Dispersed fluorescence spectroscopy of primary and secondary alkoxy radicals

Jin Jin, Ilias Sioutis, György Tarczay, Sandhya Gopalakrishnan, Andrew Bezant, and Terry A. Miller

Citation: The Journal of Chemical Physics 121, 11780 (2004); doi: 10.1063/1.1814104

View online: http://dx.doi.org/10.1063/1.1814104

View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/121/23?ver=pdfcov

Published by the AIP Publishing

Articles you may be interested in

An experimental and ab initio study of the electronic spectrum of the jet-cooled F2BO free radical

J. Chem. Phys. 140, 164302 (2014); 10.1063/1.4871010

Laser-induced fluorescence and dispersed fluorescence spectroscopy of jet-cooled 1-phenylpropargyl radical

J. Chem. Phys. 130, 144313 (2009); 10.1063/1.3110682

The ground state energy levels and molecular structure of jet-cooled HGeCl and DGeCl from single vibronic level

emission spectroscopy

J. Chem. Phys. 124, 124320 (2006); 10.1063/1.2181142

Observation of a new phosphorus-containing reactive intermediate: Electronic spectroscopy and excited-state

dynamics of the HPBr free radical

J. Chem. Phys. 123, 144304 (2005); 10.1063/1.2055227

Laser induced fluorescence and resonant two-photon ionization spectroscopy of jet-cooled 1-hydroxy-9,10-

anthraquinone

J. Chem. Phys. 122, 034304 (2005); 10.1063/1.1829977

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

JOURNAL OF CHEMICAL PHYSICS

VOLUME 121, NUMBER 23

15 DECEMBER 2004

Dispersed ﬂuorescence spectroscopy of primary and secondary alkoxy

radicals

a)

¨

Jin Jin, Ilias Sioutis, Gyorgy Tarczay, Sandhya Gopalakrishnan, Andrew Bezant,

and Terry A. Miller

Laser Spectroscopy Facility, Department of Chemistry, The Ohio State University, 120 West 18th Avenue,

Columbus, Ohio 43210

͑Received 16 July 2004; accepted 17 September 2004͒

Dispersed ﬂuorescence ͑DF͒ spectra of 1-propoxy, 1-butoxy, 2-propoxy, and 2-butoxy radicals have

been observed under supersonic jet cooling conditions by pumping different vibronic bands of the

˜ ˜

B-X laser induced ﬂuorescence excitation spectrum. The DF spectra were recorded for both

conformers of 1-propoxy, three conformers of the possible ﬁve of 1-butoxy, the one possible

conformer of 2-propoxy, and two conformers of the possible three of 2-butoxy. Analysis of the

˜

spectra yields the energy separations of the vibrationless levels of the ground X and low-lying A

electronic state as well as their vibrational frequencies. In all cases, the vibrational structure of the

DF spectra is dominated by a CO stretch progression yielding the ␯ stretching frequency for the

CO

˜

X state and in most cases for the A state. In addition to the experimental work, quantum chemical

˜

calculations were carried out to aid the assignment of the vibrational levels of the X state and for

˜

some conformers the A state as well. Geometry optimizations of the different conformers of the

isomers were performed and their energy differences in the ground states were determined. The

˜

results of the calculation of the energy separations of the close-lying X and A states of the different

Institute of Physics. ͓DOI: 10.1063/1.1814104͔

alkoxy radicals^5,6yielded detailed information about the

I. INTRODUCTION

˜

structure and geometries of both the B and X states. It was

found that various bands of the LIF spectra uniquely corre-

lated with different conformers of a given structural isomer.

However, information about the excitation energy of the

The oxidation of hydrocarbons is among the most impor-

tant of chemical processes, with combustion and the degra-

dation of volatile organic compounds injected in our atmo-

sphere being the prime examples. The reactions of the

simplest oxygen-containing organic radicals, e.g., alkoxy,

RO, and peroxy, RO₂radicals, affect the yield of ozone, air

toxics, and organic aerosols in the atmosphere.¹Combustion

of hydrocarbon fuels at temperatures р1000 K share many

mechanistic details with their atmospheric oxidation. The ex-

istence of multiple reaction pathways for these radicals com-

plicates efforts for the understanding of the potential oxida-

tion scheme of hydrocarbons, as it is the competition among

the various radical reactions pathways that determines the

end products of the oxidation.²Hence elucidation of their

chemistry is paramount for obtaining a better understanding

of combustion processes and the impact of hydrocarbons in

the atmosphere.

˜

low-lying A state as well as the ground X state’s vibrational

structure and dynamics has been relatively sparse. A suitable

technique for investigating this area is laser excited, dis-

persed ﬂuorescence ͑DF͒ spectroscopy. It has been used to

probe the vibronic levels in the ground state of CH₃O⁷and

C₂H₅O.⁸However, DF studies have not been extended to

larger alkoxy radicals thus far.

The alkoxy radicals can be viewed as derivatives of

methoxy. While the nominal point-group symmetry of meth-

oxy is C_3v, for all other alkoxy radicals the symmetry is

lower ͑except possibly for t-butoxy͒. Upon the lowering of

the point-group symmetry, the ground electronic state of

2

methoxy E is resolved into a pair of states with the ground

˜

state designated as X and the lowest lying excited state as A.

