High resolution cavity ringdown spectroscopy of jet-cooled deuterated methyl peroxy (CD 3 O 2 ) in the near IR Shenghai Wu, Patrick Rupper, Patrick Dupré and Terry A. Miller Laser Spectroscopy Facility, Department of Chemistry, The Ohio State University, Columbus, OH 43210

Alkyl peroxy radicals play a key role as intermediates in oxidation of hydrocarbons (atmospheric as well as combustion chemistry) Methyl peroxy is simplest alkyl peroxy radical → starting point for spectroscopic characterization Ambient cell cavity ring-down spectroscopy (CRDS) Several peroxy radicals have been studied in our lab → near IR A – X electronic transition is sensitive, species-specific diagnostic Rotational structure is only partially resolved (congestion due to different rotational lines and different conformers) High resolution, rotationally resolved IR CRDS of methyl peroxy under jet-cooled conditions would be of great value Jet-cooled Peroxy Radicals (CD 3 O 2 ) Motivations ~~

Strong, but non-selective electronic transition in the UV (B 2 A’’ – X 2 A’’) Weak, very selective transition in the near IR (A 2 A’ – X 2 A’’) Room temperature CRDS (pulsed 1 and cw 2 ) spectrum at moderate resolution (photolysis of acetone) → overlap of several rotational transitions (congestion), no spin-rotational structure resolvable Negative-Ion PES 3 (instrument resolution 40 cm -1 ) Recently ionization detection techniques: TOFMS with moderate resolution laser (supersonic jet expansion 4 and effusive beam 5 ) → parent cation only stable for CH 3 O 2 Methyl Peroxy: Spectroscopic Background 1 Miller group and Y. P. Lee group (JCP 112 (2000) 10695, JCP (2007)) 2 D. B. Atkinson, J. L. Spillman, JPCA 106 (2002) G. B. Ellison, M. Okumura and coworkers, JACS 123 (2001) Bernstein group (H. B. Fu, Y. J. Hu, and E. R. Bernstein, JCP 125 (2006) ) 5 G. Meloni, P. Zou, S. J. Klippenstein, M. Ahmed, S. R. Leone, C. A. Taatjes, and D. L. Osborn, J. Am. Chem. Soc. 128 (2006) ~~ ~~

Ti:Sa ring cw laser Ti:Sa Amplifier (2 crystals) Nd:YAG pulse laser Raman Cell PD InGaAs Detector Ring-down cavity with slit-jet (absorption length ℓ = 5 cm) L = 67 cm Vacuum Pump 1 m single pass, 13 atm H nm,  ~ 1 MHz mJ  ~ MHz (FT limited) ℓ Nd:YAG cw laser 1 st Stokes, ~ 1.3  m (NIR), ~ 2 mJ  SRS ~ 220 MHz (limited by power and pressure broadening in H 2 ) Δ Doppler (slit jet) ~ 250 MHz R ~ – 1.3  m SRS (stimulated Raman scattering) 20 Hz, ns, 350 mJ slit-jet: longer absorption path-length less divergence of molecular density in the optical cavity S. Wu, P. Dupré and T. A. Miller, Phys. Chem. Chem. Phys. 8 (2006) 1682 P. Dupré and T. A. Miller, Rev. Sci. Instrum. 78 (2007) Experimental Setup

IR Beam 9 mm -HV radical densities of molecules/cm 3 (10 mm downstream, probed) rotational temperature of K plasma voltage ~ 500 V, I  1 A (~ 400 mA typical), 100 µs length dc and/or rf discharge, discharge localized between electrode plates, increased signal compared to longitudinal geometry Previous similar slit-jet designs: D.J. Nesbitt group, Chem. Phys. Lett. 258 (1996) 207 R.J. Saykally group, Rev. Sci. Instrum. 67 (1996) 410 Pulsed Supersonic Slit-jet and Discharge Expansion 5 cm 5 mm 10 mm Electrode carrier gas (300 – 700 Torr Ne) + precursor CD 3 I (1%) and O 2 (10%) Viton Poppet 00 l eq α min Δ(αl)√ΔtΔ(αl)√Δt (ringdown time) (absorption equivalent length) (experimental sensitivity (standard deviation in one pass absorbance)) (noise equivalent absorption)  300 µs  6 km 0.02 ppm 4.5 ppb Hz -1/2 Main reactions in our discharge environment CD 3 I → CD 3 + I (discharge) CD 3 + O 2 + [M] → CD 3 O 2 + [M] CD 3 + CD 3 → C 2 D 6 O 2 → 2 O (discharge) CD 3 + O → D 2 CO + D CD 3 + CD 3 O 2 → 2 CD 3 O 2 CD 3 O 2 → products

