3......“ 1...».

 

III.’....!¢I.
no .I 1" I\blu.
.. .A~I.-k..lix

 

            

    

              

    

.
.
.-. . .a..u.rvl|ht.-h.ournr. (It
. 1.. .1 . i . . . : c. «v .. v. 00 1.14,}.ollllopr'.
.. , I. .Ilnll ‘3‘ . .. . . . .- -. . - . . .. Iavllcn 11.1.: . Itzlfic
Q‘. x 7'1) . I I‘LEI‘ a 3'55; .’ . . . . . . . ... .. . . v - . . o . . . £133.: .‘gao 411...!
:‘F‘? a . at)!” Lat-t: 7.5.... sultgvtnu; U. ... R‘r.‘ it ... . .. . ... . a . . .. . .. .. . 1. : . . . - . . . . .. - . . 5'12}. ’L.‘!? 5‘ .
‘39“ MM V...rur...\‘.p. it . .Llp‘vqkwpvi 5...... fl. .. {v.7}. Quinn! . :55}. . . . . . . . . , . . . . . - ._ .. . oI. .InI 1..! (J. II“. {I .
.1... ...I. .u. I, ..tr.I-I....-av..,i.. 5-H...” .. ....u.t...r.é.t... .it .. 4...... . .. . , I..- .1500.-.
. Fl I 73”.. .. 4.3.... Knit J? JW.5|‘VI1.’L\‘\GI - I‘JN'I".I:I: .
. l . u .l I . L .:E[ tA. .l H. 1' I: y will : 4" i!""'
, -.. II...FII..§H .u0\ yaw-0.3 6AM. f.\-\§$I. 3.9.33.5. . (V7.1... . .. . - . - - . 1 . olraifl'lvtlvl’ltlloflt’ H
II...\\ I 1..! 39.1.... . I. .- . . I I. . . . I . . . . .. ... ., .. H . . . . -.. . .. » . at- 9 .! . claw"?
. us'..{.-.\l1la‘..l1 .. . - .. , 1... .. .n . .. .. - .. .u. .. .. . .. . . (OI )p I v If; 41!?- h-. .7" h
. 2d. I

, “via-alum

                        

                  
            

         

     

     

t... .F S

(I: It: I- x: I

{$1.'II.P_<7 blI'....hs. v
. .liool.llpr.xrl‘. {-¢.b|o'. I

(y. . .4 400.00. .I III! 5.2:..-1‘6?! \. CC.) . a v. .
...- .. .. . 7...... ....-l.IoJI... .vIIIO 01.1 tn!!! on
- - . . . . $7719.47... I). III“. I ... 2.. It’d
. ..- . . . . ,-!....I........oL..¥.. IF... I .3!) I

.IT .5. II... 19.7.17: .335‘ .20. . ... . .. . .4.5r..-.Dtpl.vilo|.r.Ib..O"II-l(nr t

u. .. . ~w-5:..I..|..tat.)uf{ . .17-L . A - . ...I.I .K‘...’ .vtvllv..ot.u.\’.lp.-vlucl'£ p . O.)
7...: I?..9. ., .\ r... «Ca-v. . I . . . - . {1.9.12}, 5. III:!I.\ ..Ib|..l.. .rll‘l 0 ..y ‘i p I
I s\. .GN‘CA furl. .r ‘\§.EVLVI. . gun II; . . . . . If. ‘.rl.lrl\ . ‘« ..vlrtchlv .I‘au. :Irjr.m.hfla'flllhr!. Hal
. WI 3... .K-t};$-l. .JV -1 v. .. it...” .nPrLO.rvrIb.r)i‘.--I.ItIvllr'O nih.icol '0'.-
. .t 15.. . o. «.1 ..... '. I 1..-1..] ‘lI:I.lfID'F-.I.Drb.l’ISII....

., v. .4? Hun. 3 7 HI... . 1er ’3’ .5 lrlo'vitxfii. a 0{|.c‘... 12:7)

. f-0.w.l.. |.v

 

Ifiz‘

 

   

 

            

    

l. .. .
. :-
. .1!!! . .. is... . _ .
.11.. 1‘3. 3.... Ufl!3n~9ulht: .. . . . .. . .
. C. 515‘. . .. . . .- .
.|.ud.vp §hufl1§fif€i I .3 I . 7 - v I .... I..,.,.A..v. ....l\... t. 6.. It... :...t$>bl5ul|.fobutf...-FII$|9|
i ..' 51.1....‘h? .~ . . . -DY A.-. «P I . . .u....!|.(lc$ If- 1' .br- .nn..¥.1(|:ft....I.‘I.§ln.¥ll$l’ll'1;’)l
I12... . ... . . . .I \- -I.. -1. 9.-..‘0 .15}. I. 4.. If}. 1 £121.03 1. It’s“).
.. . . 1..... ., . -.(.-I--.I..7..I.I¢.xcr5r--.I........r.(.....t.. 1...“?!an
. . . - . .. In I 0:. k I ' .I.ltlll I .IévItv."€ lo. \rrIPvlnvl‘ro’VQI ’1’..st
- Pg!

.57

I I in! II. IIAr.l»$;§-S .. .. ....vv..':flo.va.!' - .i

       

  

(I: I,

' ‘ 3112::

1 4.253%?
.9.

.v
-t
t

 

V... II ..I..~ : ... I
. .IIS . 1.7a“... JIIIIDIF...LI\. .9. .. .

.. . 1.--.9.-. .I. I.-. )1
.uIII, .. :11...” 3.1.». 7..-.-.) : .

    
                   

. .l’llvlclI‘lt . Y...
. .. . . .4) 3!.5 .p’cl.t‘.lnll.1l|'.tlofn
. ((A1 .. .3 a...:...!..'.. {Iv-‘12.}? . .c.tl|It.I'OII\I.-.llfi¢z
. . .1 I‘li?l._-«.Illa-J.o"r.L‘l’tli‘)1.{’£, .ig

I I!!! 3 1... .a.“.x’|i“.‘ov 'rl.vl|- 9).-

. .
I . .. .«I ...;. III. .l’tf} Clix-ill

 

  

.Jn. . . . cm“ ....u...nuI.n.. 0.1%....

     

 

   
                 

    

 

v. . nLI -. . - .Dll fl .. Ill; flnI| Ii.
\‘ IN! '3 n .IV“"L'L.|\\>-N . 1.1.3: I» - > - .lzi.‘II .l .il‘v § ‘0 IV ‘ta’ut‘l I.“
Q. II . I4 ‘ 0 s . v . . \ k)- .IlC. ‘rIIII . MIA-.13}, Lit .II.J)I . I 11!... .41.; ..... I 169?. «I:
It)? fulrIIA (Ju‘oIfiv‘l. .‘Il‘..?o. Intuit,“ Jrcl \thI .... 1.... . aux... .. nl£l!I-(iII.' “NH. 5.15.3.3. . (91:21.10 13 rfitaxl«H-I.P .l 0417.; ‘13.? v loirlcci. lei].
3.!va TV“ I.-. fall“??? .1399“ 1:21.. tL..|..E....J.lr.lt. ... I .9. . :I W.” 147-20 :- 3.1....fiwhfflnflx.{iiz . II ... 5...?1: 5.}! ..I...I....-.oll..lp‘ll. .5. If- :1 .. .-. .
. .5 ill { .u ! biliv h.[~|€x 5). -\.~. :~ (IV.\I El lo.‘ ...I - ...\/.3wc\. I. . I .
U V 4" I) 9 0 v .1. ck‘l
.Muvsb‘fidliinl ...v , . s57». . I. . - ... s
- 3.1L. .. . .. - . I.
, . . . . 1-.... .273...- %I
V u» .. .61? 3-11 ..... L '5 5.3.3.. u‘xnn..l...u...ob. In :. .u r}, «it. t I. L ‘Ill' u! ’.v.b~(.\l. 3!. V n
. . - .. . 2.. .. . .2. . . -
.n. I In. :7 . . .Il . .. .-.. 17...]; $342.41... I”... . .. .. . . I , v I
, u N -.. . . .L§ . . .7 .
. l 1 ‘ . I . . III, '| .II I'.‘ -‘
- .ll" ll'

   

' ‘
. 7.398423

MICHIGAN STATE UNIVERSITY \UMIES‘

|\\\ Will\\l\\\\\\\\\\\\\\\\\\l\Wfl “MW

3 1293 00788 6017

l
1’

 

 

 

 

 

lLllHRJllI!’
Hickman State
L University

*

 

 

J

This is to certify that the

dissertation entitled
I'Investigation of Low—Lying Excited Electronic States
of Perylene; Two—Photon Excitation Spectroscopy of
Perylene in Solution and in Supersonic Jet"

presented by

Jeonga Yu

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Chemistry

 

 

I Major Jofessor

Date March 19, 1990

MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

PIACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due. .

ll DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU I: An Affirmative ActiorVEquel Opportunity Inditution
cMmS-o:

INVESTIGATION OF LOW-LYING EXCITED ELECTRONIC STATES OF
PERYLENE; TWO-PHOTON EXCITATION SPECTROSCOPY OF
PERYLENE IN SOLUTION AND IN A SUPERSONIC JET

BY

JEONGA YU

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Chemistry

199)

ABSTRACT
INVESTIGATION OF LOW-LYING EXCITED ELECTRONIC STATES OF
PERYLENE; TWO-PHOTON EXCITATION SPECTROSCOPY OF
PERYLENE IN SOLUTION AND IN A SUPERSONIC JET
BY
J conga Yu

Two-photon excitation spectroscopy has been utilized to study the low-
lying excited electronic states of perylene. The fluorescence from the two-
photon induced electronic excited states of perylene in n-hexane solution at
room temperature has been measured using two-photons from the same
excitation laser under both linear and circular polarization conditions over
the incident dye laser wavelength spectral region 560 - 840 nm (two-photon
energy 35,700 - 23,800 cm‘l); a portion of the two-photon excitation (TPE)
spectrum, over the region 650 - 840 nm (two-photon energy 30,770 - 23,800
cm'l), was also obtained from a 77 K isopentane glass. From the observed
peaks and the ratio of the two-photon cross sections for circularly and
linearly polarized incident excitation beams (Q = 5d, / 81in), low-lying rur*
electronic states of gerade parity in the 23,800 - 35,700 cm“1 range, as well
as a dipole-forbidden B3“ state below 26,500 cm‘l, have been located and
characterized. The results are compared with values of perylene excited
state origins predicted by calculations, or observed by various alternative

experimental methods.

Jeonga Yu

In order to provide more definitive assignments and more direct
comparison to theory, two-photon excitation of perylene in a supersonic jet
was attempted. To characterize and optimize a home-built supersonic jet
system, the one-photon spectra of perylene over the spectral region 410 - 420
nm, of iodine in the X(12'.) -’ 3(31'1) transition region (578 - 596 nm), and of
[2,2]paracyclophane (1 (So -* 31) = 280 — 320 nm) were obtained under
various conditions. The preliminary results of the two-photon excitation of
supersonic jet-cooled perylene, over the region 576.8 - 577.8 nm (two-photon
energy 34,670 - 34610 cm'l), is reported.

ACKNOLEDGMENTS

I would like to thank professors G. E. Leroi and D. G. Nocera for
their help and advice throughout my graduate work, and professors G. T.
Babcock and S. R. Cruoch for serving on my guidance committee.

I am also grateful for the opportunity to get to know many special
people during the past years at Michigan State University. Their concern
and friendship have made this experience pleasant.

More importantly, I want to thank my parents and family. Without
them this all would have not been possible.

Financial support from Michigan State University and National

Science Foundation is also gratefully acknowledged.

TABLE OF CONTENTS

Page
LIST OF TABLES ...................... vi
LIST OF FIGURES ............... . ...... vii
CHAPTER I INTRODUCTION ............. 1
I. Linear Spectroscopy ......... 2
II . Nonlinear Spectroscopy ........ 5
A. Two-Photon Absorption . . . . . . 7
B. Characteristic Properties of
Two-photon Transitions ...... 11
1. Intensity Dependence ....... 11
2. Polarization Dependence . . . . . 12
3. Resonance Efl'ect ......... 16
C. Experimental Techniques . . 22
1. Fluorescence Measurement . 24
2. Detection of TP Absorption
by MP1 of Molecules ....... %
III. Motivation ......... . . . 26

CHAPTER II TWO-PHOTON SPECTROSCOPY OF PERYLENE

IN SOLUTION . . A ............. 35
I. Experimental ............ 2%

A. Materials ............ :3

B. Absorption and Emission Spectra . . 36

C. Two-photon Excitation Spectra . . .
1. General Considerations .
2. Polarization Measurement
Procedure ...........
3. Dye Laser Wavelength Scan . . .

II. Results and Discussion . . .

CHAPTER III TWO-PHOTON SPECTROSCOPY IN
A SUPERSONIC JET ............

I. Introduction ............

II. Supersonic Expansion . . . .
' A. Translational Cooling . . . . .
B. Vibrational and Rotational Cooling .

C. Condensation and Complex
Formation ............

III. Spectroscopy in Supersonic Jets . .
A. The Number of Excited Molecules . .
B. Detection ..... . ........

IV. Limitations of Supersonic Jet
Spectroscopy ............
A. Vaporization ...........
B. Sensitivity ............

V . Experimental .............

A. Jet Generating System .......
B. Excitation Source .........
C. Detection .............

iv

82} 88

83

101

. 105

109

111
112

119
119
122

123
123
131

REFERENCES

VI.ResultsandDiscussion . . . . . . . . 135

A. Pressure Effect .......... 136

1. Chamber Pressure ....... 136

2. Carrier Gas Pressure . . . . . . 139

B. Distance Effect .......... 142

C. Nozzle Size Effect ......... 145

D. Sensitivity ............ 145

E. TPE of perylene .......... 148
..................... 157

LIST OF TABLES

Tables Page
1-1 Examples of Four-Wave Mixing .......... 6
1-2 Values of the Polarization Variables F, G,

and H for Two-Photon Transitions .......... 15
13 Examples of the Two-Photon Tensor Patterns

for Groups .................... 17
2-1 Observed and Calculated In-Plane Vibrational

Frequencies (cm'l) of Perylene ........... 81
2-2 Comparison of Experimental and Calculated

Values of Transition Energies to Low-Lying

Symmetry-Forbidden Excited States of Perylene . . . 90

Figures

1-1
1-2
1-3

1-4

2-1
2-2

2-3

2-5

2-6

2-7

LIST OF FIGURES

Various multiphoton absorption processes .
The two-photon absorption processes of a molecule

UV-visible electronic absorption spectrum of
perylene (10'4 M) in n-hexane solution . . . .

The nine Kekule structures of perylene
classified according to symmetry .

Bond lengths (A) in the ground and excited
state of perylene ................

Two-photon transition processes . .

A schematic diagram of the two-photon excitation of
fluorescence measurement . . .

The florescence measurement sample chamber
for two-photon excitation . . . .........

Dispersed fluorescence spectra: (a) A“ = 365 nm:

(101” = 584 nm .................

Power dependence of the two-photon induced
fluorescence of perylene («'10-4 M in n-hexane)
at different initial laser wavelengths . . .

The tuning range and relative output power of dyes
used in the TPE of perylene ..........

The spectral response of the photodiode . .

Page

41

. 46

51

2—9
2-10
2-11

2—12

2-13

2-14

2-15

2-16

2—17

2-18

3-1

3-2

3—3

3-5

Liquid nitrogen dewar and high-vacuum cell
for 77 K TPE measurement ..........

The alignment of the polarization changing optics .
Representation of the electric field vector in space

The electronic circuit diagram of the stepping
motor interface ............. .

The symmetries of the n—electron cloud of the ground
state and the electronic excited states of perylene .

One-photon UV-visible absorption and two-photon
excitation spectra of perylene in n-hexane solution .

TPE of perylene in n-hexane solution (~10'4 M) and
polarization ratio, 9, plotted as a function of
the incident dye laser wavelength . .

TPE spectrum of perylene in isopentane at 77 K .

Observed and calculated energy level diagram of
the excited electronic states of perylene
near the peak at 26,500 cm'1 . . . .......

Two-photon transitions to various vibrational levels
of a 35,, electronic symmetry state . . .

Energy diagram of the lower excited electronic
states of perylene ..............

Velocity distribution in an effusive and in a
supersonic beam ........

Shock structure surrounding the expanding
gas of a free jet ...... _ ........

The geometric cooling of the translational
degrees of freedom in relation to the expansion aids

Vibrational and rotational cooling via collisions
in a jet .................

Various two-beam double resonance spectroscopies

71

102

106

. 117

3—7

3-9
3-10
3-11

3-12

313

3-14

3-15

3-16
3-17

The experimental set-up of a time of flight mass
spectrometer coupled with laser desorption into
a supersonic jet followed by laser MPI .

Schematic block diagram of the supersonic
jet experiment ................

Schematic diagram of the chamber employed for
supersonic expansion experiments .

A side-view cross section of the sample compartment
Schematic diagram of the detection system .

The LIF spectra ofjet-cooled iodine molecules
at different chamber background pressure .

The LIF spectra of perylene in a jet as a function
of the carrier gas pressure . .

The LIF spectra of perylene in a jet at different
distances from the nozzle to the laser intersection

The effect of nozzle size on the internal
temperature cooling ........

The LIF from [2,2]paracyclophane expanded through

a supersonic jet . . ........... _ . . .
TPE spectrum of perylene in a supersonic jet .

The relaxation processes following two-photon
excitation in solution and in a jet . . . .

a

a ass i

E

.149

151

CHAPTERI

INTRODUCTION

I. LINEAR SPECTROSCOPY

When Maxwell formulated the electromagnetic theory of light, 'he
made the approximation that the dielectric susceptibility (x) and magnetic
permeability (n) to the medium were independent of the strengths of the
applied fields, and thus the dielectric polarization (P) and magnetization
(M) were linearly proportional to the field amplitudes (E,H)[1],

P=so x-E;M=(n-l )-H (l-l)
where so is the permittivity of free space. The susceptibilities (X), in which
the resonant structure of the spectra are contained, must be regarded as

complex numbers in order to describe the behavior at all frequencies; then

the optical frequency fields need to be written in complex notation

E(r,t) = % {E(r)e-i(mt-k-t) + E*(r)ei(cot-k-t)}, (1-2)

where k is the wave vector and (o is the frequency of the wave. Thus, the

wave equation becomes

1 62E 41: 62?
V2E - 23w: 32-32-21 (1-3)

I which relates the wave vector, k, to the susceptibility, x. The disparate

optical behavior of linear media is contained in the relative magnitudes of

3

the various constitutive coefficients and in their behavior as a function of
frequency.

