Acquisition, Computation, and Assignment of
the S1 Acrolein Vibronic Spectrum

Kevin W. Paulisse, Tyson O. Friday, Margaret L. Graske, and William F. Polik

Hope College, Holland, MI 49423

http://www.chem.hope.edu/~polik/poster/acro98.htm

Abstract

Acrolein (CH2CHCHO) is a prototypical conjugated organic molecule with interesting spectroscopic properties. The S1 vibronic spectrum of acrolein was obtained in a molecular beam using Cavity Ring Down spectroscopy. Four techniques are used to assign the experimental spectrum. The Gaussian 94 software program is used to compute the vibrational spectrum of S1 acrolein to assist in assignment of normal modes, with the results being scaled to a similar calculation performed on formaldehyde. Franck-Condon progressions are observed for several normal modes, most significantly for the out-of-plane skeletal torsion vibration (n18). An analysis of cold bands versus hot bands using spectra collected at different temperatures allows assignment of transitions originating from excited vibrational levels in the ground state. Parallel and perpendicular band structures, which arise from out-of-plane vibrations versus in-plane vibrations, are well-defined and observed. The current set of assignments increases the number of known assignments nearly three-fold and corrects several misassignments in previous literature.

Introduction

Acrolein is a prototypical conjugated bond system which absorbs in the near ultraviolet.

Since acrolein does not fluoresce, Cavity Ringdown (CRD) spectroscopy is used to measure the absorption spectrum of acrolein in a molecular beam.

The S1 vibronic spectrum consists of transitions from vibrational levels of the ground electronic state (S0) to vibrational levels of the first excited singlet electronic state (S1).

The spectrum is assigned by predicting the fundamental transitions using computational chemistry software, analyzing Franck-Condon progressions, comparing spectra taken at different temperatures, and observing rotational band structure.

Cavity Ringdown (CRD) Spectroscopy

A pulse of laser light is directed into an optical cavity which has highly reflective mirrors on each end. The light bounces back and forth through the sample thousands of times. The small intensity of light that escapes through the second mirror is monitored by a photomultiplier tube (PMT). The ringdown time of the cavity is a measure of absorbance by sample contained between the end mirrors. The large optical pathlength results in a highly sensitive absorption technique.

Experimental Setup

A Nd:YAG laser emits 532 nm light pumping a dye laser, which is tuned over a wavelength range of 680-780nm. The red light passes through a doubling crystal which converts it into ultraviolet (UV) light. The beam translation introduced by the angle-tuned doubling crystal is counteracted with a quartz compensating block. The mixture of UV and residual red light is separated by a pair of 60° prisms, and a second pair of prisms compensates for the beam translation. Some of the red light is directed into a hollow cathode lamp for calibration. The pure UV is directed through a spatial filter, which consists of a focusing lens, pinhole, and collimating lens. The high quality light is sent to the molecular beam chamber where it interacts with the sample. The signal detected by the PMT is averaged by the oscilloscope and sent to the computer, where the lifetime is fit to an exponential decay to monitor absorbance.

Results

The room temperature spectrum of acrolein is very congested with overlapping peaks on a steeply sloped baseline. The flatter baseline and improved resolution in the molecular beam spectrum is an indication that the acrolein has been cooled, simplifying the spectrum.

Computation of Acrolein Spectrum

The Gaussian 94 program1 was used to predict the frequencies of the S1 normal modes.

Gaussian 94 was first used to compute the S1 normal mode frequencies of formaldehyde (H2CO). Because the systems are similar, frequency scaling factors were calculated by comparing the H2CO results to experimental values from the literature.

Gaussian 94 was then used to compute the S1 vibrational spectrum of acrolein using the RCIS method with the 6-311+G(3DF,2PD) basis set.

Mode Description Calculated Frequency
(cm-1)		 Observed Frequency
(cm-1)
n1 Asym vinyl CH stretch

3097

n2 Symm vinyl CH2 stretch

3017

n3 Vinyl CH stretch

3048

n4 Formyl CH stretch

2866

n5 C=O stretch

1177

1262

n6 C=C stretch

1498

1406
n7 CH2 bend

1316

n8 Formyl CH rock

1180

1132
n9 Vinyl CH rock

1316

1279
n10 Vinyl CH2 rock

867

647
n11 C-C stretch

1050

n12 CCO bend

468

483
n13 CCC bend

301

302
n14 Vinyl C=C torsion

613

n15 Formyl CH wag

393

n16 Vinyl CH2 wag

889

916
n17 Vinyl CH wag

985

n18 Skeletal torsion

148

252

1 Frisch, M., et al. Gaussian 94, Rev. E.2; Gaussian, Inc.: Pittsburgh, PA, 1995.

Franck-Condon Progressions

Franck-Condon progressions are series of peaks resulting from changes of a vibrational quantum number in a normal mode. The n18 mode shows a number of Franck-Condon progressions in the room temperature spectrum.