The ﬁrst excited state of methoxy is electronically nondegen-

This paper concerns itself with the alkoxy radicals.

There have been extensive spectroscopic investigations, via

3

˜

˜ ˜

erate and corresponds to a state designated as B for the

alkoxies larger than methoxy ͑except t-butoxy͒. The LIF ex-

the B-X electronic transition, of the smaller alkoxy radicals

as well as some studies of larger ones.⁴Laser-induced ﬂuo-

rescence ͑LIF͒ excitation spectroscopy of jet-cooled radicals

has provided a good deal of information about the excited

states. Very recently, high-resolution rotationally resolved

spectroscopy of a number of bands of jet-cooled primary

˜ ˜

citation spectrum involves the B X transition while the DF

˜ ˜

˜

spectra involve the B→A and B→X emission. The B state

excitation is well known to be nominally a one-electron ex-

citation of a p-␴ bonding electron to ﬁll the half-ﬁlled p-␲

orbital localized on the O atom.

a͒

In the present work, LIF spectroscopy was combined

with supersonic free jet expansion techniques to record the

Permanent address: Department of General and Inorganic Chemistry,

¨

Eotvos University, P. O. Box 32, H-1518 Budapest 112, Hungary.

0021-9606/2004/121(23)/11780/18/$22.00

11780

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11781

TABLE I. Summary of the conditions used for the experiments with the

primary and secondary alkoxies.

by a rotary oil pump. To produce the alkoxy radicals, pho-

tolysis of the precursors just above the throat of the nozzle

was performed using the output of a tripled Nd:yttrium alu-

minum garnet ͑YAG͒ laser ͑Quanta-Ray DCR-2, 355 nm͒.

This produced the desired alkoxy radicals which were then

probed about 10 mm downstream from the photolysis laser.

The probe beam was the frequency doubled output ͑Inrad

Autotracker II͒ of a tunable dye laser ͑SpectraPhysics

PDL-3͒ pumped by the second harmonic of a Nd:YAG

͑Quanta-Ray DCR͒ laser at 532 nm. This work required the

use of three laser dyes ͑LDS 698, 750, and 751͒ which

spanned appropriate spectral regions for the DF excitation

frequencies. The delay time between the ﬁring of the pho-

tolysis and dye lasers, along with the time delay and the

opening time of the nozzle, were controlled by a pulse gen-

erator ͑Stanford Research DG535͒.

Alkoxies

Primary

Secondary

Temperature of the sample bomb (°C)

Backing pressure ͑psi͒

10

Ϫ5

90

60

15

40

40,80^a

150^b

30–35

60^c

Slit width ͑␮m͒

Resolution (cm^Ϫ1

)

Accumulation time ͑min͒

Grating ͑grooves/mm͒

1800

^aThe backing pressure in the 2-propoxy experiment was kept at 40 psi and

for 2-butoxy at 80 psi.

^bThe DF emission following the LIF excitation of the ﬁrst member of the

CO stretch progression of 2-propoxy was collected with an entrance slit

width of 50 ␮m.

^cFor 2-butoxy the accumulation time for the DF collection via bands A

through D was extended at least 3 h for the ﬁrst 2000 cm^Ϫ1redshifted from

the pumping frequency.

The photolysis and probe beams counterpropagated

through the vacuum chamber. The ﬂuorescence signal was

collected perpendicularly to the laser beams. On one side of

the chamber, the ﬂuorescence emission was collimated by an

1 in. diameter lens (f/1) and a second lens focused the ra-

diation onto the photocathode of the photomultiplier tube

͑EMI 9659Q͒ connected to an ampliﬁer. The ampliﬁed LIF

signal was integrated by a boxcar averager, digitized by an

A/D converter and fed into a personal computer for further

data processing. On the other side of the chamber, the ﬂuo-

rescence emission was dispersed by a Spectra Pro 300i

monochromator using a 1800 grooves/mm grating and then

imaged onto an intensiﬁed charge-coupled device. The detec-

tion of the dispersed photons for the secondary alkoxies was

done exclusively with the Princeton Instruments PI-

MAX512HB camera. In the case of the primary alkoxies two

cameras were used, the above one and a second one, Princ-

eton Instruments PI-MAX512RB. The emission frequencies

reported for the DF spectra of the alkoxies were obtained by

calibrating them against a Fe-Ne lamp, or known laser fre-

quencies.

˜

DF spectra from a number of B state levels of the 1- and

2-isomers of the propoxy and butoxy radicals. These spectra

˜ ˜

can yield the direct measurements of the A-X electronic state

separation. Emission was observed primarily to the CO vi-

brationless level and excited stretch and CCO vibrational

˜

modes ͑or modes coupled with these motions͒ of both the X

˜

and A states. The DF data also serve to support the identiﬁ-

cation of multiple conformations of the alkoxy radicals.

Extensive quantum chemical computations were carried

out as an aid to the assignment of the observed vibronic

˜ ˜

˜

bands. Both the A-X energy gap and the X and A vibrational

frequencies can be used as benchmarks against which theo-

retical treatments of these complex radicals can be measured.

