Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

8.5 W VECSEL output at 1270 nm with conversion efficiency of 59 %

Open Access Open Access

Abstract

We report on 1270 nm vertical-external-cavity surface-emitting lasers (VECSELs) with up to 59% conversion efficiency and maximum output power of 8.5 W (pump limited), at 5°C heat sink temperature. These VECSELs comprised wafer-fused gain mirrors in the flip-chip (thin-disk) heat dissipation scheme. The reflected pump light from the gain mirror surface was found to depend considerably on temperature and pump power.

© 2015 Optical Society of America

1. Introduction

Semiconductor vertical-external-cavity surface-emitting lasers (VECSELs) (also called semiconductor disc lasers, SDLs) can efficiently convert low-cost laser diode pump light into high quality laser beams at selected wavelength. They combine the emission wavelength flexibility of semiconductor lasers based on bandgap engineering with the benefits of high output power with high beam quality of solid state lasers [1]. These features have made VECSELs useful in a wide variety of applications, ranging from spectroscopy [2], to pumping other solid-state lasers [3], and microwave photonics [4]. Extension of the emission wavelength range through higher harmonics yields VECSELs operating at visible wavelengths, which are attractive for medical applications [5] and laser projectors [6].

High performance VECSELs have been reported for wavelengths between 800 and 1200 nm [7–9] and above 2 µm [10]. VECSELs emitting around 1300 nm can fill in the gap of available high power solid state lasers of this spectral range, and offer an attractive approach for red light high power lasers using intra-cavity second harmonic generation. The InP-based material system is typically used for 1300 nm edge-emitting lasers due to its high optical gain. Yet, the low contrast in refractive index and low thermal conductivity in this material system results in poor performance of InP-based distributed Bragg reflectors (DBRs), an essential component in a VECSEL structure. Common approaches for GaAs-based VECSELs in this wavelength range employ AlGaAs/GaAs DBRs combined with dilute-nitride GaInNAs quantum wells (QWs) [11] or quantum-dot based active regions [12] in their gain mirrors, but suffer from low optical gain at the longer wavelengths part of the spectrum. An alternative technique for making VECSELs emitting in the 1200–1600 nm wavelength range is by fusing together gain mirrors composed of AlGaAs DBRs and InP-based QW active cavities, using wafer bonding [13]. With such composite gain mirrors, one can combine the best gain material for this wavelength range, InP-QW-based, with GaAs-based DBRs for good thermal management. Wafer-fused VECSELs have achieved Watt-level output powers both for the 1300 nm and the 1500 nm wavelength range [14,15]. For 1300 nm emission, employing intra-cavity diamond heat-spreaders and the flip-chip (thin disk) mounting, record output powers of 7.1 W [16] and 6.1 W [17], respectively, have been reached. Further progress in this area requires improving thermal management at the gain chip level, especially with larger pumped area, as well as more efficient absorption of the pump beam by the active region (e.g. 50% single-pass 980 nm pump light absorption was demonstrated in [17]).

In this work we increased the pump absorption efficiency in wafer-fused GaAs-InP gain mirrors by increasing the reflectivity for the pump light at the DBR-metal interface, similar to the reports in [18, 19]. This interface exists in the flip-chip dissipation scheme where the DBR-side of the gain mirror is bonded via metal layer to the heat-spreader. Applying this modification resulted in increasing the maximum output power of the 1270 nm VECSEL to 8.5 W. In addition, we measured systematically the pump power reflected off our gain mirrors. We observed that the reflectivity is not constant with respect to either absorbed power or temperature. By taking into account the measured reflectivity, we estimate the conversion efficiency to be as high as 59% for these VECSELs.

