We report on a III-nitride vertical-cavity surface-emitting laser (VCSEL) with a III-nitride tunnel junction (TJ) intracavity contact. The violet nonpolar VCSEL employing the TJ is compared to an equivalent VCSEL with a tin-doped indium oxide (ITO) intracavity contact. The TJ VCSEL shows a threshold current density (Jth) of ∼3.5 kA/cm2, compared to the ITO VCSEL Jth of 8 kA/cm2. The differential efficiency of the TJ VCSEL is also observed to be significantly higher than that of the ITO VCSEL, reaching a peak power of ∼550 μW, compared to ∼80 μW for the ITO VCSEL. Both VCSELs display filamentary lasing in the current aperture, which we believe to be predominantly a result of local variations in contact resistance, which may induce local variations in refractive index and free carrier absorption. Beyond the analyses of the lasing characteristics, we discuss the molecular-beam epitaxy (MBE) regrowth of the TJ, as well as its unexpected performance based on band-diagram simulations. Furthermore, we investigate the intrinsic advantages of using a TJ intracavity contact in a VCSEL using a 1D mode profile analysis to approximate the threshold modal gain and general loss contributions in the TJ and ITO VCSEL.

Vertical-cavity surface-emitting lasers (VCSELs) are used in many applications due to their highly directional, high power density emission, and the ability to fabricate high density 2D arrays.1,2 The small active region volume of these devices also allows one to achieve low threshold currents and high modulation frequencies, thereby reducing the required input power and increasing data transmission rates, respectively.3 The ability to fabricate single mode VCSELs allows for the production of narrow line-width resonators, which are desirable for many applications, such as atomic clocks, gyroscopes, or magnetometers.4–7 Furthermore, III-nitride VCSELs have great potential for visible light communication systems (VLC), where the current state-of-art uses micro-light-emitting diode (LED) arrays, which are limited by the broad spectral bandwidth of LEDs and carrier lifetime limited frequency response.8,9 The advantages of GaAs- and InP-based VCSELs, which are widely commercialized, carry over to III-nitride VCSELs; however, the large difference in emission wavelengths (red-infrared vs. UV-green) makes III-nitride VCSELs unique.

The mature nature of GaAs- and InP-based VCSELs demonstrates the potential for these devices; however, III-nitride VCSELs research is in its infancy and poses a number of interesting challenges. Many of these challenges are related to the difficult nature of III-nitride epitaxial growth, where growing highly conductive and reflective epitaxial distributed Bragg reflectors (DBRs) is non-trivial.10–14 Furthermore, the inability to grow high reflectivity p-side and n-side epitaxial DBRs leads to significant thermal challenges in flip-chip, dual dielectric DBR VCSELs, as dielectrics have poor thermal conductivity. Additionally, the low electrical conductivity of p-type III-nitride materials prevents current spreading, making intracavity contacts a necessity. The most obvious choice for such a contact is a transparent conductive oxide (TCO) such as tin doped indium oxide (ITO). However, TCOs can have absorption coefficients exceeding 1000 cm−1,15 which can cause the intracavity contact absorption loss to dominate the total loss in III-nitride VCSELs. In GaAs- and InP-based VCSELs, intracavity contact schemes have also been tested.16 Similar to the III-nitrides, InP-based VCSELs also suffer from low p-type conductivity,17 thus researchers have employed epitaxially grown tunnel-junction (TJ) contacts, allowing them to achieve efficient current spreading and a low internal loss.16 

Here, we present a III-nitride VCSEL using a III-nitride TJ intracavity contact and compare the results to a VCSEL with an ITO intracavity contact. We discuss the general challenges and advantages of TJ VCSELs, building on insights from band diagram and 1D cavity mode simulations. The TJ VCSEL shows a 324% increase in differential efficiency and a 56% decrease in threshold current density (Jth) compared to the ITO VCSEL. However, the TJ VCSEL also displays a ∼1.5 V increase in the forward voltage, and both devices have filamentary lasing in the aperture.