˜ ˜

Information about the A-X energy separation is important for

a better understanding of the vibronic interactions between

˜ ˜

these close-lying, lowest energy states. The A-X separation

is also very important for understanding the kinetics and de-

˜

tailed dynamics of the alkoxy radicals as the degree of A

state participation in thermal reactions is dependent upon its

excitation energy. The separation is also critical to an under-

standing of the ground state spin-rotation splittings.

III. RESULTS

DF spectra were observed for both the propoxy and bu-

toxy radicals. There are two isomers for propoxy; the linear

chain isomer 1-propoxy and the branched chain one denoted

as 2-propoxy, often referred to as isopropoxy. In the case of

the butoxy radicals, there exist four possible isomers two of

which will be the focus of this work, namely, the 1-butoxy

and 2-butoxy species.

The remainder of this paper is organized as follows. In

Sec. II an outline of the experimental details is given. The

analysis of the experimental spectra takes place in Sec. III.

Section IV centers on a number of aspects pertaining to the

structure and spectra of the primary and secondary alkoxies.

The paper’s conclusions are presented in Sec. V.

It is known that different conformations can exist for a

given structural isomer. Therefore, theoretically speaking,

one should anticipate two different conformers for 1-propoxy

and one possible conformation for isopropoxy. Five confor-

mations are possible for 1-butoxy and three for 2-butoxy.

High-resolution spectroscopic studies have identiﬁed,^5,6via

detailed analysis of the rotational structure of the vibronic

bands, both conformers of 1-propoxy and three of 1-butoxy

populated in the free jet expansion. The conformers for

1-propoxy are denoted as gauche G and trans T. G is a

C₁-symmetry structure and T is of C_ssymmetry. For

1-butoxy the corresponding conformations ͑point-group

symmetries͒ are G₁T₂(C₁), T₁T₂(C_s), and T₁G₂(C₁).

II. EXPERIMENTAL DETAILS

The alkyl nitrite precursor molecules were prepared by

the dropwise addition of the concentrated sulfuric acid to a

saturated solution of NaNO₂and the corresponding alcohol.⁹

A few torr of the alkyl nitrite vapor was entrained into the jet

ﬂow by passing helium at a suitable backing pressure over

the liquid contained in a stainless steel reservoir maintained

at a suitable temperature depending upon the vapor pressure.

A summary of the experimental conditions is given in Table

I. The seeded ﬂow was then expanded through a 300 ␮m

standard pulsed nozzle ͑general valve͒ into the jet chamber

which was evacuated by a mechanical booster pump backed

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

11782

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Jin et al.

Ϫ1

˜ ˜

TABLE III. Experimental and calculated A-X energy separation ͑in cm

)

of a number of open-chain alkoxy radicals.

Past and present studies

Alkoxies^a

Calculated^b

Experimental

d

Ethoxy (²A )

314͓380͔

225͓369͔

͓8͔

364,^c355͑10͒

Ј

e

1-propoxy, T (²A )

321͑10͒

214͑10͒

Љ

e,f

1-propoxy, G (²A)

d,f

2-propoxy (²A )

101

68͑10͒,^e,f1225͑65͒

Ј

e

1-butoxy, T₁G₂(²A)

¯

271͑10͒

292͑10͒

129͑10͒

55͑10͒

e

1-butoxy, T₁T₂(²A )

169

Љ

FIG. 1. Newman projections of three unique conformational structures of

2-butoxy at their local minima. The designation of the conformers is based

on the value of the dihedral angle ␾ which is formed by C₂–C₁–O and

C₁–C₂–C₃planes; rotation about the C₁–C₂bond changes ␾ from 60°

͑gauche clockwise Gϩ), to 180° ͑trans T) and Ϫ60° ͑gauche counterclock-

wise GϪ).

e,f

1-butoxy, G₁T₂(²A)

2-butoxy, ‘‘B’’ ( ²A)

2-Butoxy, ‘‘A’’ ( ²A)

¯

e,g

¯

e,g

¯

125͑10͒

^aThe symmetries of the computed ground electronic states of the alkoxy

radicals are included in parenthesis. The primed symmetry labels refer to

the nominal C_spoint group whereas the unprimed are under C₁symmetry.

b

˜ ˜

The adiabatic A-X energy separations were calculated at the B3LYP/6-31

The reﬂection plane of symmetry in 2-propoxy classiﬁes it as

a C_s-symmetry structure. 2-Butoxy possesses the gauche

clockwise conformer Gϩ (C₁), the trans T (C₁) conformer,

and the gauche counterclockwise conformation GϪ (C₁)

͑see Fig. 1͒. Because of the chiral stereocenter in 2-butoxy,

enantiomers are expected in each conformer. They should

give identical spectra, thus there are only three unique con-

former spectra possible for 2-butoxy.

*

ϩG level of theory, except for those included in brackets which were

done at the EOM-EE-CCSD/cc-pVDZ level of theory. The energy separa-

tion is calculated as the difference between the C₁energy minimum of the

˜

X state and the C_senergy minimum of the A state.