2. Gain mirrors

The structure of the gain mirrors employed is illustrated in Fig. 1 (left). It consisted of an InAlGaAs/InP-based QW active region that includes InAlGaAs gain section, and InP spacer and cap layers. It was designed for pumping at 980 nm and was grown by metallorganic vapor phase epitaxy (MOVPE). The InAlGaAs QWs, with a photoluminescence (PL) peak wavelength of 1250 nm (room temperature), were placed at the antinodes of the optical field inside the active region with a 3-3-2-2 distribution of the number of QWs in each group. They were separated by InAlGaAs barriers and spacers, characterized by a PL spectrum peaked at ~1 µm (room temperature), where the pump-light should be absorbed. InP layers are placed on both sides of the InAlGaAs QW section in order to confine the photo-generated carriers. The resulting active region is about 3 λ thick. This relatively thin sub-cavity design results from a trade-off between the absorption efficiency and the low thermal conductivity of the InAlGaAs segment. The active region was wafer-fused with a MOVPE grown DBR, comprising 24 quarter-wavelength GaAs/AlAs layer pairs. The wafer fusion process was previously described in [13–17].

 figure: Fig. 1

Fig. 1 Left: A schematic cross-section of the gain mirror structure. Varying the thickness of the Ti adhesion layer changes the optical performance. Right: Direct cavity VECSEL in flip-chip configuration on a water-cooled heat sink (HS), with the reflected light being discarded in a beam dump (BD). We detect, in situ, the pumped (with detector DP), and reflected (DR) beams, using partially reflecting beam samplers, as well as the emitted (E) power (see text for details).

Download Full Size | PDF

After selective etching of the GaAs substrate, the DBR-side of the fused structure was metalized with Ti/Au, the wafer was cleaved into chips, and then thermo-compression-bonded to a metalized chemical vapor deposited (CVD) diamond surface. On one part of the wafer we deposited a standard metallization scheme with ~50 nm Ti and ~150 nm Au, providing a reference sample. On another part we deposited only ~5 nm of Ti with again ~150 nm Au. The Ti is necessary for bonding the weakly reactive Au onto the DBR. However, Ti absorbs the pump light and thus acts as additional thermal load. A thin Ti layer is preferred, resulting in this interface reflecting the residual pump light back into the active region [18, 19].

3. Optical measurements setup

The gain mirrors were mounted in a direct cavity VECSEL configuration, schematized in Fig. 1 (right), with an output coupler (OC) consisting of a spherical mirror of radius of curvature of 50 mm and 2.5% output coupling. The 980 nm diode pump light was delivered via a 200 µm diameter multi-mode fiber, and imaged onto the sample employing an achromatic lens of focal length f1 = 45 mm in combination with a plano-convex lens of f2 = 75 mm; the distance between the two lenses being L = 120 mm (see Fig. 1). This resulted in a pump spot of ~333 µm diameter, elongated due to an incidence angle of approximately 36°.

We measured, in situ, the pumped (P) and reflected (R) power, using partially reflecting beam samplers, as well as the emitted power (E). The absorbed power referred to hereafter is the difference between pump and reflection, A = P-R. As Hader et al. [20] have pointed out, this definition of absorbed power yields an overestimated value, since surface scattering losses and additional photoluminescence losses are neglected. For the current report we did not estimate the contributions of these loss channels. However, we have performed in situ measurements of the reflected pump power for all VECSEL operation conditions.

4. VECSEL performance

Figure 2 shows the light-light characteristics for the two gain mirrors described above. Heat sink temperature and maximally achieved emission power are stated in the insets. The performance of the reference sample, Fig. 2 (left), is comparable to previously reported results obtained in the flip-chip configuration [17]. The pump-limited maximum output power values for the thin Ti layer sample, Fig. 2 (right), at 5°C and 15°C of 8.5 W and 7.7 W, respectively, both exceed the previously set benchmark of emitted 7.1 W at 7°C with a 300 µm spot in the intra-cavity diamond heat-spreader approach [16]. These results demonstrate for the first time gain mirrors in the 1300 nm waveband in the flip-chip dissipation scheme that surpass the performance of previously published results based on the intra-cavity arrangement.

 figure: Fig. 2

Fig. 2 Light-light characteristics for gain mirrors with thick (left) and thin (right) Ti layer, respectively. The insets list the maximally achieved output powers and conversion efficiencies corresponding to the data at different temperatures.