The general structure of the devices demonstrated here is similar to those reported in Ref. 15, with minor variations in the n-GaN and p-GaN thickness to account for the reduced number of quantum wells (QWs) (7 instead of 10). A representative device schematic can be found in Ref. 15. Here, we also used an n-Al0.4Ga0.8N top-down photoelectrochemical (PEC) etch-stop layer,18,19 and bonded to a Ti/Au coated copper block. The device employing the ITO intracavity contact used a multi-layer ITO electron-beam evaporation method,15,20 and incorporated a 1/8th-wave Ta2O5 spacer between the ITO and the 16 period (16P) Ta2O5/SiO2 p-side DBR (p-DBR). The device with the n-GaN TJ was fabricated by growing a ∼141 nm n-GaN TJ via ammonia molecular beam epitaxy (MBE) (details below), after metal-organic chemical-vapor deposition (MOCVD) of the other III-nitride epitaxial layers on free standing m-plane GaN substrates.21,22 Because the TJ does not introduce high levels of loss to the cavity, unlike ITO, one can grow the TJ thick in order to reduce the current spreading resistance. Furthermore, the low loss introduced by the TJ implies that there is no need for the 1/8th-wave Ta2O5 spacer, which serves to minimize the internal loss by aligning a null of the mode to the ITO layer. Both devices had a 12P Ta2O5/SiO2 n-DBR.

While III-nitride LEDs with TJ contacts have been reported, no III-nitride VCSEL has incorporated a TJ intracavity contact.23–30 The 1st generation of III-nitride VCSELs10–13,15,18,19,31–34 did not use a TJ because MOCVD grown TJs have been shown to yield highly resistive contacts.23–27 This is due to hydrogen repassivation of p-GaN during the MOCVD n-GaN TJ growth, and the intrinsic doping limits of MOCVD grown n-GaN.35 Furthermore, the growth of this n-GaN layer after the p-GaN growth prevents Mg activation.36 Previous reports of TJs show an increase in the turn-on voltage and differential resistance, compared to conventional TCO current spreading layers. More recently, TJs grown on c-plane GaN have used a thin AlN layer,37 InGaN layer,28,29 or GdN nanoislands30 between the n++GaN and p++GaN layers to reduce the tunneling barrier. These TJs are also grown using MBE, which reduces hydrogen repassivation and allows activation of p-GaN before regrowth and/or during MBE p-GaN growth. On nonpolar or semipolar planes, the effects of the intrinsic polarization present on c-plane will not be seen by the junction, thus only the InGaN or GdN layers would be expected to enhance performance.

In this demonstration, we employ ammonia MBE with solid source effusion cells for Ga and Si to regrow an n-GaN TJ on m-plane VCSEL epitaxial layers grown by MOCVD. Like MOCVD, ammonia MBE uses thermally cracked NH3 as a source of active nitrogen for GaN growth; however, hydrogen levels present during growth are much lower (∼10−6 Torr) and the regrowth does not result in hydrogen passivation of the MOCVD grown p-GaN layers. Before growth of the TJ, the MOCVD p-GaN was activated, then a mesa etch was performed, followed by a Ti/Au hardmask deposition and Al ion implantation to define the aperture.15 The Ti/Au hardmask was then removed in aqua regia, a deionized (DI) water rinse was performed, followed by a solvent clean, prior to loading into the MBE, and baking at 400 °C for 1 hr. The TJ regrowth was then performed at 750 °C, as measured by pyrometry. The Ga flux during growth was on the order of 10−7 Torr with NH3 flow rate of 200 sccm. The presence of a streaky reflection high energy electron diffraction pattern during growth indicated smooth, two dimensional regrowth of the TJ. The TJ consisted of an n++GaN (39.6 nm)/n-GaN (39.6 nm)/n++GaN (39.6 nm)/n-GaN (22.1 nm) stack with the n++GaN layers having a Si concentration of 1.1 × 1020 cm−3 and the n-GaN layers having a concentration of 1 × 1019 cm−3, while the MOCVD grown p++GaN (14 nm) had a Mg concentration of 2.5 × 1020 cm−3. The step-function doping was used to reduce free carrier absorption in the TJ by aligning the n++GaN layers to nulls in the mode. Using the Hall method on a set of test samples, the resistivity and carrier concentrations of the n++GaN and n-GaN films were measured to be 4 × 10−4 Ω cm and 4.4 × 10−3 Ω cm, respectively.

Figure 1 shows SiLENSe38 simulations of the TJ contact. Here, we see that the depletion width is predicted to be ∼7.95 nm, with 6.25 nm of depletion on the n-side. This is a large depletion width, suggesting that tunneling would not occur in such a junction. However, because we see a fairly small voltage increase (∼1.5 V) and no change in the differential resistance (Fig. 3), it is likely that regrowth interface or defect states assist in carrier transport across the junction.39–41 Also, SiLENSEe simulations predict the electric field at the junction to be two times the breakdown field for GaN (5 MV/cm),42 which may play some role in the tunneling process. Further investigations are necessary to understand the detailed nature of tunneling in III-nitride homojunctions.

FIG. 1.

