier density and mirror loss with various $Q$ factors (dashed lines). (d) Calculated threshold carrier density and enhancement factor as a function of the $Q$ factor. (e) Calculated threshold current density as a function of the $Q$ factor at various defect densities.

Download Full Size | PDF

To achieve light amplification, the optical gain must exceed the optical loss in the active material. We therefore performed theoretical simulations to calculate the optical gain using Fermi’s golden rule [18,52], as well as the optical loss resulting from free-carrier absorption (FCA) and intervalence band absorption (IVBA) of the GeSn active layer at $T = 300 K$ using an empirical linear absorption cross-section model [38,53]. Figure 6(b) shows the calculated material gain and optical loss under various injected carrier densities. As the injected carrier density increases, a sufficient number of electrons can populate the direct conduction band to achieve population inversion, thereby generating optical gain. At the same time, the optical loss also increases with increasing injected carrier density. However, the optical gain can surpass the optical loss; thus, net optical gain from the GeSn active layer is achievable.

Although resonant EL was observed in the GeSn cavity, lasing action was not achieved, which is likely due to the low $Q$ factor caused by the low reflectivity of the FP cavity. The $Q$ factor can be significantly enhanced to 100–10,000 by employing distributed Bragg reflectors [52,54], which can consist of alternating Si and ${SiO}_{2}$ layers. Thus, the mirror loss of the cavity can be significantly reduced. As the net modal gain can overcome the mirror loss, lasing action can occur. We then calculate the net optical gain as a function of the injected carrier density and the mirror loss with different $Q$ factors to evaluate the threshold carrier density. The mirror loss of a Fabry–Perot cavity can be calculated using [52]

α_{m} = \frac{1}{L_{eff}} \ln (\frac{1}{r_{1} r_{2}}),

where

r_{1} r_{2}

for a given

Q

factor can be calculated using Eq. (4). The spectral enhancement factor can then be calculated using Eq. (3). Figure 6(c) shows the calculated net optical gain as a function of the injected carrier density and mirror loss with different

Q

factors at

T = 300 K

. As the injected carrier density increased, the net optical gain increased. Meanwhile, as the

Q

factor of the cavity increases, the mirror loss is significantly reduced. The interaction of the net optical gain and the mirror loss represents the threshold carrier density. Figure 6(d) shows the calculated threshold carrier density and spectral enhancement factor as a function of the injected carrier density. It can be seen that the threshold carrier density can be considerably reduced with an increasing

Q

factor, making it promising to achieve lasing action at room temperature. Meanwhile, as the

Q

factor increases from 50 to 1000, the spectral enhancement factor can be significantly increased from 3.66 to 70.7, highlighting the impact of the

Q

factor on the light emitting efficiency. With the threshold carrier density, the threshold current (

I_{th}

) can then be further estimated using [52]

I_{th} / (q V_{0}) = A n_{th} + B n_{th}^{2} + C n_{th}^{3},

where

V_{0}

is the volume of the active region;

q

is the elementary charge; and

A

,

B

, and

C

are, respectively, the Shockley–Read–Hall (SRH) recombination coefficient, the radiative recombination coefficient, and the Auger recombination coefficient. The threshold current density (

J_{th}

) can be converted from the expression

J_{th} = I_{th} / A_{c}

. Due to the lack of experimental coefficients for GeSn alloys, the radiative recombination coefficient and Auger recombination coefficient used in the calculations are approximated by these of Ge (

B = 8.7 \times 10^{- 14} {cm}^{3} S^{- 1}

and

C = 1.42 \times 10^{- 32} {cm}^{6} S^{- 1}

[55]). The SRH recombination coefficient is a strong function of material quality and defect density (

N_{D}

), which can be estimated using

A = G N_{D}

[54,56], where

G

is a material parameter related to the defect density. Due to the lack of experimental data in GeSn alloys, the coefficient was approximated by that of Ge (

G = 0.19 {cm}^{2} / s

) [57,58]. Figure 5(e) shows the calculated threshold current density and threshold current as a function of the

