New Techniques in Chirped Pulse FTMW Spectroscopy Illustrated

  • Slides: 33
Download presentation
New Techniques in Chirped Pulse FTMW Spectroscopy Illustrated by Measurements of the MW Spectrum

New Techniques in Chirped Pulse FTMW Spectroscopy Illustrated by Measurements of the MW Spectrum of Suprane (C 3 H 2 F 6 O) Richard D. Suenram, 1 Steven T. Shipman, 1 Brooks H. Pate, 1 and Gordon G. Brown 2 University of Virginia 2 Coker College 1

Overview • Chirped-Pulse Fourier Transform Microwave Spectroscopy (CP-FTMW) • 2 – 8 GHz and

Overview • Chirped-Pulse Fourier Transform Microwave Spectroscopy (CP-FTMW) • 2 – 8 GHz and 800 – 2200 MHz Configurations • MW – MW Double Resonance • Strategies to Reduce Measurement Time and Sample Consumption

Balle-Flygare Cavity Fourier Transform Microwave (FTMW) Spectrometer Advantages: • High sensitivity using cavity •

Balle-Flygare Cavity Fourier Transform Microwave (FTMW) Spectrometer Advantages: • High sensitivity using cavity • 4. 5: 1 S/N for 18 O 13 C 34 S* 0. 000094% (5000 shots, 1% in Ne) Disadvantages: • limited bandwidth • Slow scan speed!! 14 hours / 11 GHz • Difficult to obtain accurate intensities • Requires sub-m. W MW power for most molecules (m ~ 0. 1 D) T. J. Balle and W. H. Flygare, Rev. Sci. Instrum. 52, 33 (1981). *R. D. Suenram, J. U. Grabow, A. Zuban, and I. Leonov, Rev. Sci. Instrum. 70, 2127 (1999)

True Broadband Rotational Spectroscopy The Goal: • Acquire a multi-GHz spectrum with every valve

True Broadband Rotational Spectroscopy The Goal: • Acquire a multi-GHz spectrum with every valve shot to gain a multiplex advantage for use with other techniques Two issues to resolve: • Generating broad (in frequency), short (in time) pulses • Getting these pulses in and out of the spectrometer Arbitrary Waveform Generator MW Synthesizer Replacement (Image from www. testequity. com) Standard Gain Horn Mirror Replacement (Image from www. atmmicrowave. com)

Chirped Pulse Excitation (Linear Frequency Sweep) J. C. Mc. Gurk, T. G. Schmalz, and

Chirped Pulse Excitation (Linear Frequency Sweep) J. C. Mc. Gurk, T. G. Schmalz, and W. H. Flygare, “Fast passage in rotational spectroscopy: Theory and experiment”, J. Chem. Phys. 60, 4181 (1974). Chirped Pulse Instantaneous Frequency: Need: 11, 000 MHz/1 ms Synthesizer: 300 MHz/1 ms Use arbitrary waveform generator as the frequency source a = sweep rate

Suprane Rotational Constants and Dipole Moments Conformer II A (MHz) 2330. 14750(9) 2414. 7604(2)

Suprane Rotational Constants and Dipole Moments Conformer II A (MHz) 2330. 14750(9) 2414. 7604(2) B (MHz) 865. 37070(4) 752. 2603(2) C (MHz) 753. 18859(5) 722. 9995(2) m. A (D) 1. 476(1) 0. 7258* m. B (D) 0. 758(3) 1. 3953* m. C (D) 0. 235(1) 0. 8011* 1 – 0 cm-1 * B 3 LYP/6 -311 G(d, p) 2 – 408 cm-1

7. 5 – 18. 5 GHz CP-FTMW Spectrometer 1) AWG generates a chirped pulse

7. 5 – 18. 5 GHz CP-FTMW Spectrometer 1) AWG generates a chirped pulse that is upconverted to 7. 5 – 18. 5 GHz and amplified. 2) The pulse is broadcast into the vacuum chamber where it interacts with molecules in a pulsed jet. 3) The FID is amplified, mixed down, and finally digitized on a fast oscilloscope. For more information, see: Rev. Sci. Inst. 79, 053103 (2008).