SILENSe simulations of the TJ employed in the reported TJ VCSEL. (a) shows the ionized donor and acceptor concentrations, assuming a donor ionization energy of 5 meV and an acceptor ionization energy of 165 meV. (b) Band diagram of the TJ contact. The total depletion width is ∼7.95 nm, with 6.25 nm of depletion on the n-side.

FIG. 1.

SILENSe simulations of the TJ employed in the reported TJ VCSEL. (a) shows the ionized donor and acceptor concentrations, assuming a donor ionization energy of 5 meV and an acceptor ionization energy of 165 meV. (b) Band diagram of the TJ contact. The total depletion width is ∼7.95 nm, with 6.25 nm of depletion on the n-side.

Close modal
FIG. 3.

LIV and LJV characteristics of the 6.95λ TJ VCSEL and ITO VCSEL with 12 μm aperture diameters measured under pulsed operation (0.3% duty cycle, 100 ns pulse width). The ITO VCSEL shows a threshold current of ∼9 mA (8 kA/cm2), while the TJ VCSEL shows a threshold current of ∼4 mA (3.5 kA/cm2). The TJ VCSEL shows significantly higher differential efficiency due to the large reduction in internal loss; however, the TJ results in an ∼1.5 V increase in the forward voltage.

FIG. 3.

LIV and LJV characteristics of the 6.95λ TJ VCSEL and ITO VCSEL with 12 μm aperture diameters measured under pulsed operation (0.3% duty cycle, 100 ns pulse width). The ITO VCSEL shows a threshold current of ∼9 mA (8 kA/cm2), while the TJ VCSEL shows a threshold current of ∼4 mA (3.5 kA/cm2). The TJ VCSEL shows significantly higher differential efficiency due to the large reduction in internal loss; however, the TJ results in an ∼1.5 V increase in the forward voltage.

Close modal

To gain insight into the optical advantages introduced by using a TJ vs. ITO intracavity contact, we carried out 1D transmission matrix method (TMM) simulations of each device and calculated the loss introduced by the DBRs (mirror loss, αm), intracavity contact (αi,ITO and αi,TJ), all other III-nitride layers (αi,III-Nitrides), and the threshold modal gain (Γgth). Table I shows the constants used in the 1D TMM simulations. To approximate the threshold current density (Jth), we used the 405 nm m-plane edge-emitting laser g(J) data from Ref. 43 and the confinement factor from TMM simulations to calculate the modal gain vs. current density for 405 nm VCSELs with ∼7λ thick cavities and a QW number ranging from 3 to 10 QWs. Specifically, the modal gain equation used has the form Γg(J)NwΓ1g0ln[(J+Js)/(NwJtr1+Js)],44,45 where Nw is the number of QWs, Γ1 is the average confinement factor per well (approximated using the 1D TMM simulations), Γ is the total confinement factor, g0 is the empirical gain coefficient,43Jtr1 is the transparency current density per well,43 and Js is a linearity parameter.43 Fig. 2 shows the modal gain vs. current density, for cavity designs with different numbers of QWs, overlaid with the breakdown of the sources of loss in the ITO and TJ VCSEL. Comparing the Γgth values, we see that by using the TJ instead of ITO the Γgth is reduced from ∼41.6 cm−1 to ∼14.1 cm−1. Looking at the breakdown of the sources of loss, we see that for the ITO VCSEL the internal loss from the ITO is ∼20 cm−1 higher than the internal loss from the III-nitride layers, resulting in the intracavity contact contributing to 74% of the total internal loss. In contrast, the internal loss from the TJ is lower than the internal loss from all other III-nitride layers. Observing the modal gain vs. current density for devices with 3–10 QWs, we see that for the case of an ITO VCSEL, 10 QWs would be expected to give the lowest Jth; however, for the TJ VCSEL, 7 QWs would yield the lowest Jth. Reducing the number of QWs can be advantageous because, as the number of QWs increases for VCSELs, the barrier width and QW width must be decreased in order to achieve a higher enhancement factor. However, if the barrier width is too small, carriers may not be efficiently confined, leading to a reduction in the radiative recombination efficiency and gain per well. Thus, a smaller number of QWs gives one a larger design space over which the barrier and QW width can be optimized. Next, reducing the number of QWs reduces the chance of some poorly populated QWs contributing to absorption loss in the device.46 Finally, for long wavelength III-nitride VCSELs (>450 nm), a lower number of QWs can mitigate morphological growth issues, such as relaxation, associated with highly strained InGaN QWs.47 Beyond the improvements in Jth that a TJ design offers, we also expect a large improvement in differential efficiency. Using the 1D TMM model for the 7 QW VCSEL design and assuming the top-side of the device emits 99.989% of the light, with an injection efficiency of 65%,43 we calculate the TJ VCSEL to have a top-side differential efficiency (ηd,top) of ∼3%, whereas the ITO VCSEL has an ηd,top of ∼1.1%.