Q

factor with different defect densities in the range of

N_{D} = 10^{6} - 10^{9} {cm}^{- 2}

at

T = 300 K

. The results show that the threshold current density significantly decreases with an increasing

Q

factor because a higher

Q

factor can significantly lower the mirror loss, and thus the threshold carrier density. In addition, it is also found that as the defect density decreases from

10^{9} {cm}^{- 2}

to

10^{7} {cm}^{- 2}

, the SRH nonradiative recombination is reduced, thereby significantly reducing the threshold current density. When the defect density further decreases to

10^{6} {cm}^{- 3}

, the reduction in the threshold current density is less. Thus, it is essential to improve the material quality to achieve lasing at lower thresholds. This concept has been experimentally validated in defect-engineered optically pumped GeSn lasers on GeSn-on-insulator platforms [35,36]. However, from Fig. 5(d), it is found the threshold current density is much higher than the maximum injected current in our devices. The results may explain why lasing action was not achieved in our devices.

In addition to enhancing the $Q$ factor, it is equally important to enhance the directness of the energy band structure of the GeSn active region. This can be done by: (i) further increasing the Sn content in the GeSn active layer, and/or (ii) introducing tensile strain into the GeSn active region [33]. The Sn content in the GeSn can be actively effectively increased to higher than 14% by using composition-graded growth methods with MBE [59] and CVD [32] approaches. Furthermore, the use of p-i-n heterojunctions rather than homojunctions could improve the carrier confinement, thereby also improving emission efficiency and lowering the threshold [43]. Improving the material quality and reducing defect densities via proper approaches such as thermal annealing [47,60] and graded GeSn buffer technologies [32] are also necessary to lower the lasing threshold. Therefore, it is anticipated that our proposed device with further increased Sn content and optimized device structure is promising to achieve lasing at room temperature.

6. CONCLUSION

In this study, we demonstrated room-temperature light emission from a GeSn horizontal p-i-n RCLED on an SOI platform. Sn (4.3%) was introduced to the GeSn active layer to reduce the energy difference between the direct and indirect conduction bands to enhance direct-gap interband transitions. The design of lateral p-i-n structures can reduce the optical loss caused by the metal pads to improve the emission efficiency. A vertical Fabry–Perot cavity was created to enhance the light emission efficiency. Resonant electroluminescence was observed at room temperature under CW current injection; several emission peaks with adequate quality factors corresponding to the Fabry–Perot cavity modes were identified. The envelope of the resonant emission peaks was $\sim 1960 nm$ and had a high radiative efficiency and a good spectral enhancement factor. Theoretical analysis showed that the observed electroluminescence originates from the $c Γ \to HH$ direct-gap transition, and a net optical gain can be achieved at room temperature. We believe that our demonstration of GeSn lateral p-i-n diodes provides a promising solution for electrically injected GeSn lasers operating at room temperature for EPICs.

Funding

Ministry of Science and Technology, Taiwan (MOST 110-2636-E-194-002, MOST 110-2923-M-002-011, MOST 111-2636-E-194-002); Taiwan Semiconductor Research Institute (JDP111-Y1-031).

Acknowledgment

The authors acknowledge support from the Advanced Institute of Manufacturing with High-Tech Innovations (AIM-HI) at the National Chung Cheng University from the Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan.

Disclosures

The authors declare no conflicts of interest.

Data Availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

REFERENCES

1. M. J. Deen and P. K. Basu, Silicon Photonics: Fundamentals and Devices (Wiley, 2012).

2. R. Soref, “The past, present, and future of silicon photonics,” IEEE J. Sel. Top. Quantum Electron. 12, 1678–1687 (2006). [CrossRef]

3. R. Soref, “Mid-infrared photonics,” in Optical Fiber Communication Conference (OFC) (Optica Publishing Group, 2015), paper W4A.4.

