Frequency Band Performance Comparisons for Room-Temperature Chirped Pulse Millimeter Wave Spectroscopy Justin L. Neill, Brent J. Harris, Robin L. Pulliam, Matt T. Muckle BrightSpec, Inc., 770 Harris St. Suite 104b, Charlottesville, VA, USA. Brooks H. Pate Department of Chemistry, University of Virginia, McCormick Road, Charlottesville, VA, USA. June 24, 2016 FC10
Overview Goal of the study: To offer a quantitative performance comparison between room temperature chirped-pulse millimeter wave spectrometers operating in different frequency bands for rapid analytical characterization of complex mixtures. Criteria: Correct identifications (needs: high resolution, good frequency coverage) Particular challenge is finding a weak component in a strong ‘forest’ Range of molecular coverage (needs: bandwidth, matched to molecules of interest) Sensitivity (needs: high power, well matched to strong transitions) Ability to do advanced pulse sequences (need: high power to reach p/2, p pulse conditions) Acknowledgement: This material is based upon work supported by the U.S. Army under Contract No. W31P4Q-15-C-0019. 6/24/2016
The Theory The basic population/frequency factors favor higher frequency…. Simulations at 313 K Propylene Acrolein Anisole Note: Very small molecules (e.g. NH3, H2O, CO, NO….) have wider frequency spacings and do not emit in every band. 6/24/2016
The Instruments ….but other band-dependent effects also play a role. Specification 75-110 GHz (W-Band) 260-290 GHz 520-580 GHz Source power (typ.) 30 mW* 30 mW 2 mW SSB Conversion Loss 10 dB 8.5 dB 11.5 dB Beam Size (FWHM) 4.0 cm 1.8 cm 1.2 cm Typical Transmission Efficiency (power) 40% 35% 15% Typical FID dephasing time (1/e2, Doppler limited, OCS) 3 ms 900 ns 450 ns *More power available (Gallium Nitride pulsed amplifiers) Beam size is set by parabolic mirrors – and limited by desire to have Rayleigh length of ~80 cm The bigger beam size at W-band allows us to interact with more molecules and achieve better sensitivity – signal proportional to A1/2L – if we can achieve the same Rabi flip angle excitation (and slower dephasing at W-band allows this!) 6/24/2016
The Measurements Broadband Survey: Segmented Chirped Pulse Millimeter Wave Spectroscopy Chirped pulses generated by 12 GS/s arbitrary waveform generator (Keysight M8190, typ. -60 dBc spur-free dynamic range) Detection made on 4 GS/s digitizer w/real time signal accumulation (Keysight U1084A) Targeted Monitoring: Achieve “p/2” resonance condition for optimum sensitivity Can perform double resonance measurements to recover selectivity in complex mixtures Time (ms) Intensity (V) J.L. Neill et al., Opt. Express 2013, 21, 19743. A.L. Steber et al., 68th ISMS, 2013, Talk WH12. B.J. Harris et al., 68th ISMS, June 2013, Talk WH13. B.J. Harris et al., Proc. SPIE 9362, Terahertz, RF, MMW and Sub-MMW Technology and Applications VIII, 936215, Feb 2015. 6/24/2016
The Measurements Optical: Stainless steel sampling chamber, 65 cm length CF63 end ports (I.D. 6 cm), Teflon windows Dursan coated, heating capability Off-axis parabolic mirrors, 75 mm eff. focal length 20-25 dBi standard gain feed horns 75-110 GHz FT-MRR Spectrometer in Discovery form factor (Reconfigurable) Calibrated mixture – six components at 1% each, balance N2 6/24/2016
Partial Pressure 3s Sensitivity (mTorr) The Results Single Component Reference Samples Example: Ethanol 75-110 GHz, 2.4 mTorr 260-290 GHz, 1.3 mTorr H2O 520-580 GHz, 2 mTorr Molecule Partial Pressure 3s Sensitivity (mTorr) 10 min broadband survey 75-110 GHz 260-290 GHz 520-580 GHz CH3CN 0.11 0.006 OCS 0.16 0.019 1.1 MeOH 2.0 1.2 4.1 EtOH 2.5 0.91 1.00 THF 11 1.9 6.2 Acetic Acid 3.3 0.52 1.65 Acrolein 0.72 0.12 2.6 10 minute high-dynamic-range surveys (segment bandwidths ~30 MHz) Signal to noise ratio ~2,000:1 (s ~ 2x10-4 mV) 6/24/2016
Quantifying Spectral Confusion Most molecules (especially as they get bigger) can occupy a lot of channels….this can impose a practical detection limit in trace scenarios – finding a weak component in a dense forest. Complexity metric: fraction of channels occupied as a function of dynamic range Ethanol Apodization window: Kaiser-Bessel (b = 8) Linewidths (FWHM): 75-110 GHz: 0.55 MHz 260-290 GHz: 1.2 MHz 520-580 GHz: 2.8 MHz 6/24/2016