TABLE I.

Summary of VCSEL material constants used in a 1D transmission matrix method (TMM) simulation of the 405 nm cavity mode. The absorption coefficients are rough estimates meant to give a general idea of the internal loss in the structures.15,49–51

LayerThickness (nm)Index, nAbs. Coeff., α (cm−1)
Air (λ/4) 101.25 
12P n-DBR (SiO2 + Ta2O566.8 + 45.6 2.22 + 1.52 
n-Al0.4Ga0.615 2.414 10 
n-GaN 762.65 2.557 10 
7× MQWλ = 405 GaN 2.557 
InGaN 2.72 −gth 
GaN 2.557 
p-Al0.25Ga0.75N (EBL) 2.456 40 
p-GaN 62.2 2.557 40 
P++GaN 14 150 
Intracavity contact ITO (λ/4) 46.7 2.17 2000 
TJ n++GaN 39.6 2.557 100 
n-GaN 39.6 15 
n++GaN 39.6 100 
n-GaN 22.1 15 
Ta2O5 (λ/8) (ITO VCSEL only) 22.802 2.220 0 
16P p-DBR (SiO2 + Ta2O566.8 + 45.6 2.22+1.52 
SiO2 (λ/5.2) 51.37 1.516 
Ti 10 2.046 893 740 
Au (λ/4) 61.37 1.65 607 141 
LayerThickness (nm)Index, nAbs. Coeff., α (cm−1)
Air (λ/4) 101.25 
12P n-DBR (SiO2 + Ta2O566.8 + 45.6 2.22 + 1.52 
n-Al0.4Ga0.615 2.414 10 
n-GaN 762.65 2.557 10 
7× MQWλ = 405 GaN 2.557 
InGaN 2.72 −gth 
GaN 2.557 
p-Al0.25Ga0.75N (EBL) 2.456 40 
p-GaN 62.2 2.557 40 
P++GaN 14 150 
Intracavity contact ITO (λ/4) 46.7 2.17 2000 
TJ n++GaN 39.6 2.557 100 
n-GaN 39.6 15 
n++GaN 39.6 100 
n-GaN 22.1 15 
Ta2O5 (λ/8) (ITO VCSEL only) 22.802 2.220 0 
16P p-DBR (SiO2 + Ta2O566.8 + 45.6 2.22+1.52 
SiO2 (λ/5.2) 51.37 1.516 
Ti 10 2.046 893 740 
Au (λ/4) 61.37 1.65 607 141 
FIG. 2.

Modal gain vs. current density for 405 nm VCSELs with 6.95λ cavity thicknesses and a variable number of QWs. The breakdown of the sources of loss in the cavity is overlaid on the plot, with the calculated threshold modal gain for the ITO intracavity contact VCSEL (Γgth, ITO VCSEL) and III-nitride TJ VCSEL (Γgth, TJ VCSEL) shown as well. Going from the ITO VCSEL to the TJ VCSEL, the Γgth is reduced considerably, suggesting that 10 QWs may be optimal for the ITO VCSEL, but that 7 QWs may be optimal for the TJ VCSEL.

FIG. 2.

Modal gain vs. current density for 405 nm VCSELs with 6.95λ cavity thicknesses and a variable number of QWs. The breakdown of the sources of loss in the cavity is overlaid on the plot, with the calculated threshold modal gain for the ITO intracavity contact VCSEL (Γgth, ITO VCSEL) and III-nitride TJ VCSEL (Γgth, TJ VCSEL) shown as well. Going from the ITO VCSEL to the TJ VCSEL, the Γgth is reduced considerably, suggesting that 10 QWs may be optimal for the ITO VCSEL, but that 7 QWs may be optimal for the TJ VCSEL.