4. J. Wang, F. Sciarrino, A. Laing, and M. G. Thompson, “Integrated photonic quantum technologies,” Nat. Photonics 14, 273–284 (2020). [CrossRef]

5. J. Liu, J. Michel, W. Giziewicz, D. Pan, K. Wada, D. D. Cannon, S. Jongthammanurak, D. T. Danielson, L. C. Kimerling, J. Chen, F. O. M. Ilday, F. X. Kärtner, and J. Yasaitis, “High-performance, tensile-strained Ge p-i-n photodetectors on a Si platform,” Appl. Phys. Lett. 87, 103501 (2005). [CrossRef]

6. G. E. Chang, S. W. Chen, and H. H. Cheng, “Tensile-strained Ge/SiGe quantum-well photodetectors on silicon substrates with extended infrared response,” Opt. Express 24, 17562–17571 (2016). [CrossRef]

7. S. Ghosh, K. C. Lin, C. H. Tsai, K. H. Lee, Q. Chen, B. Son, B. Mukhopadhyay, C. S. Tan, and G. E. Chang, “Resonant-cavity-enhanced responsivity in germanium-on-insulator photodetectors,” Opt. Express 28, 23739–23747 (2020). [CrossRef]

8. S. Su, B. Cheng, C. Xue, W. Wang, Q. Cao, H. Xue, W. Hu, G. Zhang, Y. Zuo, and Q. Wang, “GeSn p-i-n photodetector for all telecommunication bands detection,” Opt. Express 19, 6400–6405 (2011). [CrossRef]

9. Y. H. Peng, H. H. Cheng, V. I. Mashanov, and G. E. Chang, “GeSn p-i-n waveguide photodetectors on silicon substrates,” Appl. Phys. Lett. 105, 231109 (2014). [CrossRef]

10. Y. H. Huang, G. E. Chang, H. Li, and H. H. Cheng, “Sn-based waveguide p-i-n photodetector with strained GeSn/Ge multiple-quantum-well active layer,” Opt. Lett. 42, 1652–1655 (2017). [CrossRef]

11. H. Tran, T. Pham, J. Margetis, Y. Zhou, W. Dou, P. C. Grant, J. M. Grant, S. Al-Kabi, G. Sun, R. A. Soref, J. Tolle, Y. H. Zhang, W. Du, B. Li, M. Mortazavi, and S. Q. Yu, “Si-based GeSn photodetectors toward mid-infrared imaging applications,” ACS Photon. 6, 2807–2815 (2019). [CrossRef]

12. W. T. Hung, D. Barshilia, R. Basu, H. H. Cheng, and G. E. Chang, “Silicon-based high-responsivity GeSn short-wave infrared heterojunction phototransistors with a floating base,” Opt. Lett. 45, 1088–1091 (2020). [CrossRef]

13. J. Liu, M. Beals, A. Pomerene, S. Bernardis, R. Sun, J. Cheng, L. C. Kimerling, and J. Michel, “Waveguide-integrated, ultralow-energy GeSi electro-absorption modulators,” Nat. Photonics 2, 433–437 (2008). [CrossRef]

14. Y. H. Kuo, Y. K. Lee, Y. Ge, S. Ren, J. E. Roth, T. I. Kamins, D. A. B. Miller, and J. S. Harris, “Strong quantum-confined Stark effect in germanium quantum-well structures on silicon,” Nature 437, 1334–1336 (2005). [CrossRef]

15. M. A. Van Camp, S. Assefa, D. M. Gill, T. Barwicz, S. M. Shank, P. M. Rice, T. Topuria, and W. M. J. Green, “Demonstration of electrooptic modulation at 2165 nm using a silicon Mach-Zehnder interferometer,” Opt. Express 20, 28009–28016 (2012). [CrossRef]

16. Y. D. Hsieh, J. H. Lin, R. Soref, G. Sun, H. H. Cheng, and G. E. Chang, “Electro-absorption modulation in GeSn heterostructures for wide-spectrum mid-infrared silicon photonics,” Commun. Mater. 2, 40 (2021). [CrossRef]