Suprane: 7. 5 – 18. 5 GHz Region Overall agreement with SPCAT is good;

Suprane: 7. 5 – 18. 5 GHz Region Overall agreement with SPCAT is good; intensity falloff above 16 GHz is due to horn coupling efficiency and power droop in the TWTA.

2 – 8 GHz Direct Detect CP-FTMW Spectrometer Horns 2. 5 – 7. 5

2 – 8 GHz Direct Detect CP-FTMW Spectrometer Horns 2. 5 – 7. 5 GHz, 15 d. Bi gain Amplifier 300 W TWTA 4 W SSA Repetition rate 5 Hz (20 GS/s, 20 ms FID) 20 Gs/s Oscilloscope No mixers needed!

Suprane: 2. 5 – 7. 5 GHz Region Data scaled to match noise in

Suprane: 2. 5 – 7. 5 GHz Region Data scaled to match noise in overlap region – intensities are still excellent! For optimal polarization conditions, inherent sensitivity is the same as 7. 5 – 18. 5.

Extension to Very Low Frequency 0. 8 – 2. 0 GHz CP-FTMW 1618. 56

Extension to Very Low Frequency 0. 8 – 2. 0 GHz CP-FTMW 1618. 56 MHz 101 - 000 * GSM Cell Phone Bands Amplifier: 10 W, 800 – 2200 MHz from Mini-Circuits. Note: switches can survive a 10 W, 16 ms pulse when cooled!

Extension to Very Low Frequency 0. 8 – 2. 0 GHz CP-FTMW 849. 42

Extension to Very Low Frequency 0. 8 – 2. 0 GHz CP-FTMW 849. 42 MHz 515 - 505 * GSM Cell Phone Bands

Suprane from 800 MHz to 18. 5 GHz

Suprane from 800 MHz to 18. 5 GHz

MW-MW Double Resonance Scheme 505 404 303 20 Gs/s Oscilloscope 202 515 414 313

MW-MW Double Resonance Scheme 505 404 303 20 Gs/s Oscilloscope 202 515 414 313 212 413

MW-MW Double Resonance First (chirped) pulse polarizes rotational transitions over a large bandwidth. 1

MW-MW Double Resonance First (chirped) pulse polarizes rotational transitions over a large bandwidth. 1 A second narrowband pulse pumps a single transition. This destroys coherences with connected levels, giving intensity modulations in the detected FID. 2 1 2

Suprane Double Resonance Pump pulse is tuned to 404 – 303 at 6413. 32

Suprane Double Resonance Pump pulse is tuned to 404 – 303 at 6413. 32 MHz. Every molecular transition with >75% modulation involves either the 404 or the 303 level. Only false positive is due to interference from UVa wireless network.

Multiple Nozzles: Linear Scaling Signals increase linearly with # of molecules (field detection). Noise

Multiple Nozzles: Linear Scaling Signals increase linearly with # of molecules (field detection). Noise is set by amplifier. To reach a given sensitivity, N nozzles give N 2 reduction in measurement time, and N reduction in sample consumption!

Multiple Nozzles: S/N Equivalency Nozzles should be separated by at least 8”. Different nozzles

Multiple Nozzles: S/N Equivalency Nozzles should be separated by at least 8”. Different nozzles can have a surprising degree of variability.

Multiple FIDs per Valve Pulse Valve pulses last for ~600 ms, but FID only

Multiple FIDs per Valve Pulse Valve pulses last for ~600 ms, but FID only lasts for 40 ms. Can potentially save a factor of 30 in sample consumption and 90 in time!

Multiple FIDs per Valve Pulse Some decrease in intensity by FID 10, but would

Multiple FIDs per Valve Pulse Some decrease in intensity by FID 10, but would probably still see gains up to 15. Key reasons this works: • Large open volume with perpendicular nozzle orientation • Polarizing pulse is in the weak pulse limit Currently, software limits preclude averaging in this mode, but this will hopefully change in the near future.