Close modal

In our experimental investigation, we compare an ITO VCSEL and TJ VCSEL with 7 QWs for each design, to eliminate any device performance changes resulting from using different active region designs. Fig. 3 shows the LIV characteristics for the TJ and ITO VCSELs measured under pulsed operation (0.3% duty cycle, 100 ns pulse width) at room temperature. We see a ∼1.5 V increase in the forward voltage going from the ITO VCSEL to the TJ VCSEL. Recent investigations into reducing this voltage penalty suggest that optimizing the surface treatment process before regrowth may improve the voltage by ∼0.5–1 V. Comparing the differential resistance (Rd) for each device, we see that the TJ does not add any series resistance to the device (Rd = 37 Ω), which is in contrast to what is observed in the literature.23–30 Looking at the output power characteristics, we see that the threshold current density is reduced from 8 kA/cm2 (9 mA) for the ITO VCSEL to 3.5 kA/cm2 (4 mA) for the TJ VCSEL. The TJ VCSEL shows a differential efficiency of 0.262%, while the ITO VCSEL differential efficiency is 0.062%. Both of these values are much lower than what is predicted by simulations; however, this is commonly observed in III-nitride VCSELs.10–13,15,18,19,31,32,34 It is likely that this large discrepancy in the differential efficiency is due to the filamentary nature of the lasing in the aperture (details below (Fig. 4)). Both devices were designed to lase at 405 nm; however, a Ta2O5 spacer was accidently deposited on the n-side of the devices, which shifted the cavity resonance wavelength to 410 and 417 nm for the ITO and TJ VCSEL, respectively. It is of note that this may have also led to a misalignment of the peak gain and the cavity resonance wavelengths, which can lead to an increase in the threshold current density. Both devices show a spectrometer resolution limited FWHM of ∼2 nm and a slight increase in the peak wavelength with increasing current (∼0.005 nm/mA).

FIG. 4.

Optical microscope images taken under low gain to prevent detector saturation. (a) shows the ITO VCSEL, while (b) shows the TJ VCSEL operating at various current densities. Both cases exhibit filamentary lasing in the aperture. The non-circular emission pattern in the ITO case is a result of partial over-etching during the top-down PEC etch.

FIG. 4.

Optical microscope images taken under low gain to prevent detector saturation. (a) shows the ITO VCSEL, while (b) shows the TJ VCSEL operating at various current densities. Both cases exhibit filamentary lasing in the aperture. The non-circular emission pattern in the ITO case is a result of partial over-etching during the top-down PEC etch.

Close modal

Optical microscope (near-field) images taken as a function of current density for both devices are shown in Fig. 4. The ITO VCSEL (Fig. 4(a)) and the TJ VCSEL (Fig. 4(b)) display filamentary lasing. This filamentation results in large areas of the aperture not contributing to the stimulated output power. The origin of this filamentation is not well understood; however, in our previous report, we eliminated a number of potential sources, suggesting that it may be a result of non-uniform current spreading, contact resistance, absorption loss, and lateral index fluctuations.15 In the case of the ITO VCSEL, the polycrystalline nature of the ITO was proposed to be creating a spatially varying absorption loss,15 leading to filamentation. However, here we see filamentation in the ITO VCSEL and the TJ VCSEL, where the TJ VCSEL is epitaxially grown, and thus the filamentation is not a result of spatial variations in the intracavity contact absorption loss. Related to this is the consideration of variations in contact resistance and current spreading (local current density) across the aperture. The polycrystalline nature of ITO contacts makes it possible for the contact resistance to vary from grain to grain. Additionally, recent investigations on the MBE regrown TJs employed here have seen large variations in the emission intensity across large area LEDs employing such TJs. Thus, for the ITO and TJ intracavity contacts, local variations in the contact resistance may play an important role in filamentation. With a variation in contact resistance, one would expect a local variation in current density and heating, inducing a change in the local refractive index and loss, which may then induce filamentary lasing. In early reports on GaAs-based lasers, filamentary lasing was also observed, which was predominantly attributed local built-in gain (loss) and refractive index variations.48 Though we have proposed a number of potential causes for filamentation, more rigorous investigations into this phenomenon are necessary to fully understand its origin.

In summary, we have demonstrated a III-nitride VCSEL employing a III-nitride TJ intracavity contact. We discussed the unexpected electrical performance characteristics of such a TJ, building on band-diagram simulations, and comparing fully processed VCSELs with a TJ and ITO intracavity contact. A 1D mode profile simulation was used to reveal the intrinsic threshold modal gain enhancements achieved by a TJ, where we see that the internal loss from the intracavity contact can be reduced to a value below that of all the other III-nitride layers by using a TJ rather than ITO. Comparing the LJV characteristics of a TJ VCSEL to an ITO VCSEL, we see that the TJ VCSEL shows dramatic improvements in threshold current density (3.5 vs. 8 kA/cm2) and differential efficiency (0.262% vs. 0.062%), due to the significantly reduced internal loss. Finally, by comparing the near-field profiles in the TJ and the ITO VCSEL, we suggest that filamentary lasing may be predominantly related to variations in contact resistance and non-uniform current spreading, both of which may then create local variations in the refractive index and free carrier absorption.