Hamiltonian Unpaired electron → coupling of electron spin with molecular rotation Hund’s case b (absence of external magnetic fields, ignoring hyperfine interaction) Structure due to internal rotation (tunneling) not resolvable for CD 3 O 2 → consider only overall rotation (asymmetric rigid rotor) and spin-rotation N = rotational angular momentum S = electron spin angular momentum A,B,C = rotational constants in the principal axes of inertia (a,b,c)  = reduced spin-rotation tensorial components (C s symmetry → 4 independent components) ~ ~

Spectra of CD 3 O 2 (RT and Jet-cooled) room temperature jet-cooled C s symmetry → pure c-type transition moment close to a prolate symmetric top ( ΔKa ΔJ Ka” ) [CD 3 O 2 ] = 3.7(3.1) x cm -3 → interplay between RT (cross-section) and jet-cooled investigation

Jet-cooled CRDS Spectrum of CD 3 O 2 10 % O 2 and ~ 1% CD 3 I in Ne dc discharge: 350 mA stepsize: 50 MHz, RD time average: 4 A 2 A’ ← X 2 A”, vibrationless band ~~ r Q0Q0 p Q1Q1 p Q2Q2 p Q3Q3 r Q1Q1 r Q2Q2 spread out over ~ 30 cm -1 > 1000 lines, 350 of which due to single transition small background (0.2 ppm) → due to precursor molecule

Jet-cooled CRDS Spectrum of CD 3 O 2 A 2 A’ ← X 2 A” simulation 1 using 15 fitted parameters T = 15.5 K Voigt profile (Lorentzian 330 MHz, Gaussian 390 MHz) vast majority could be simulated ( r P 1 enhanced at cm -1 ) spin-rotation doublet well resolved opposite extension of J progression (B,C rotational constants) 1 SpecView simulation package, V.L.Stakhursky, T.A.Miller, 56th Symposium, 2001 pQ1pQ1 rQ0rQ0 J”= J”= ~ ~ J”=N”+1/2 J”=N”-1/2

Jet-cooled CRDS Spectrum of CD 3 O 2 A 2 A’ ← X 2 A”, p P 1 band J’’=1.5 ~ ~ two other branches in this region (not labelled) p Q 2 around 7370 cm -1 r P 0 around cm -1 J”=N”-1/2 J”=N”+1/2

Assignment of the CRDS Spectrum wave numbers / cm -1 ppm –  K = 1 –  K = -1 |K”| = 0 |K”| = 1 |K”| = 2 |K”| = 3 |K”| = 4 Experiment Simulations at T = 15 K Q Q P P R R

Spectroscopic Constants of CD 3 O 2 cm -1 XA A B C ~~ 1 M. B. Pushkarsky, S. J. Zalyubovsky, T. A. Miller, JCP 112 (2000) (15)T (15) (15)ε cc (10)1.2932(10)A (11) (11)B (11) (11)C (15) (15)ε aa (14) (14)ε bb (22)0.0138(22)|ε ba +ε ab |/2 AX CD 3 O 2 cm -1 ~~ ~ ~ ~ ~~ 350 lines have been used in the fit (with N up to 10 and K up to 4) Standard deviation of the fit is cm -1 T 00 = (15) cm -1 is consistent with the value determined from room temperature 1 spectra, i.e., T 00 = (8) cm -1 Rotational constants in remarkable good agreement with those derived from room temperature spectra and also from ab initio (< 6 %) → benchmark calculations |ε” aa |  |ε’ aa | only change in sign, ε bb, ε cc are small, in agreement with c-type transitions ε << A,B,C → spectrum dominated by rotational structure 350 lines have been used in the fit (with N up to 10 and K up to 4) Standard deviation of the fit is cm -1 T 00 = (15) cm -1 is consistent with the value determined from room temperature 1 spectra, i.e., T 00 = (8) cm -1 Rotational constants in remarkable good agreement with those derived from room temperature spectra and also from ab initio (< 6 %) → benchmark calculations |ε” aa |  |ε’ aa | only change in sign, ε bb, ε cc are small, in agreement with c-type transitions ε << A,B,C → spectrum dominated by rotational structure ~~ ~ ~~