In transparent dielectrics (or at frequencies far removed from a
transition resonance), E and H are in phase but the ratio of amplitudes is
changed from free space. The resultant relationship between P and E
contains a phase lag which is experimentally observed as the refractive
index (n = k/ko), where n2 = 1 4- Rex (Rex means the real part of x) and k0
is the wave vector in vacuum. The observable intensity of the wave is

defined as
= ‘33 |E(r) I 2, (14)
81:

where c is the speed of the light in the vacuum.

At frequencies near resonance, the electric field is absorbed to an
extent determined by the extinction coefficient K, where 2nK = Im)(
(imaginary part of x); the imaginary term in the wave vector introduces an
attenuation factor exp(-xkor) on the electromagnetic wave. The attenuation
(extinction coefficient) in absorbing media obeys the Beer-Lambert law

(k-VH = -2KI. (1-5)

Early workers in quantum mechanics recognized that variations of
the index of refraction and attenuation coefficient contained crucial
information about the energy levels of the medium. Bohr related the
attenuation coefficient to the probability of a transition between energy
levels separated by the energy of a quantum of light, E = hv. Fermi

expressed the transition probability per unit time, Fij, in terms of the

4

matrix element of the dipole moment operator u = er which connected the

two levels
rij = 2313*“”Emu”2 p(Ei — Ej - hv), (16)

where the factor p(E) is a density-of-states function that reflects the observed
lineshape and e is the electronic charge. If the refractive index of the
medium is different from one, the dipole moment operator needs to be
corrected for the local field, u -t ((n2+2)/3) er. Einstein calculated the

probability for spontaneous light emission per unit time

I 2

Aha = 3546, (Eb—Ea)3 (1-7)

and pointed out that the transition rate in the Fermi equation applies
equally to absorption and stimulated emission processes. Kramers and
Kronig showed that the variation of the index of refraction was related to
the absorption. Thus, the quantum mechanics of the classical methods of
spectroscopy - emission, absorption, and dispersion - was .well understood
early in the twentieth century[2-5]. The fourth method of linear
spectroscopy - light scattering - required more complicated methods and
was explained by using a vibrating model[6,7].

These classical phenomena of linear spectroscopy contributed
immeasurably to the growth of quantum mechanics and our
understanding of atoms and molecules. The experimental limitations,
however, that resulted from the unfavorable selection rules, DOppler
broadening and other complicating phenomena, were recognized as

intrinsic to linear spectroscopy.

II. NONLINEAR SPECTROSCOPY

The term nonlinear spectroscopy refers to those phenomena
involving light where the induced polarization, P, of an atomic or
molecular electron cloud is not at the same frequency as that of the driving
electric field, E. For centrosymmetric molecules, P is related to E by the

expression
P =80 E( x1+ X3E2+ ....) (1-8)

where so is the permittivity of free space and Xi are the susceptibility
tensors of i with rank (i+1)[8].
Linear spectroscopy occurs when the susceptibilities are space and

time independent, and when X1>>X 3. Therefore,
Hwaammxb ‘um

where a) is the frequency of the driving radiation. In systems with
inversion symmetry, the lowest-order nonvanishing nonlinear
susceptibility is of third order, x3. The third-order polarization can be
expanded into the form:

P3(coi) = x3(mi,coj,(ok,m1) E6) E(k) E(1), (1-10)

where four waves are being mixed by X 3. Several common combinations of

frequencies are shown in Table 1-1. Like the first order polarization, the

6

Table 1-1. Examples of Four-Wave Mixing.

 

 

Resultant Mixed Resonance Phenomenon
Frequency Frequencies Frequency
3031 031, col, (01 «~- Frequency tripling
(01 all, (01*, m1 -- Self-focusing
col — we Self-induced
transparence
(oz (01, (01*, m2 -- Intensity-dependent
refractive index
(01 + (02 = we Two-photon absorption
(01 - (02 = (0,, Stimulated Raman
emission
012 - (01 = 0),, . Inverse Raman

2“’1 " “‘2 0’1» 0’1»‘°2* ""

(”l-“’fmv

absorption

Four wave mixing

CARS

* = the complex conjugate of the electric field strength.

a) e = an electronic transition.
CDV = a vibrational transition.

7

third order (nonlinear effect) can be used to obtain structural information[9-
11].

The possibility of observing optical frequency nonlinear phenomena
requires high field strength. An upper bound for the field strength can be
estimated by the field, E0, binding an electron to the nuclei, and leads to the
approximate rule that X1/X3 ~ E0[12]. A typical value for E0 would be 3 x 108
V cm'1 which corresponds to an intensity of ~ 1015 W cm'z. The
experimental dilemma was alleviated when the first laser was designed by
T. Maiman[13]. After that, nonlinear optical effects (resulting from the
higher order terms) including stimulated Raman gain[14], multiphoton
absorption[15], second harmonic generation[16], sum and difference
frequency generation[17], saturated absorption[18], etc., were discovered

and verified.

A. Two-Photon Absorption

Multiphoton absorption, one of the nonlinear effects, basically
consists of the excitation of molecules by two or more photons. The upper
state of the transition decays like any excited state, either nonradiatively to
the ground state, or to a fluorescing state and then to the ground state.
Alternatively, it may undergo photochemical change. New vibronic and
electronically excited states, which are not accessible via one-photon
spectroscopy, can be observed over a wide range, from lower excited states
. to ionized continua (several examples are shown in Figure 1-1). As the
simplest example of a multiphoton transition, the two-photon transition

has several features that are important for spectroscopy[19].

 

Figure 1-1.

Various multiphoton absorption processes: a) two-photon
transition with resonant intermediate; b) two-photon
transition with nonresonant (virtual) intermediate; c)
stepwise three-photon transition; and d) three-photon

ionization process with resonant intermediate states.

10

First, the selection rules for two-photon transitions differ from those
for one-photon transitions. For centrosymmetric molecules the one-photon
transition only has significant oscillator strength when the parity of the
electronic wave function changes upon absorption (g—u or u—g). The two-
photon absorption process can be thought of as two simultaneous one-
photon absorptions, with the first photon achieving an intermediate state
and the second photon taking the molecule to the final state. Therefore, in
centrosymmetric cases, two-photon excited states will have opposite parity
to the one-photon excited states. When the ground state has gerade
character, the transition moment is nonzero if the intermediate state has
ungerade character and the final state is gerade. Since the intermediate
“state differs from the initial state by one orbital, the final state may differ by
two orbitals. Therefore the final state can result from two promotions of one
electron or one promotion of two electrons. The latter is strictly forbidden in
one-photon spectroscopy.

Second, during two-photon excitation molecular states may be excited
that in the one photon process require a shorter-wave photon to be absorbed;
"vacuum UV" energies (200 - 100 nm) are available using photons in the
more accessible range 400 to 200 nm. Moreover, additional information,
especially about intermediate states, can be obtained from two-photon
spectra because the absorption coefficient is measured as a function of two

frequencies.

11

B. Characteristic Properties of'l‘wo-pboton Tramitions

Two-photon transitions have characteristic features such as
radiation field intensity dependence, polarization dependence, and

resonance enhancement[20].

1. Intensity Dependence

The multiphoton transition probability is formulated according to
time-dependent perturbation theory[21]. When two-photons are from the
same laser (031 = (02 = (or), for a two-photon absorption from state a to state 11

the transition probability W2 is given as

. 2 snianlzsmluLaz. (1,11)
W2 (1 I 2 Am _ mt
m

where I is the intensity of the radiation field, 111 the virtual intermediate
state, Am the energy difl'erence between the intermediate and initial states,
it the dipole moment, and cor is the radiation frequency. This equation
shows that the two-photon probability is proportional to the square of the
radiation field intensity. More generally, the n-photon transition
probability is proportional to In. If saturation is absent, one determines the
order of a multiphoton transition from the slope of the log-log plot of the
transition probability as a function of radiation field intensity,

an(n) = nlnI + c. (1-12)

12

2. Polarization Dependence

One of the most interesting features of two-photon spectroscopy is the
potentiality to identify the nature of the excited state from the polarization
studies. The polarization dependence of two-photon transitions in crystals
was considered by Bader and Gold[22], and also by Inoue and Toyozawa[23].
For randomly-oriented, nonrotating molecules the exposition was given by
Monson and McClain[24,25], who derived an expression for the averaged
transition rate and showed how to distinguish states of different
symmetries using an appropriate combination of photons with different
polarizations. The effects of molecular rotation were considered by Bray
and Hochstrasseli26].

Polarization studies in random molecules are uninformative in one-
photon absorption because of an averaging out of polarization effects. The
one-photon transition probability a from ground state lg> to final state <f|
is proportional to IA-jlfgl‘2 where A is the polarization vector of the absorbed
photon and pf: is the transition vector between states g and f. In randomly
oriented systems, 0' must be averaged over all orientations of the molecules
to obtain <0>. The result is «3) = 1/3(k-K*Xufg-tlfg*). Since kl“ = 1 for all
polarizations, the absorption is completely independent of the polarization
of the incident beam.

In two-photon absorption, the transition probability, 8, is given by

5 = nesabaf-u l 2, (1-13)

13

where 7t and u are the polarization vectors of the two absorbed photons and
S°f is the two-photon transition tensor between states <o| and |f>. When
the two photon transition of a fluid is considered, 8 is averaged over all

orientations of the molecule to yield <5>,
<8> = OFF 4- figG + SHH (1-14)
The terms F, G and H depend only on the polarization of the incident

photons, and the terms 5F,5G and 81.1 are molecular parameters

independent of the polarization of the photons where

 

 

3
8F: 233:? (1-15a)
a=b=1
3
5G= Elsie” (1-151))
(a.b=l)
3
5H: zswsb; (1-150)
(a,b:l)
S‘f =<0|Sablf> ' (1-15d)
[pali><ilpb] [pbli><ilpa]
and . Sab= 2|: (Ei-El) + (Eng) ]. (1-15e)

1
In these equations a,b are coordinate axes in the molecular frame, pa is the
operator for total momentum in the a direction, li> is the intermediate
state, E is the transition energy from lo> to li>, and E1 and E2 are the
energies of the two absorbed photons.
The functions F, G and H are given by

l4

F=4Ii-uI2-1—Ii.u*12 (1-16a)
=-I>.-uI2+4-— Ii.u*l2 (1-16b>
=-Ii-uI2—1+4Ii-u*I2 (1-16c)

and it follows that <8> is a linear function of the lk-u I 2 and lx-u” l 2.
Therefore 8p, 83 and 83 can be independently determined with carefully
designed experiments which require the use of linear and circular
polarized light. The values for F, G, and H under different beam
polarizations are shown in Table 1-2.

For the "single beam experiment", which we applied to the TPE
measurement of perylene, the two-photons come from the same laser beam,
(E1 = E2, 1. = n). In this case only two independent parameters, 8F and 8G
(since 8G = 8H), remain which are determined from the two measurements

of 8 with linearly and circularly polarized light. From Table 1-2,

51in = 2 517(0)) + 4 59(0)). (1-17a)
8d,. = -2 55(0)) + 6 8g(m). (1-17b)

The two-photon polarization ratio 9(a)) is defined as

8 - 3 5
mm) = m =, '5‘“ + G (1-18)
lin 5r + 2 5G.

The value is bound to the interval

0 S Q S §2_ (1-19)

15

AVE 8.88%." 895

 

 

 

I IN. N HI I I
a m a a a H H a m
m a m. m m a a a O
I m. m N I I
m .e. a a a H a a a
awe. Awe. «Iv if.» are. dis IT... «TIT 83>
m a e a a. a a H assure?
ammo

 

6.85329 :Sonmég you m can .0 .m 3:35; newsman—om e5 me 333’ .NA 039—.

16

and is, in symmetric molecules, closely related to the symmetry of the
excited state; it corresponds to the "degree of polarization" in Raman
spectroscopy. Therefore 52 can serve as a symmetry-indicator (Table 1-3).
Even if the theoretical limit is not seen in a real spectrum, due to the
influence of vibronic coupling or overlapping of peaks arising from
different symmetry states, the relative trends in (2(a)) provide valuable
information on the symmetries of the excited states.

3. Resonance Efi’ect

The intermediate state in two-photon absorption can be off-resonance
from all ungerade states, so called a virtual state (Figure 1-2), although the
total energy of the two photons must satisfy the resonance condition. When
the excitation laser. is tuned and its frequency approaches a real
intermediate electronic state, the two-photon absorption signal will

dramatically increase (resonance enhancement),

 

w2 ~ [Mi 1} (1-20)
where W2 is the probability of the two-photon transition, I‘ is the
homogeneous half-width of the transition and Ad) is the energy difference
between the laser and real intermediate level. For nonresonant two-photon
transitions, the vibronic structure is mainly determined by the Franck-
Condon vibrational overlap between the initial and final states, because the
energy mismatch to the intermediate level is so large that the vibronic

structure of the intermediate state can be neglected. On the other hand, the

17
Table 1-3. Examples of the Two-Photon Tensor Patterns for Groups.

 

 

Tensor Pattern Point Group Polarizaion
Irreducible Representations Ratio
Dz 02v 132-.
0,30 0 2
[0 Oyyo ] g
0 0 on
A—A Al-Al Ag-Ag
31-31 Az-Az Big-313
32-32 81-81 Bag-Ba
33-33 132-82 Bag-Bag
Au‘Au
Blu"'Bln
Hal-Ba.
Baa-Baa
l" “‘"l "’
nyo 0 :2'
0 O 0
A—Bl Al-Az Ag-Blg
82-33 31-32 Bag-Ba
Au-Blu
Bau-
0 0 on 3
[. . . J ,
030 O
Az-Bz Al—Bi Ag-B
BIB—'B38
Au-Bza
Bin-Bax
0 0 0
[W]
0 oz, 0
A-Ba Al-Bz Ag-ng
3,..3, Aa-Bi Big-Ba
AU-B3U

18
Table 1-3. (continued)

 

 

Tensor Pattern Point Group Polarizaion
Irreducible, Represents Ratio
04v» D4» Dad D41:
<3xx 0 0
0 CW 0 < 1
0 0 022
A1 ‘ A1 Alg " A13
A2 - A2 A28 — A2g
B1 - Bl Blg - Big
32-32 BZg" 323
33¢ " 33:
Alu " Alu
A2u - Aeu
Blu " Bln
Bzu- Ba:
on 0 0 3
0 on 0 §
0 O 0
A1 - B1 A1g - 318
A2 ‘32 Azg-Bzg
All! ' Blu
A2u - Ban
0 ex}, 0 3
“y: 0 0 E
0 0 0
A1 - B2 A1; - 32g
Az-Bl Azg-Big
Alu " 32.1

A2!) ' Blu

 

l9

 

 

 

Table 1-3. (continued)
Tensor Pattern Point Group Polarizaion
Irreducible Representations Ratio
C4v’ D4’D21 D41
0 0 on 0 0 3
[0 0 0 ]+ 0 0 <3sz '2'
02:0 0 0 °zy 0
A1 - E A1g - Eg
A2-E Azg-Eg
Bl—E 318- 1338
Bz-E %-Eg
Alu'Eu
Blu'Eu
Bzu-Eu
{on 0 0
0 oyy 0 +
K0 0 on]
(on 0 0 J [0 axyo ]
0 «:30 + (3ny 0 <1
\0 0 0 0 0 0
E—E Engg

 

20

Figure 1-2. The two-photon absorption processes of a molecule.

21

 

 

Nonresonant Intermediate Resonant Intermediate

Figure 1-2.

22

vibronic structure of resonant two-photon transitions cannot be explained
by simple Franck-Condon overlap of the vibrational wave functions of the

initial and final states, but depends also on the intermediate state.

C. Experimental Techniques

In the first article on two-photon absorption, Maria Goeppert-Mayer
briefly discussed the feasibility of experimentally detecting two-photon
processes[27]. She wrote that because of the quadratic intensity
dependence, it would be dificult to observe two-photon absorption. Hence,
no progress was made in two-photon absorption experiments over a period
of some thirty years after Goeppert-Mayer's prediction. Compared with a

0‘17 cm2), the cross

one-photon cross section for a typical molecule (~1
section of multiphoton transitions are 40450 cm4s and ~10‘82 cmfis2 for two-
photon and three-photon transitions, respectively. The first successful
visible-region experiment was performed by Kaiser and Garrett[28] using a
high power ruby laser beam at 1. = 6943 A focused into a Eu2+-doped CaF2
crystal. The emission of blue light from the two-photon excited state was
observed.

Since two-photon absorption is a second-order process, it is weak at
moderate light intensities. This demands a sensitive technique for
detecting two-photon absorption in a sample. In principle, higher light
intensities could be used, but then higher-order processes become more
probable and the measured spectrum may be a superposition of two-, three-,

or more than three-photon spectra. Indeed, even ionization and

fragmentation of a molecule is possible at high radiation intensities. For

23

these reasons highly sensitive detection techniques and moderate light
intensities are more favorable than high light intensities and nonsensitive
detection techniques.

Several experimental techniques have been developed to measure
two-photon spectra. These include, direct two-photon measurement[29,30] ,
indirect fluorescence excitation measurement[31,32], thermal
blooming{33], and multiphoton ionization[34,35]. Only direct two-photon
absorption yields absolute cross section 8-values, although with very low
accuracy and sensitivityl36,30(b)l.

The absorption law obeyed by tWo-photon transitions is similar to that
for one-photon absorption, except that the absorbance is intensity

dependent. The fraction of light absorbed can be determined by
AP 8P
-f-.- [1"] t 0, (1-21)

where P is the incident power, AP is the change due to absorption, 8 is the
two-photon cross section, C is the concentration, C is the path length, and A
is the transverse area of the incident beam. For the case where
fluorescence from same excited state follows two-photon absorption, it can
be seen fi'om eq. 1-21 that the fluorescence intensity is proportional to AP; Pf
= q AP, and is therefore proportional to P2, where q depends on the
fluorescence quantum yield and on geometrical factors of the experimental
set-up[37,38]. Monitoring the intensity of the fluorescence initiated by‘ two-
. photon absorption as a function of the laser frequency can allow the 8-value
measurements to be made over a wide spectral range, although only the

relative values of 8 at different laser wavelengths are obtained. Because of

24

its sensitivity and convenience, the indirect fluorescence measurement is
the most widely utilized method to obtain (relative) 8-values.