17. T. Hu, B. Dong, X. Luo, T. Y. Liow, J. Song, C. Lee, and G. Q. Lo, “Silicon photonic platforms for mid-infrared applications,” Photon. Res. 5, 417–430 (2017). [CrossRef]

18. G. E. Chang and H. H. Cheng, “Optical gain of germanium infrared lasers on different crystal orientations,” J. Phys. D 46, 065103 (2013). [CrossRef]

19. G. E. Chang, S. W. Chang, and S. L. Chuang, “Strain-balanced Ge_zSn_1−z–Si_xGe_ySn_1−x−y multiple-quantum-well lasers,” IEEE J. Quantum Electron. 46, 1813–1820 (2010). [CrossRef]

20. R. A. Soref and C. H. Perry, “Predicted band gap of the new semiconductor SiGeSn,” J. Appl. Phys. 69, 539–541 (1991). [CrossRef]

21. R. Chen, H. Lin, Y. Huo, C. Hitzman, K. I. Kamins, and J. S. Harris, “Increased photoluminescence of strain-reduced, high-Sn composition Ge_1−xSn_x alloys grown by molecular beam epitaxy,” Appl. Phys. Lett. 99, 181125 (2011). [CrossRef]

22. M. Oehme, K. Kostecki, M. Schmid, F. Oliveira, E. Kasper, and J. Schulze, “Epitaxial growth of strained and unstrained GeSn alloys up to 25% Sn,” Thin Solid Films 557, 169–172 (2014). [CrossRef]

23. I. S. Yu, T. H. Wu, K. Y. Wu, H. H. Cheng, V. Mashanov, A. Nikiforov, O. Pchelyakov, and X. S. Wu, “Investigation of Ge_1−xSn_x/Ge with high Sn composition grown at low-temperature,” AIP Adv. 1, 042118 (2011). [CrossRef]

24. V. R. D’Costa, C. S. Cook, A. G. Birdwell, C. L. Littler, M. Canonico, S. Zollner, J. Kouvetakis, and J. Menéndez, “Optical critical points of thin-film Ge_1-ySn_y alloys: a comparative Ge_1-ySn_y/Ge_1-xSi_x study,” Phys. Rev. B 73, 125207 (2006). [CrossRef]

25. S. Assali, J. Nicolas, S. Mukherjee, A. Dijkstra, and O. Moutanabbir, “Atomically uniform Sn-rich GeSn semiconductors with 3.0–3.5 μm room-temperature optical emission,” Appl. Phys. Lett. 112, 251903 (2018). [CrossRef]

26. M. A. Mircovich, C. Xu, D. A. Ringwala, C. D. Poweleit, J. Menéndez, and J. Kouvetakis, “Extended compositional range for the synthesis of SWIR and LWIR Ge_1–ySn_y alloys and device structures via CVD of SnH₄ and Ge₃H₈,” ACS Appl. Electron. Mater. 3, 3451–3460 (2021). [CrossRef]

27. Y. Miao, G. Wang, Z. Kong, B. Xu, X. Zhao, X. Luo, H. Lin, Y. Dong, B. Lu, L. Dong, J. Zhou, J. Liu, and H. H. Radamson, “Review of Si-based GeSn CVD growth and optoelectronic applications,” Nanomaterials 11, 2556 (2021). [CrossRef]

28. S. A. Ghetmiri, W. Du, J. Margetis, A. Mosleh, L. Cousar, B. R. Conley, L. Domulevicz, A. Nazzal, G. Sun, R. A. Soref, J. Tolle, B. Li, H. A. Naseem, and S.-Q. Yu, “Direct-bandgap GeSn grown on silicon with 2230 nm photoluminescence,” Appl. Phys. Lett. 105, 151109 (2014). [CrossRef]