Comparison between CD 3 O 2 and CH 3 O 2 See Patrick Dupré’s talk (RF10)

Conclusions Studied A – X vibrationless transition (weak) of deuterated methyl peroxy radical (CD 3 O 2 ) via pulsed CRDS in the near IR (1.36 µm) Observed the spectrum under jet-cooled conditions (T rot ~ 15 K) by combining a narrow-bandwidth laser source (~ 220 MHz) with a supersonic slit-jet expansion and electric discharge (dc or rf) Rotational and spin-rotational structure have been resolved in the spectra and corresponding spectroscopic constants (15 for ground and excited state) were determined ~ ~

Acknowledgments Prof. Terry A. Miller Shenghai Wu, Patrick Dupré Gabriel Just and Prof. Anne B. McCoy Miller group members Machine shop: Jerry Hoff, Larry Antal, Joshua Shannon NSF for funding

Peroxy Transition a // a/a/ yy  xx eV O2O2 CH 3 O 2 CH 3 O 2 : a 1  g - X 3  g  A2A/A2A/ ~ X 2 A // ~ CH 3 O 2 : RO 2 - R perturbation HOMO (non bonding on O atoms, in plane) SOMO (antibonding (  *), out of plane)

Cavity Ringdown Absorption Spectroscopy R L A = Nσl A = L/c τ absorber - L/c τ 0 Time Intensity 00  absorber τ absorber lN σ + = cL)/( R1 - ( ) 0 τ c L )/ ( R1 - = CRDS: - absorption technique with good sensitivity - immune to the noise caused by source fluctuations - absolute determination of the absorption cross-section l

Direct Absorption Measurement t (  s) Transmission signal  0 = ring-down time for empty cavity  ( ) = ring-down time in presence of absorbing medium  = absorption,  min = minimum detectable absorption L = length of cavity, l = medium absorption length R = mirror reflectivity, c = speed of light

Non-exponential Decay  laser > =  Doppler The beginning of the decay reflects the medium absorption The end of the decay reflects the empty cavity absorption Ringdown time depends on angular frequencies of the incoming EM field The non-linear response of the absorption medium The absorption is saturated at the very beginning of the decay The later part of the decay is approximated by the linear absorption The chemical or physical dynamics faster than   Multi-exponential decay To analyze the decay as a function of time

Number Density Observed number density: Minimum detectable number density: Abs I observed rotational (or vibrational) integrated absorption per pass, i.e. the integrated area of the spectral line (or spectral band)   integrated absorption cross-section, (  A ref ( ) d / A ref ( 0 )) *  peak example CH 3  I = 2.1 x cm/molecule l = 5 cm Abs I = 13 x cm -1 → N obs = 1.2 x cm -3

Rotation-vibration Hamiltonian H R : rotational Hamiltonian (asymmetric top, rigid rotor) H RT : spin-rotation Hamiltonian z, a  p  = internal rotation angular momentum  = internal rotation angle F = internal rotational constant A,B,C = rotational constants N = rotational angular momentum  = axis of internal motion Structure due to internal rotation not resolvable for CD 3 O 2 → consider only overall rotation (asymmetric rigid rotor) and spin-rotation Nuclear spin statistics: 11:16 for A and E levels

CH 3 O 2 : Internal Rotation 5E 4E 4A 1 3A 2 0E, 0A 1 1E, 1A 2 barrier to internal rotation around the C-O bond is substantially higher in the A than the X state (~1100 cm -1 vs. ~300 cm -1 ) small methyl torsional frequencies in the ground state (132 cm -1 for CH 3 O 2 and 107 cm -1 for CD 3 O 2 ): location of vibrational levels lead to typical sequence and also some atypical transitions ~ ~