The details of the experimental aspects of the above methods are well
reviewed in the literature. Because they are the most common methods,
fluorescence excitation and multiphoton ionization are the measurement

techniques reviewed here.

1. Fluorescence Measurement

The most convenient and sensitive method for measurement of two-
photon spectra is the detection of the fluorescence from the excited
molecules. Two-photon excitation (TPE) of fluorescence spectroscopy is
based on the detection of the fluorescence as a function of the incident laser
energy, and is the technique used to determine the low-lying one-photon
forbidden gerade state of the perylene described in this dissertation. In this
method the sample to be studied is irradiated with one or two laser pulses,
of which at least one must be tunable. The two-photon induced fluorescence
is collected by an optical system after appropriate filtering (chemical or
optical filters, and/or monochromator), and detected by a photomultiplier.
For a "single-beam" experiment the signal from the detector is normalized
at each wavelength to the square of the incident laser power. The TPE
method requires that emission from same excited state follows the two-

photon absorption by the excited states of interest.

25

2. Detection of Two-photon Absorption by Multiphoton Ionization of

Molecules

The multiphoton ionization technique for obtaining information from
excited molecules was demonstrated for the first time by Lineberger and
Patterson in 1972[34] and developed by J ohnson[35]. This method involves
collecting electrons released from molecules after irradiation by one or
more laser pulse(s). After a molecule has been excited by two-photon
absorption, it absorbs subsequent photons, which finally results in
ionization. The ion signal is recorded as a function of the laser frequency.
The sensitivity of ion detection is quite high compared with that of
absorption. However, this method is more complicated in interpretation
because difl‘erent absorption steps may overlap, giving rise to congested
spectra from different transitions. An interesting feature of multiphoton
ionization is the signal enhancement by resonant intermediate states.
Since the ionization efficiency is strongly enhanced when the photon energy
is in resonance with a real intermediate state, scanning the wavelength of
the laser leads to a modulation of the ion current, which reflects the
spectrum of the intermediate states. This method is sometimes abbreviated
as REMPI or RE2PI. A new branch of multiphoton ionization called
multiphoton ionization mass spectroscopy (MPIMS), in which a mass
spectrometer is used to identify the ionization products, has been
developed[39,40]. MPIMS yields important information on the dynamic
behavior involving both energy relaxation and fragmentation reactions that

take place in the excited states of neutral and ionic molecules.

26

For comparison, the difi'erences between the spectra from above two
methods are summarized as follows: 1) the ionization technique is more
sensitive to states with high excitation energies near the ionization
continuum that barely fluoresce; 2) the ionization technique is especially
sensitive for detection of Rydberg states, whose relaxation rates are
supposed to be not as fast as relaxation from valence states; 3) with
fluorescence detection one preferentially obtains pure two-photon spectra,
since fluorescence from higher, three-photon states that might be
simultaneously excited by the absorption of a third photon, is not very
probable; and 4) in many cases the ionization technique may produce
overlapping spectra from two-, or three-photon intermediate states, and

therefore assignment is not always unambiguous.

III. MOTIVATION

Considerable attention has been paid‘ to the electronic spectroscopy of
condensed aromatic compounds because of their importance as
chromophores in photochemical and biological systems, such as
carcinogens and laser dyes, etc., as well as their intrinsic interest[41-44].

As a model of centrosymmetric polycyclic aromatic compounds,
perylene is an interesting molecule from the spectroscopic point of view.
Unlike other condensed aromatic compounds the lowest singlet one-photon
transition (weak in most other molecules of this type) of perylene is the
strongest band. Also, the two lowest electronic transitions observed in the
one-photon spectrum are separated by an unusually large energy spacing

(>10,000 cm‘l, see Figure 1-3). In the gas phase and in solution perylene

27

Figure 1-3. UV-visible electronic absorption spectrum of perylene (10‘4

M) in n-hexane solution.

28

 

 

 

 

 

 

 

 

use can com com
_ a a _ . I_III.|qIIII4.- -I....I 4 _ 1 J a A
p . — p _ h n — m n P . — P p h L
7:33 cocmm 88m .259” _ 88v

3138

29

has a high molecular point symmetry, D211» and the transitions in the
visible and near UV region can be correlated only to n-electron excitations of
the aromatic rings. Negligible conjugation is indicated by the consideration
of Kekule structures for the two central bonds joining the two naphthalene
group rings (Figure 1-4), and both theoretical calculations and
experimental results indicate that the n-electron bond order changes with
electronic excitation (Figure 1-5)[45].

In addition to investigations in the condensed phase, perylene has
been well studied even in supersonic jets with regard to energy transfer[46],
intramolecular vibrational relaxati0n[47], and eximer formation[48], as
well as the absorption and emission spectra[49]. Most of the studies,
however, are limited only to the first excited state, and only one two-photon
measurement (at the 14,400 cm'1 fundamental of the ruby laser for gas
phase perylene) has been reported. However, theoretical calculations[50],
dichroism analysis of perylene embedded in a stretched polymer[50(a), 51],
fluorescence analysis in the gas phase[52], and low temperature absorption
spectra[53] all indicate that one-photon forbidden gerade states, as well as a
B;1 state forbidden in conventional spectra by the + - pseudo-parity
selection rule[54], exist between the first two absorption bands.

Transitions to gerade excited states are normally forbidden by one-
photon selection rules because the electronic ground state is Ag, and only
u-g transitions are permitted for centrosymmetric chromophores in
conventional (one-photon) spectroscopy. The unique even-parity selection
rule and the dependence of the fluorescence intensity on the polarization Of
the excitation beam in two-photon spectroscopy reveal important

information regarding such "hidden" states.

30

Figure 1-4. The nine Kekule structures of perylene classified according
' to symmetry.

31

ca
30 A“

:0 O:
:0 O:
rarswa%
OO 00 O

Figure 1.4.

 

 

 

32

Figure 1-5. The bond length (A) in the ground and excited state of perylene.
From ref. [45a].

33

 

 

 

 

Bond | 1.2 2-3 13-6 17-9 14-6 16-20 4.17

 

 

GnmndState 1.45 1.38 1.50 1.44 1.38 1.45 1.42

1", Excited State 1.39 1.40 1.43 1.42 1.42 1.42 1.418

 

 

Figure 1-5.

 

34

In this dissertation, the observation and assignment of hidden
electronic states of perylene by two-photon excitation spectroscopy will be
described. This work contains two parts. Firstly, the TPE spectra of
perylene in n-hexane solution at room temperature over the dye laser
wavelength spectral region 560 — 840 nm, and in isopentane solution at 77 K
over the region 650 - 840 nm have been measured. These spectra were
obtained using two photons from the same laser. For polarization analysis,
the ratio of the fluorescence intensity generated by circularly and linearly
polarized laser beams was also measured. Symmetry assignments are
given to the observed features based on the polarization ratio, 9 = Icir/ Ilin'
Secondly, to obtain highly resolved spectra, TPE of perylene in a jet was
attempted. As an initial step, the one-photon spectra of perylene in the Ag
-’ Bzu transition region (~410 - 420 nm), of iodine over the spectral region
578 - 596 nm (X02) -* B(3I'I)), and of [2,2]paracyclophane (Ag 4 Blg A. (280
— 320 nm)) were obtained.

CHAPTERII

TWO-PHOTON EXCITATION SPECTROSCOPY OF PERYLENE IN
' SOLUTION

35

I. EXPERIIVIENTAL
A. Materials

Perylene (99.+ % purity) and spectroscopic grade n-hexane were
purchased from Aldrich Chemical Co. and EM Science, respectively, and
used without further purification. The high purity isopentane used for low-
temperature experiments was prepared by degassing isopentane, obtained
from Burdick and Jackson, by repeated freeze-pump-thaw cycles and then
vacuum distilling onto NaK alloy contained in a high vacuum flask.

B. Absorption and Emission Spectra

UV-visible absorption spectra of perylene were recorded on a Perkin-
Elmer Lambda Array 3840 spectrometer before and after two-photon
excitation spectra were obtained to ensure that no photochemistry had
occurred due to laser irradiation. The sample was dissolved in the
spectroscopic grade n-hexane (~10"4 M) and placed in a square quartz cell
with 1 cm pathlength. The dispersed emission spectrum was obtained by
using a home-built emission spectrometer which is described in detail in

reference 55, housed in N ocera's laboratory at Michigan State University.

36

37

C. Two-photon Excitation Spectra

1. General Considerations

Two-photon excitation (TPE) spectrum can be obtained by (1) two-
beam two-photon excitation ((01 at (02), or (2) one-beam two-photon excitation
(col = 002). In method (1). the first photon, of energy (1)1, can be in resonance
with a real electronic excited state of the molecule; the two-photon cross
section is thus much higher, and this method strongly enhances the
emission intensity in comparison to method (2). The latter involves a
process in which the intermediate state is a virtual state (see eq 1-20). Also,
with the use of two different beams more information can be obtained, since
more combinations of the polarization of each beam can be applied.
However, the two-beam method could not be applied in experiments with
perylene because a series of red and near infrared tunable lasers (not
available in our laboratory) would be required for the tunable beam, 012, in
order to cover the two-photon region of interest, and the fluorescence signal
monitored is the emission from the first excited state of perylene (Figure 2-
1). If two beams are used to achieve the two-photon transition, the pump
beam would be in resonance with the first excited state, which emits
strongly. Thus it would be extremely difficult to distinguish the signal
which results only from the two-photon transition.

In general, a two-photon excitation spectrum is displayed as a TPE
profile, which consists of a plot of the two-photon induced fluorescence
intensity divided by the product of the powers of the two exciting photons
(method 1) or the square of the exciting laser power (method 2) versus

excitation energy. In order to perform polarization analysis, the TPE

38

Figure 2-1. Two-photon transition processes, where In is an intensity of
fluorescence from the two-photon induced state, and Ta is an

intensity from the one-photon induced excited state.

39

In

 

 

 

 

 

 

(a) (b)

One-beam two-photon excitation Two-beam two-photon excitation

Figure 2-1.

4o

profile must be measured under several different conditions of incident
laser polarization.

The experimental apparatus is schematically depicted in Figure 2-2.
A NszAG (Quanta-Ray DCR-2) pumped tunable dye laser (PDL-2) was
used as the excitation source. The Nd:YAG laser was operated at 15 Hz
with average pulse energy of approximately 1 mJ/pulse. The output from
the dye laser was completely linearly polarized with a Clan-Thompson
prism P1, and a set of three Fresnel rhombs (P2 and P3) were used to
prOduce linearly or circularly polarized light without changing the beam
intensity and position. The use of these rhombs to control the polarization
of the laser beam in the TPE experiment is described in detail in the section
on polarization measurement procedure. The outcoming beam from the
third rhomb was focussed onto the sample (1 cm pathlength quartz cell)
using a glass focussing lens L (f = 10 cm).

The fluorescence measurement sample chamber, shown in the
Figure 2-3, contained two quartz spherical lenses which collected the
emitted light at 90° from the incident dye laser beam (f = 10 cm each). These
lenses focused the two-photon induced fluorescence onto the entrance slit of
a Spex 1680A monochromator, M, after it passed through a combination of
UV-transmitting filters F2 (Corning 5-57 or 4-96). The wavelength of the
monochromator was set at 470 i 5 nm. The energy region selected contains
the strong 0 - 1 emission band (0 - 1298 cm’1 and 0 — 1369 cm“1
florescence). The signal was detected using a Hamamatsu R928
photomultiplier tube, D1, which is characterized by low dark current, fast
rise time (~2 ns), and high spectral response in the region of interest. The
PMT was mounted in a Pacific Instruments 3150RF housing at room

temperature and normally operated at 960 V. The intensity of each pulse of

41

Figure 2-2. A schematic diagram of the two-photon excitation of
fluorescence measurement. Here SH is a second harmonic
generator, M is a mirror, BS a beamsplitter, OP is opal glass,

and Osc. is an oscilloscope.

42

 

P1

 

inmmqul Nd/YAG laser J

 

 

 

 

 

 

 

 

Pam

L

 

 

 

L

{:5
O

m\111111

 

Figure 2-2.

 

43

Figure 2-3. The fluorescence measurement sample chamber for TPE.

 

 

 

 

 

 

 

Figure 2-3.

 

 

 

45

the exciting laser was measured by deflecting a small portion of the beam
with a beamsplitter through opal glass to a fast photodiode (Hewlett
Packard 5082-4200) after it had passed through the sample. The photodiode
is indicated as PD in Figure 2-2. This arrangement provided for accurate
relative measurement of the reference power, since less than 1 x 10‘5 of the
total number of photons in an excitation pulse are absorbed by the sample in
a virtual two-photon process. Before the collection of TPE data, the pure
solvent background was measured. Also, the dispersed emission spectrum
induced by the two-photon transition (Figure 2-4) and the necessary
quadratic intensity dependence of the TPE fluorescence on the exciting
laser power (Figure 2-5) were confirmed.

Under control of the trigger from the variable output of the Nd:YAG
laser, peak signals from the sample and the reference photodiode (i.e., the
peak height was assumed to be proportional to the peak intensity of the TPE)
Were integrated (gatewidth set at 10 ns) and 30 of these integrated signals
were averaged for each data point by the boxcar integrator/averager
(Stanford SR 232). To prevent signal ringing through the. cable connections,
50-9 terminators were first attached to the PMT output and the averaged
outputs from the boxcar integrator/averager. The averaged outputs from
the boxcar integrator were collected, and the signal from the sample was
normalized to the square of the exciting laser intensity by a personal
computer (PC). This procedure was followed because the signal from the
two-photon excited state (In) is proportional to the square of the exciting
laser power, <8> ~ <Ifl>/<I02>. Under computer control, the wavelength
setting of the dye laser was changed and the process repeated for each
excitation point of the spectrum within the tuning range of the dye. Nine

dye solutions were used to cover the spectral range from 560 to 840 nm.

46

Figure 2-4. Dispersed fluorescence spectra: (a) one-photon excitation, 16x
= 365 nm: (b) two-photon excitation, 1.6x = 584 nm.

47

 

470 nm
/ 470 nm

1 J

xex = 365 nm

xex = 584 nm

 

 

 

IntenSIty
g
‘2;

 

 

 

 

 

 

 

 

 

420 520 420 520
Wavelength (nm)

(a) (b)

Figure 2-4.

48

Figure 2-5. Power dependence of the two-photon induced fluorescence of
perylene (~10‘4 M in n-hexane) at two-different initial laser

wavelengths.

Logl

1.3

1.2

1.1

1.0

49

 

  
   
 
    
 
 

 

 

- 595 nm; Slope = 1.91
1
L 570 nm; Slope =1.85
1.3 1.4 1 5 1.6 1.7
Log!

Figure 2.5.

 

1.8

50

They were selected in such a way that the tuning ranges overlapped by at
least 5 nm. Since it is not possible to measure the absolute 8 values from the
fluorescence measurements, the partial spectra that were measured with
different dyes needed to be matched in the overlapping range. The ranges
covered by the dyes used in our experiment are shown in Figure 2-6. The
TPE spectrum measured was not corrected for variations in the
fluorescence quantum yield or the spectral response of the photodiode as a
function of excitation energy (Figure 2-7). Data from the literature[56, 57]
indicate that the quantum yield varies only little with excitation energy for
perylene in the energy range 435 - 250 nm.

. The two-photon excitation spectrum of perylene at 77 K over the 780 -
650 nm dye laser wavelength region was also obtained as described below.
The sample solution was prepared by using a high-vacuum cell consisting
of a 1 cm square quartz cuvette attached to a sidearm terminating with a 10
ml round bottom flask (Figure 2-8). These two parts were separated from
one another with a high vacuum valve. Perylene was placed in the cuvette
and isolated from the side arm with the valve. High purity isopentane was
degassed by freeze-pump-thaw cycles and transferred to the side-arm
round bottom flask by bulb-to-bulb vacuum distillation, and then mixed with
the perylene. To make a homogeneous glass solution, the sample cell was
slowly immersed in the liquid nitrogen contained in a silver-coated Dewar,
also shown in Figure 2-8, having three quartz windows located along
orthogonal axes (i.e., "entrance" and "exit" windows along an excitation
axis and one window along a "fluorescence monitoring" axis). Nitrogen
gas was blown across the surface of the windows to prevent moisture
condensation. The procedure for obtaining the 77 K spectrum was the same

as the room temperature measurement, except that the reference beam

51

Figure 2-6. The tuning range and relative output power of dyes used in
the two-photon excitation of perylene.

Relative Energy

Output

 

52

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LDS 320/821

 

 

 

 

 

 

1.05750
LD$751
i—Ql \
...._.I. L. .—
I-—--1 1
t l -
530— 590 520 coo 100 no no sea

Wavelength (nm)

Figure 2-6.

 

53

Figure 2-7. The spectral response of the photodiode.

Response

54

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2.7,

1.0 I
0.8 ‘...ooeoe+..
0.6 (a. ‘i
O. ...
0.4 .7. .7.
.: ,2
l
0.2 .t.
‘.
PhOtOns :
0.1 g:
.08 E
.06 Z
':
.04 a
: L
’ I
it . T
.01
400 600 800 1000 1200
Wavelength (nm)

Ellen-Guiana“.

55

Figure 2-8. Liquid nitrogen Dewar and High-vacuum cell for 77 K TPE

measurement.

56

$5 Espoe>hmmm

 

 

 

 

 

on: 8533/ on.

as 25E

EBaQ some”: Z 3534

 

 

 

OocoumouosE

 

 

57

intensity was monitored before the beam had passed through the Glan-

Thompson prism, P1.

2. Polarization Measurement Procedure

The polarization ratio of interest is the ratio of the two-photon cross—
sections for circularly and linearly polarized incident beams: Q = Scirl 51in.
To obtain the polarization ratio, 9, one has to measure the normalized
fluorescence intensity under both polarizations of the incident light, Ici,
and Ilin- Two different experimental methods can be used. One is the
simultaneous measurement in two different samples, and the second
utilizes two consecutive runs with the same sample, with the polarization
of the laser light changed between the two measurements[59-61]. In the
first case, all optical and opto-electronic parts of the experimental setup
must have the same. excitation energy (co) dependence, and in the second
case the light source and detector system must have a very good long term
stability. For the two-photon excitation 0 measurement of perylene the
second method was used.

