THz QCL sources for operation above cryogenic temperatures Mikhail Belkin

THz QCL sources for operation above cryogenic temperatures Mikhail Belkin Department of Electrical and Computer Engineering University of Texas at Austin IQCLSW, Monte Verita, Switzerland 008

Need for semiconductor THz source Frequency, Hz Spectroscopy Local oscillator for THz heterodyning Remote sensing, screening, inspection, etc. THz frequency amplification

THz Quantum Cascade Lasers Maximum operating temperature VS frequency BS Williams, Nature Photonics, 57-55 (007) + 008 data of Capasso, Colombelli, and Linfield groups

Upper laser state lifetime vs T LO phonon Electron lifetime in state, sec. µ 0-8. µ 0-6. µ 0-4. µ 0 -. µ 0 - k // ν=thz ν=thz ν=3thz ν=4thz ν=5thz ν=6thz LO phonon scattering of thermal electrons: τ upper (E // < E LO -hν)=0ps τ upper (E // >E LO -hν)=0.5ps Assume D Fermi distribution in state and low population in state E LO =36meV 8.7 THz 0 0 50 00 50 00 50 300 350 Electron temperature, K

RT THz QCLs: very difficult Upper state lifetime <ps for temperature >00K This lifetime is similar to that in mid-ir QCLs BUT, THz QCLs have higher w/g losses and smaller injection efficiency Mid-IR QCLs THz QCLs T max (pulsed) 30 300 80 60 40 0 00 80 60 40 0 00 80 60 40 0 0 First demonstration 994 995 996 997 998 Year T max (pulsed) 30 300 80 60 40 0 00 80 60 40 0 00 80 60 40 0 First demonstration 00 003 004 005 006 007 008 009 Year Nevertheless, no fundamental limit on RT operation

Room-temperature THz source?. Investigate alternative designs, improve GaAs/AlGaAs AlGaAs quality, try different materials (InGaAs/AlInAs,, Si/SiGe SiGe, AlGaN/GaN), low- dimensional systems (quantum dots, wires), etc.. Develop semiconductor THz sources that do not require population inversion across the THz transition

Difference Frequency Generation Difference-frequency generation (DFG) occurs in a medium with second-order nonlinear susceptibility χ () ω ω ω THz =ω -ω Pumps THz QCL source using intra-cavity DFG Dual-frequency mid-infrared QCLs with χ () THz radiation is generated via intra-cavity DFG Widely tunable THz source at RT ω ω ω THz

Challenges for intra-cavity THz DFG ( ) ( ω ) χ I ( ω ) I ( ω ) l coh I THz l coh = ( ) ( ) ) k k k + α THz THz Solid-state systems QCLs I(ω ), I(ω ) 00-000 MW/cm -0 MW/cm l coh t0 mm d0.mm χ () ~00 pm/v can be designed Need giant χ ()

Giant χ () in quantum wells χ z z z 3 ( ) e n nn' n' = Ne + h ε0 n, n' nn' nn' n' n' n n ( ω ω + iγ ) ( ω + ω + iγ ) ( ω ω + iγ ) Sirtori et al., APL (994)

Giant χ () in quantum wells ω ω ω 3 THz χ z z z 3 ( ) e n nn' n' = Ne + h ε0 n, n' nn' nn' n' n' n n ( ω ω + iγ ) ( ω + ω + iγ ) ( ω ω + iγ ) Giant absorption!