The authors would like to thank Mitsubishi Chemical Corporation for providing high-quality free-standing m-plane GaN substrates. This work was funded in part by the King Abdulaziz City for Science and Technology (KACST) Technology Innovations Center (TIC) program, and the Solid State Lighting and Energy Electronics Center (SSLEEC) at the University of California, Santa Barbara (UCSB). Partial funding for this work came from Professor Boon S. Ooi at King Abdullah University of Science and Technology (KAUST), through his participation in the KACST-TIC program. A portion of this work was done in the UCSB nanofabrication facility, with support from the NSF NNIN network (ECS-03357650), as well as the UCSB Materials Research Laboratory (MRL), which was supported by the NSF MRSEC program (DMR-1121053).

1.
J.
Wallace
,
Laser Focus World
51
,
75
(
2015
).
2.
A.
Pruijmboom
,
Laser Focus World
50
,
21
(
2014
).
3.
K. L.
Lear
and
A. N.
Al-Omari
,
Proc. SPIE
6484
,
64840
J (
2007
).
4.
D. K.
Serkland
,
K. M.
Geib
,
G. M.
Peake
,
R.
Lutwak
,
A.
Rashed
,
M.
Varghese
,
G.
Tepolt
, and
M.
Prouty
,
Proc. SPIE
6484
,
648406
(
2007
).
5.
M. J.
Miah
,
A.
Al-Samaneh
,
A.
Kern
,
D.
Wahl
,
P.
Debernardi
, and
R.
Michalzik
,
IEEE J. Sel. Top. Quantum Electron.
19
,
1701410
(
2013
).
6.
R.
Lutwak
,
A.
Rashed
,
M.
Varghese
,
G.
Tepolt
,
J.
Leblanc
,
M.
Mescher
,
C.
Stark
,
D. K.
Serkland
,
K. M.
Geib
, and
G. M.
Peake
, in
39th Annu. Precise Time Time Interval Meet.
(
The Institute of Navigation (ION)
,
2007
), pp.
269
290
.
7.
S.
Knappe
,
V.
Shah
,
P. D. D.
Schwindt
,
L.
Hollberg
,
J.
Kitching
,
L. A.
Liew
, and
J.
Moreland
,
Appl. Phys. Lett.
85
,
1460
(
2004
).
8.
J. J. D.
McKendry
,
D.
Massoubre
,
S.
Zhang
,
B. R.
Rae
,
R. P.
Green
,
E.
Gu
,
R. K.
Henderson
,
A. E.
Kelly
, and
M. D.
Dawson
,
J. Lightwave Technol.
30
,
61
(
2012
).
9.
J. J. D.
McKendry
,
R. P.
Green
,
A. E.
Kelly
,
Z. G. Z.
Gong
,
B.
Guilhabert
,
D.
Massoubre
,
E.
Gu
, and
M. D.
Dawson
,
IEEE Photonics Technol. Lett.
22
,
1346
(
2010
).
10.
T.-C.
Lu
,
C.-C.
Kao
,
H.-C.
Kuo
,
G.-S.
Huang
, and
S.-C.
Wang
,
Appl. Phys. Lett.
92
,
141102
(
2008
).
11.
T.-C.
Lu
,
J.-R.
Chen
,
S.-W.
Chen
,
H.-C.
Kuo
,
C.-C.
Kuo
,
C.-C.
Lee
, and
S.-C.
Wang
,
IEEE J. Sel. Top. Quantum Electron.
15
,
850
(
2009
).
12.
T.-C.
Lu
,
S.-W.
Chen
,
T.-T.
Wu
,
P.-M.
Tu
,
C.-K.
Chen
,
C.-H.
Chen
,
Z.-Y.
Li
,
H.-C.
Kuo
, and
S.-C.
Wang
,
Appl. Phys. Lett.
97
,
071114
(
2010
).
13.
T.-C.
Lu
,
T.-T.
Wu
,
S.-W.
Chen
,
P.-M.
Tu
,
Z.-Y.
Li
,
C.-K.
Chen
,
C.-H.
Chen
,
H.-C.
Kuo
,
S.-C.
Wang
,
H.-W.
Zan
, and
C.-Y.
Chang
,
IEEE J. Sel. Top. Quantum Electron.
17
,
1594
(
2011
).
14.