The polarization changing device (Figure 2-9) consists of a fixed
Clan-Thompson prism (linear polarizer) P1, a double-Fresnel rhomb
configured such that it can be rotated around the optical axis of the laser
beam (achromatic 7J2 retarder) P2, and a fixed Fresnel rhomb (achromatic
M4 retarder) P3. The laser output fi'om the dye laser is completely linearly
polarized with the Clan-Thompson prism P1. If the double-Fresnel rhomb,
P2, is rotated around the axis of the light propagation, the vector of the
outcoming light is rotated with respect to the linearly polarized incident
beam. Therefore rotation of P2 changes the angle between the orientation of

58

Figure 2-9. The alignment of the polarization changing optics: here P1 is a
Clan-Thompson prism, P2 is a double-Fresnel polarization

rotator, P3 is a Fresnel rhomb, and M is a mirror.

 

 

 

 

60

the linearly polarized light and the optical axis of the third Fresnel rhomb,
P3. The light leaving P3 is linearly polarized every 450 increment of rotation
about the optic, and circularly polarized light is obtained halfway between
two positions which yield linearly polarized beams (Figure 2-10). To
confirm whether the light leaving the third Fresnel rhomb is completely
circularly polarized, the outcoming beam was reflected back through the
incoming path using mirror M (Figure 2-9), and the intensity of the
reflected image was monitored in front of the linear polarizer P1. When the
outcoming beam was completely circularly polarized, no reflected beam
was observed at position A. The polarization ratio, 9, was measured every
1 nm, and each point represents the average over 1000 laser shots at each

polarization.

3. Dye Laser Wavelength Scan

To scan the two-photon excitation spectrum the grating position of
the dye laser was computer controlled. The dye laser scanning procedure
is as follows. The variable output from the Nd:YAG laser triggers the
boxcar integrator to start collecting data. The busy output signal from the
SR232 triggers the boxcar interface. After the user-specified number of
busy output pulses are received, the interface sends a signal to the stepping
motor transformer. Then the stepping motor attached at the dye laser
scans the wavelength a preset amount, and data collection is initiated
again. To scan the stepping motor two signals are required. One is the

TTL pulse for the stepping and the other is a logic level high (5 V) or low (0

61

Figure 2-10. Representation of the electric field vector in space, (a) and
(b),and the projection, (c), of the electric field vector of the
initial incoming beam and the electric field after the rotator.
Here Z is the optical axis of the rotator, E0 is the incoming
beam strength; Ex and 1532 are the x and 2 components of the
electric field, respectively; E21 and E41 are the electric field
vectors of the outcoming beam from the rotator and the third
Fresnel rhomb, respectively, Z0 is the optical axis of the third
Fresnel rhomb; and a is the rotation angle of the rotator with
respect to the incoming beam and the optical axis of the third

Fresnel rhomb.

62

225° Rotation 45° Rotation

Z=Optical Axis Z=Optical Axis

E
22.5“ ‘
0

 

 

(3) 0° E

 

Figure 2-10.

63

V) for the stepping direction. The electronic circuit diagram of the stepping
motor transformer designed by Ralph Thiim[58] is shown in Figure 2-11.

II. RESULTS AND DISCUSSION

Assuming that perylene, l, in solution has D2h molecular symmetry

(the long and short in-plane axes being y (B2u) and x (B3u), respectively),

 

Perylene

the two-photon active electronic states that can couple with the totally
symmetric ground state, A; belong to the Ag, Blg, BZg’ and B3g
symmetries. Additionally, making the reasonable assumption that the

lower excited states which lie in the visible and the near UV region are m“,

Figure 2-11. The electronic circuit diagram of the stepping motor interface.

65

.2.“ «Sufi

 

 

 

. r ,l,
hum—99% c3951 3 «a r. 093.5
.53 an 558580
>3 3231 an 3 II 323% .2502
gee—ml S 3 .l com
H.” ,[l
EH
.4.
OH Au .0
w a a H H - a
6: mm : almoA a
H . v w
m Ana 2 .0 a «ml 1
N. m w N o m V m
e: a . a3 a
Na «a
«a
z a M241. .. a . . ya
a,” mm:
x a 38m? . 2 II V 2
a a v m a .0 a II .2
H p v a
.582 a. a . am am
, + a
>m+ >n 3% a
Q >9. £8325

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

66

i.e. having in-plane transition moments, the two-photon activity is further
restricted to Ag and B1g(xy) states (Figure 2-12).

The normal one-photon absorption spectrum of perylene in n-hexane
at room temperature is shown in Figure 2-13. The two-photon fluorescence
excitation spectrum, obtained under linear polarization of the incident
laser radiation is superimposed.

The UV-visible absorption spectrum of perylene has a strong first
absorption band (log a ~5) in the visible region (~435 nm), which has been
assigned to the A; (So) -' B;(Sl) transition, polarized along the long
molecular axis[50(a)]. The second strong absorption band (loge ~4) isin the
UV region (~250 nm) and has been assigned to the A; (So) -> Bags“)

transition, which is polarized along the molecular short axis. At the same
time, perylene has a fluorescence quantum yield of about ~1 over the entire
spectral range 435 - 250 nm in solution at room temperature. It is
noteworthy that the energy region below ~13,000 cm‘1 (laser wavelength
above 750 nm) does not produce notable two-photon absorption within our
experimental error limit, although the one-photon spectrum has strong
absorption in the region of twice this energy.

To reveal more detail, the TPE intensity and a point-by-point plot of
the polarization ratio (2 vs. incident laser wavelength are reproduced in
Figure 2- 14. In addition, the TPE spectrum of perylene in isopentane at 77K
over the two-photon energy region of 23,600 - 35,700 cm‘1 (dye wavelength
780 - 650 nm) is shown in Figure 2-15.

Seven major absorption maxima are observed in the two-photon
spectrum at room temperature, being located at 26,530 (754 nm), 27,860 (718
nm), 28,950 (690 nm), 30,300 (660 nm), 32,000 (625 nm), 33,900 (590 nm), and

67

Figure 2-12. The symmetries of the n-electron cloud of the ground state
A and the low-lying excited electronic states of perylene.

68

z (3m)

 

 

 

 

 

 

 

“‘1’st : : . 2Au'3322’
. 3B3g9 2Blu ; J 333? 2Blu
*

 

 

2A“,3B2¢, { 44"— l ‘ ‘ 2Au,3B2¢,

 

 

 

 

 

 

Ground State (Ag) Excited State (A, 3,“, 133,, 31,)

Figure 2- 12

69

Figure 2-13. One-photon UV-visible absorption and two-photon excitation

spectra of perylene in n-hexane solution.

7O

Ea? omv

 

d -

/

Rel. e

 

 

 

 

 

 

 

 

 

 

mm om
Mucmmuoam 58:98:

(gr/”1) 30'1 ‘Kusnezur 5m. 1921

71

Figure 2-14. TPE of perylene in n-hexane (10'4 M) and polarization ratio,
9, plotted as a function of the incident dye laser wavelength.

Rel. TPE Intensity, Log (In/Ii)

72

Two-Photon Energy/RX

36 34 32 30

 

[3] [2]

 

llllllllllllll

'I‘I’"""Y11r"""‘I'V'"""'I‘Y'Ir'

l

 

l

 

I

[1]

ll]:

 

 

1.5 [rulir1llvlurvlllII11ilIlitltjriIIIIIII]IIUI' I1

1.0 "'

 

 

05 l l l l l l 1

Figure 2-14.

 

 

o
———

 

73

Figure 2-15. TPE spectrum of perylene in isopentane at 77 K. The
polarization ratio over a somewhat narrower incident laser

wavelength range is also included.

Rel. TPE Intensity, Log (1“ H02 )

74

 

 

R 27,860 m'1

 

26,530 cm'1

I l L l

L

0.0

 

 

650 670

690 710 730

7» (nm)

Figure 2-15.

750

770

1.5

1.0

0.5

75

34,900 cm‘1 (573 nm); the two-photon induced fluorescence intensity is
strongly enhanced as one moves from lower to higher energies. The two-
photon polarization parameter, (2, starts with a value close to the
theoretical upper limit of 1.5[25] in the region of 26,530 cm'1 (754 nm) and
reaches a value of 1.35 in the vicinity of the maximum of the peak at 27,860
cm'1 (718 nm). The polarization ratio remains between 1.25 - 1.00 in the
region 23,170 — 32,790 cm'1 (710 - 610 nm). The a curve drops from ~12 to
~0.85 near 34,500 cm“1 (580 nm) and shows a distinct decrease below 570
nm, indicating either a superposition of two different symmetry transitions
or vibronic coupling. In the energy. region above 28,570 cm‘1 (700 nm), the
vibronic structure of the spectrum and the values of the polarization ratio
become complex.

On the basis of the intensity of the TPE and the value of the
polarization ratio, (2, it is convenient to consider the TPE spectrum in three
parts (Figure 2-14).

The first band system I I] in the twoophoton excitation (TPE) spectrum
of perylene, in the region 780 - 705 nm (25,600 cm-1 - 28,400 cm-1 two-
photon energy) shows distinct sub-structures: a sharp maximum at 758 nm
(26,380 cm-l); and a shallow maximum at 716 nm (27,900 cm-l). The
structure is better resolved at 77 K, where two relatively strong and
comparatively broad (~280 cm"1 FWI-IM) peaks are revealed at about 26,530
cm‘1 and 27,860 cm’l. The polarization ratio at 758 nm is close to the
theoretical limiting value of 1.5, which indicates that this peak is the result
of a transition from the A; ground state to an excited state of BL vibronic
symmetry. From the bandwidth and enhanced intensity of the peaks at
liquid nitrogen temperature, it is likely that the bands at 716 nm and 758

nm belong to the same electronic state.

76

The experimentally and/or theoretically obtained energy levels of the
excited states of perylene in the energy region near 26,500 cm’1 are depicted
in Figure 2-16 [50(a)-(c)]. Here B; is the one-photon allowed lowest singlet
excited electronic state (SI), B31 is the experimentally and theoretically
predicted second excited electronic state of perylene[50(a), 50(c), 52, 53, 61],
B'1'g is the predicted and observed lowest "g" state, and the dotted line
represents our observed peak at ~26,500 cm'l. The one-photon allowed
lowest singlet excited state, B3“, is observed at 23,000 cm‘1 (435 nm). It
seems highly questionable to assign the band at 26,530 cm'1 (754 nm) as a
vibronically induced 1321:1331; (= B 1g) false origin. In this case, it would be
necessary to assume for the inducing b3u vibrational mode in the excited
state a value of ~3,500 cm‘1 above the 23,000 cm"1 B31 origin, which is
unreasonable. A BYE allowed transition in this spectral region also seems
to be excluded. The experimentally-observed first Big state occurs at ~29,000
cm‘1[50(a)-(c)] although the two-photon data do not provide sufficient
information to confirm the reliability of those experiments. Also,
theoretical calculations predict the first B]: state at 29,000 — 30,600 cm‘l.
On the other hand, several calculations predict a B3, state at ~27,400 cm"1
[50(a), 50(c)]. From these facts it seems likely that the bands are
vibronically-induced transitions to the B31 state. The somewhat high
intensity for a vibronically allowed transition is probably due to the near-
resonance with allowed transitions.

Having an almost vanishing oscillating strength and being hidden
under the strong A; -* B; transition, the B31 state was not observed in the
conventional one-photon spectroscopy of perylene. However, the existence
of the A; -’ Baltransition in the absorption and photoelectron spectra of the

perylene was already inferred by Clar[53, 62]. A linear dichroism

77

Figure 2-16. Observed and calculated energy level diagram of the excited
electronic states of perylene near the peak at 26,500 cm'l.

78

 

Bi" (29,000-30,600 cm'l)

3;“ (27,400-2s,500 om“)

 

lllllllllIllIlllllllllllllllllllllllllll"IIIIIIIIIIIlllllllllllllllll 26,500 cm'l

~3,500om"

23,000 om‘1

 

Figure 2-16.

79

measurement on perylene embedded in stretched polymer sheets has
revealed a weak short-axis polarized contribution in the high energy region
of the first absorption band[50(a)]. An indication of the B31 state at «365 nm
is also found in polarization spectrum of perylene in the gas phase[52].
According to the pseudoparity selection rules of alternant hydrocarbon
theory (- -> - is forbidden for one-photon transitions)[54], a B31 state is
expected in two-photon absorption(A'g'-+ 3.3“) to be vibronically allowed.
Perylene (C20H12) has ninety normal vibrations which are classified

as follows:
l'vib: 16ag+ 15b1g+ 15b2u+ 15b5h+7au+8b1u+6b2g+8b3g

where as, blg, b2“, and b3,1 are the symmetries of the planar vibrations.
Because of the g-rg selection rule for two-photon absorption, the active
vibrations in the BE, electronic state will be of "u" symmetry. All such

vibrations: blu, b2“, b3u, and an my be active since

33.3%: blu' b2u’ an) = (Ag » 32g 31g, 33 ):

respectively. Vibrations of symmetries all and b1u are out-of-plane motions,
and can be ignored for the vibronic mirdng of 1m" states. On the basis of the
polarization ratio value in the region of the lowest-energy peak (~1.5), it is
likely that a higher excited B- ,g state lends intensity to the transition
through vibronic coupling. And the somewhat high intensity for a
vibrationally-induced transition is probably due to the small energy gap

between the 331 and B-lg states (Figure 2-16). Therefore, the active vibrations

80

for the vibronic coupling are likely to be of b2“ symmetry. From these
considerations, the lowest-energy peak in our two-photon spectrum, at
26,530 cm‘l, is assigned to a "false origin" of the symmetry forbidden
transition A;-+ B" , whose true electronic origin is located at a lower
energy. The observed transition terminates in a state of B1g vibronic
symmetry, lying above the origin by the energy of a b2u vibrational mode.
Bands of b2u symmetry ranging from 300-3,000 cm‘1 have been observed in
the ground state vibrational spectrum of perylene (see Table 2-1) obtained
from ir and Raman measurements[52, 63-65], and the B31 origin would
thus be placed in the 23,500-26,200 cm‘1 range.

The vibrational mode which contributes to the vibronically induced
transition at 27,860 cm’1 (718 nm) can be a b2u vibrational mode which has
an energy difference of ~1,330 cm'1 from the b2u vibration which induces
the first 37,.sz“ band at 754 run, an odd overtone band of b2u symmetry (33,"
hm“), or a combination band involving an ag (IE-31* bzu" ag) vibrational mode.
From Table 2-1, the first possibility can be excluded. If one of the overtone
bands of a b2u mode is responsible for the transition at 27,860 cm‘l, then it
cannot be a second overtone of b2u, which would have B;1 vibronic symmetry
[351* b2u” b2u = 3'5“] and thus be forbidden in two-photon spectroscopy
(Figure 2-17). Iszun is responsible for the observed peak, it must be a third
overtone {BE} bzu" bzu“ b2u = BL], and the frequency of the responsible b2u
mode should be half of the vibrational interval from the band at 26,530 cm"1
(754 nm) to the band at 27,860 cm'1 (718 nm), ~665 cm‘l. No b2u mode near
665 cm“1 is listed among the fundamental frequencies in Table 2-1.
Because'distortion of the ring can be ignored for the lower 1m" excited state
of perylene, the vibrational frequencies in the excited electronic state will be

little shifted. Thus, the possibility of a third overtone is also excluded. The

 

81

 

 

 

82 .32 an moo mg . meme
mm: .35 an moo he: 22 .22
SS .8: an moo no: mo: .3:
82 .3: one moo no: .63.: 3:
media We. oo .moo .ooo
mags o oo .moo .ooo 68:
Em .mmm an oo «mm 8m 58
m8 .mem no moo .ooo as am
3.8m “an oo .moo .ooo . :38
238 He oo .moo .ooo
one .mem n moo .ooo am am .88
En .mme e ooo an «me me.
«8 an. an moo .ooo
a... .8... Men moo .ooo 8m .93
En .mmn Mn moo .ooo
9.1.2 o ooo Be .3 8m
8m.mmm we moo .ooo menace m8
m9. .mmm an ooo 5.
8m 58 an moo .ooo
5 no... no ooo so new
one. anions;
.3 «93855 532m 23 no 8.8m .3 deg Bo .35 Emmoocoomouosrm

.ocoinom «a $188 momocoououm 355395 285-5 woes—330 98 25830 .né 035—.

 

82

 

 

 

88 .88 an no 88 .88
88 .88 an no 88 .88
88.88 on no 88
88 .88 Man ooo .88 .oo 88 88 .88
88 .888 We. moo .oo
88 .88 we _ moo .oo 88 88
88.88 a e moo .oo 88 .888 88
88 .88 8e. ooo .88 .oo
88 .88 man ooodoo .oo 88 .88
88 .88 We. moo .oo
88 .88 e ooo .88 .oo
:8 .88 an moo .oo 88 .88
88 .88 we moo .oo 8.88 .88 88
88 .88 e moo .oo 88 .88 88
88 .88 wan moo .oo 88 .88
88 .88 Mn oo . .
88 .88 e moo 88 .88 $8
88 .88 Man moo 88 .88
88 .88 8n moo
88. 883858
.3 .8585. baoéhm 05 «o 8.8m H3 .8855 Bo .35 83838335

 

8055889 .nfi 033—.

 

 

83

Figure 2-17. Two-photon transitions to various vibrational levels of a Bgu
electronic symmetry state. The right sides of (a) and (b)
represent the symmetry of the vibrational levels, and left
sides levels show the corresponding total vibronic
symmetries. A.g is the symmetry of the ground state of
perylene.

84

 

 

 

 

 

(a)

Figure 2-17.

 

 

 

29°
#8825;

\/
/\

 

 

 

 

.5".

(b)

85

band at 718 um should probably be assigned to combination of the (BE‘ b2u)

level with an a vibration, which likely corresponds to the 1,370 cm‘1

8
ground-state frequency (C=C, C=C-H ring stretching motion) measured
from Raman spectroscopy.