χ () with population inversion 3 ω ω... ω THz χ z z z 3 ( ) e n nn' n' = Ne + h ε0 n, n' nn' nn' n' n' n n ( ω ω + iγ ) ( ω + ω + iγ ) ( ω ω + iγ ) Laser action instead of absorption! Gmachl, Belyanin, et al., JQE (003)

χ () with population inversion 3 ω ω... ω THz χ z ( ω ω + iγ ) ( ω + ω + iγ ) ( ω ω + iγ ) Laser action instead of absorption! Active region design z z 3 ( ) e n nn' n' = Ne + h ε0 n, n' nn' nn' n' n' n n Section, χ () and ω Section, ω ω ω ω THz

χ () () -section design Bound-to-continuum active region for λ 9µm [Faist et al. IEEE JQE (00)] Injection 3 4 5 Extraction

χ () () -section design 3 4 5 3 4 5

χ () () -section design -4 Laser gain 3 4 5 - -3-5 Assuming ÑΓ=0meV λ =8.9 µm 00 0 40 60 80 00 Ñω, (mev) 3 4 5

χ () () -section design 3 4 5 30 cm - - -3-4 -5 Laser gain Assuming ÑΓ=0meV N e 0 5 cm -3 λ =8.9 µm 00 0 40 60 80 00 Ñω, (mev) 3 4 5

χ () () -section design 3 4 5 30 cm - - -3-4 -5 Laser gain Assuming ÑΓ=0meV N e 0 5 cm -3 λ =8.9 µm 00 0 40 60 80 00 Ñω, (mev) 3 4 5 ω THz λ =8.9 µm λ THz =60 µm λ =0.5µm χ () 3 0 4 pm/v

Waveguide design y x z 0µm Mode H x, a.u. 0.35 0.30 0.5 0.0 0.5 0.0 λ =8.9 µm λ =0.5 µm 4 0 8 6 4 Refractive index 0.05 30 stages of each QCL structure z y x 0. mm.8 mm 0-30 µm 60 µm Mode H x, a.u. 0.00 0 4 6 8 0 4 0.4 0. 0.0 0.08 0.06 0.04 0.0 z, µm λ THz =60 µm 0 4 0 8 6 4 Refractive index 0.00 0 4 6 8 0 4 z, µm 0 l coh =70µm

Device performance: mid-ir Spectra Power (combined) Intensity, a.u. 300K 50K 50K Peak power, W.5.0.5.0 80K 50K 00K 50K 300K 80K 0.5 900 950 000 050 00 50 00 Wavenumbers, cm - 0.0 0.5.0.5.0.5 3.0 3.5 Peak current, A 0.0 5-µm-wide, tapered to 50-µm-wide, -mm-long, back facet HR coating. Testing in pulsed mode (60ns pulses at 50kHz).

Device performance: mid-ir Spectra Power (combined) Intensity, a.u. 300K 50K 50K Short pass filter Peak power, W.5.0.5.0 80K 50K 00K 50K 300K 80K 0.5 900 950 000 050 00 50 00 Wavenumbers, cm - 0.0 0.5.0.5.0.5 3.0 3.5 Peak current, A 0.0 5-µm-wide, tapered to 50-µm-wide, -mm-long, back facet HR coating. Testing in pulsed mode (60ns pulses at 50kHz).

Product of the pump powers ( ) ( ω ) χ W ( ω ) W ( ω ) W THz l eff W(ω )xw(ω ), W.6.4..0 0.8 0.6 80K 50K 00K 50K 300K 0.4 0. 0.0 0.0 0.5.0.5.0.5 3.0 3.5 Peak current, A 5-µm-wide, tapered to 50-µm-wide, -mm-long, back facet HR coating. Testing in pulsed mode (60ns pulses at 50kHz).

Terahertz emission 80K 4 0 4 0 6 Experiment (zoom-in) Signal, a.u. 6 8 4 8 4 0 00 0 40 60 80 00 8 x 09 7 6 5 N970A uppper 80K DFG signal simulation (incoherently) Wavenumber, cm - Simulations from mid-ir data 0 00 400 600 800 000 00 Wavenumber, cm - signal 4 3 0 00 0 0 30 40 50 60 70 80 90 00 wavenumber (cm-) 5-µm-wide, tapered to 50-µm-wide, -mm-long, back facet HR coating. Testing in pulsed mode (60ns pulses at 50kHz).