Download Full Size | PDF

In addition, we have systematically recorded the reflected pump power off the gain mirror surface. The reflectivity, r = R/P, was found to be not constant, neither with respect to heat sink temperature, nor with respect to absorbed power. As one can observe in Fig. 3 a decrease in heat sink temperature and an increase in pump power was accompanied by an increase in reflectivity. One possible explanation for these findings is that, since the InAlGaAs light absorbing spacers have a nominal room temperature PL peak at ~1 µm, the 980 nm pump operates near the absorption edge. By decreasing the temperature, the absorption edge is blue-shifted, resulting in a decrease in absorption coefficient for the 980 nm pump, accompanied by a corresponding increase of internally reflected residual pump light. On the other hand, by increasing the pump power we reach the absorption saturation in the InAlGaAs material, which leads to the same effect. We also performed reflectivity measurements of the gain mirror after removing the output coupler – see inset Fig. 4 (left). We observed the saturation in the lasing configuration to occur at lower pump power levels. This is probably related to a different carrier dynamic under lasing and non-lasing conditions. Said effect is less pronounced at higher temperatures: there pump absorption in InAlGaAs spacers and barriers is increased. Following the saturation onset, the thermal effect takes over, letting the reflectivity curves to converge, as shown in Fig. 3. To understand whether the peaks in Fig. 3 relate with the roll over in Fig. 2, further investigations are necessary.

 figure: Fig. 3

Fig. 3 Reflectivity (r = R/P) of samples with thick (left) and thin (right) Ti layer, respectively, versus absorbed power

Download Full Size | PDF

 figure: Fig. 4

Fig. 4 Summaries of base reflectivity (indicated by dot-dashed lines in Fig. 3) (left) and conversion efficiencies (dashed lines of Fig. 2) (right), versus heat sink temperature. Expressions for the linear fit are also shown. Inset: Reflectivity versus absorbed power of the thick Ti layer sample with (black) and without (grey, line) output coupler, for three temperatures.

Download Full Size | PDF

As illustrated in Fig. 3, the measured reflectivity, r = R/P, of the thin Ti layer sample (right) is significantly higher than that of the reference (left). This higher reflectivity is due to larger absorption of the pump light by the thicker Ti layer in the reference sample. The low pump regime “base” reflectivity (indicated in the figures with dot-dashed lines) exhibits a linear variation with heat sink temperature, summarized in Fig. 4 (left). As a result of this higher reflectivity, the conversion efficiency in the sample with the thinner Ti layer is also higher, as shown in Fig. 4 (right), reaching a maximum value of 59% at 5°C heat sink temperature. In this work, conversion efficiency represents the gradient between absorbed and emitted power, as indicated by the dashed lines in Fig. 2. The improved conversion efficiency is because of better recycling of the pump power, reflected from the Ti layer and passing a second time in the QW absorbing layers. Naturally, the calculated maximum conversion efficiency with respect to pump power rather than absorbed power is lower, namely 20% and 14% for the thin and thick Ti layer gain mirrors, respectively.

Figure 5 presents the emission spectra of the two VECSEL types, corresponding to the maximum output power conditions of the different heat sink temperatures. The values for the thick Ti layer reference sample with heat sink temperatures of 28°C and above refer to the roll over point of the light-light characteristics. For this sample, the longest emitted wavelengths appear to converge to a common value (~1272 nm). Due to pump power limitations, a similar conclusion cannot yet be reached for the thin Ti layer case. Further work is needed to understand the implications of the spectral behavior regarding the thermal resistance of the structure [20]. The absence of intensity modulation across the lasing spectrum illustrates the advantage of the flip-chip mounting over the intra-cavity diamond configuration [16].

 figure: Fig. 5

Fig. 5 Emission spectra of the VECSELs employing the thick (left) and thin (right) Ti layers in the gain mirrors, for different heat sink temperatures, measured under conditions where the peak output is achieved. The values of these peak powers are indicated for each heat sink temperature.

Download Full Size | PDF

5. Conclusion

The pumping efficiency of gain mirrors for 1300 nm-range VECSELs employing wafer-fused InP-based QW active material and GaAs-based DBRs was increased by lowering the Ti content in the metal interface between DBR and heat-spreader, resulting in enhanced reflectivity of the pump beam back into the optical gain element. With such gain mirrors, arranged in a flip-chip geometry, record output power of 8.5 W and 59% efficiency at 5°C were achieved at 1270 nm emission wavelength. Detailed comparison of gain mirrors with different reflectivities at the metal surface evidence the importance of reducing absorption of the pump beam at the metal layer. The flip-chip gain mirror mounting configuration is shown to be very effective in removing heat and increasing the overall power conversion efficiency.