IR Range Coverage from Ti:Sa

Ti:Sa ring cw laser Ti:Sa Amplifier (2 crystals) Nd:YAG pulse laser Raman Cell (H 2 ) PD InGaAs or InSb Detector Ring-down cavity with slit-jet (absorption length ℓ = 5 cm) L = 67 cm Vacuum Pump 1 m single pass nm,  ~ 1 MHz mJ  ~ MHz ℓ Nd:YAG cw laser 1 st Stokes, ~ 1.3  m (NIR), ~ 2 mJ 2 nd Stokes, ~ 3  m (MIR), ~ 200  J  SRS : ~ 180 NIR ~ 220 MIR (limited by power and pressure broadening in H 2 ) R ~ 1.3  m ~ 3  m SRS (stimulated Raman scattering) ns, 350 mJ slit-jet: longer absorption path-length less divergence of molecular density in the optical cavity  Doppler (slit jet) ~ 100 MIR ~ 200 NIR D. Anderson, S. Davis, T. Zwier and D. Nesbitt, Chem. Phys. Lett. 258, 207 (1996) S. Wu, P. Dupré and T. A. Miller, Phys. Chem. Chem. Phys. 8 (2006) 1682 P. Dupré and T. A. Miller, Rev. Sci. Instrum. 78 (2007) Experimental Setup

High-Resolution Ti:Sa Laser System quadrupole amplification (in a seeded cavity)

Longitudinal Discharge - HV Design adapted from: D.J. Nesbitt group Chem. Phys. Lett. 258 (1996) 207 R.J. Saykally group Rev. Sci. Instrum. 67 (1996) 410 Discharge Expansions Transversal Discharge

Characteristics of CRDS setup

A 2 A' - X 2 A" Transition of CH 3 O 2 and CD 3 O 2 ~ ~

Jet-cooled CRDS Spectrum of CD 3 O 2 A 2 A’ ← X 2 A”, r R 0 band ~ ~ 3.5 J”=2.5 other branch in this region (not labelled) → r R 1 J”=N”-1/2 J”=N”+1/2

C.-Y. Chung et al., JCP, accepted, 2007 COO bend O-O stretch N v’’ v’ origin methyl torsion A 2 A' - X 2 A" Transition of CH 3 O 2 Under Ambient Conditions (RT) ~ ~

Rotational Contour of the Origin Band of CH 3 O 2 (RT) overlap of several rotational transitions - no spin-rotational structure resolvable - high-resolution spectra of methyl peroxy under jet-cooled conditions would be of great value

Rotational Temperature we can influence the rotational temperature in the jet by precursor seeding concentration we can vary T rot by a factor of two (15 K to 28 K) “over-seeding” increases density number, which is correlated to a background level increase T rot ~ 28 K T rot ~ 15 K

Rotational Temperature 350 lines have been used in the fit (with N up to 10 and K up to 4) Standard deviation of the fit is cm -1 T 0 = (15) cm -1 is consistent with the value determined from room temperature 1 spectra, i.e., T 0 = (8) cm -1 Rotational constants in remarkable good agreement with those derived from room temperature spectra Linewidth (Voigt profile) 330 MHz Lorentzian → finite lifetime ~ 1.5 ns of electronic transition 390 MHz Gaussian → Doppler (~ 300 MHz) plus source (~ 250 MHz) linewidth 350 lines have been used in the fit (with N up to 10 and K up to 4) Standard deviation of the fit is cm -1 T 0 = (15) cm -1 is consistent with the value determined from room temperature 1 spectra, i.e., T 0 = (8) cm -1 Rotational constants in remarkable good agreement with those derived from room temperature spectra Linewidth (Voigt profile) 330 MHz Lorentzian → finite lifetime ~ 1.5 ns of electronic transition 390 MHz Gaussian → Doppler (~ 300 MHz) plus source (~ 250 MHz) linewidth

Kinetics - detection capability of produced molecules in the MIR (CH 3, C 2 H 6, H 2 CO) and NIR (CH 3 O 2 ) - using literature absorption cross-sections and experimental absorbances → molecular densities - using literature reaction rate constants for kinetics studies Reaction time / µs Density number / cm -3 [CD 3 O 2 ] = 3.7(3.1) x cm -3 (probe region) CH 3 C2H6C2H6 C2H6C2H6 H 2 CO CH 3 O 2 no oxygen present with oxygen present