G.
Cosendey
,
A.
Castiglia
,
G.
Rossbach
,
J.-F.
Carlin
, and
N.
Grandjean
,
Appl. Phys. Lett.
101
,
151113
(
2012
).
15.
J. T.
Leonard
,
D. A.
Cohen
,
B. P.
Yonkee
,
R. M.
Farrell
,
T.
Margalith
,
S.
Lee
,
S. P.
DenBaars
,
J. S.
Speck
, and
S.
Nakamura
,
Appl. Phys. Lett.
107
,
011102
(
2015
).
16.
R.
Michalzik
,
VCSELs: Fundamental, Technology and Applications of Vertical-Cavity Surface-Emitting Lasers
, 1st ed. (
Springer
,
Berlin, Heidelberg
,
2013
).
17.
F. A.
Kish
and
R. M.
Fletcher
,
Semicond. Semimetals
48
,
149
226
(
1997
).
18.
C. O.
Holder
,
J. T.
Leonard
,
R. M.
Farrell
,
D. A.
Cohen
,
B.
Yonkee
,
J. S.
Speck
,
S. P.
DenBaars
,
S.
Nakamura
, and
D. F.
Feezell
,
Appl. Phys. Lett.
105
,
031111
(
2014
).
19.
C. O.
Holder
,
J. S.
Speck
,
S. P.
DenBaars
,
S.
Nakamura
, and
D.
Feezell
,
Appl. Phys. Express
5
,
092104
(
2012
).
20.
J. T.
Leonard
,
D. A.
Cohen
,
B. P.
Yonkee
,
R. M.
Farrell
,
S. P.
Denbaars
,
J. S.
Speck
, and
S.
Nakamura
, “Smooth e-beam-deposited tin-doped indium oxide (ITO) for III-nitride vertical-cavity surface-emitting laser intracavity contacts,”
J. Appl. Phys.
(unpublished).
21.
A.
Hirai
,
Z.
Jia
,
M. C.
Schmidt
,
R. M.
Farrell
,
S. P.
DenBaars
,
S.
Nakamura
,
J. S.
Speck
, and
K.
Fujito
,
Appl. Phys. Lett.
91
,
191906
(
2007
).
22.
R. M.
Farrell
,
D. A.
Haeger
,
X.
Chen
,
C. S.
Gallinat
,
R. W.
Davis
,
M.
Cornish
,
K.
Fujito
,
S.
Keller
,
S. P.
DenBaars
,
S.
Nakamura
, and
J. S.
Speck
,
Appl. Phys. Lett.
96
,
231907
(
2010
).
23.
M.
Diagne
,
Y.
He
,
H.
Zhou
,
E.
Makarona
,
A. V.
Nurmikko
,
J.
Han
,
T.
Takeuchi
, and
M.
Krames
,
Phys. Status Solidi
188
,
105
(
2001
).
24.
S. R.
Jeon
,
Y. H.
Song
,
H. J.
Jang
,
G. M.
Yang
,
S. W.
Hwang
, and
S. J.
Son
,
Appl. Phys. Lett.
78
,
3265
(
2001
).
25.
I.
Ozden
,
E.
Makarona
,
a. V.
Nurmikko
,
T.
Takeuchi
, and
M.
Krames
,
Appl. Phys. Lett.
79
,
2532
(
2001
).
26.
T.
Takeuchi
,
G.
Hasnain
,
S.
Corzine
,
M.
Hueschen
,
R. P.
Schneider
, Jr.
,
C.
Kocot
,
M.
Blomqvist
,
Y.
Chang
,
D.
Lefforge
,
M. R.
Krames
,
L. W.
Cook
, and
S. A.
Stockman
,
Jpn. J. Appl. Phys.
40
,
861
(
2001
).
27.
M.
Diagne
,
Y.
He
,
H.
Zhou
,
E.
Makarona
,
a. V.
Nurmikko
,
J.
Han
,
K. E.
Waldrip
,
J. J.
Figiel
,
T.
Takeuchi
, and
M.
Krames
,
Appl. Phys. Lett.
79
,
3720
(
2001
).
28.
S.
Krishnamoorthy
,
F.
Akyol
, and
S.
Rajan
,
Appl. Phys. Lett.
105
,
141104
(
2014
).
29.
Z. H.
Zhang
,
S.
Tiam Tan
,
Z.
Kyaw
,
Y.
Ji
,
W.
Liu
,
Z.
Ju
,
N.
Hasanov
,
X.
Wei Sun
, and
H.
Volkan Demir
,
Appl. Phys. Lett.
102
,
193508
(
2013
).
30.
S.
Krishnamoorthy
,
T. F.
Kent
,
J.
Yang
,
P. S.
Park
,
R. C.
Myers
, and
S.
Rajan
,
Nano Lett.
13
,
2570
(
2013
).
31.
Y.