The TPE in region [2], 28,400 (705 nm) — 33,000 cm"1 (605 m), is
more intense than that in region [1] and continues to rise slowly. It
consists of a broad band with maxima at 28,990 (690 nm) and 30,300 cm"1
(660 nm), and a weaker shoulder at 32,000 cm‘1 (625 nm) having three sub-
structures, at 32,790 (610 nm), 323.5" (620 nm), and 31,750 car-1 (630 nm).
Above 28,400 cm"1 (below 705 nm) at 77 K, the vibrational structure and the
polarization ratio value in the spectrum become complex (Figure 2-15). It is
reasonable to assume that the discontinuity corresponds to the onset of a
different electronic transition. However, the mean value of the polarization
ratio stays approximately constant at ~1.2 throughout this region,
indicating that all the bands likely belong, whatever their origin, to the
same electronic system. These bands are not seen in the one-photon
solution spectrum in the same energy region. From this fact and the
relatively large intensity it is suggested that the observed bands are due to
symxhetry-allowed g -> g electronic transitions.

Several theoretical calculations place the lowest Big state at 29,000 -
30,600 (690 - 654 nm) cm'1[50(a)-(c)]; other BL states may be nearby[50(a),
50(c)]. We also suggest that BYE electronic states are responsible for the
fluorescence obtained following TPE in this region. At the origin, and for
other Blg vibronic levels, 0 should be 1.5. In-plane blg vibrations will have
Ag vibronic symmetry, for which 9 = 0.67 is expected[25]. In low resolution
spectra the measured polarization ratio will reflect the vibronic symmetry

of the specific state being accessed (3/2 or 2/3 in the case of perylene), plus

86

contributions from overlapping bands, as well as underlying levels built on
lower-energy origins, which have other symmetry. Thus deviation from
the mean (2 ~1.15 may be more diagnostic of the symmetry of a state
reached by two-photon excitation than the numerical value is. On this
basis, we tentatively assign the origin of the first ng electronic state to the
lowest-energy component of the 77 K triplet obtained with laser radiation
near 685 nm, at 28,950 cm‘l, where Q exhibits a local maximum.

The TPE spectrum at 77 K in the wavelength region below 705 nm
reveals several peaks [28,410 (704 nm), 28,650 (698 nm), 28,940 (691 nm),
29,150 (686 nm), and 29,410 cm"1 (680 nm)). Since two different electronic
states (B31 and 133;) contribute here, definitive assignments cannot be made
from our low-resolution spectrum.

From the remaining structure in the 31,750 cm"1 (630 nm) region, it
is not clear how to make an assignment from the observed polarization
ratio at room temperature; however, the presence of another BE state near
32000 cm‘l, where a polarization maximum is obtained, cannot be ruled
out.

In region [3], we observe a very strong band with maxima at 33,900
cm'1 (590 nm) and 35,100 cm‘1 (570 nm). On the long wavelength side of the
33,900 cm‘1 peak a shoulder is discerned. In the region of this shoulder, (2
drops down to ~1 and then rises up to 1.2 at the maximum near 33,900 cm’1
(590 nm). Regardless of the differing levels of sophistication, all
calculations predict a A; electronic state in the neighborhood of 34,000 cm‘1
[50(a)-(d)]; there may be other g-states nearby[50(a)]. CARS[50(d)] and
picosecond[66] dynamics measurements on perylene in the S1 (B'gu) state
both show a strongly-allowed one-photon transition to a (presumably A?

state which lies 35,300 — 38,400 cm‘1 above the ground state. On the basis of

87

the observed polarization ratio, however, the absorption maximum at ~590
nm should be assigned to a B1g vibronic symmetry state rather than to an
origin of Ag symmetry. Thus we place the origin of the lowest A; excited
state at 33,450 cm‘1 where a polarization minimum is obtained in n-hexane
solution at room temperature. The peak at 590 nm (~330 cm"1 higher than
the assigned A; state origin) may be due to presence of a big vibration in the
A; excited electronic state, corresponding to the big mode at 316 cm‘1 in
ground state. On the other hand, one calculation[50(a)] Predicts a B3; state
close to the A; state (~200 cm'l), so the polarization maximum near 33,900
cm"1 could signify the origin of a B; excited state. In the region of 35,090
cm“1 (570 nm), (2 drops from 1.2 to 0.8, indicating that most of the TPE
intensity above 570 nm is related to final states of symmetry Ag. However, it
is not clear whether the peak at 570 nm is due to an electronic origin of
symmetry Ag or to vibronic perturbation; there are several strong 1300 "
1600 cm'1 ag vibrations in the ground state. Approximately 330 cm"1
higher than this value there is local maximum in the Q-curve. Again, this
may be due to the presence of a blg vibration with frequency similar to the
ground state mode at 316 cm'l.

The results of this work are shown in Figure 2-18, and collected in
Table 2-2, where they are compared with values of perylene excited state
origins predicted by various calculations, or observed by one-photon
experiments in low temperature glasses, stretched films, etc. The
theoretical predictions are generally slightly higher than the experimental
values in n-hexane solution, but the correlation is surprisingly good. In
order to obtain better resolved spectra, which would provide more definitive

assignments and compare more directly to theory, we tried to obtain the

88

Figure 2-18. Energy diagram of the lower excited electronic states of
perylene.

89

 

40,000 cm"

A; C» 34,900 (:m'1

/ 3:36?) 33, 780 cm"
"""""""""" A; 33,450 cm"

""""""""" BI, (?)32,000 cm“

"""""""" BI, 28,950 em"

"""""""" 13;; 23,500 - 26,200 cm"

 

23,000 cm"

Figure 2-18.

Table 2-2. Comparison of Experimental and Calculated Values (cm-1) of
Transition Energies to Low-Lying Symmetry-Forbidden
Excited States from the Ground (A;) State of Perylene.

91

.onom uou 5 63m: 9828:: 3308235 05 .8.“ con/mm mm 8.25m o: ”Ssofiazmmofi SE mo 33m 628 Beams

2.53 are e

 

 

 

2838.8 23.. 2f?
5.53 8%” 8o .8 .653 83m 83m m< 38 us me + we.
2838.3” :eeasfie 8% amuse; +m<
28982.3 5538.2: m< 84.8 m<
H352 awaken ,
25833.8 .8898. 958.8 mm
:38. 2.98 :3 8.48 88.2 . 825 we .25 + mm
23832.8 ”8.8.33.8“..8 mm e88 mm
H352 omodm
8&8 me + 5: tmm
:88... ee + an tmm
28.0388 63°85 e38 antmm
6538.5. .3538qu .mm §q8-§.8 .mm
.0823. naaoatomxm anoaammmm< aaaoataxm acofinmmmfiw
acaonmdqo

 

2.3.562.

.N& 039—.

 

92

TPE spectrum of perylene vapor via supersonic jet spectroscopy; this is

described in Chapter III.

CHAPTERIII

TWO-PHOTON SPECTROSCOPY IN A SUPERSONIC JET

93

I. INTRODUCTION

One of the most powerful methods for investigating the microsc0pic
properties of a compound is molecular spectroscopy. In principle, the more
complex the molecular spectrum is, the more information about the
molecule it contains. However the spectrum of (a molecule can be so
complex that analysis is impossible.

The laws of quantum mechanics require that the frequency of an
absorbed photon is proportional to the difference in energy between the
initial state and the final state of a molecule. If the molecules are found in
many different initial energy states, the possible energy transitions can be
extremely large. On the other hand, if the number of initial energy states is
limited, the number of possible transitions is reduced. The most practical
way to limit the number of initial states is to lower the temperature of the
molecules. At low temperature, however, molecules tend to condense into
the liquid or the solid state, and the interactions of the molecules with one
another bring about significant changes in the spectrum. It is also
important to study the properties of the individual molecules, free from
interaction with their surroundings. Such a study is best carried out in the
gas phase. Therefore the problem is to make a gas cold enough to limit the
number of initial energy states without allowing the gas to condense and
form a liquid or a solid.

In order to study molecules in the gas phase at low temperature,
Levy, Wharton and co-workers used a refrigerant technique which cools the

molecules very quickly and then thins them out before they can

94

 

 

95

condense[69, 70]. The refi'igerant is an uncollimated, free jet of gas ejected
into an evacuated chamber under conditions where the flow is many times
faster than a sound wave in that medium; the technique is called
supersonic free jet spectroscopy; the speed of mass flow is larger than the
speed of sound not because the speed of the mass flow is large, but because
the local speed of sound (proportional to Tm) is so low. If the flow is
collimated with devices downstream from the nozzle, a supersonic
molecular beam is produced. Therefore the use of a supersonic jet or beam
allows the preparation of internally cold, isolated, gas phase molecules and
thereby retains the advantages of solid state or matrix isolation
spectroscopy without the disadvantage of molecular perturbations.

| As a molecular beam source, the supersonic jet was proposed by
Kantrowitz and Grey to increase the intensity of the molecular beam in
1951(71]. However, its implementation was prevented by insufficient pump
capacity. Then the realization that supersonic expansion may also be
useful in spectroscopic applications, due to the partial cooling of rotational
and vibrational degrees of freedom, was reported by several workers[72-73].
In 1974, Levy, Wharton and co-workers applied the supersonic jet technique
to molecular spectroscopy; they took advantage of the rotational cooling
effect of seeded continuous nozzle expansion to dramatically simplify the
fluorescence excitation spectra of N 02 (seeded in helium)[70].

II. SUPERSONIC EXPANSION

An effusive beam is one in which the hole diameter (D) is smaller

than the mean free path (10) of the molecules in the reservoir; no collisions

96

between gas molecules downstream of the orifice are present, and the
distribution of internal molecular states is identical with that in the
reservoir. The velocity distribution is given by the Maxwell-Boltzmann
distribution characteristic of the reservoir temperature (Figure 3-1). On the
other hand, in a gas that has expanded through a supersonic nozzle, where
the hole diameter is larger than the mean free path of the molecules in the
vicinity of the hole, there will be many collisions among molecules in the
vicinity of the hole area. The expansion converts the enthalpy associated
with random motion into direct mass flow and therefore causes the mass
flow velocity to increase. Moreover, the effect of a large number of collisions
in the orifice vicinity is to make the velocity uniform and cause the cooling.
The velocity distribution in the direction of the mass flow is also shown in
Figure 3-1. Because of the increase in the velocity, the peak of the
distribution shifis and because the gas has cooled, the distribution narrows.
In supersonic nozzle expansion the reservoir pressure or the orifice size is
increased to the point where D >> 1.0.

The feature of a supersonic expansion that distinguishes it from
ordinary cooling is the fact that the molecules of interest are only in contact
with the cold bath for a limited time, and therefore the final state of the
system is determined more by kinetics than by thermodynamics. Since the
cooling is due to kinetic phenomena, all particles will cool in such a free
expansion regardless of their ideality. In fact, the most ideal gas, helium,
was found to attain the lowest temperature yet achieved in a free jet
expansion, with a translational temperature of 0.006 K[74].

The importance of establishing a supersonic flow condition is that as
the gas propagates supersonically, a shock wave due to collisions, with

background gas molecules is established around the flow, Figure 3-2. The

 

Figure 3-1.

97

Velocity distribution in an effusive molecular beam(a), and a

supersonic beam(b).

98

 

 

 

 

 

 

Figure 3-1.

 

 

Figure 3-2.

Shock structure surrounding the expanding gas of a free jet.
Nozzle pressure is Po, the pressure in the expansion
chamber is P1, D is the diameter of the nozzle, Xm is the
distance to the Mach disk from the nozzle. From ref. [131].

100

Jet Boundary Barrel Shock

xix/fl

I / Reflected Shock

"U

0

D Isentropic Core <— Mach Disk

 

 

I

I
I

I
1

Figure 3-2. .

101

effect of this shock wave is to develop a region around the beam protected
from intrusion by warm background gas in the vacuum chamber. Thus
one can take full advantage of the cooling effects of the expansion for
spectroscopy. The components of the shock wave are referred to as the
barrel shock which extends outward from the orifice to the Mach disk,
which is a circular shaped area perpendicular to the flow direction.

A. Translational Cooling

In a supersonic expansion the molecules of interest, either pure or
seeded in a carrier gas, are expanded through a small nozzle. As the
expansion proceeds, molecules with small perpendicular velocities tend to
congregate close to the center line of the jet, while molecules with large
perpendicular velocities tend to move far away from the center line (Figure
3-3)[75]. Since the translational temperature is determined by the width of
the velocity distribution not by the position of the peak, the narrowing of the
velocity distribution which takes place in a supersonic expansion is called
translational cooling. Therefore, a cooling of the translational degrees of
freedom can be achieved without collisions, and the cold translational bath
acts as a refrigerant for the other degrees of freedom; however, the cooling
of internal degrees of freedom cannot be obtained without collisions.
Through binary collisions, various degrees of freedom (translational,
vibrational and rotational) are equilibrated with the cold bath by means of
cooling of the perpendicular velocity, which is transferred to the parallel

component via the collisions.

 

Figure 3-3.

102

The geometric cooling of the translational degrees of freedom

in relation to the expansion axis.

103

 

 

 

 

 

Figure3-3.

104

A supersonic expansion requires a reservoir with gas at a relatively
high pressure Po, a small orifice of diameter D, an expansion volume at
lower background pressure P1, and pumps to maintain the downstream
pressure in the presence of the gas discharge through the nozzle. By using
the assumptions that the expansion is isentropic (without viscous forces,
heat conductivity, shock waves, and heat sources or sinks such as chemical

reactions) and that the gas mixture is an ideal gas, the relations are[76]
T P
To ( Po) p0 ( )

where To, Po, and 00 are the temperature, pressure, and density in the
reservoir, T, P, and p are the same quantities in the beam and y is the heat
capacity ratio Cp/Cv. The Mach number, M, is the ratio of the flow velocity,
u, to the local speed of sound, a, where for an ideal gas the speed of sound is
given by a =[flCp-CV)T]1/2. For a continuous flow, the Mach number is

given by[77]
1
M = Mg)“ — (32)

where x is the downstream distance, D is the nozzle diameter, and A is a
constant that depends on y and is 3.26 for a monatomic gas (7 = 5/3).

Eqs 3-1 and 3-2 suggest that if one took a spectrum sufficiently far
downstream, an arbitrarily low translational temperature could be
achieved. However, there are fundamental limitations to the amount of
translational cooling attainable. As the expansion proceeds, the density of
the gas drops and eventually becomes too low to provide sufficient

collisions, and the Mach number and temperature are frozen from this

105

point on downstream. Anderson and Fenn[7 8] have predicted that the
terminal Mach number, Mt, is

1.
Mt = e (%)( WY (3'3)

where A-is the mean free path in the nozzle and s is a constant which is
characteristic of the gas in question. For a monatomic gas the terminal
Mach number, Mt, is a function only of the product (POD) and is, using an
experimentally determined value of s for argon, given by[78]

Mt: 133mm)“ (34)

In the case of helium, quantum effects cause the collisional cross-section
to increase with decreasing relative energy, and this leads to very much

larger terminal Mach numbers than those predicted by eq 3-4[79].

B. Vibrational and Rotational Cooling

As described above, the primary process in a supersonic jet
expansion is the cooling of the translational degrees of freedom. Near the
orifice, where the pressure of the gas in the jet is relatively high, the atoms
in the jet collide with one another continually. These collisions allow full or
partial equilibrium between the transverse velocity components and the
other degrees of freedom in the motion of the molecules (Figure 3—4).

As long as the molecules collide at low velocity with the cold atoms,

the energy of rotational and vibrational motion of the molecules is

106

Figure 3-4. Vibrational and rotational cooling via collisions in a jet.

107

 

‘F‘ Jet Centerline

 

 

 

 

 

Figtn'e3-4.

108

transferred to the translational cold bath. As the gas flows downstream
from the nozzle, the density of the gas drops and eventually becomes too low
to provide the collisions necessary to connect the internal degrees of
freedom (vibrational and rotational) with the translational cold bath. Thus,
cooling of the molecules caused by the exchange of energy in collisions is a
self-limiting process, and the degrees of freedom which equilibrate with the
translational bath are the coldest at the point where the expansion runs out
of collisions.

The actual state of the beam depends on the rate at which the
internal degrees of freedom come into equilibrium with the, cold
translational bath. The cooling attained in the expansion for any particular
degree of freedom depends on the collisional cross-section for relaxation for
that degree of freedom. For the rotational degrees of freedom this cross—
section is roughly equal to gas kinetic cross-section. In general, the rate of
equilibrium between translations and rotations is relatively fast, and
therefore extensive rotational cooling takes place before the molecules enter
the collision-free region. The vibrational degrees of freedom have
collisional relaxation cross-sections considerably smaller than those of gas
kinetics. Thus vibrational excitation is more difficult to cool in a free jet
expansion, and usually cooling occurs to a temperature below the original
temperature but above the final translational temperature. However, the
vibrations having the smallest cross-sections have a high vibrational
frequency and therefore have a small Boltzmann population in the excited
states.

In order to further cool a molecular gas of interest, a small amount
of this gas is mixed with a large quantity of a monatomic carrier gas,

which cannot store energy in internal degrees of freedom. Although the

109

lowest translational temperature from a pure expansion has been attained
with helium, it does not always produce the most efficient cooling of seed
molecules. For seeded beams with large differences in size and weight
between the seed (sample) and carrier molecules (e.g. mhav/mfim) ~50),
there is a velocity slip effect[80-83]. For large, heavy, rigid molecules
Amirav et a1. [82] found that cooling varied in the order Xe > Kr > Ar > Ne >
He. However, the frequency of three-body collisions, which lead to
condensation, depends on the mass and interatomic potentials of the atoms,
and lighter species have been found to create lower terminal temperature

because of the reduced rate of condensation[84].

C. Condensation and Complex Formation

The main advantage of the supersonic jet as a refrigerant is that the
molecular motions can be cooled for spectroscopic purposes before the
molecules condense. An important factor controlling the cooling in a
supersonic expansion is the possibility of condensation. Compared to
translational cooling, condensation is slow because its initiation requires
simultaneous collisions among three or more particles. Unlike cooling,
which requires only that energy be exchanged between two colliding
particles, condensation requires that at least two particles are bound
together. To form a bond, a third particle must collide with the first two
particles during the time of the two-body collision so that the third particle
can carry off the binding energy of the pair as translational energy. Three-
body collisions (roughly proportional to PozD [85]) require much higher

density in the gas than two-body collisions, and normally the cooling is

110

complete before a significant number of three-body collisions can take place.
Nevertheless, three-body collisions do occur; the heat of formation of these
complexes is liberated into the translational degrees of freedom of the
expanding gas, and this limits the cooling that is ultimately attained. One
explanation for the fact that helium attains lower energy in a free
expansion than any other gas might be that the ground state potential of the
Hez molecule is thought to be too weakly attractive to support a bound
state[86]. Therefore, condensation must initiate with the far more unlikely
occurrence of four-body collisions in the expansion to form He3[87].