^cSee Ref. 8. We have reassigned the transition previously assigned to ␯ to

18

˜

the vibrationless level of the A state. The average value of the frequency

was recalculated using the raw data in Table 2 of the reference.

^dSee Ref. 19.

e

˜ ˜

The A-X separation obtained from the present DF experiments.

^fSee text for possible caveats.

^gAverage of two values, see text for details.

A. Quantum chemistry calculations

To aid the assignment process we have carried out a

number of quantum chemistry calculations. These included

͑i͒ the geometry optimization of the different conformers of

the primary and secondary radicals and the determination of

their relative energy difference in their ground electronic

state; ͑ii͒ the calculation of the energy separations of the

package¹¹͒ for selected alkoxy radicals performed at the

equation-of-motion excitation energy coupled cluster singles

12

and doubles ͑EOM-EE-UCCSD͒ level of theory using the

cc-pVDZ basis set. Tables IV–VII refer to only one confor-

mation of each structural isomer, while Tables I–XII of the

Supplementary Material¹³contain a complete summary of

the calculated harmonic vibrational frequencies for all of the

experimentally observed conformations in their ground elec-

tronic state. ͑In the case of 2-butoxy, the vibrational frequen-

cies of all three of the conformers are reported, although only

two conformations are identiﬁable in the DF spectra.͒ The

˜

close-lying A and X states for each conformer; ͑iii͒ the cal-

culation of the harmonic vibrational frequencies of the con-

formers in their ground state. These calculations have been

performed by the GAUSSIAN 98 program package¹⁰at the

*

B3LYP/6-31ϩG level of theory. The geometry optimiza-

tion of all of the conformers was done at C₁symmetry.

Tables II, III, and IV–VII summarize the computational re-

sults ͑i͒, ͑ii͒ and, in part ͑iii͒, respectively. Table III also

includes some results ͑obtained by the ACESII program

˜

calculations of the A state vibrational frequencies for the C_s

conformers are also summarized in the Supplementary Ma-

terial. Tables IV–VII include all frequencies lower than

Ϸ1500 cm^Ϫ1labeled following the Mulliken notation

¨

TABLE II. Energy differences ͑in cm^Ϫ1) relative to the energy of the lowest

energy conformation of the primary and secondary radicals in their ground

scheme¹⁴for a C₁molecule. These lower frequency vibra-

tions are the ones relevant to the interpretation of the DF

spectra; the remaining higher vibrational frequencies are, as

noted, given in the Supplementary Material.

*

electronic state as obtained at the B3LYP/6-31ϩG level of theory.

Alkoxy

Conformer

E^a

The ﬁrst two columns of the tables label and give the

value of the harmonic frequencies obtained at the

C₁-symmetry minimum while their vibrational character is

described in terms of ͑i͒ CO stretch and ͑ii͒ CCO deforma-

tion, and ͑iii͒ other motions in the remaining three columns.

These determinations are based upon visual inspections of

the vibrational motion as displayed by GAUSSVIEW ͑Gauss-

ian, Inc.͒. It is a useful guide for the assignment process to

approximately know the degree of participation of the CO

stretch and the CCO deformation in the vibrational character

of each mode, since for these directions alone are there sub-

1-propoxy

G

T

0

20

1-butoxy

G₁T₂

T₁T₂

T₁G₂

0

62

339

2-propoxy

2-butoxy

0

254

314

Gϩ

T

GϪ

^aNo zero-point energy correction was applied. For all of the alkoxy radicals,

the geometry optimization led to distinct unique conformations and was

done at the C₁-symmetry.

˜

stantial differences in the X and B geometry leading to sig-

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11783

TABLE IV. The calculated lowest 17 vibrational frequencies ͑in cm^Ϫ1) of the T conformer of 1-propoxy radical

*

in its ground electronic state as obtained at the B3LYP/6-31ϩG level of theory.

˜

X,T

Assignment B3LYP^a

CO stretch

CCO deformation^b

Small

Q^c

␯

1436(a )

1414(a )

1377(a )

1328(a )

1305(a )

1238(a )

1128(a )

1074(a )

1034(a )

967(a )

889(a )

771(a )

475(a )

307(a )

270(a )

230(a )

Small

Medium

Ј

Љ

Ј

Љ

Ј

Љ

Ј

Љ

Ј

Љ

CH₃, CH₂rock, CC stretch

CH₂rock, CC stretch

CH₃, CH₂rock

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

Small

CH₃, CH₂rock

CH₃, CH₂rock, CH₂torsion

CH₃, CH₂rock, CC stretch, CCC motion

CH₃, CH₂rock, CC stretch

CC stretch

Medium

Large

Medium

CH₃, CH₂rock

Medium

Large

CH₃, CH₂rock, CC stretch

CH₃, CH₂rock, CH₂torsion

CCC bend

CH₂torsion

CCC bend

CH₃, CH₂torsion

Backbone ﬂex

Medium

119(a )

Љ

^aThe unscaled harmonic frequencies were obtained at the C₁-symmetry minimum and their symmetries under

¨

the C_spoint group are included in parenthesis. The labeling of the vibrational modes follows the Mulliken

notation scheme ͑Ref. 14͒ assuming C₁symmetry.