Temperature Effects

The Boltzmann factor predicts that only the low energy vibrational levels are populated at low temperatures. Consequently, if a band appears in the room temperature spectrum (295 K) but disappears in the molecular beam spectrum (70K), the corresponding transition must have originated from an excited vibrational energy level.

Transitions originating from the ground vibrational level are called "Cold Bands," while transitions originating from excited vibrational levels are called "Hot Bands."

Rotational Band Structure

"Parallel Bands" occur when the change in out-of-plane vibrational quanta (n14-n18) is odd. "Perpendicular Bands" occur when the change in out-of-plane vibrational quanta is even.

Parallel bands appear in the spectrum as narrow bands with a sharp high-energy edge, and perpendicular bands appear as broader bands. The narrow parallel bands arise from the D K=0 rotational selection rule (Q-branch), and the broad perpendicular bands arise from the D K=± 1 selection rule (P,R-branches).

Assignments

The following are the assignments for the acrolein S1 vibronic spectrum. Calculated combination band frequencies are based on an anharmonic oscillator model.

Indicated are the molecular beam S1 energy (MB), room temperature S1 energy (RT), rotational band symmetry (Sym), calculated position (Calc), difference between calculated and observed position (Diff), and the present assignment. The symmetry labels suggested by Brand and Williamson2 are used (a=parallel band, c=perpendicular band).

MB

(cm-1)

RT

(cm-1)

Sym

Calc

(cm-1)

Diff

(cm-1)

Assignment

-573

c

-564

-9

12(0,1)

-328

a

-313

-15

12(0,1) 18(1,0)

-154

a

-160

6

18(0,1)

0

0

c

0

0

Origin

94

c

91

3

18(1,1)

190

c

188

2

18(2,2)

252

250

a

251

0

18(1,0)

297

300

c

297

0

13(1,0)

334

334

a

331

3

18(2,1)

416

a

416

0

18(3,2)

487

481

c

487

0

12(1,0)

487

481

c

491

-4

18(2,0)

514

c

514

0

X(1,0)

564

c

576

-12

12(1,0) 18(1,1)

647

648

c

647

0

10(1,0)

719

720

a

719

0

18(3,0)

732

733

a

736

-4

12(1,0) 18(1,0)

764

764

a

765

-1

X(1,0) 18(1,0)

810

a

814

-4

12(1,0) 18(2,1)

888

a

896

-8

12(1,0) 18(3,2)

911

912

a

911

0

16(1,0)

974

965

c

974

0

12(2,0)

995

c

1001

-6

X(1,0) 12(1,0)

1135

1126

c

1135

0

8(1,0)

1204

a

1199

5

12(1,0) 18(3,0)

1263

1262

c

1263

0

5(1,0)

1294

c

1294

0

9(1,0)

1370

a

1370

0

8(1,0) 18(1,0)

1408

1404

c

1408

0

6(1,0)

1512

a

1514

-2

5(1,0) 18(1,0)

1545

1543

a

1545

0

9(1,0) 18(1,0)

1593

a

1594

-1

5(1,0) 18(2,1)

1616

1615

c

1622

-6

8(1,0) 12(1,0)

1662

1664

a

1659

3

6(1,0) 18(1,0)

1697

1700

c

1705

-9

6(1,0) 13(1,0)

1746

1746

c

1746

0

5(1,0) 12(1,0)

1775

1761

c

1776

-1

9(1,0) 12(1,0)

1896

1895

c

1896

0

6(1,0) 12(1,0)

1985

1984

a

1982

3

5(1,0) 18(3,0)

2017

2017

a

2013

4

9(1,0) 18(3,0)

2063

2057

c

2055

8

6(1,0) 10(1,0)

2120

2120

a

2127

-7

6(1,0) 18(3,0)