The practical limit of the cooling in a supersonic expansion is the
requirement of adequate pumping capacity to handle the gas discharge
through the nozzle. Since the amount of complex formed in the expansion
is proportional to POZD, the cooling attained in the expansion is proportional
to POD, and the total mass throughput from the nozzle is proportional to P0
D2, in order to increase the cooling attained in the expansion by a factor of A
while ratio Pam is adjusted to keep a constant value of PozD (the amount of
complexes formed constant) the throughput would have to be increased by a
factor of (A)3[120].

To overcome the pumping limitation, pulsed nozzles and moderate
background pressure augmentation (Fla-10"2 to 1 Torr; Campargue-type)
have been used. In the first method, the gas flows only a small fraction of '
time, so the use of a pulsed nozzle increases the effective pumping speed in
a small apparatus. The effective pumping speed of the system will be the
true speed of the pumping system multiplied by the inverse of the pulsed
nozzle duty cycle. The second method was developed by Camparguef88-90].
The main idea is that the chamber pressure is increased at a constant

PolPl ratio, instead of decreasing the chamber pressure; pumps of a fixed

111

volumetric pumping speed have a mass throughput that increases linearly
with pressure. At higher background pressures, the deve10ping free jet is
encased in a shock-wave structure. The shock wave extends from the
nozzle tip to the Mach disk, and it is thick and dense enough to prevent
penetration of the core of the jet by warm background gas. By increasing
the pressure in the nozzle Po, it is possible to reach a mode of operation
where the expanding gas in the region upstream of the shock structure is
unaffected by the background gas and is equivalent to the same region of a
free jet expanding into an infinite vacuum. This is true even though the
increase in the nozzle pressure has brought about a corresponding increase
in the pressure in the expansion chamber. In contrast to the Campargue-
type expansion, the Fenn-type[91, 92] expansion utilizes a low background
pressure (10'3 to 10'6 Torr). Due to the decrease of the collision frequency in
a jet of relatively low density, the expansion becomes frozen at a distance
from the nozzle much smaller than the distance Xm, at which the Mach
disk of a free jet shock wave structure would be formed. Nevertheless,
downstream distances where significant “molecular cooling has occurred
are accessible to interrogation by focussed, transverse laser beams. All the
measurements described in this dissertation were taken under low

background pressure (P1 <10'3 Torr) conditions.

III. SPECTROSCOPY IN SUPERSONIC JE'IS

Supersonic jet spectroscopy has been applied to the study of a various
molecules, from diatomics to large organic molecules such as

porphyrins[93]. Although detection methods and the molecule sizes vary

112

somewhat, laser spectroscopic studies in supersonic jets fall into four
categories: 1) Studies of molecular spectra to provide information about the
molecule itselfI94-98]; 2) Studies of the physical phenomena of supersonic
expansion[99-104]; 3) Dynamic studies of intramolecular dynamics and
radiationless processes[105-112]; and 4) Investigations of formation and
structure of molecular complexes, such as van der Waals and hydrogen-
bonded complexes[114-119].

The first two categories, 1 and 2, are generally applied to small-size
or medium-size molecules, for which we can obtain resolved rotational
structure in the electronic spectrum; two-photon spectroscopy of perylene is
in this first category. Taking advantage of the collision-free conditions and
low temperatures, categories 3 and 4 represent more recent, and very
promising applications of supersonic jet spectroscopy. Subjects of such
experiments include energy flow from one particular state to another,
preparation of an intense beam of highly reactive and unusual free
radicals, and formation of highly selective photofragments. Also, if the
molecules of interest are cooled sufficiently, then van der Waals complexes
may form which are too weak to form stable bonds under ordinary
laboratory conditions. In this dissertation studies in the latter two

categories will not be reviewed.

A. The Number of Excited Molecules

When a molecule is mixed with monatomic gas at a number density

ratio c, and expanded from a supersonic nozzle of diameter D at a pressure

113

P0 and a temperature T0 the number density of the molecule in the nozzle is

Po
po = c FT: (3-5)

where k is Boltzmann's constant. As the gas expands into the vacuum

chamber, the density falls as[77]
2
p on po (1%) , (3-6)

where x is the distance downstream from the nozzle. If a laser beam
crosses the molecular beam at a distance x downstream of the nozzle, a
column of excited molecules will be produced in the molecular beam. The

column length L is given by[120]
L on x (3-7)

If the laser pulse duration is much shorter than the time required for
a molecule in the supersonic beam to pass. across the laser beam then,
neglecting saturation effects, each laser will produce a total of Ne excited
molecules[120],

Ne=LpEO (3-8)

where p is the number density of the absorbing molecule, E is the number of
laser photons per pulse, 0 is the absorption cross section of the molecules
for these photons, and all possible decay routes out of the excited state have

been neglected. Thus,

114

1
N, a 3;; 69)

Ne depends only on the inverse power of the distance from the nozzle even
though the jet density falls as x‘z. Usually NO is sufficiently large, even for
moderate laser powers, for spectroscopic measurements to be made.
Therefore there are no advantages obtained by decreasing the laser beam
size, and in fact this can only increase the possibility of saturation. For
long pulses or continuous lasers, however, if the laser beam is focussed at
the interaction region with the molecular beam then the time dependence

and coherence of the excitation can be reduced.

EDechiion

Although a jet-cooled molecule is collision-free and in a well defined
state, the concentration of molecules is so low that normal spectrosc0pic
methods such as direct absorption spectroscopy are difficult to apply. As an
indirect method, molecular fluorescence (total undispersed and dispersed
fluorescence) has been the principal method for detecting the formation of
excited states and for characterizing the structure and time evolution of
those states in supersonic jets. In laser fluorescence excitation
spectroscopy, which was applied for all measurements described in this
chapter, the wavelength of the exciting light is tuned and the total
undispersed fluorescence is collected. In dispersed fluorescence
spectroscopy, the intensity of the emission is recorded as a function of the

wavelength of the emitted light. In contrast to the fluorescence spectrum of

115

molecules in a matrix, that of an isolated molecule can be strongly
dependent on the particular excited state that is originally populated when
the laser light is absorbed.

However, even with carefully designed detection systems only ~5 % of
the fluorescence can be collected, and to obtain this value a detector with
excellent quantum efficiency is required. For molecules which do not
fluoresce with good quantum yields within a sufficiently short time to be
detected (with a space-fixed detector), such that the excited molecule passes
‘ out of the field of view, other detection methods must be used. Basic
requirements for laser induced fluorescence (LIF) analysis of large
molecules are considered to be an emission quantum yield exceeding 10'3,
and a volatility such that a vapor pressure > 10’5 Torr (10"7 Torr for a slit
nozzle) can be attained[121].

One approach to avoiding these limitations is mul ti photon
ionization(MPI). In the simplest case MP1 is an n-photon absorption

process leading to ionization and obeys the condition
“Em > IP (3-10)

where Eph is the photon energy and IP is the molecular ionization potential.
The signal is provided by ions generated by multiphoton ionization of a
molecule, and ions are detected as the final product in a mass
spectrometer. Thus the ionization cross section reflects the absorption-
excitation spectrum of the intermediate states. However, such a process is
entirely nonselective; any sample which can satisfy eq 3- 10 will be ionized.
When one of the absorption processes terminates in a real excited state of

the molecule, the ion signal intensity of that molecule can be greatly

116

enhanced (see eq 1-20), and through the use of a time-of-flight mass
spectrometer (TOFMS, in which ions starting at different points will reach
the detector at different times) in conjunction with the jet this problem can
be solved[121].

Direct absorption can be utilized as a detection method for molecules
in supersonically cooled samples, although the sensitivity is low[122-124].
Under normal expansion conditions, the concentration of molecules in a jet
is extremely low (< 1015 molecules within the laser/jet intersection) and the
optical pathlength of the jet available for the absorption measurement is
generally too short to make the method viable. To improve the number
density of molecules, Jortner and co-workers[122] developed the planar
supersonic expansion technique, in which a pulsed slit nozzle (90 mm
length, 0.27 mm width) is employed. The jet emanating from the slit,
produced by the supersonic expansion, provided sufficient optical
pathlength for absorption measurements in jets, and increased the
sensitivity by up to two orders of magnitude.

Recently, highly excited electronic 'states of molecules have been
extensively studied by the use of double resonance excitation. A jet-cooled
molecule is excited to an allowed excited state by the absorption of one
photon of laser frequency v1, then the excited state is further promoted to a
higher state by the absorption of another photon of frequency v2. Various
methods for the detection of highly excited states by double resonance
excitation are shown in Figure 3-5. The principles, advantages, and
disadvantages of these double resonance spectroscopies are described well

in reference [125].

Figure 3-5.

117

Various two-beam double resonance spectroscopies: (a) two-
color multiphoton ionization spectroscopy, (b) two-color MPI
assisted by autoionization, (c) two-color MPI assisted by
collision, (d) two-color fluorescence dip spectroscopy, (e) two-
color ionization dip spectroscopy, (f) stimulated emission
pumping, (g) stimulated emission spectroscopy utilizing the
ion dip technique, and (h) population labeling spectroscopy.
Here X is the ground electronic state, A and R are resonant
excited electronic states, and I marks the onset of the

ionization continuum. From ref. [125].

118

autoionfi‘zation
V V
R or 2 I
V2
A A
# v 1
X X

ion
V2
V1
9.) (1))
ion
2
V1
(c)

H

(
m. I W

collmon R

0 R
1L1 v V2
*1 by A A
:wo- l V1
fluo.dip

sion X ((1) X

the ion dip
my. I
R
,ant v2 V1
the A
i
X
(6)
ion dip

    

 

 

Figure 3-5.

119

IV. LIMITATIONS OF SUPERSONIC JET SPECTROSCOPY
A. Vaporization

The jet technique requires the vaporization of the molecule, yet many
molecules of interest have very low volatility. Such molecules are often
thermally labile and cannot be vaporized by the simple expedient of
increased heating. Several techniques have been developed to achieve the
vaporization of large molecules, such as laser desorption[121] and
supercritical fluids[126, 127].

I Laser desorption involves the use of a pulsed laser to induce rapid
heating effects such that the molecules desorb from a surface before they
have time to decompose. In the laser desorption process, both ions and
neutrals are formed, in a ratio that depends on the surface temperature
induced by the laser. Thus the power density of the laser at the surface is
the most critical parameter in desorption efficiency, and the desorption
process is not strongly wavelength dependent. The limitation of laser
desorption is the formation of ions. However, often desorption conditions
can be arranged so that primarily neutral molecules are desorbed and then
entrained in a supersonic flow. A native chlorophyll with relative
molecular mass of 892 was obtained by Grotemeyer et al.[128], and adenine
and cytosine were were studied by Tembreull and Lubman by laser
desorption-jet expansion-MPI[121]. Figure 3-6 shows a schematic
representation of the experimental set-up for the laser desorption-jet
expansion-TOFMPI apparatus.

Another method of introducing low volatility, thermally labile species

into jets involves supercritical fluids, where organic solvents are used as

Figure 3-6.

120

The experimental set-up of a time of flight mass spectrometer
coupled with apparatus for laser desorption into a supersonic
jet followed by laser MPI. Here CV is a gate valve, and LN2 is
a liquid nitrogen trap. From ref. [121].

 

121

 

* to Monitor

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure3-6.

122

carriers. The solubility of large molecular weight organic species is
considerably higher in such a fluid than in a rare gas. Additionally, in the
expansion process the supercritical fluid carrier will itself rapidly vaporize
as the pressure drops from high pressure (~100 atm) to low pressure (< 1
atm). However the cooling in a supercritical fluids is considerably'less
than that in rare gases, and it is dificult to remove solvent molecules from
the sample cluster. Polyatomic molecules have internal degrees of freedom
which need to be cooled, and the high pressure differential required for the
expansion to obtain the lowest translational temperatures is not within the

experimentally convenient range.

B. Sensitivity

Although the supersonic expansion places the molecule to be studied
in a collision-free environment and in well defined rotational and
vibrational states (even for large molecules), the number density of the
molecules is very low, which limits jet spectroscopy. Detection limits in the
10-100 ng range for naphthalene (1013 — 1014 molecules, = ~10'4, 4),, ~ 0.2)
were demonstrated by Hayes and Small[129], who used a continuous jet.
Perylene (f = 0.332), with its near-unity quantum yield ca, lies in the high-
sensitivity end of the above range. Pepich et al.[130] coupled a pulsed
expansion to CC and laser induced fluorescence detection, and molecules
with fluorescence quantum yields similar to the molecules studied by Hays
and Small could be detected in the picogram range.

123

V. EXPERIMENTAL

Figures 3-7 and 3-8 are schematic block diagrams of the instrument
used for our experiment. The experimental components used to observe the
signal from molecules expanded in a supersonic jet can be separated into
three main parts: the apparatus for generating the jet, the laser system for
exciting the gas molecules in the jet, and the detector system utilized to
obtain the signal from the excited molecules. (In our experiments,

fluorescence from the laser-irradiated molecules was detected.)

A. Jet Generating System

The jet generating system consists of a sample compartment and a
vacuum pumping system. The sample compartment is composed of a
sample cell, a heating device, and a nozzle (Figure 3-9). The carrier gas
(argon) at pressure P0 (0.5 - 2 atm) is passed through a quartz sample tube,
C, packed with solid sample 81 (iodine, perylene, or [2,2]paracyclophane),
and the resultant gas mixture is expanded into the vacuum chamber
through a supersonic continuous nozzle, N1. To prevent dispersion of the
unvaporized solid sample into the vacuum chamber, the sample is packed
with glass wool. To produce the jets, perylene (Aldrich, mp 277 - 279 °C)
and [2,2]paracyclophane (Aldrich, mp. 285 - 288 °C) were heated to 220 °C
‘ (60 V variac setting) and 150 °C (50 V variac setting), respectively. The
temperature is monitored by an iron-constantan thermocouple (Omega J
type, TC in Figure 3-8, with an upper temperature limit of 600 °C) attached

to the surface of the heating tape. The continuous nozzle N1 is placed

124

Figure 3-7. Schematic block diagram of the supersonic jet experiment.
' Here B are bafiles, T is a telescope, and BD the beam dump.

125

Trigger

 

 

__1

ummmmul {)ye Il[ Nd/YAG laser I
aser Sample

 

 

 

Carrier gas
PMT
T B

 

 

A

Chamber

 

s
sf“ \
s“ .
\s‘ Experimental
% Pump

 

 

 

 

 

 

 

 

 

 

I Plotter l

 

Figure 3-7.

126

Figure 3-8. Schematic diagram of the chamber employed for supersonic

expansion experiments.

 

127

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

nail—"d
*----------

 

 

b-------------------------Q

i .

Interlock System

 

Figure 3-8.

Figure 3-9.

128

A side-view cross section of the sample compartment. C is
the sample chamber, SI the sample, 82 the sample cell.
holder, N1 the continuous nozzle (comprised of an
interchangeable pinhole orifice), N2 the nozzle holder, G is
glass wool, and M a magnet.

129

 

 

 

 

‘IIII
Q
‘0
:-

1

'\

I ‘ g ”' '1.)
OBS". "“9102“
«‘2‘ (0) M

O
n
o

1

Figure 3-9.

130

between the nozzle holder and sample cell holder. The nozzle holder, N2,
shown in Figure 3-9 is made from stainless steel with a round-shaped end
plate having an ~3 mm diameter hole and a 30° conical tip. The holder
threads onto sample cell holder Sz, which is made from a length of
stainless steel tubing, also having an approximately 3 mm diameter hole.
A magnet M, which holds the nozzle N1 in place, is mounted on the exterior
of the sample holder. Various diameter nozzles (15-1000 um) purchased
from Ealing are used for different experiments. All the measurements
described in this Chapter were made with a continuous nozzle. The XY Z
translator shown in Figure 3-8 is used for smooth translation of the nozzle
in all directions to permit the supersonic jet to be centered, and to control
the distance from the nozzle to the laser intersection point while the
experiment is in progress.

The expansion chamber is evacuated by Varian VHS 10" oil diffusion
pump with pumping speed 5300 Us for air, backed by a 500 l/min
mechanical pump (Varian Hi-vac-45). In our experiments typical
background (chamber) pressures of < 10'4 Torr are attained with the jet
running (without jet : ~2 x 10'6 Torr). The pumping system is protected by
an interlock system which monitors the flow rate of cooling water (set at 1.5
1/min.), the foreline pressure (set at 600 mTorr), and the air pressure on the
gate valve (set at 75- 120 lb). Ifthe preset conditions are not maintained, an
electropneumatic gate valve automatically seals the vacuum chamber and
disconnects the power to the diffusion pump.' The interlock control circuit
was designed and implemented by Marty Rabb, the MSU Chemistry
Department's Electronic Designer.

131

B. Excitation Soume

A Nd:YAGpumped dye laser was used as the excitation light source
for fluorescence excitation spectra of jet-cooled perylene and iodine. For
[2,2]paracyclophane, the output from the dye laser was frequency doubled
by a doubling crystal (Spectra Physics WEX). The dyes R590, S420, and
DCM were used for iodine and two-photon excitation of perylene, for one-
photon excitation of perylene, and for [2,2]paracyclophane, respectively; all
dyes were purchased from Exciton. The procedure used to scan the dye
laser wavelength is described in Chapter II. At the same time that the
wavelength of the dye laser is stepped the Q-switch output from the
Nd:YAG laser triggers the WEX to rotate the doubling crystal. The scan
rate of the PDL must not exceed 0.05 nm/s. However, if the scan rate is
slow, e.g. a PDL scan rate less than 0.1 nm/min, then the system loses its
feedback lock, which will result in the detuning of the crystal due to
ambient temperature change.

To minimize the extent of laser light scattering into the fluorescence
detection system, the output from the dye laser or WEX is focussed into the
vacuum chamber with a telescope; the laser beam has ~1 mm diameter as
it intersects the supersonic jet. Also the laser beam is baffled within the
entrance and exit arms, each of which carries a quartz window set at
Brewster's angle at its extremity. The purpose of the baffles (B in Figure 3-
7) is to reduce scattered laser light. The baffle apertures define the laser
beam which intersects the sample beam; uncollimated laser light reflects
from the baffles rather than hitting the nozzle and being scattered. The
Brewster angle windows help reduce reflected laser light from the exit and

the entrance windows, and also remove any unpolarized portion of the laser

132

beam since only parallel polarized laser beams would be transmitted by

windows set at this angle.

