Laser Offset Stabilization for Terahertz (THz) Frequency Generation Kevin Cossel Dr. Geoff Blake California Institute of Technology Kevin Cossel Dr. Geoff Blake California Institute of Technology

What is Terahertz Spectroscopy?  ~1x x10 13 Hz or ~ Terahertz (THz)  ~ cm -1  ~ µm  Also known as far-infrared (FIR) or sub-millimeter spectroscopy  Study low-energy processes both in the laboratory and in remote sensing applications  ~1x x10 13 Hz or ~ Terahertz (THz)  ~ cm -1  ~ µm  Also known as far-infrared (FIR) or sub-millimeter spectroscopy  Study low-energy processes both in the laboratory and in remote sensing applications

Why study Thz region?  Many uses  High-resolution spectroscopy  Vibration-rotation coupling  Lower spectral density expected  Remote sensing  Astronomy :  Matched to emission from cold dust clouds  Characterize organic material (especially amino acids) present in the interstellar medium  Lower spectral density expected  SOFIA & Herschel  Need lab data first  Many uses  High-resolution spectroscopy  Vibration-rotation coupling  Lower spectral density expected  Remote sensing  Astronomy :  Matched to emission from cold dust clouds  Characterize organic material (especially amino acids) present in the interstellar medium  Lower spectral density expected  SOFIA & Herschel  Need lab data first

THz sources  Existing sources have problems  Solid-state electronic oscillators  Power drops above 200 MHz  Doubling/tripling not good above 1 THz  Lasers  Low frequency = long lifetime, no direct bandgap lasers  Quantum cascade lasers – >3 THz, 10 Kelvin, narrow tunability  THz Time Domain Spectroscopy  Probe with sub-picosecond pulses  Gate detector with laser  Limited resolution  Optical-heterodyne  Existing sources have problems  Solid-state electronic oscillators  Power drops above 200 MHz  Doubling/tripling not good above 1 THz  Lasers  Low frequency = long lifetime, no direct bandgap lasers  Quantum cascade lasers – >3 THz, 10 Kelvin, narrow tunability  THz Time Domain Spectroscopy  Probe with sub-picosecond pulses  Gate detector with laser  Limited resolution  Optical-heterodyne

Purpose  Develop a spectrometer that can be used to characterize the spectra of molecules in the range of ~ Terahertz (THz)  Need THz source  Inexpensive  Multiterahertz bandwidth  Accurate  Low linewidth (<10 MHz)  High-stability  Develop a spectrometer that can be used to characterize the spectra of molecules in the range of ~ Terahertz (THz)  Need THz source  Inexpensive  Multiterahertz bandwidth  Accurate  Low linewidth (<10 MHz)  High-stability

Frequency Modulation Change current = change laser frequency The same as adding frequency components Then scan the laser What’s happening?

Frequency Modulation Spectroscopy of HDO

Diode laser locking  Use feedback to reduce wavelength fluctuations (reduce linewidth)  FMS signal is error signal  Negative error increases wavelength  Use PID controller: Feedback = P + I + D P = proportional to error signal I = integrate error (remove offset) D = derivative (anticipate movement)  Use feedback to reduce wavelength fluctuations (reduce linewidth)  FMS signal is error signal  Negative error increases wavelength  Use PID controller: Feedback = P + I + D P = proportional to error signal I = integrate error (remove offset) D = derivative (anticipate movement) 0 Locking Range Error Wavelength

Tunable locking  Lock laser 1 to HDO line  Generate offset between laser 1 and laser 2  Lock offset  Lock laser 3 to different HDO line  Output is difference between laser 2 & laser 3  Narrow tune = offset  Wide tune = lock to different lines  Lock laser 1 to HDO line  Generate offset between laser 1 and laser 2  Lock offset  Lock laser 3 to different HDO line  Output is difference between laser 2 & laser 3  Narrow tune = offset  Wide tune = lock to different lines

FMS Locking  Electro-optic modulator provides frequency modulation  Photodetector varying intensity beat note  Mix with driving RF DC output  Feedback DC error signal to PID controller  Controls piezo which adjust wavelength  Electro-optic modulator provides frequency modulation  Photodetector varying intensity beat note  Mix with driving RF DC output  Feedback DC error signal to PID controller  Controls piezo which adjust wavelength