Higuchi
,
K.
Omae
,
H.
Matsumura
, and
T.
Mukai
,
Appl. Phys. Express
1
,
121102
(
2008
).
32.
K.
Omae
,
Y.
Higuchi
,
K.
Nakagawa
,
H.
Matsumura
, and
T.
Mukai
,
Appl. Phys. Express
2
,
052101
(
2009
).
33.
D.
Kasahara
,
D.
Morita
,
T.
Kosugi
,
K.
Nakagawa
,
J.
Kawamata
,
Y.
Higuchi
,
H.
Matsumura
, and
T.
Mukai
,
Appl. Phys. Express
4
,
072103
(
2011
).
34.
T.
Onishi
,
O.
Imafuji
,
K.
Nagamatsu
,
M.
Kawaguchi
,
K.
Yamanaka
, and
S.
Takigawa
,
IEEE J. Quantum Electron.
48
,
1107
(
2012
).
35.
S.
Fritze
,
A.
Dadgar
,
H.
Witte
,
M.
Bügler
,
A.
Rohrbeck
,
J.
Bläsing
,
A.
Hoffmann
, and
A.
Krost
,
Appl. Phys. Lett.
100
,
122104
(
2012
).
36.
Y.
Kuwano
,
M.
Kaga
,
T.
Morita
,
K.
Yamashita
,
K.
Yagi
,
M.
Iwaya
,
T.
Takeuchi
,
S.
Kamiyama
, and
I.
Akasaki
,
Jpn. J. Appl. Phys.
52
,
08JK12
(
2013
).
37.
J.
Simon
,
Z.
Zhang
,
K.
Goodman
,
H.
Xing
,
T.
Kosel
,
P.
Fay
, and
D.
Jena
,
Phys. Rev. Lett.
103
,
026801
(
2009
).
38.
STR Gr. (
2015
).
39.
X. a.
Cao
,
E. B.
Stokes
,
P. M.
Sandvik
,
S. F.
LeBoeuf
,
J.
Kretchmer
, and
D.
Walker
,
IEEE Electron Device Lett.
23
,
535
(
2002
).
40.
J. B.
Fedison
,
T. P.
Chow
,
H.
Lu
, and
I. B.
Bhat
,
Appl. Phys. Lett.
72
,
2841
(
1998
).
41.
K.
Sakowski
,
L.
Marcinkowski
,
S.
Krukowski
,
S.
Grzanka
, and
E.
Litwin-Staszewska
,
J. Appl. Phys.
111
,
123115
(
2012
).
42.
T. P.
Chow
and
M.
Ghezzo
, in
III-Nitride, SiC, and Diamond Materials For Electronic Devices
, edited by
D. K.
Gaskill
,
C. D.
Brandt
, and
R. J.
Nemanich
(
Mater. Res. Soc. Proc.
,
1996
), Vol. 423, pp.
69
73
.
43.
R. M.
Farrell
,
D. A.
Haeger
,
P. S.
Hsu
,
K.
Fujito
,
D. F.
Feezell
,
S. P.
Denbaars
,
J. S.
Speck
, and
S.
Nakamura
,
Appl. Phys. Lett.
99
,
171115
(
2011
).
44.
R. M.
Farrell
,
Growth, Fabrication, and Characterization of Continuous-Wave AlGaN-Cladding-Free M-Plane InGaN/GaN Laser Diodes
(
University of California
,
Santa Barbara
,
2010
).
45.
L. A.
Coldren
,
S. W.
Corzine
, and
M. L.
Mašanović
,
Diode Lasers and Photonic Integrated Circuits
(
John Wiley & Sons, Inc.
,
Hoboken, NJ, USA
,
2012
).
46.
J.
Piprek
,
Appl. Phys. Lett.
105
,
011116
(
2014
).
47.
P.
Shan Hsu
,
M. T.
Hardy
,
E. C.
Young
,
A. E.
Romanov
,
S. P.
DenBaars
,
S.
Nakamura
, and
J. S.
Speck
,
Appl. Phys. Lett.
100
,
171917
(
2012
).
48.
F. R.
Nash
,
J. Appl. Phys.
44
,
4696
(
1973
).
49.
E.
Kioupakis
,
P.
Rinke
, and
C. G.
Van de Walle
,
Appl. Phys. Express
3
,
082101
(
2010
).
50.
E.
Kioupakis
,
P.
Rinke
,
A.
Schleife
,
F.
Bechstedt
, and
C. G.
Van De Walle
,
Phys. Rev. B
81
,
241201(R)
(
2010
).
51.
M.
Kuramoto
,
C.
Sasaoka
,
N.
Futagawa
,
M.
Nido
, and
A. A.
Yamaguchi
,
Phys. Status Solidi A
192
,
329
(
2002
).