C. Detection

The undispersed total fluorescence from the sample was collected for
all measurements reported here. Figure 3-10 shows the detection system
used to collect and detect the fluorescence from the laser-illuminated free
jet. Lens L (d = 5 cm, fl = 5 cm, glass) collected the fluorescence from the
sample and focused it onto the photocathode of an electronically cooled
Hamamatsu R562 photomultiplier tube. The PMT is mounted in a
magnetically and electronically shielded Pacific Instrument housing, and
was normally cooled to -20 °C. The distance between the collecting lens
and the crossed beams can be adjusted (from 45 to 75 mm) from inside the
vacuum chamber. Since the largest source of noise was from scattered
laser light, various filters were placed in front of the PMT. A slit in front of
the PMT was used to transmit the fluorescence and to block the light
scattered from other parts of the vacuum chamber. (To more effectively
reject light scattered from other parts of the apparatus and fluorescence
from molecules not in the cold portion of the molecular beam, an imaging
optical system can be used[131].) In order to normalize the fluorescence
signal to the exciting laser intensity, the laser beam reflected from the
Brewster angle entrance window was collected by a photodiode, PD. [See
Figure 3-7.] The remainder of the data processing procedure was the same

as that utilized in the two-photon excitation fluorescence measurement of

133

Figure 3-10. Schematic diagram of the detection system. W represents a

glass window.

134

PMT
W

1.25m I

 

l

 

 

 

 

V

CHAMBER

1

 

 

1.3-3.0 in.
S)

 

 

Figure 3-10.

135

perylene described in Chapter II, except that the signal was normalized to
the laser intensiton for one-photon excitation, instead of to 102.

A typical experimental procedure can be described as follows: 1) The
laser beam is aligned at the center line of a nozzle. 2) The vacuum chamber
is sealed and the mechanical pump starts to evacuate the chamber. Only
valve 1 in Figure 3-8 is opened. 3) When the pressure in the chamber
reaches < 50 mTorr, valve 1 is closed and valve 2 is opened to evacuate the
diffusion pump. 4) After the pressure inside the diffusion pump falls below
50 mTorr, the cooling water of the diffusion pump is circulated and the air
pressure of the gate valve is set between 75 - 120 lb (above 125 lb the piston of
the gate valve can be damaged). 5) The diffusion pump is turned on and
. the gate valve is then opened. If preset conditions of the interlock are not
maintained, the diffusion pump power will not be connected. 6) When the
pressure in the vacuum chamber is < 10"5 Torr and stabilized, the jet is
generated. 7) The laser wavelength is set at a known peak position of the
sample and the carrier gas pressure is changed until a signal from the
laser-irradiated molecules is detected with the PMT. 8) The conditions of
the jet and the detection system are optimized; to prevent hitting the nozzle
surface with the exciting laser beam, all the measurements are started at
the minimum value of the distance between the nozzle and the laser/jet

intersection region, and the distance is increased.

VI. RESULTS AND DISCUSSION

To utilize the home-built supersonic jet (for which plans were drawn

by Tun-Li Shen of Prof. Leroi's group), we tested and characterized the jet

“Wm-Jr”- ' 'H
-. !

136

system with one-photon spectra of iodine, perylene and
[2,2]paracyclophane. Since the translational, and to a slightly lesser extent
the vibrational and rotational, cooling depends mainly on the reservoir
pressure and diameter of the nozzle, and on the distance from the nozzle to
the laser intersection, these parameters in particular were examined.
Also, two-photon excitation of perylene with a single laser was attempted.
Measurements reported in Chapter III monitored undispersed fluorescence
as the signal; all spectra are plotted as the normalized total fluorescence vs.

the incident dye laser wavelength.

A. Pressure Efi'ect
1. Chamber Pressure

Figure 3-11 shows a portion of the fluorescence excitation spectrum of
a supersonic jet containing molecular iodine (sample tube held at room
temperature, vapor pressure ~0.2 Torr) over the 578 - 596 nm excitation
range (X12 -’ B311). In scan (a) the Ar carrier gas pressure is 0.5 atm, the
nozzle diameter is 100 um (Km 2 100) and chamber pressure is 2 x 10‘4
Torr. The upper portion of this figure, scan (b), shows the LIF (laser
induced fluorescence) spectrum over the same spectral range at a chamber
pressure of 2 x 10"1 Torr. The translational temperature obtained from eq
3-1 was 1.4 K. From spectrum (a) we can deduce that the rotational
temperature is nearly equilibrated with the translational temperature.
From the relative Frank-Condon factors reported in the literature[132] the
vibrational temperature is calculated to be about ~150 K. When the

137

Figure 3-11. The laser induced fluorescence spectra of jet-cooled iodine
molecules at difi‘erent chamber background pressures. (a) P1
= 2 x 10-4 Torr. (b) P1 = 2 x 10'1 Torr.

138

 

 

 

 

 

 

Ll inrrllllllrllmml

596 594 592 590 588 586 584 582 580 578

 

Wavelength (nm)

Figure 3-11.

139

chamber pressure is increased to 200 mTorr (mechanical pump only)
under otherwise the same conditions, the spectrum becomes congested and
the unresolved vibrational and rotational structure reproduced in (b) is

measured.

2. Carrier Gas Pressure

The effect of the carrier gas pressure was measured with perylene
over the dye laser spectral range 417 - 413 nm. The laser induced
fluorescence excitation spectra of perylene at different argon carrier
pressures (P = 150 - 760 torr) are shown in Figure 3-12. The carrier gas
pressure was increased from traces (1) to (5). The nozzle diameter is 100
um and the laser beam crossed the supersonic jet ~7 mm down stream
from the nozzle, X/D ~70. By increasing the carrier gas pressure, the
cooling of the vibrational and rotational degrees of freedom is increased and
the spectrum becomes better resolved. On the other hand, the number
density of perylene molecules is decreased, and as a result the intensity of
the signal is decreased. The most intense peak, at 415.7 nm, is the 0-0
origin of the Ag 4 B3,1 transition of perylene. Above a certain argon carrier
pressure, the appearance of new spectral features is observed, exemplified
by the marked peak in trace (5). This additional feature, appearing on the
low energy side of the 0-0 transition of the bare perylene molecule (~50 cm‘“1
lower than the 0-0 transition) is assigned to a vibration in the perylene-Ar
1:1 van der Waals complex[133]. Generally, the position of the peaks
attributed to van der Waals complexes depend on the identity of the carrier

gas (the positions of fluorescence from isolated molecules are independent

 

140

Figure 3-12. The laser induced fluorescence spectra of perylene in a jet as
a function of the carrier gas pressure. The argon carrier
pressure is increased from (1) to (5). The arrow represents a

vibration of a perylene-Ar vdW complex.

 

 

Intensity

141

 

   
 

 

4“.-

Baseline
l l | l

 

 

Wavelength (nm)

Figure 3-12.

418

142

of the carrier gas) and the relative intensity of peaks attributed to (carrier
gas)n-sample vdW complexes exhibit a dependence of the form P2n on the
carrier gas pressure P. Above an argon backing pressure of 760 Torr, with
a 100 um diameter nozzle, the pumping speed of the pump could not handle
the throughput from the nozzle, and backstreaming of diffusion pump oil

occurred.

BDishneeEfl'ect

The energy of rotational and vibrational motion of a polyatomic
sample molecule is transferred to the translational cold bath through
collisions. As long as the molecules make a collision, the number of
collisions is a function of the distance from the nozzle to the laser cross
section, X; the cooling of the internal degrees of freedom increases as the
distance increases. The LIF spectra of perylene plotted in Figure 3-13 were
obtained at different dietances X, between 5 mm and 10 mm (D = 100 pm).
The intensity of the scattered laser beam is increased by decreasing the
distance, but the signal is increased because the number density of the
sample also increases. X/D = 50 is the minimum distance under our
experimental conditions at which the scattered laser light could be
sufficiently discriminated. From the spectra shown in Figure 3-13 the
dependence of sample cooling on the distance, X, is not clear; there is little
difference among the three traces. The likely reason for this observation is
that cooling is already complete even at the shortest distance (M = 50)

accessible in these measurements.

 

143

Figure 3-13. Laser induced fluorescence spectra of perylene in a jet at
different distances from the nozzle to the laser intersection;
X/D = 100(a), 70(1)), and 50(0). The arrow designates a hot
band.

 

144

 

0-0

 

 

Intensity

 

 

 

 

 

 

415 416 417

Wavelength (nm)

Figure 3-13.

145

C.NozzleSizeEflect

The number of collisions is proportional to the product of stagnation
pressure Po and the diameter of the nozzle D, POD. The effect of the nozzle
size was measured by using various nozzle diameters (50 and 100 pm),
under constant backing pressure Po -.- 0.5 atm. Under otherwise identical
experimental conditions, as the nozzle size is increased a better resolved

spectrum is obtained (Figure 3-14).

D. Sensitivity

As described above, the efi'ect of the scattered laser light was serious
for the desired quantitative measurements. However, the one-beam two-
photon excitation measurement of jet-cooled perylene may be possible .
despite the scattered laser light, providing the two-photon cross-section is
sufficiently high. Unlike the one-photon excitation of perylene, which
involves resonance emission from the excited state, the fluorescence from
the two-photon induced excited state has higher energy than the exciting
laser beam. Thus the scattered laser light can be reduced by filtering
methods (see Chapter II).

The number of perylene molecules entrainable in a jet is very low,
even when a large nozzle diameter is employed; a nozzle diameter of 200 pm
(the maximum diameter for continuous expansion attainable with our
pumping system in order to maintain a pressure of < 5 x 10‘4 Torr in a jet
with 1 atm argon expansion pressure) at a sample temperature of 220 °C

provides ~1012 molecules in the laser/jet interaction cross-section. The

146

Figure 3-14. The effect of nozzle size on the internal temperature cooling:
iodine molecules expanded through nozzle diameter D = 100
pm (a) and 50 um (b), over the excitation region 57 8 - 586 nm.

 

147

 

(b)

 

 

 

Intensity

(a)

 

 

 

 

l l l l 1 LI 1 l

 

585 583 581 579

Wavelength (nm)

Figure 3-14

148

number of perylene molecules in 10‘4 M solution within the laser cross-

014 molecules;

section under the experimental conditions of Chapter II is >1
the number of molecules in the jet is thus at least two orders of magnitude
lower.

To test the sensitivity of our detection system, the one-photon
excitation spectrum of [2,2]paracyclophane (PCP) was obtained over the
excitation range 330 - 280 nm (Ag -+ Blg)' The oscillator strength of PCP is f
= 0.001 at 310 nm[134] and that of perylene is f = 0.332 at 415.7 nm[50(a)]. A
portion of the excitation spectrum for LIF from supersonic-jet expanded
PCP is shown in Figure 3- 15. (The inset is an absorption spectrum obtained
at room temperature in isopentane solution; the arrows denote the spectral
region reproduced in Figure 3-15 for the LIF of PCP' in the jet.) PCP was
heated to 150 °C (1 Torr, variac setting 50 V). The nozzle diameter was 200
pm, the argon carrier pressure was 1 atm and the PMT power was 760 V.
In the wavelength region above 300 nm in the jet spectrum, the baseline
continues to rise, because of the increasing spectral response of the
collecting glass lens and windows. Compared to the spectrum in reference

[135], the cooling efl‘ect in our PCP LIF spectrum is poor, but the sensitivity
of the detection system is comparable.

E. Two-photon Excitation of Perylene

The preliminary result of the two-photon excitation of supersonic jet-
cooled perylene over the region 576.8 - 577.8 nm (two-photon energy 34,670 -
34,610 cm'l, where the solution TPE spectrum shows strong enhancement;

see Figure 2- 14) is shown in Figure 3-16. This spectrum was obtained with

149

Figure 3-15. The laser induced fluorescence from [2,2]paracyclophane
. expanded through a supersonic jet, obtained over the
excitation range 312 - 307 nm. The inset is a broader-range

absorption spectrum obtained in isopentane solution at room

temperature.

150

H: \

[2,2]paracyclophane

 

 

 

 

 

 

 

 

Intensnty
(f3—

 

 

 

 

 

 

 

 

 

 

l l l L L l l l l
307 308 309 310 311 312
Wavelength (nm)

Figure 3-15.

151

Figure 3-16. Two-photon excitation spectrum of perylene in a supersonic
jet over the dye laser wavelength 576.8 — 577.8 nm (two-photon
energy 34,670 - 34,610 cm’l).

.2.)

Rel. Intensity [In/I

152

 

 

 

577.0 nm 57'” nm

44 /

 

577.0 ' 577.5

Wavelength (nm)

Figure 3'16.

 

153

two-photons from same laser, and linear polarization of the excitation
radiation. The nozzle diameter is 200 pm, the argon carrier pressure is 0.5
atm, the distance X is 8 mm (X/D = 40), and the PMT power is 1190 V.
. Much of the scattered laser light was removed by a Corning 5-57 filter in the
detection optics. The one-photon spectrum of perylene in the jet was
obtained before and between the measurements of the TPE, in order to
check the alignment of the dye laser beam and of the detection system in
relation to the molecular beam. Although the signal-to-noise level is low
(~'1.5), the peaks at about 577 .0 and 577.1 nm from the two-photon induced
excited state in the jet are reproducible and their intensity depends on the
argon carrier pressure. Since both Ag and Blg symmetry states are
expected in this region according to the two-photon excitation spectrum of
perylene in n-hexane solution, without a wider range spectrum and
polarization analysis it is not possible to provide an assignment for these
peaks.

Several reasons are suggested for the low signal—to-noise ratio in the
TPE spectrum: 1) the number density of perylene molecules is low; 2) the
alignments of the detection system and of the laser beam to the jet are not -
optimum; 3) glass windows in the detection system cut off emission in the
wavelength region below 300 nm; and 4) unlike the relaxation in solution,
following laser irradiation the contribution of collisions to the nonradiative
relaxation process can be neglected in the jet. Thus the fluorescence from
the two-photon induced electronic excited state could be one of transitions
from the gerade excited state to vibronically ungerade ground state levels,
or alternative radiative transitions could be involved in the decay processes
(Figure 3-17). Therefore the cross-section for the emission from a two-

photon induced excited state in a jet can be quite different from that in

154

Figure 3-17. The relaxation processes following two-photon excitation in

solution (a), and in a jet, (b) and (c).

155

.2..." 2:3..—

E47:

 

 

........-............—._.r..-...-.... ....-..-.....-.— -.-.-...-.u..c..-.—-.—r—-..-.-.-—.- .--.-.-...-_...

 

 

 

 

   

ZOEAOM Zn

A5

 

-.-..-..-...-..-.-.fi...-.-......-. .-._-.-......-..-.

 

 

156

solution. Nevertheless, it was promising that a signal from the two-photon
induced electronic excited state of perylene could be detected with our home-
built supersonic jet system.

To improve the signal-to-noise level, it is proposed that methods be
utilized which can increase the number density of perylene [such as a slit
nozzle, or a pulsed nozzle with large diameter (> 200 um)], and that the
measurements be systemized with one-photon excitation of supersonic jet-
cooled perylene to optimize the alignment. With more carefully designed
experiments, quartz optics and various filters, and time-resolved
measurements with ultrashort (< 1 ns) laser pulses, it may be possible to
understand the relaxation processes which follow the two-photon excitation

of perylene excited states.

r.‘

 

REFERENCES

157

REFEREN.

1. M. D. Levenson and S. S. Kano, "Introduction to Nonlinear Laser
Spectroscopy", Boston, Academic Press, 1988.

2. a) R. C. Hilborn, Am. J. Phys., 50, 982-986 (1982).

b) L. I. Schiff, "Quantum Mechanics", 3rd ed. McGraw-Hill, New York,
1968.

3. R. L. Kronig, J. Opt. Soc. Amer., 12, 547 (1926)

4. O. Howarth, " Theory of Spectroscopy" Halsted, New York, 1973.

5. A. Yariv, "Quantum Electronics", 2nd ed., p165, Wiley, New York, 1975.

6. A. Herzberg, "Infrared and Raman Spectra", Von Nostrand-Reihold,
Princeton, New Jersey, 1945; D. A. Long, "Raman Spectroscopy",
McGraw-Hill, New York, 1977.

7. H. O. Schrotter and H. W. Klockner, "Raman Spectroscopy of Gases and
Liquids",C. A. Weber, ed. 1 (Topics in Current Physics), Springer-
Verlag, Berlin, 1979.

8. F. Zernike and J. E. Midwinter, "Applied Nonlinear Optics", Wiley-
Interscience, New York, 1973.

9. MS. Feld and V. S. Letohov, eds. "Coherent Nonlinear Optics" (Topics
on Current Physicss 27). Springer-Verlag, Berlin, 1980.

10. N. Bloembergen, "Nonlinear Optics". Benjamin, New York, 1965, and
references therein.

11. R. G. Brewer, Coherent Optical Spectroscopy, in "Nonlinear
Spectroscopy" (Proc. Int. School Phys., Enrico Fermi, Course 64) (N.
Bloembergen, ed.) North-Holland, publ., Amsterdam, 1977.

12. N. Bloembergen, in "Quantum Electronics; A Treatise", H. Robin and
C. L. Tang, ed., Academic Press, New York, 1975.

158

159

13. T. H. Maiman, Nature, 187, 493 (1960).

14. P. D. Maker and R. W. Terhune, Phys. Rev., 137, 801 (1965).

15. W. L. Peticolas, I. P. Goldsborough and K. E. Rieckhoff', Phys. Rev.
Lett., 10, 43 (1963).

16. P. A. Franken, A. E. Hill, C. W. Peters and G. Weinreich, Phys. Rev.
Lett., 7, 118 (1961).

17. M. Bass, P. A. Franken, A. E. Hill, C. W. Peters and G. Weinreich,
Phys. Rev. Lett., 8, 18 (1962).

18. a) A. Szoke and A. Javan, Phys. Rev. Lett., 10, 521 (1963).
b). R.A. McFarlane, W.R. Bennet, Jr. and W. E. Lamb, Jr., Appl. Phys.
Lett., 17,53 (1970)

j 19. a) A. M. Bonch—Bruerich and V. V. Khodovoi, Sov. Phys-usp., 8,1 (1965).
b) W. L. Peticolas, Annu. Rev. Phys. Chem, 18, 233 (1967).
c) H. Mohr, in "Quantum Electronics: A Treatise" (H. Robin and C. L.
Tang, eds.), Vol. 1, Academic Press, New York, 1975.

