High-Resolution UV Spectroscopy

Introduction

Samples are heated to 300 K to 700 K and expanded through a 125 µm nozzle. Four cm downstream of the nozzle, the molecular beam is skimmed before passing into a second differentially-pumped chamber. One meter downstream, the beam is orthogonally crossed with the UV laser beam. The UV fluorescence is spatially filtered and collected at the crossing point with 20 % efficiency using two spherical mirrors, and fluorescence photons are detected on a cooled photomultiplier tube interfaced to a photon counting system. Background counts are < 500 s^-1 mW^-1. Additional hardware is included to aid in the analysis of complex spectra. A microwave guide is inserted near the beam crossing region for high-precision (< 100 kHz) rotational excitation in excited electronic states and a liquid-He-cooled bolometer monitors changes in the molecular-beam energy from absorption of UV and/or microwave radiation in metastable states.

A unique calibration system is used with a precision of 200 kHz at 1000 THz (2 parts in 10¹⁰). The main components include a frequency-stabilized HeNe laser, an actively stabilized confocal reference cavity, and an acoustic-optic modulator (AOM) for generation of a tunable sideband on a small portion (< 1 mW) of the dye-laser fundamental. The frequency-stabilized HeNe laser has a day-to-day frequency drift rate of < 200 kHz based on comparisons with sub-Doppler spectra of 1-fluoronaphthalene (FWHM = 3 MHz). The control system consists of a lock-in amplifier and PID servo electronics to control the dye laser so that the sideband remains fixed on a peak maximum of the reference cavity at all times. The double-pass configuration of the AOM permits tunability over a full free spectral range of the reference cavity enabling continuous scans over any desired frequency interval with a digital step resolution of 60 kHz.

The computer interface controls the laser system. In addition to acquiring the fluorescence photon count and the UV power, the reference cavity transmission signals, the PID dye laser error signal, and the AOM ramp voltages are simultaneously digitized at rates from 500 Hz to 5 kHz. These high digitization rates aid in the removal of low frequency noise from mechanical pump vibrations and acoustic disturbances. This is accomplished by "binning" photons relative to both the digitally recorded and calibrated ramp and lock-in error signals, with the natural consequence of having fully linearized scans. The transmission fringe voltage provides a real-time diagnostic of the dye laser performance. Typically, 1.2 cm^-1 single-mode scans are done in < 20 min and are overlapped at integral fringe spacings after linearization.

To test the precision of the instrument, the rotationally resolved fluorescence-excitation spectrum of the S₁ S₀ electronic origin of 2-chloronaphthalene has been acquired. More than 1000 well-resolved rotational lines are recorded near the band origins of the ³⁵Cl and ³⁷Cl isotopic forms over a 3.5 cm^-1 spectral region at a rotational temperature of 10 K. At the finer level of detail shown in the lower panels, additional level structure is observed, particularly for low J transitions near the band origin. The highly asymmetric line shapes are due to the nuclear quadrupole hyperfine interaction splitting each rotational line. From a least-squares fit of the line profiles at the full measurement precision of 200 kHz, the nuclear quadrupole coupling constants, eQq_zz and are determined for the first time in the electronically excited state of an aromatic molecule. The lower traces illustrate the fitted lineshapes and the residuals from the analysis.

To assess the calibration errors of the spectrometer, the differences between all transitions having a common upper state are generated. These values are compared with the "exact" ground-state level differences calculated from microwave data and plotted as a function of the UV energy difference. A non-zero slope indicates a systematic error in the relative calibration. The combination-difference analysis of the "corrected" UV data is shown at the above. From the lack of curvature in this plot, an estimate of the thermal drift rate is < 400 kHz/hour. The scatter in the data is < 1 MHz or twice the 500 kHz digital resolution of the data. The overall standard deviation is < 400 kHz which is less than 10 % of the linewidth and reflects a measurement precision of 4 parts in 10¹⁰ at 318 nm.

A computer program has been written for the graphical analysis of complex spectra. Three different levels of horizontal-scale expansion are possible to permit easy and quick access to various parts of the spectrum at full experimental resolution. It is also possible to display simultaneously two simulated spectra for visual evaluation of the validity of theoretical models. Each simulated line may be convoluted with a Lorentzian and/or Gaussian lineshape function to adequately match the experimental data. Quantum number labels for one or more transitions are accessed via the mouse with single-step capacities to assign experimental frequencies for refinement of model parameters via least-squares. Approximate methods are also implemented to rapidly simulate spectra after changes in the model parameters. Track-bars are included for variation of rotational constants, quadrupole coupling constants and transition moment components.

Access to parameters for non-rigid rotor models and nuclear quadrupole coupling are provided in dialog format. These dialogs are front ends to several console-based programs for calculating theoretical spectra for 1-state (microwave) and 2-state (vibrational and electronic) systems. Least-squares analyses of assignment files by variation of any or all parameters provide for assessment of the significance of theoretical models by means of the parameter uncertainties and the observed-minus-calculated standard deviation of the assigned line set. Together with the approximate methods, the user can rapidly explore parameter space to test theoretical models and assign complex molecular spectra. Methods are now under development to combine the power of these two techniques.

• "High Resolution Studies of Tropolone in the S₀ and S₁ Electronic States: Isotope Driven Dynamics in the Zero-Point Energy Levels,"
  Keske, J.C., Blake, T.A., Lin, W., Pringle, W.C., Novick S.E., and Plusquellic, D.F.,
  J. Chem. Phys. (in press, 2006).



• "High Resolution Spectroscopic Studies of 1-(1-naphthyl)ethylamine in S₀ and S₁: Exploring the Dependence of Circular Dichroism on Conformational Structure,
  Plusquellic, D.F., Lavrich, R.J., Petralli-Mallow, T., Davis, S.R., Korter, T.M., and Suenram, R.D.,
  Chem. Phys. 283, 355 (2002).



• "Probing Nuclear Quadrupole Interactions in the Rotationally Resolved S₁ S₀ Electronic Spectrum of 2-Chloronaphthalene,"
  Plusquellic, D.F., Davis, S.R., and Jahanmir, F.,
  J. Chem. Phys. 115, 225 (2001).



• "High resolution optical spectroscopy in the UV,"
  Majewski, W.A., Pfanstiel, J.F., Plusquellic, D.F., and Pratt, D.W.,
  in Laser techniques in chemistry. XXIII, 101-148 (1995).



• "Acid-base chemistry in the gas phase. The cis- and trans-2-naphthol-NH₃ complexes in their S₀ and S₁ states,"
  Plusquellic, D.F., Tan, X.Q., and Pratt, D.W.,
  J. Chem. Phys. 96, 8026-8036 (1992).



• "Exploiting quantum interference effects for the determination of the absolute orientation of an electronic transition moment vector in an isolated molecule,"
  Plusquellic, D.F. and Pratt, D.W.,
  J. Chem. Phys. 97, 8970-8976 (1992).



• "Methyl group torsional dynamics from rotationally resolved electronic spectra. 1- and 2-methylnaphthalene,"
  Tan, X.Q., Majewski, W.A., Plusquellic, D.F., and Pratt, D.W.,
  J. Chem. Phys. 94, 7721-7733 (1991).
  

OTD Home Page   |   Technical Inquiries   |   Site Comments
Online: March 1999   -   Last updated: January 2006