Offset Locking  Laser 1 locked to HDO  Lasers 1 and 2 combined on fast (40 GHz) photodetector  Output difference frequency  Mix with tunable RF source Output 0-1 GHz  Send to source locking counter  Feedback to laser 2, offset locking up to ±20 GHz  Laser 1 locked to HDO  Lasers 1 and 2 combined on fast (40 GHz) photodetector  Output difference frequency  Mix with tunable RF source Output 0-1 GHz  Send to source locking counter  Feedback to laser 2, offset locking up to ±20 GHz

Results – FMS locking  2 hours  Free-running (blue)  47 MHz standard deviation  4.9 MHz RMSE  2 MHz/second drift  Locked (red)  Mean 20 kHz  3.5 MHz standard deviation  5x10 -5 MHz/second drift  2 hours  Free-running (blue)  47 MHz standard deviation  4.9 MHz RMSE  2 MHz/second drift  Locked (red)  Mean 20 kHz  3.5 MHz standard deviation  5x10 -5 MHz/second drift  10 seconds  Free-running (blue)  30 MHz peak-peak deviations  5.5 MHz standard deviation  Locked (red)  10 MHz peak-peak  3 MHz standard deviation  10 seconds  Free-running (blue)  30 MHz peak-peak deviations  5.5 MHz standard deviation  Locked (red)  10 MHz peak-peak  3 MHz standard deviation

Results – Offset locking  Difference frequency  Two free-running (blue, left):  300 MHz drift  5 MHz RMSE  One laser PID locked (red)  PID + offset locking  1.3 MHz standard deviation (over 75 seconds)  Mean accurate to 260 kHz  <1x10 -6 MH/second drift (stable for 15 hours)  Difference frequency  Two free-running (blue, left):  300 MHz drift  5 MHz RMSE  One laser PID locked (red)  PID + offset locking  1.3 MHz standard deviation (over 75 seconds)  Mean accurate to 260 kHz  <1x10 -6 MH/second drift (stable for 15 hours)

Discussion  Currently:  PID lock  20 kHz accuracy  3 MHz linewidth  Low drift  Offset (Lasers 1 & 2)  ±20 GHz (easily changed to ±40 GHz)  300 kHz accuracy  Very stable  High spectral density of HDO  Predicted: >3 THz bandwidth, 8 MHz linewidth, 300 kHz accuracy  Work to lower linewidth/improve accuracy  Currently:  PID lock  20 kHz accuracy  3 MHz linewidth  Low drift  Offset (Lasers 1 & 2)  ±20 GHz (easily changed to ±40 GHz)  300 kHz accuracy  Very stable  High spectral density of HDO  Predicted: >3 THz bandwidth, 8 MHz linewidth, 300 kHz accuracy  Work to lower linewidth/improve accuracy

Conclusion  Developed a technique for generating a tunable THz difference between two lasers with a final linewidth of <10 MHz  Combine lasers on ErAs/InGaAs photomixer to generate THz radiation  Other techniques could provide higher stability at the cost of tunability or wide bandwidth but limited resolution  Compromise system  Working on improving linewidth (hopefully 1 MHz) and bandwidth (up to 15 THz)  Tunability/linewidth combination already useful for spectroscopy (developing Fourier transform terahertz spectrometer)  Developed a technique for generating a tunable THz difference between two lasers with a final linewidth of <10 MHz  Combine lasers on ErAs/InGaAs photomixer to generate THz radiation  Other techniques could provide higher stability at the cost of tunability or wide bandwidth but limited resolution  Compromise system  Working on improving linewidth (hopefully 1 MHz) and bandwidth (up to 15 THz)  Tunability/linewidth combination already useful for spectroscopy (developing Fourier transform terahertz spectrometer)

Acknowledgements Dr. Geoff Blake Rogier Braakman Matthew Kelley Dan Holland NSF Grant Dr. Geoff Blake Rogier Braakman Matthew Kelley Dan Holland NSF Grant