^bFor a symmetry modes the CCO deformation is an in-plane bend and for a symmetry modes it is an

Ј

Љ

out-of-plane twist.

^cThe contributions of various internal coordinates Q to the normal mode indicated in the ﬁrst column are listed,

other than the CO stretch or CCO deformation which are listed in the preceding two columns with a qualitative

indication of their magnitude.

niﬁcant Franck-Condon factors for progressions in modes

containing these motions. The tables contained in the

Supplementary Material show that generally there appear to

be few pronounced differences among the vibrational fre-

quencies of the possible conformers for the alkoxy radicals.

Table II gives the calculated energies of the conformers.

It is worthwhile noting that the conformer that is calculated

to be lowest in energy for each isomer has always been ob-

served experimentally. However, the conformer populations

are certainly not in equilibrium at the rotational temperature

͑1.0–1.5 K͒.

the size of the molecule increases. For the aforementioned

C_sstructures, we conclude from the calculations no serious

distortion of the PES of the ground state is expected, hence

one would not expect vibrational progressions other than the

totally symmetric ones to be promoted signiﬁcantly in the

excited electronic state.

Finally it is worth mentioning a possibly obvious but

quite important point. In percentage terms we expect the cal-

˜

culation of the vibrational frequencies of both the X and A

˜ ˜

states to be much more accurate than the A-X separation.

The problem with the latter prediction, of course, is that the

˜ ˜

Table III includes experimentally determined A-X en-

˜ ˜

A-X difference is very small compared to the computed ab-

ergy separations. The smallness of these separations for the

C_s-symmetry structures calls for further investigation about

possible pseudo-Jahn-Teller interactions, since it is possible

for the two potential energy surfaces ͑PESs͒ to distort sub-

stantially upon consideration of such vibronic effects. For

2-propoxy, the T conformer of 1-propoxy and the T₁T₂con-

former of 1-butoxy the geometry optimizations started from

C₁-symmetry structures and resulted in 0.4, 1.9, and

7.4 cm^Ϫ1stabilization, respectively, as compared to the op-

solute energies of the states, which means that even if the

absolute error in the states’ energies is small the error in the

˜ ˜

difference can be relatively large. To get accurate A-X values

would require a very large basis set with high-quality ͑cor-

relation͒ methods. One could also have to take into account

all the auxiliary corrections ͑e.g., relativistic, Born-

Oppenheimer breakdown, core-correlation, anharmonic zero-

point energy corrections͒.

˜

timized C_s-symmetry point of the X potential energy curve.

B. DF observations

Although single-reference theories are not accurate enough

to reliably calculate the stabilization energy, owing to artiﬁ-

cial symmetry breaking problems,^15,16these results may be

used as a means to obtain a rough qualitative picture of

trends in the alkoxy radicals. Methodologies like the

equation-of-motion ionization potentials coupled cluster

singles and doubles^17,18avoid problems of artiﬁcial symme-

try breaking but they become computationally expensive as

Figures 2 and 3 show survey scans of LIF spectra of the

primary and secondary alkoxy radicals, respectively. Laser

excited DF spectra were recorded for a number of vibronic

bands labeled by the letters shown in the corresponding LIF

traces.

In principle, the DF spectra provide a direct experimen-

˜

tal way of obtaining the energy difference between the A and

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

11784

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Jin et al.

TABLE V. The calculated lowest 23 vibrational frequencies ͑in cm^Ϫ1) of the T₁T₂conformer of 1-butoxy

*

radical in its ground electronic state as obtained at the B3LYP/6-31ϩG level of theory.

˜

X,T₁T₂

Assignment

B3LYP^a

CO stretch

CCO deformation^b

Q^c

␯

1435(a )

CC stretch, CH₃rock

CH₃, CH₂rock, CC stretch

CH₃, CH₂rock, CCC bend

CH₃, CH₂rock

Ј

14

1413(a )

Small

Ј

15

1393(a )

Ј

16

1369(a )

Ј

17

1339(a )

␯

Љ

18

1304(a )

␯

Љ

CH₃, CH₂rock

19

1282(a )

Small

␯

Ј

CH₃, CH₂rock

20

1227(a )

␯

Љ

CH₃, CH₂rock

21

1133(a )

Medium

␯

Ј

CH₃, CH₂rock, CCC bend

CH₃, CH₂rock, CC stretch

CC stretch

22

1079(a )

Large

Small

Ј

23

1061(a )

Ј

24

␯

1006(a )

Small

Large

Medium

Љ

CH₃, CH₂rock, CC stretch

CH₃, CH₂rock

25

995(a )

Medium

␯

Ј

26

914(a )

Ј

27

849(a )

Љ

28

746(a )

␯

Љ

CH₃, CH₂rock

29

467(a )

Large

Medium

CH₂torsion, CCC bend

CH₃, CH₂torsion

␯

Ј

30

391(a )

Ј

31

361(a )

Љ

32

244(a )

␯

Љ

CH₃, CH₂torsion

33

185(a )

Backbone ﬂex

␯

Ј

34

␯

114(a )

Backbone ﬂex

Љ

35

␯

106(a )

Backbone ﬂex

Љ

36

^aThe unscaled harmonic frequencies were obtained at the C₁-symmetry minimum, and their symmetries under

¨

the C_spoint group are included in parenthesis. The labeling of the vibrational mode follows the Mulliken

notation scheme ͑Ref. 14͒.