2174

2175

a

2174

0

5(1,0) 16(1,0)

2209

2208

a

2205

4

9(1,0) 16(1,0)

2395

2395

c

2398

-3

5(1,0) 8(1,0)

2419

c

2429

-10

8(1,0) 9(1,0)

2544

2540

c

2544

0

5(1,0) 9(1,0)

2544

2540

c

2543

1

6(1,0) 8(1,0)

2632

2629

a

2633

-1

5(1,0) 8(1,0) 18(1,0)

2662

2656

a

2657

5

5(1,0) 12(1,0) 16(1,0)

2669

2661

c

2669

0

5(1,0) 6(1,0)

2685

2684

c

2689

-4

6(1,0) 9(1,0)

2793

a

2795

-2

5(1,0) 9(1,0) 18(1,0)

2872

2873

c

2881

-9

5(1,0) 8(1,0) 12(1,0)

2901

c

2911

-10

8(1,0) 9(1,0) 12(1,0)

2921

a

2920

1

5(1,0) 6(1,0) 18(1,0)

2948

a

2940

8

6(1,0) 9(1,0) 18(1,0)

3022

3020

c

3021

1

5(1,0) 9(1,0) 12(1,0)

3148

3156

c

3152

-4

5(1,0) 6(1,0) 12(1,0)

3168

c

3172

-4

6(1,0) 9(1,0) 12(1,0)

3456

3459

a

3455

1

5(1,0) 9(1,0) 16(1,0)

3603

3604

a

3600

3

6(1,0) 9(1,0) 16(1,0)

3672

c

3679

-7

5(1,0) 8(1,0) 9(1,0)

3811

c

3804

7

5(1,0) 6(1,0) 8(1,0)

3928

c

3936

-8

5(1,0) 6(1,0) 9(1,0)

2Brand, J.; Williamson, G. Discussions Faraday Soc. 1963, 35, 184-191.

Fit Parameters

Transition frequencies were calculated using the anharmonic oscillator model with the following fit parameters.

Parameter S0 (cm-1) S1 (cm-1)
w5   1263
w6   1405
w8   1135
w9   1294
w10   647
w12 564 487
w13   297
w16   911
w18 168 257
wX   514
x5,6   -2.5
x5,9   -13.0
x5,12   -4.5
x6,9   -13.0
x6,12   1.0
x8,18   -16.0
x12,18   -2.1
x18,18 -8.3 -5.8

Acrolein S1 Spectra

The observed bands in the room temperature and molecular beam spectra are assigned as shown.

Discussion

The acrolein S1 vibronic spectrum was assigned considering the computed spectrum, Franck-Condon progressions, temperature effects, and rotational band structure. This allows for the correction of previous literature assignments2 and the identification of previously unassigned bands.

The band previously assigned as 14(1,0) is actually a hot band assigned here as 18(2,1). Consequently, many more combination bands involving n18 are observed than previously believed, allowing for a better understanding of the S1 n18 potential energy surface. The band previously identified as 16(1,0) is also a hot band, but is not yet identified.

The band previously assigned as 15(1,0) is actually a perpendicular band, not a parallel band. This band is assigned here as 10(1,0) based on its computed frequency and rotational band structure.

The band previously assigned as 15(2,0) is assigned here as 9(1,0), as observed progressions indicate two strong fundamental bands in the 1260-1300 cm-1 range.

Conclusion

The present work assigns 65 bands, whereas previous literature2 identified 27 bands, 16 of which were corrected in the present assignment. The present work identifies 9 fundamental transitions in the acrolein spectrum. The identification of numerous other combination bands increases confidence in the assignment of fundamental transitions.

The molecular beam spectrum was critical to correctly assigning the S1 vibronic spectrum by differentiating between hot and cold bands and by simplifying the spectrum to identify bands not observed in the room temperature spectrum.

The computation of the acrolein spectrum provided reasonable predictions of the locations of fundamental transitions. The acrolein results can be used to compute scaling factors for other similar conjugated systems in the future.

Acknowledgments

We gratefully acknowledge Dr. Marsha Lester (University of Pennsylvania) and Dr. Kevin Lehmann (Princeton University) for useful advice on Cavity Ringdown spectroscopy. We also thank Brad Mulder for fabrication of the apparatus. This research has been sponsored by National Science Foundation grant CHE-9157713 and a research fellowship from Pfizer, Inc.