Acknowledgments

This work was supported by the Swiss Nanotera Valorization Fund.

References and links

1. M. Kuznetsov, F. Hakimi, R. Sprague, and A. Mooradian, “High-power (0.5-W CW) diode-pumped vertical-external-cavity surface-emitting semiconductor lasers with circular TEM00 beams,” IEEE Photon. Technol. Lett. 9(8), 1063–1065 (1997). [CrossRef]  

2. P. Čermák, J. Hovorka, P. Veis, P. Cacciani, J. Cosléou, J. El Romh, and M. Khelkhal, “Spectroscopy of 14NH3 and 15NH3 in the 2.3 µm spectral range with a new VECSEL laser source,” J. Quant. Spectrosc. Radiat. Transf. 137, 13–22 (2014). [CrossRef]  

3. L. Fan, M. Fallahi, J. Hader, A. R. Zakharian, M. Kolesik, J. V. Moloney, T. Qiu, A. Schulzgen, N. Peyghambarian, W. Stolz, S. W. Koch, and J. T. Murray, “Over 3 W high-efficiency vertical-external-cavity surface-emitting lasers and application as efficient fiber laser pump sources,” Appl. Phys. Lett. 86(21), 211116 (2005). [CrossRef]  

4. G. Baili, L. Morvan, M. Alouini, D. Dolfi, F. Bretenaker, I. Sagnes, and A. Garnache, “Experimental demonstration of a tunable dual-frequency semiconductor laser free of relaxation oscillations,” Opt. Lett. 34(21), 3421–3423 (2009). [CrossRef]   [PubMed]  

5. S. B. Brown, E. A. Brown, and I. Walker, “The present and future role of photodynamic therapy in cancer treatment,” Lancet Oncol. 5(8), 497–508 (2004). [CrossRef]   [PubMed]  

6. J. L. Chilla, H. Zhou, E. Weiss, A. L. Caprara, Q.-Z. Shou, S. V. Govorkov, M. K. Reed, and L. Spinelli, “Blue and green optically pumped semiconductor lasers for display,” Proc. SPIE 5740, 41–47 (2005).

7. K. S. Kim, J. Yoo, G. Kim, S. Lee, S. Cho, J. Kim, T. Kim, and Y. Park, “920-nm vertical-external-cavity surface-emitting lasers with a slope efficiency of 58% at room temperature,” IEEE Photon. Technol. Lett. 19(20), 1655–1657 (2007). [CrossRef]  

8. B. Heinen, T.-L. Wang, M. Sparenberg, A. Weber, B. Kunert, J. Hader, S. W. Koch, J. V. Moloney, M. Koch, and W. Stolz, “106 W continuous-wave output power from vertical-external-cavity surface-emitting laser,” Electron. Lett. 48(9), 516–517 (2012). [CrossRef]  

9. E. Kantola, T. Leinonen, S. Ranta, M. Tavast, J.-P. Penttinen, and M. Guina, “1180nm VECSEL with 50 W output power,” Proc. SPIE 9349, 93490U (2015).

10. S. Kaspar, M. Rattunde, T. Töpper, U. Schwarz, C. Manz, K. Köhler, and J. Wagner, “Electro-optically cavity dumped 2m semiconductor disk laser emitting 3ns pulses of 30 W peak power,” Appl. Phys. Lett. 101(14), 141121 (2012). [CrossRef]  

11. J. M. Hopkins, S. A. Smith, C. W. Jeon, H. D. Sun, D. Burns, S. Calvez, M. D. Dawson, T. Jouhti, and M. Pessa, “0.6 W CW GaInNAs vertical external-cavity surface emitting laser operating at 1.32 um,” Electron. Lett. 40(1), 30–31 (2004).

12. M. Butkus, J. Rautiainen, O. G. Okhotnikov, C. J. Hamilton, G. P. A. Malcolm, S. S. Mikhrin, I. L. Krestnikov, D. A. Livshits, and E. U. Rafailov, “Quantum dot based semiconductor disk lasers for 1-1.3 μm,” IEEE J. Sel. Top. Quantum Electron. 17(6), 1763–1771 (2011). [CrossRef]  

13. A. Sirbu, K. Pierscinski, A. Mereuta, V. Iakovlev, A. Caliman, Z. Micovic, N. Volet, J. Rautiainen, J. Heikkinen, J. Lyytikainen, A. Rantamaki, O. G. Okhotnikov, and E. Kapon, “Wafer-fused VECSELs emitting in the 1310 nm waveband,” Proc. SPIE 8966, 8966OG (2014).