^bFor a symmetry modes the CCO deformation is an in-plane bend and for a symmetry modes it is an

Ј

Љ

out-of-plane twist.

^cThe contributions of various internal coordinates Q to the normal mode indicated in the ﬁrst column are listed,

other than the CO stretch or CCO deformation which are listed in the preceding two columns with a qualitative

indication of their magnitude.

1. 1-Propoxy

˜

X states. However, to accurately determine the origin of the

˜

˜ ˜

A state, one needs to uniquely identify the B-A origin band

For 1-propoxy, the DF spectra were recorded by pump-

ing bands A and B shown in Fig. 2. Band A has been iden-

tiﬁed by detailed rotational analysis⁵as the origin of the G

conformer and band B is the origin of the T conformer.

These DF spectra are shown in Fig. 4, while Fig. 5 shows an

expanded view of their DF spectra within 1700 cm^Ϫ1red-

shifted from the pumping frequency.

among bands belonging to the vibrationally excited levels of

˜ ˜

the X (A) states, the latter of course giving valuable infor-

mation about the vibrational frequencies of these states. Our

assignments were aided with quantum chemical predictions

of both the energy separation of these two states and the

vibrational levels for the ground electronic state. From the

present and previous¹²computational work, we have found

The

T conformer of 1-propoxy is a nominally

˜

the PES for the close-lying A state to be similar to that of the

C_s-symmetry structure and therefore one would expect the

totally symmetric fundamental frequencies to dominate in

the DF spectrum. It is the CO stretch motion that is expected

to exhibit the strongest ﬂuorescence intensity, thus the stron-

gest transitions should be assigned to CO stretch vibrations.

ground state. Hence the vibrational frequencies are not ex-

pected to differ greatly between these two electronic states,

with the exception of some possible frequencies that exhibit

pseudo-Jahn-Teller activity for the C_sconformers.¹²

It is pertinent to mention the following additional prin-

ciples upon which we base our assignments of the vibrational

˜

A strong progression in the X state of the T conformer can

immediately be identiﬁed with a 1-0 vibrational interval of

˜

1050 cm^Ϫ1redshifted from the X state origin located at

levels of the X and A states:͑i͒ under C_s-symmetry ͑or near

C_ssymmetry͒, only the totally symmetric fundamentals are

electronically allowed and therefore these are expected to

exhibit the strongest ﬂuorescence intensity; ͑ii͒ only vibra-

tions of CO stretch or CCO deformation ͑bend or twist͒, or

vibrations coupled to those, are anticipated to have signiﬁ-

˜

0 cm^Ϫ1. We attribute this main vibrational progression to

␯

₁₈, as this mode has the strongest CO stretch character

among all of the modes ͑see Table IV͒. Mode ␯ involves

signiﬁcant but less participation of the CO stretch motion

than ␯₁₈and that is analogously reﬂected in the DF spectrum

via the ﬂuorescence intensity of the corresponding 1-0 band

assigned to the 1017 cm^Ϫ1peak in Table VIII.

19

˜ ˜

cant nondiagonal Franck-Condon factors, since the B, A,

˜

and X potentials are quite similar along other modes, as the

electronic transition is localized on the O atom.

Based upon the match with the theoretically calculated

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11785

TABLE VI. The calculated lowest 16 vibrational frequencies ͑in cm^Ϫ1) of 2-propoxy radical in its ground

*

electronic state as obtained at the B3LYP/6-31ϩG level of theory.

˜

X

Assignment

B3LYP^a

CO stretch

(C)₂CO deformation^b

Q^c

␯

1428(a )

CH₃rock, CC stretch

CH₃, CH rock, CC stretch

CH₃, CH rock

Ј

12

1404(a )

Љ

13

1261(a )

Large

Ј

14

1207(a )

Medium

CH₃, CH rock

␯

Ј

15

1165(a )

CH₃, CH rock

␯

Љ

16

1069(a )

Large

CH₃rock, CC stretch

CH₃, CH rock

CH₃rock

␯

Љ

17

1054(a )

Small

Large

Ј

18

984(a )

Ј

19

921(a )

CH₃rock, CC stretch

CH₃, CH rock

␯

Љ

20

888(a )

Љ

21

809(a )

Medium

CC stretch

␯

Ј

22

␯

453(a )

390(a )

Љ

Medium

Large

CCC bend

Ј

23

24

␯

359(a )

244(a )

Ј

Small

CCC bend

CH₃torsion

Ј

25

26

197(a )

Љ

27

^aThe unscaled harmonic frequencies were obtained at the C₁-symmetry minimum and their symmetries under

¨

the C_spoint group are included in parenthesis. The labeling of the frequencies follows the Mulliken notation

scheme.