Quantifying Spectral Confusion Most molecules (especially as they get bigger) can occupy a lot of channels….this can impose a practical detection limit in trace scenarios – finding a weak component in a dense forest. Complexity metric: fraction of channels occupied as a function of dynamic range N,N-Dimethylacetamide (Common diluent for analysis of solid samples by headspace) Apodization window: Kaiser-Bessel (b = 8) Linewidths (FWHM): 75-110 GHz: 0.55 MHz 260-290 GHz: 1.2 MHz 520-580 GHz: 2.8 MHz 6/24/2016
Cell Pressure Dependence Optimal Pressure for Nitrogen/Air-Based Mixtures 75-110 GHz, 4 us gate (‘high-res’) 75-110 GHz, 2 us gate (‘low-res’) 260-290 GHz, 1.8 us gate An advantage of FT spectroscopy with apodization windows – the lineshape is nearly pressure invariant at the peak sensitivity point 6/24/2016
Broadband Mixture Spectra ACN (off screen) ACN (off screen) 6/24/2016
Sensitivity – Broadband FTmmW Concentration detection limits in N2 (10 minute scan) Molecule 75-110 GHz 260-290 GHz Sensitivity Ratio Simulation Ratio (calc.) High-Res, 10 mTorr 30 mTorr Acetonitrile 14 ppm 0.7 ppm 19 31 Methanol 140 ppm 34 ppm 4.2 11 Ethanol 270 ppm 70 ppm 4.0 12 Dichloromethane 315 ppm 22 ppm 14 21 2-Propanol 900 ppm 160 ppm 5.4 Tetrahydrofuran 1800 ppm 175 ppm 10 9.8 *Higher power amplifiers available at W-band, potential for factor of ~2 improvement *Resolved hyperfine structure at W-band that collapses at 260-290 GHz 6/24/2016
Sensitivity – Targeted FTmmW For single lines, we can tailor pulse length to p/2 condition for each transition – measure nutation curves directly in the matrix Acetonitrile Methanol Sensitivity Results – Targeted Spectroscopy (1 minute per component) Molecule 75-110 GHz 260-290 GHz Ratio Low-Res, 20 mTorr 30 mTorr Acetonitrile 0.63 ppm 0.10 ppm 6.2 Methanol 4.0 ppm 2.1 ppm 1.9 Ethanol 19.3 ppm 4.4 ppm 4.4 Dichloromethane 14.4 ppm 1.6 ppm 9.1 2-Propanol 24 ppm 12 ppm 2.1 Tetrahydrofuran 43 ppm 18 ppm 2.4 Note: These molecules are polar enough that p/2 condition can be reached in a time < T2; so a more powerful source would not improve sensitivity. For molecules with dipole moments < 0.5 D it would improve sensitivity. 6/24/2016
Approximate LDL in N2, Targeted, 1 minute Higher Frequencies 520-580 GHz gives the best coverage for small molecules: Molecule Approximate LDL in N2, Targeted, 1 minute 75-110 GHz 260-290 GHz 520-580 GHz Other transitions H2O -- 40 ppb 22 GHz + other low-m transitions NH3 60 ppb 24 GHz H2S 520 ppb 168, 216 GHz NO 2 ppm 150, 250 GHz CO 1 ppm 115, 230 GHz PH3 200 ppb 340 ppb H2CO >1 ppm 14 ppb HCN ~200 ppb 6 ppb 10 ppb But we are power starved for almost all molecules: HCN (m = 2.98 D) PH3 (m = 0.57 D) 6/24/2016
Conclusions Optimal sensitivity for most molecules suitable for room temperature FTmmW spectroscopy is found between 200-300 GHz, but… Room-temperature FT-millimeter wave spectroscopy at lower frequencies can give competitive sensitivity performance for many molecules, and: -Better complex mixture performance (greater dynamic range before confusion) -Lower cost Working at higher frequencies can give access to smaller molecules while maintaining reasonable sensitivity to larger molecules – however, less suited to complex mixtures (broader linewidths, higher line density, harder to achieve p/2 or p excitation for advanced measurements) The best band depends on the application! 6/24/2016
Thank you! Millimeter-Wave FT-MRR Analyzer Microwave FT-MRR Analyzer 770 Harris St. #104b | Charlottesville, VA 22903 Ph: (434) 202-2391 http://www.brightspec.com/ Millimeter-Wave FT-MRR Analyzer (Fourier Transform-Molecular Rotational Resonance) Microwave FT-MRR Analyzer Discovery Series First Installation, U. Valladolid Pulsed Jet CP-FTMW Spectroscopy Enantiomeric Excess (by 3-Wave Mixing) BrightSpec ONE D. Patterson et al., Nature 2013, 497, 475-478 S. Lobsinger et al. J. Phys. Chem. Lett. 2015, 6, 196-200 Room Temperature CP-mmW 6/24/2016