Deep Averaging One of the key advantages to coherent detection is the lack of

Deep Averaging One of the key advantages to coherent detection is the lack of any apparent noise floor. The exact phase of each pulse is extremely reproducible. We have co-added data spanning days with no loss of signal. Extracting assignable spectra becomes increasingly difficult, particularly as the double 13 C threshold is reached.

106 shot spectrum of suprane Have currently assigned – all 13 C species, 18

106 shot spectrum of suprane Have currently assigned – all 13 C species, 18 O species of Conf I, NS of Conf II, and 20 Ne- and 22 Ne-Conf I clusters. S/N of 10, 000: 1 is a “complexity limit” beyond which life is difficult. (Here: double 13 C on Conf I, 13 C on Ne clusters, 13 C on Conf II, and more!)

106 shot spectrum of 1 -hexanal x 1000 To date 10 conformers have been

106 shot spectrum of 1 -hexanal x 1000 To date 10 conformers have been assigned, along with 22 13 C species.

106 shot spectrum of 1 -heptene x 150 To date 12 conformers have been

106 shot spectrum of 1 -heptene x 150 To date 12 conformers have been assigned, along with 14 13 C species.

Residuals from 106 shot measurements (x 50 from 1 M) (x 12 from 1

Residuals from 106 shot measurements (x 50 from 1 M) (x 12 from 1 M) Sample impurities are almost certainly a problem at this level. Definitely some trace Cl-containing species in the 1 -heptene. There’s still work to be done!

Summary of Total Gains Assume S/N for normal species is 3: 1 on a

Summary of Total Gains Assume S/N for normal species is 3: 1 on a single valve pulse # of nozzles 13 C 3: 1 (cycles) (time) 13 C 3: 1 (sample) 1 10, 000 16. 7 min 10, 000 shots 3 1111 1. 9 min 3, 333 shots 3 (10 FIDs) 111 11 sec 333 shots Cavity FTMW Comparison: 14 hours for 11 GHz coverage at the same sensitivity* 220, 000 valve shots required *R. D. Suenram, J. U. Grabow, A. Zuban, and I. Leonov, Rev. Sci. Instrum. 70, 2127 (1999)

Collaborators Steve Cooke (University of North Texas) Brian Dian (Purdue) Kevin Douglass (NIST) Scott

Collaborators Steve Cooke (University of North Texas) Brian Dian (Purdue) Kevin Douglass (NIST) Scott Geyer (MIT) Lu Kang (Union College, KY) Zbigniew Kisiel (Institute of Physics, Polish Academy of Sciences) Alberto Lesarri (University of Valladolid, Spain) Helen Leung and Mark Marshall (Amherst) John Muenter (Rochester) David Pratt (Pittsburgh) Nick Walker (University of Bristol, UK) Li-Hong Xu (University of New Brunswick, Canada)

Acknowledgements The Pate Lab Leonardo Alvarez-Valtierra Matt Muckle Justin Neill Sara Samiphak Funding NSF

Acknowledgements The Pate Lab Leonardo Alvarez-Valtierra Matt Muckle Justin Neill Sara Samiphak Funding NSF Chemistry CHE-0616660 NSF CRIF: ID CHE-0618755 Special Thanks: Tom Fortier and Tektronix

Suprane PES and Conformers 5 – 1495 cm-1 B 3 LYP/6 -311 G(d, p)

Suprane PES and Conformers 5 – 1495 cm-1 B 3 LYP/6 -311 G(d, p) 1 – 0 cm-1 2 – 408 cm-1 4 – 883 cm-1 3 – 816 cm-1

Accurate Relative Intensities

Accurate Relative Intensities

6. 5 – 12. 2 GHz CP-FTMW Spectrometer Mixers removed from back-end setup to

6. 5 – 12. 2 GHz CP-FTMW Spectrometer Mixers removed from back-end setup to avoid problems with “ghost peaks”.

Suprane Residuals from 40 k shot measurement

Suprane Residuals from 40 k shot measurement