14. J. Lyytikäinen, J. Rautiainen, L. Toikkanen, A. Sirbu, A. Mereuta, A. Caliman, E. Kapon, and O. G. Okhotnikov, “1.3-µm optically-pumped semiconductor disk laser by wafer fusion,” Opt. Express 17(11), 9047–9052 (2009). [CrossRef]   [PubMed]  

15. J. Rautiainen, J. Lyytikäinen, A. Sirbu, A. Mereuta, A. Caliman, E. Kapon, and O. G. Okhotnikov, “2.6 W optically-pumped semiconductor disk laser operating at 1.57-µm using wafer fusion,” Opt. Express 16(26), 21881–21886 (2008). [CrossRef]   [PubMed]  

16. A. Sirbu, A. Rantamäki, E. J. Saarinen, V. Iakovlev, A. Mereuta, J. Lyytikäinen, A. Caliman, N. Volet, O. G. Okhotnikov, and E. Kapon, “High performance wafer-fused semiconductor disk lasers emitting in the 1300 nm waveband,” Opt. Express 22(24), 29398–29403 (2014). [CrossRef]   [PubMed]  

17. A. Rantamäki, A. Sirbu, E. J. Saarinen, J. Lyytikäinen, A. Mereuta, V. Iakovlev, E. Kapon, and O. G. Okhotnikov, “High-power flip-chip semiconductor disk laser in the 1.3 μm wavelength band,” Opt. Lett. 39(16), 4855–4858 (2014). [PubMed]  

18. J. Hader, T.-L. Wang, J. M. Yarborough, C. A. Dineen, Y. Kaneda, J. V. Moloney, B. Kunert, W. Stolz, and S. W. Koch, “VECSEL optimization using microscopic many-body physics,” IEEE J. Sel. Top. Quantum Electron. 17, 1753 (2011).

19. M. Devautour, A. Michon, G. Beaudoin, I. Sagnes, L. Cerutti, and A. Garnache, “Thermal management for high-power single-frequency tunable diode-pumped VECSEL emitting in the near- and mid-IR,” IEEE J. Sel. Top. Quantum Electron. 19(4), 1701108 (2013). [CrossRef]  

20. J. Hader, T.-L. Wang, J. V. Moloney, B. Heinen, M. Koch, S. W. Koch, B. Kunert, and W. Stolz, “On the measurement of the thermal impedance in vertical-external-cavity surface-emitting lasers,” J. Appl. Phys. 113(15), 153102 (2013). [CrossRef]  

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (5)

Fig. 1
Fig. 1 Left: A schematic cross-section of the gain mirror structure. Varying the thickness of the Ti adhesion layer changes the optical performance. Right: Direct cavity VECSEL in flip-chip configuration on a water-cooled heat sink (HS), with the reflected light being discarded in a beam dump (BD). We detect, in situ, the pumped (with detector DP), and reflected (DR) beams, using partially reflecting beam samplers, as well as the emitted (E) power (see text for details).
Fig. 2
Fig. 2 Light-light characteristics for gain mirrors with thick (left) and thin (right) Ti layer, respectively. The insets list the maximally achieved output powers and conversion efficiencies corresponding to the data at different temperatures.
Fig. 3
Fig. 3 Reflectivity (r = R/P) of samples with thick (left) and thin (right) Ti layer, respectively, versus absorbed power
Fig. 4
Fig. 4 Summaries of base reflectivity (indicated by dot-dashed lines in Fig. 3) (left) and conversion efficiencies (dashed lines of Fig. 2) (right), versus heat sink temperature. Expressions for the linear fit are also shown. Inset: Reflectivity versus absorbed power of the thick Ti layer sample with (black) and without (grey, line) output coupler, for three temperatures.
Fig. 5
Fig. 5 Emission spectra of the VECSELs employing the thick (left) and thin (right) Ti layers in the gain mirrors, for different heat sink temperatures, measured under conditions where the peak output is achieved. The values of these peak powers are indicated for each heat sink temperature.
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.