29. S. Wirths, R. Geiger, N. von den Driesch, G. Mussler, T. Stoica, S. Mantl, Z. Ikonic, M. Luysberg, S. Chiussi, J. M. Hartmann, H. Sigg, J. Faist, D. Buca, and D. Grutzmacher, “Lasing in direct-bandgap GeSn alloy,” Nat. Photonics 9, 88–92 (2015). [CrossRef]

30. J. Margetis, S. Al-Kabi, W. Du, W. Dou, Y. Zhou, T. Pham, P. Grant, S. Ghetmiri, A. Mosleh, B. Li, J. Liu, G. Sun, R. Soref, J. Tolle, M. Mortazavi, and S. Q. Yu, “Si-based GeSn lasers with wavelength coverage of 2–3 μm and operating temperatures up to 180 K,” ACS Photon. 5, 827–833 (2015). [CrossRef]

31. J. Chrétien, N. Pauc, F. A. Pilon, M. Bertrand, Q. M. Thai, L. Casiez, N. Bernier, H. Dansas, P. Gergaud, E. Delamadeleine, R. Khazaka, H. Sigg, J. Faist, A. Chelnokov, V. Reboud, J. M. Hartmann, and V. Calvo, “GeSn lasers covering a wide wavelength range thanks to uniaxial tensile strain,” ACS Photon. 6, 2462–2469 (2019). [CrossRef]

32. Y. Zhou, W. Dou, W. Du, S. Ojo, H. Tran, S. A. Ghetmiri, J. Liu, G. Sun, R. Soref, J. Margetis, J. Tolle, B. Li, Z. Chen, M. Mortazavi, and S.-Q. Yu, “Optically pumped GeSn lasers operating at 270 K with broad waveguide structures on Si,” ACS Photon. 6, 1434–1441 (2019). [CrossRef]

33. A. Elbaz, D. Buca, N. von den Driesch, K. Pantzas, G. Patriarche, N. Zerounian, E. Herth, X. Checoury, S. Sauvage, I. Sagnes, A. Foti, R. Ossikovski, J.-M. Hartmann, F. Boeuf, Z. Ikonic, P. Boucaud, D. Grützmacher, and M. El Kurdi, “Ultra-low-threshold continuous-wave and pulsed lasing in tensile-strained GeSn alloys,” Nat. Photonics 14, 375–382 (2020). [CrossRef]

34. H.-J. Joo, Y. Kim, D. Burt, Y. Jung, L. Zhang, M. Chen, S. J. Parluhutan, D.-H. Kang, C. Lee, S. Assali, Z. Ikonic, O. Moutanabbir, Y.-H. Cho, C. S. Tan, and D. Nam, “1D photonic crystal direct bandgap GeSn-on-insulator laser,” Appl. Phys. Lett. 119, 201101 (2021). [CrossRef]

35. A. Bjelajac, M. Gromovyi, E. Sakat, B. Wang, G. Patriarche, N. Pauc, V. Calvo, P. Boucaud, F. Boeuf, A. Chelnokov, V. Reboud, M. Frauenrath, J.-M. Hartmann, and M. El Kurdi, “Up to 300 K lasing with GeSn-on-insulator microdisk resonators,” Opt. Express 30, 3954–3961 (2022). [CrossRef]

36. J. Chrétien, Q. M. Thai, M. Frauenrath, L. Casiez, A. Chelnokov, V. Reboud, J. M. Hartmann, M. El Kurdi, N. Pauc, and V. Calvo, “Room temperature optically pumped GeSn microdisk lasers,” Appl. Phys. Lett. 120, 051107 (2022). [CrossRef]

37. H. H. Tseng, K. Y. Wu, H. Li, V. Mashanov, H. H. Cheng, G. Sun, and R. A. Soref, “Mid-infrared electroluminescence from a Ge/Ge_0.922Sn_0.078/Ge double heterostructure p-i-n diode on a Si substrate,” Appl. Phys. Lett. 102, 182106 (2013). [CrossRef]