^bFor a symmetry modes the (C) CO ͑or HCO͒ deformation is an in-plane bend and for a symmetry modes

Ј

Љ

2

it is an out-of-plane twist.

^cThe contributions of various internal coordinates Q to the normal mode indicated in the ﬁrst column are listed,

other than the CO stretch or CCO deformation which are listed in the preceding two columns with a qualitative

indication of their magnitude.

TABLE VII. The calculated lowest 22 vibrational frequencies ͑in cm^Ϫ1) of the Gϩ conformer of 2-butoxy

*

radical in its ground electronic state as obtained at the B3LYP/6-31ϩG level of theory.

˜

X,Gϩ

Assignment

B3LYP^a

CO stretch

(C)₂CO deformation

Q^b

␯

1432

1418

1353

1311

1244

1194

1165

1067

1059

1030

1000

936

927

794

773

468

431

365

CH₃, CH₂rock, CC stretch

CH₃, CH₂rock

15

␯

16

Small

Large

Small

17

18

CH₃, CH₂rock, CC stretch

CH₃, CH₂rock, CC stretch, CCC bend

CH₃, CH₂rock, CC stretch

CCC bend

19

20

21

22

Medium

Large

23

Large

24

25

Large

Medium

26

Medium

27

28

Medium

29

Large

30

CCC bend

31

␯

32

␯

247

CCC bend

33

␯

216

208

CH₃torsion, CCC bend

CH₃torsion

34

35

95

Backbone ﬂex

␯

36

a

¨

Unscaled harmonic frequencies. The labeling of the vibrational modes follows the Mulliken notation scheme

͑Ref. 14͒ for a C₁molecule.

^bThe contributions of various internal coordinates Q to the normal mode indicated in the ﬁrst column are listed,

other than the CO stretch or CCO deformation which are listed in the preceding two columns with a qualitative

indication of their magnitude.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

11786

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Jin et al.

FIG. 4. DF spectra of the T and G conformations of 1-propoxy radical. The

top trace was obtained by pumping band B ͑Fig. 2͒, the origin of the T

conformer, at 29 218 cm^Ϫ1while the bottom trace was obtained by pumping

band A ͑Fig. 2͒, the origin of the G conformer, at 28 634 cm^Ϫ1. The x axis

indicates ͑red͒ shift from the pump frequency. Assigned members of the

FIG. 2. Survey scans of the LIF spectra of the primary radicals. The line

labeled B in 1-propoxy extends off-scale so as to show more clearly the

weaker transitions.

predominant CO stretch progression, ␯ (T and G), are indicated by ticks

18

˜ ˜

on the top horizontal bar while the * indicates the line assigned to the A-X

origin transition. The intensities are not corrected for the grating efﬁciency

and the camera sensitivity curves. The experimental resolution was

15 cm^Ϫ1

.

frequencies ͑see Table VIII͒, we have assigned several other,

˜

lower intensity, totally symmetric fundamentals to the X

state. The bands at 274 (␯^X₂₅), 527 (␯₂^X₃), 895 (␯₂^X₁), and

1060 (␯^X₁₇) cm^Ϫ1have been assigned to levels with reason-

ably large CCO internal coordinate contributions. This as-

signs all totally symmetric fundamentals calculated to be be-

low 1200 cm^Ϫ1. The theoretical calculations have shown

that within 1300–1450 cm^Ϫ1there are four totally symmet-

ric vibrational modes that all involve some CO stretch and/or

CCO bend movement in their motion. Only two of those,

X

likely ␯ and ␯₁₁, appear obviously in the spectrum and

12

their peaks are at 1426 and 1460 cm^Ϫ1, respectively. Their

proximity may result in intensity borrowing between these

bands. Nontotally symmetric fundamentals are not allowed

FIG. 3. Survey scans of the LIF spectra of the second-

ary alkoxies. Note the difference in frequency scales.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11787

TABLE VIII. Assignment of the DF spectrum of the T conformer of

1-propoxy radical. The vibrational assignment is made for the emission

pumping bands shown in Fig. 5.

Frequency

Assignment

Predicted^a

Experimental^b

˜

A

X

0

274

␯

0

X

270

25

A

321

␯

0

340^c

527

X

475

614

␯

2␯

23

^A₂₅ϩ␯

A

591

796

552

662

844

895

1017

1050

1060

1193

1311

1332

1352

1426

1460

0

X

24

A

0

␯

^A₂₃ϩ␯

X

889

1034

1074

1128

21

X

19

X

18

X

17

^A₂₁ϩ␯

^A₁₉ϩ␯

^A₁₈ϩ␯

^A₁₇ϩ␯

A

0

1210

1355

1395

1449

A

0

A

0

A

0

X

1414

1436

12

X

11

FIG. 5. An expanded view of the DF spectra of the T and G conformations

of 1-propoxy radical. The top trace was obtained by pumping band B ͑Fig.