Terahertz output at different T Intensity, a.u..5.0.5.0 0.5 0.0 8 6 4 0 60 40 0 0 0.3 µw 300K µw 50K 7 µw 80K 3 4 5 6 7 8 9 0 Frequency, THz Peak positions agree with mid-ir data Red-shift with temperature can also be observed in mid- IR data THz DFG signal observed up to room temperature 5-µm-wide, tapered to 50-µm-wide, -mm-long, back facet HR coating, + Silicon lens Testing in pulsed mode (60ns pulses at 50kHz).

Conversion efficiency at different T Peak THz and mid-ir power vs T Conversion efficiency vs T THz power, µw 8 6 4 0 0.0 90 0 50 80 0 40 70 300 Temperature, K.5.0 0.5 W(ω )W(ω ), W Conversion efficienty, µw/w 9 8 7 6 5 4 3 0 80 0 60 00 40 80 30 Temperature, K Conversion efficiency ~5 µw/w

Conversion efficiency: analysis THz eff S eff, l eff, refractive indices are known from waveguide calculations: n eff 3, l eff 70 µm, S eff 800 µm Estimate χ () using electron density in upper laser state from gain=loss condition: χ () 3x0 4 pm/v Uncertain parameters: Mid-IR lasing in higher order lateral modes THz wave out-coupling efficiency from QCL waveguide (~30%?)

Need for surface extraction of DFG Edge emitting THz DFG is not efficient Mid-IR pump -5 mm Active region substrate Region contributing to THz DFG output, approx. l coh ~0.mm THz DFG Only a small section ~l coh contributes to THz DFG output Surface-emitting scheme to boost THz power THz DFG Mid-IR pump Active region substrate The whole device contributes to THz DFG

Device design for surface emission 0.35 0.30 Grating groves λ =8.9 µm 4 Mode H x, a.u. 0.5 0.0 0.5 0.0 λ =0.5 µm 0 8 6 4 Refractive index 0.05 THz radiation is produced by P () P () (ω THz =ω -ω )~χ () E(ω )E(ω )e i(k -k )y For efficient vertical outcoupling, we need the grating k-vector k gr =k -k Mode H x, a.u. 0.00 0 4 6 8 0 4 0.4 0. 0.0 0.08 0.06 0.04 0.0 z, µm Grating groves 0.00 0 4 6 8 0 4 z, µm λ THz =60 µm 0 4 0 8 6 4 0 Refractive index

Surface emission Edge-emission mid-ir spectrum Surface-emission THz spectrum 5-µm-wide,with second order grating for THz DFG, -mm-long, back facet HR coating. Testing in pulsed mode (60ns pulses at 50kHz) with 3.5A current pulses at 80K.

Power output Edge emission mid-ir power Surface emission THz power 5-µm-wide,with second order grating for THz DFG, -mm-long, back facet HR coating. Testing in pulsed mode (60ns pulses at 50kHz) with 3.5A current pulses at 80K.

Future improvements Improve waveguide designs for larger l coh Improve active region designs for higher χ () Improved surface emission for more outcoupling Mode-locking for higher peak intensity mw-level THz DFG QCL sources

People Harvard: Federico Capasso, Christian Pflügl gl,, Jonathan Fan, Qijie Wang, Laurent Diehl, Sahand Hormoz. Acknowledgements Texas A&M: Alexey Belyanin, Feng Xie University of Leeds: Edmund Linfield, Suraj Khanna,, Mohamed Lachab, Giles Davies University of Paris: Raffaele Colombelli, Yannick Chassagneux ETH-Zürich: Jérôme Faist,, Milan Fischer, Andreas Wittmann Funding AFOSR FA9550-05 05-04350435

Postdoc opportunities University of Texas, Austin City of Austin mbelkin@ece.utexas.edu http://www.ece.utexas.edu/~mbelkin/