38. J. P. Gupta, N. Bhargava, S. Kim, T. Adam, and J. Kolodzey, “Infrared electroluminescence from GeSn heterojunction diodes grown by molecular beam epitaxy,” Appl. Phys. Lett. 102, 251117 (2013). [CrossRef]

39. J. D. Gallagher, C. L. Senaratne, P. Sims, T. Aoki, J. Menendez, and J. Kouvetakis, “Electroluminescence from GeSn heterostructure pin diodes at the indirect to direct transition,” Appl. Phys. Lett. 106, 091103 (2015). [CrossRef]

40. B. J. Huang, C. Y. Chang, Y. D. Hsieh, R. A. Soref, G. Sun, H. H. Cheng, and G. E. Chang, “Electrically injected GeSn vertical-cavity surface emitters on silicon-on-insulator platforms,” ACS Photon. 6, 1931–1938 (2019). [CrossRef]

41. L. Peng, X. Li, Z. Liu, X. Liu, J. Zheng, C. Xue, Y. Zuo, and B. Cheng, “Horizontal GeSn/Ge multi-quantum-well ridge waveguide LEDs on silicon substrates,” Photon. Res. 8, 899–903 (2020). [CrossRef]

42. Y. Zhou, Y. Miao, S. Ojo, H. Tran, G. Abernathy, J. M. Grant, S. Amoah, G. Salamo, W. Du, J. Liu, J. Margetis, J. Tolle, Y.-H. Zhang, G. Sun, R. A. Soref, B. Li, and S.-Q. Yu, “Electrically injected GeSn lasers on Si operating up to 100 K,” Optica 7, 924–928 (2020). [CrossRef]

43. Y. Zhou, S. Ojo, C. W. Wu, Y. Miao, H. Tran, J. M. Grant, G. Abernathy, S. Amoah, J. Bass, G. Salamo, W. Du, G. E. Chang, J. Liu, J. Margetis, J. Tolle, Y.-H. Zhang, G. Sun, R. A. Soref, B. Li, and S. Q. Yu, “Electrically injected GeSn lasers with peak wavelength up to 2.7 μm,” Photon. Res. 10, 222–229 (2022). [CrossRef]

44. Z. Liu, Y. Li, C. He, C. Li, C. Xue, Y. Zuo, B. Cheng, and Q. Wang, “Direct-bandgap electroluminescence from a horizontal Ge p-i-n ridge waveguide on Si(001) substrate,” Appl. Phys. Lett. 104, 191111 (2014). [CrossRef]

45. H. Tran, W. Du, S. A. Ghetmiri, A. Mosleh, G. Sun, R. A. Soref, J. Margetis, J. Tolle, B. Li, H. A. Naseem, and S. Q. Yu, “Systematic study of Ge_1-xSn_x absorption coefficient and refractive index for the device applications of Si-based optoelectronics,” J. Appl. Phys. 119, 103106 (2016). [CrossRef]

46. E. D. Palik, Handbook of Optical Constants of Solids (Academic, 1985).

47. J. Liu, P. Chiu, Y. Chuang, C. Liu, G. Luo, and J. Li, “High dopant activation of phosphorus in strained and relaxed GeSn by rapid thermal annealing and microwave annealing,” in Electron Devices Technology and Manufacturing Conference (EDTM) (2019), pp. 44–46.

48. E. F. Schubert, Light-Emitting Diodes, 2nd ed. (Cambridge University, 2006).

49. P. Zhou, X. Xu, S. Matsushita, K. Sawano, and T. Maruizumi, “Resonant light emission from uniaxially tensile-strained Ge microbridges,” Jpn. J. Appl. Phys. 57, 04FH10 (2018). [CrossRef]

50. G. E. Chang, R. Basu, B. Mukhopadhyay, and P. K. Basu, “Design and modeling of GeSn-based heterojunction phototransistors for communication applications,” IEEE J. Sel. Top. Quantum Electron. 22, 8200409 (2016). [CrossRef]