20. s. H. Lin, Y. Fujimura, H. J. Neusser and E. w. Schlag, "Multiphoton
Spectroscopy of Molecules", Academic Press, Orlando Fla., 1984.

21. W. M. McClain and R. A. Harris, in "Excited States", Vol. 3, ed., E. C.
Lim, Academic, New York, 197 7.

22. T. R. Bader and A. Gold, Phys. Rev., 171, 997 (1968).

23. M. Inoue and Y. Toyozawa, J. Phys. Soc. Japan, 20, 363 (1965).

24. P. R. Monson and W. M. McClain, J. Chem. Phys., 53, 29 (1970).

25. w. M. McClain, J. Chem. Phys, 55,2789 (1971).

26. R. G. Bray and R. M. Hochstrasser, Mal. Phys., 72, 3999 (1976).

27. M. Goeppert-Mayer, Naturwissenschaflen, 17, 932 (1921).

28. W. Kaiser and C. G. B. Garrett, Phys. Rev. Lett., 7, 229 (1961)

29. D. Frohlich and M. Mahr, Phys. Rev. Lett., 16, 895 (1966).

 

 

160

30. a) R. A. Swofford and W. M. McClain, J. Chem. Phys., 59, 5740 (1973).
b)R. A. Swfi‘ord and W. M. McClain, Rev. Sci. Instr., 46,246 (1975).

31. L. Wunsch, H. Neusser and E. W. Schlag, Z. Naturfarsch., 36, 1340
(1981).

32. a)A. Bergman and J. Jortner, Chem. Phys. Lett., 15, 309 (1972).
b) A. Bergman and J. Jortner, Chem. Phys. Lett., 26, 323 (1974).

33. J. R. Whinney, Acc. Chem. Res., 7, 225 (1974).

34. W. C. Lineberger and T. A. Patterson, Chem. Phys. Lett., 13, 40 (1972).

35. a)P. M. Johnson, J. Chem. Phys., 62, 4562 (1975);
b) P. M. Johnson, J. Chem. Phys., 64, 4638 (1976).

36. V. J. Kleinschmidt, W. Tottleben and S. Reutsh, Exp. Tech. Physik, 22,
191 (1974).

37. W. M. McClain, Acc. Chem. Res., 7, 129 (1974).

38. U. Fritzer, Ph. Keller and G. Schaack, J. Phys., E8, 530 (1975).

39. L. Zandee and R. B. Bernstein, J. Chem. Phys., 71, 1359 (1979).

40. U. Boesl, H. J. Neusser and E. W. Schlag, J. Chem. Phys., 72, 4327
(1980).

41. S. J. Cyvin, J. Mol. Struc., 100, 75 (1983).

42. N. L. Allinger and J. T. Sprague, J. Am. Chem. Sac., 95, 3893 (1973).

43. I. B. Berlman and M. Rokni, Chem. Phys. Lett., 22, 458, (1973); C.
Rullier, A. Declemy, and Ph. Kottis, Chem. Phys. Lett., 110, 308, (1984).

44. R. G. Harvey, ed., "Polycyclic Hydrocarbons and Carcinogenesis", ACS
Symposium Series 283, American Chemical Society, Washington D. C.,
1985.

45. a) M. A. Koner and A. Yu. Slepukhin, Tear. i Ehsperim. Khim., 9, 105
(1973).

 

161

b) B. Fourmann, C. Jouvet, A. Tramer, J. M. LeBars and Ph. Millie,
Chem. Phys., 92, 25 (1985).

46. C. T. Ryan and T. K. Gustafson, Chem. Phys. Lett., 44, 241 (1976).

47. a) C. Bouzou, C. J ouvet, J. B. LeBlond, P. Millie, A. Tramer and M.
Sulkes, Chem. Phys. Lett., 93,161 (1983). i
b) S. Schwartz and M. Topp, Chem. Phys., 83, 245 (1984)
c) B. E. Forch, K. J. Chen and E. C. Lim, Chem. Phys. Lett., 100, 389
(1983).

48. a) H. Preeskamp, A. Laufer and M. Zander, Chem. Phys. Lett., 112,
479 (1984).
b) R. M. Hochstrasser, J. Chem. Phys., 40, 2559 (1964).

49. a) V. L. Bofdanov and V. P. Klochkov, Opt. Spectrasc., 45, 51 (1978).
b) T. V. Bolotnikova and O. F. Elnikova, Opt. Spectrasc., 36, 397 (1974).
c) J. Tanaka, Bull. Chem. Sac., Japan, 36, 1238 (1963).

50. a) Y. Tanizaki, T. Yoshinaga and H. Hiratsuka, Spectracim. Acta, 34A,
205 (197 8). [See also references cited therein]
b)A. P. Blokhin, V. A. Povedailo and V. A. Tolkachev, Opt. Spectrasc.,
60, 37 (1986).
c) M. Mestechkin, L. Gutyrya and V. Poltavets, Opt. Spectrasc., 28, 244
(1970).
d) S. Matsunuma, N. Akamatsu, T. Kamisuki, Y. Adachi, S. Maeda
and C. Hirose, J. Chem. Phys., 88, 2956 (1988).

51. H. Zimmermann and N. Joop, Z. Elehtra. Chem., 65, 138 (1961).

52. N. A. Borisevich, S. P. Pliska and V. A. Tolkachev, Sav. Phys. Dakl., 26,
1145 (1981).

53. E. Clar, Spectrachem. Acta, 4, 116 (1950).

162

54. P. R. Callis, T. W. Scott and A. C. Albrecht, J. Chem. Phys., 78, 16
(1983).

55. M. Newsham, Ph. D. Dissertation, Michigan State University (1988).

56. B. Steven, Chem. Rev., 57, 439 (197 5).

57. F. Bowen and S. Veljkovic, Prac. Ray. Sac.(Landan), A236 (1956).

58. R. Thiim, M. S. Thesis, Michigan State University (1983).

59. A. P. Aleksandrov, V. P. Bredikhin and V. N. Genkin, JETP Lett., 10,
117 (1969).

60. S. Speiser and S. Kimel, J. Chem. Phys, 51, 5614(1969).

61. R. P. Drucker and W. M. McClain, J. Chem. Phys., 61, 2609 (1974).

62. E. Clar and W. Schmidt, Tetrahedron, 33, 2093 (1977).

63. R. M. Hochstrasser and C. A. Nyi, J. Chem. Phys., 72, 2591 (1980).

64. S. J. Cyvin, B. N. Cyvin and P. Klaeboe, Spectrasc. Lett., 16, 239 (1983).

65. F. Ambrosino and S. Califano, Spectrachim. Acta, 21, 1401 (1965).

66. C. Rulliere, A. Declemy and Ph. Kottis, Chem. Phys. Lett., 110, 308
(1984).

67. R. M. Hochstrasser, J. Chem. Phys., 40, 2.559 (1964).

68. S. K. Chakrabarti, J. Mal. Spectrasc., 37, 571 (1971).

69. D. H. Levy, Sci. Am., 250, 96 (1984).

70. RE. Smalley, B. L. Ramakrishna, D. H. Levy and L. Wharton, J.
Chem. Phys., 61,4363 (1974).

71. A. Kantrowitz and J. Grey, Rev. Sci. Instrum., 22, 328 (1951).

72. a) G. B. Kistiakowsky and W. P. Slichter, Rev. Sci. Instrum., 22, 333
(1951).
b) E. W. Becker and K. Bier, Z. Naturfarsch. Teil A, 9, 975 (1954).

73. J. B. Anderson, R. P. Andres and J. B. Fenn, Adv. Chem. Phys, 10,
275 (1965).

163

74. R. Campargue and R. Lebehot, 9th Symp. Rarefied Gas Dynamics,
ed. M. Becker, M. Fiebig. Porz-Wahn, DFVLR Press (1974).

75. J. P. Toennies and K. Winkelman, J. Chem. Phys., 66, 3965 (197 7 ).

76. H. W. Liepmann and A. Roshko, "Elements of Gas Dynamics", p.40
New York: Wiley (1957).

77. H. Ashkenhas and F. S. Sherman, Rarefied Gas Dynamics, 4th Symp.,
ed. J. H. de Leeuw, 2,84. New York: Academic Press (1966).

78. J. B. Anderson and J. B. Fenn, Phys. Fluids, 8, 780 (1965).

79. D. R. Miller, J. P. Toennies and K. Winkelmann, 9th Symp. Rarefied
Gas Dynamics, ed. M. Becker, M. Fiebig. Porz-Wahn, DFVLR Press

' (1974).

80. N. Abauf, J. B. Anderson, R. P. Andres, J. B. Fenn and DR. Miller,
Rarefied Gas Dynamics, 5th Symp. ed., C. L. Brundin, Academic Press,
New York (1967). g

81. R. A. Larsen, S. K. Neoh and D. R. Herschbach, Rev. Sci. Instrum., 45,
2511 (1974).

82. A. Amirav, U. Even and J. Jortner, J. Chem. Phys., 51, 31 (1980).

83. E. Kolodney and A. Amirav, Chem. Phys., 76, 5684 1982).

84. U. Even, J. Magen and J. Jortner, J. Chem. Phys. 76, 5684 (1982).

85. T. A. Milne and F. T. Greene, Adv. in High Temp. Chem. 2, 107 (1969).

86. A. L. J. Burgmans, J. M. Farrar and Y. T. Lee, J. Chem; Phys., 64,
1345 (197 6). and referenced cited therein.

87. A. P. J. van Deursen and J. Reuss, J. Chem. Phys., 63, 4559 (1975).

88. R. Campargue, Rev. Sci Instrum., 35, 111 (1964).

89. R. Campargue, Ph. D. thesis, Univ. Paris, France (1970).

90. R. Campargue, J. Chem. Phys, 52, 1795 (1970).

 

164

91. J. B. Fenn and J. Deckers, Rarefied Gas dynamics, 3rd Symposium,
vol. 1, J. A. Laurman, ed., Academic Press, New York (1963); J. B. Fenn
and J. B. Anderson, Rarefied Gas Dynamics, 4th Symposium, vol. II, J.
H. De Leeuw, ed., Academic Press, New York (1966); J. B. Fenn, J. B.
Anderson, R. P. Andres and G. Maise, Rarefied Gas Dynamics, 4th
Symposium, vol. II, J. H. De Leeuw, ed., Academic Press, New York
(1966).

92. J. B. Anderson in "Molecular Beams and Low Density Gasdynamics",
P. P. Wegener, ed., Marcel Dekker, New York ( 1974).

93. U. Even, J. Magen and J. Jortner, J. Am. Chem. Sac., 103, 4583 (1981);
U. Even, J. Magen, J. Jortner, J. Friedman and H. Levanon, J. Chem.
Phys. , 77, 4374 (1982); U. Even, J. Magen, J. Jortner and J. Friedman,
J. Chem. Phys., 77 , 4384 (1982); U. Even, J. Magen and J. Jortner, J.
Chem. Phys., 77, 4391 (1982).

94. D. K. Hsu, D. L. Monts and R. N. Zare, "Spectral Atlas of Nitrogen
Dioxide", New York: Academic, 1978.

95. D. L. Monts, T. G. Dietyz, M. A. Duncan and R. E. Smalley, Chem.
Phys., 45, 133 ( 1980).

96. C. H. Chen, S. D. Kramer, D. W. Clark and M. G. Payne, Chem. Phys.
Lett., 65, 419 (1979).

97. D. R. Preuss, S. A. Pace and J. L. Gale, J. Chem. Phys., 71, 3553 (1979).

98. D. Eisel, D. Zevgolis and W. Demtroder, J. Chem. Phys, 71,2005 (197 9).

99. A. Herrmann, S. Leutwyler, E. Schumacher and L. Waste, Chem.
Phys. Lett. , 52, 418 (1977).

100. P. Huber-Walchli, D. M. Guthals and J .W. Nibler, Chem. Phys. Lett.,
67, 233 (1979).

101. M. P. Sinha, A. Schultz and R. N. Zare, J. Chem. Phys., 58, 549 (1973).

 

165

102. T. D. Gaily, S. D. Rosner and R. I. Holt, Rev. Sci. Instrum., 47, 143
(1976).

103. a) W. Sharfin, K. E. Johnson, L. Wharton and D. H. Levy, J. Chem.
Phys., 71, 1292 (1979).
b) P. S. H. Fitch, L. Wharton and D. H. Levy, Chem. Phys. Lett.,
48, 177 (1977).

104. R. Campargue and B. Soep, Chem. Phys. Lett., 64, 469 (1979).

105. T. Chaiken, T. Benson, M. Gurnick and J. D. McDonald, Chem. Phys.
Lett., 61, 195 (1979).

106. a) J. B. Hopkins, D. E. Powers, S. Mukamel and R. E. Smalley, J.
Chem. Phys., 72, 5049 (1980).

b) T. G. Dietz, M. A. Duncan, A. C. Pulu and R. E. Smalley, J. Phys.
Chem, 86, 4026 (1982).

107. T. Suzuki, N. Mikami and M. Ito, J. Phys. Chem, 90, 6432 (1986).

108. D. H. Semmes, .1. s. Baskin and A. H. Zewail, J. Am. Chem. Sac., 109,
4104 (1987); J. L. Knee, L. R. Khundkar and A. H. Zewail, J. Chem.
Phys., 87, 115 (1987). '

109. H. G. Lohmannsroben, K. Luther and M. Stuke, J. Phys. Chem, 91,
3499 (1987). '

110. E. Heller and R. C. Brown, J. Chem. Phys., 79, 3336 (1983).

111. C. E. Otis, J. L. Knee and P. M. Johnson, J. Chem. Phys., 78, 2091
(1983); C. E. Otis, J. L. Knee and P. M. Johnson, J. Chem. Phys., 80, 13
(1984). '

112.0. A. Haynam, D. V. Brumbaugh and D. H. Levy, J. Chem. Phys., 80,
2256 (1984); C. A. Haynam, D. V. Brumbaugh and D. H. Levy, J. Chem.
Phys., 79, 1581 (1983); C. A. Haynam, C. Morter, L. Young and D. H.
Levy, J. Phys. Chem, 91, 2526 (1987); P. R. R. LangfidgeSmith, C. A.

 

166

Haynam, D. V. Brumbaugh and D. H. Levy, J. Phys. Chem, 85, 3739
(1981).

113. S. Leutwyler and J. Jortner, J. Phys. Chem, 91, 5558 (1987 ).

114.,E. A. Mangle and M. R. Topp, Chem. Phys., 112,427 ( 1987).

115. J. B. Hopkins, D. E. Power and R. E. Smalley, 85, 3739(1981).

116. K. D. Borsen, H. L. Seizle and E. W. Schlag, J. Chem. Phys., 85, 1726
(1986).

117. T. A. Stephenson and S. A. Rice, J. Chem Phys., 81, 1083 (1984); S. A.
Rice, J. Phys. Chem, 90, 3036 (1986).

118. H. Saigusa and M. Ito, J. Chem. Phys., 81, 5692 (1984); H. Abe, N.
Mikami and M. Ito, J. Phys. Chem, 86, 1766 (1982); H. Abe, N. Mikami,
M. Ito and Y. Udagawa, J. Phys. Chem, 86, 2567 (1982); Y. Tomioka,

M. Ito and N. Mikami, J. Phys. Chem, 87, 4401 (1983).

119. K. Fuke, H. Yoshiuchi and K. Kaya, J. Phys. Chem, 88, 5840 (1984); K.
Fuke, T. Yabe, N. Chiba, T. Kohida and K. Kaya, J. Phys. Chem, 90,
2309 (1986).

120. D. H. Levy, L. Wharton and R. Smalley, "Chem. Biochem. Appl.

A Lasers", 2, 1, C. B. Moore, ed., New York, Academic Press (1977).

121. R. Tembreull and D. M. Lubman, Anal. Chem, 58,1299 (1986); R.
Tembreull and D. M. Lubman, Anal. Chem, 59,1082 (1987); R.
Tembreull and D. M. Lubman, Anal. Chem, 59,1003 (1987).

122. A. Amirav, U. Even and J. Jortner, Anal. Chem, 54, 1666 (1982); M.
Sonnenschein, A. Amirav and J. Jortner, J. Phys. Chem, 88, 4214
(1984); A. Amirav, M. Sonnenschein and J. Jortner, Chem. Phys.,

102, 305 (1986).

123. V. Vaida, Acc. Chem. Res., 19, 114 (1986); D. G. Leopold, V. Vaida and

M. F. Granville, J. Chem. Phys, 81, 4210 ( 1984).

 

167

124. T. S. Zwier, M. E. Carrasquillo and D. H. Levy, J. Chem. Phys., 78,
5493 (1983).

125. M. Ito, T. Ebata and N. Mikami, Ann. Rev. Phys. Chem, 39, 123
(1988).

126. H. Fukuoka, T. Imasaka and N. Ishibashi, Anal. Chem, 58, 375
(1986); T. Imasaka, N. Yamaga and N. Ishibashi, Anal. Chem, 59, 419
(1987).

127. C. H. Sin, H. M. Pang, D. M. Lubman and J. Zorn, Anal. Chem, 58,
487 (1986).

128. J. Grotemeyer, U. Boesel, K. Walter and E. W. Schlag, J. Am. Chem.
Sac., 108, 4233 (1986).

129. J. M. Hayes and G. J. Small, Anal. Chem, 54, 1202 (1982).

130. B. V. Pepich, J. B. Callis, J. D. Sheldon Danielson and M. Gouterman,
Rev. Sci. Instrum., 57, 878(1986); B. V. Pepich, J. B. Callis, D.H. Burns,
M. Gouterman and D. A. Kalman, Anal. Chem, 58, 2825 (1986).

131. R. E. Smalley, D. H. Levy and L. Wharton, J. Chem. Phys., 64, 3266
(1976).

132. J. C. Kettley, T. F. Palmer and J. P. Simons, Chem. Phys. Lett., 115, 40
(1985). .

133. A. Amirav, U. Even and J. Jortner, Chem. Phys., 51, 31 (1980).

134. S. Iwata, K. Fuke, M. Sasaki, S. Nagakura, T. Otsubo and S. Misumi,
J. Mal. Spectrasc., 46, 1 (197 3).

135. A. Ron, M. Noble and E. K. Lee, Chem. Phys., 83, 215 (1984).