51. G. E. Chang, S. Q. Yu, J. Liu, H. H. Cheng, R. A. Soref, and G. Sun, “Achievable performance of uncooled homojunction GeSn mid-infrared photodetectors,” IEEE J. Sel. Top. Quantum Electron. 28, 3800611 (2022). [CrossRef]

52. S. L. Chuang, Physics of Photonic Devices, 2nd ed. (Wiley, 2009).

53. L. Carroll, P. Friedli, S. Neuenschwander, H. Sigg, S. Cecchi, F. Isa, D. Chrastina, G. Isella, Y. Fedoryshyn, and J. Faist, “Direct-gap gain and optical absorption in germanium correlated to the density of photoexcited carriers, doping, and strain,” Phys. Rev. Lett. 109, 057402 (2012). [CrossRef]

54. M. Kuramoto, S. Kobayashi, T. Akagi, K. Tazawa, K. Tanaka, T. Saito, and T. Takeuchi, “High-power GaN-based vertical-cavity surface-emitting lasers with AlInN/GaN distributed Bragg reflectors,” Appl. Sci. 9, 416 (2019). [CrossRef]

55. J. Jiang, M. Xue, C.-Y. Lu, C. S. Fenrich, M. Morea, K. Zang, J. Gao, M. Cheng, Y. Zhang, T. I. Kamins, J. S. Harris, and J. Sun, “Strain-induced enhancement of electroluminescence from highly strained germanium light-emitting diodes,” ACS Photon. 6, 915–923 (2019). [CrossRef]

56. S. Ghosh, H. Kumar, B. Mukhopadhyay, and G. E. Chang, “Design and modeling of high-performance DBR-based resonant-cavity-enhanced GeSn photodetector for fiber-optic telecommunication networks,” IEEE Sens. J. 21, 9900–9908 (2021). [CrossRef]

57. L. M. Giovane, H.-C. Luan, A. M. Agarwal, and L. C. Kimerling, “Correlation between leakage current density and threading dislocation density in SiGe p-i-n diodes grown on relaxed graded buffer layers,” Appl. Phys. Lett. 78, 541–543 (2001). [CrossRef]

58. Q. Chen, S. Wu, L. Zhang, W. Fan, and C. S. Tan, “Simulation of high-efficiency resonant-cavity-enhanced GeSn single-photon avalanche photodiodes for sensing and optical quantum applications,” IEEE Sens. J. 21, 14789–14798 (2021). [CrossRef]

59. M. Li, J. Zheng, X. Liu, Y. Zhu, C. Niu, Y. Pang, Z. Liu, Y. Zuo, and B. Cheng, “Sn composition graded GeSn photodetectors on Si substrate with cutoff wavelength of 3.3 μm for mid-infrared Si photonics,” Appl. Phys. Lett. 120, 121103 (2022). [CrossRef]

60. H. Li, Y. X. Cui, K. Y. Wu, W. K. Tseng, H. H. Cheng, and H. Chen, “Strain relaxation and Sn segregation in GeSn epilayers under thermal treatment,” Appl. Phys. Lett. 102, 251907 (2013). [CrossRef]

	$E_{g}^{η} (Ge)$ (eV)	$E_{g}^{η} (Sn)$ (eV)	$b_{η}$ (eV)
$L$ -valley	0.664	0.092	0.89
$Γ$ -valley	0.7985	–0.413	2.89

Mid-infrared resonant light emission from GeSn resonant-cavity surface-emitting LEDs with a lateral p-i-n structure

Abstract

1. INTRODUCTION

2. DEVICE DESIGN AND SIMULATION

3. MATERIAL GROWTH AND CHARACTERIZATION

4. DEVICE FABRICATION AND CHARACTERIZATION

A. Device Fabrication

B. Electrical Characteristics

C. Electroluminescence Characteristics

5. THEORETICAL ANALYSIS

6. CONCLUSION

Funding

Acknowledgment

Disclosures

Data Availability

REFERENCES

Data Availability

Cited By

Figures (6)

Tables (1)

Equations (7)

Photonics Research