2͒, the origin of the T conformer, at 29 218 cm^Ϫ1while the bottom trace was

obtained by pumping band A ͑Fig. 2͒, the origin of the G conformer, at

28 634 cm^Ϫ1. The x axis indicates ͑red͒ shift from the pump frequency. The

intensities are not corrected for the grating efﬁciency and the camera sensi-

a

*

Unscaled frequencies computed at the B3LYP/6-31ϩG level of theory ͑in

cm^Ϫ1). The vibrational levels of the A state were estimated by adding the

˜

A

˜

experimentally observed vibrationless level of the A state ͑denoted as ␯₀)

to the corresponding theoretically predicted frequencies under the column

of assignment. The assignment was made on the grounds that the PES for

the lowest two energy states does not change considerably.

tivity curves. The experimental resolution was 15 cm^Ϫ1

.

^bExperimentally observed vibrational bands ͑in cm^Ϫ1) relative to the vibra-

X

0

˜

tionless level of the X state ␯

.

^cPossible shoulder.

by symmetry considerations, and their overtones, though are

not prohibited, are expected to be generally weak. However,

X

the ﬁrst overtone of ␯₂₄, which is the most pseudo-Jahn-

Teller-active vibrational mode, may be a suitable candidate

A

and ␯₁₇, respectively. It should be pointed out that there are

vibrational progressions associated with ␯ and ␯ in both

for the band appearing at 662 cm^Ϫ1

.

17

19

˜

By analogy with the reasoning for the X state assign-

ments, we anticipate the totally symmetric CO stretch modes

to have large ﬂuorescence intensities in the transition to the

˜

the X and A states, probably arising via intensity borrowing

among the closely energy spaced totally symmetric ␯₁₇, ␯

,

18

and ␯ levels.

19

˜

A state as well. However, the identiﬁcation of the vibration-

The G conformer of 1-propoxy is a C₁-symmetry struc-

ture and hence one may not preclude the appearance of any

of its vibrations in the DF spectrum. However, again it is the

CO stretch character that carries most of the oscillator

strength, therefore the strongest transitions should be con-

nected with the CO stretch vibrations. In the DF spectrum of

˜

less level of the A state is the ﬁrst step for assignments of the

˜

A state vibrational levels. As Fig. 4 shows, we have identi-

ﬁed the band at 321 cm^Ϫ1as the vibrationless level of the A

˜

state. The two theoretically calculated X state vibrational fre-

quencies closest to this value that are expected to show ﬂuo-

rescence intensity correspond to 270 and 475 cm^Ϫ1͑see

Table IV͒. In addition to the mismatch in frequency ͑which is

large but not impossible given possible computational er-

rors͒, the magnitude of the observed ﬂuorescence intensity

discourages the assignment of this band to either of these

˜

the G conformer ͑see Fig. 5͒, a strong progression in the X

state can immediately be identiﬁed with a vibrational spacing

of 1066 cm^Ϫ1, which as in the T conformer, we attribute to

␯

͑see Table 2 of the Supplementary Material͒. The second

18

strongest transition below 1700 cm^Ϫ1peaks at 973 cm^Ϫ1and

is assigned to ␯₁₉, which also has signiﬁcant CO stretch

character. Table IX presents a number of assigned vibrational

bands based upon matches with calculated frequencies. In-

deed in this C₁-symmetry molecule, there appears to be a

peak assignable to every vibrational mode expected below

˜ ˜

two vibrations. Our theoretically calculated A-X separation

quoted in Table III, while as noted in Sec. III A, subject to

considerable uncertainty, is consistent with this assignment.

˜

Similar to the X state, a vibrational progression associated

˜

with the ␯ mode appears in the A state, with an energy

1200 cm^Ϫ1

.

18

interval of 1011 cm^Ϫ1from the above identiﬁed A state ori-

˜

The bands appearing at 1337, 1365, and 1546 cm^Ϫ1have

been assigned as combination bands of one quantum of ex-

˜

gin. Using the other X state assignments as a guide, we have

assigned the bands at 552, 844, 1193, 1311, and 1352 cm^Ϫ1

X

citation of ␯ with, respectively, two quanta of ␯₂₇, one

quantum of ␯₂₅, and one quantum of ␯₂₃. The corresponding

18

A

19

X

˜

to the following modes of the A state: ␯₂₅, ␯₂₃, ␯₂₁, ␯

,

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

11788

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Jin et al.

TABLE IX. Assignment of the DF spectrum of the G conformer of

1-propoxy radical. The vibrational assignment is made for the emission

pumping bands shown in Fig. 5.

˜

conformation one can easily identify the A state origin as

well as the vibrational structures built upon the bands be-

longing to it, in the G conformer it is difﬁcult to make com-

parable assignments. What is different in the later case is that

the symmetry is lower. A single argument shows that a

Frequency

Assignment

Predicted^a

Experimental^b

˜

strongly diminished A state origin intensity is possible. As-

˜

A

X

˜ ˜

X

sume for the C_s(T) conformer the B-A and B-X intensities

are exactly the same. Further assume that the electronic

eigenfunctions of the G conformer are just the symmetric

and antisymmetric linear combinations of the T state eigen-

functions. Both assumptions are qualitatively reasonable and

0

135

214

302

383

␯

0

X

125

27

A

0

X

286

305

496

25

X

24

X

485

650

23

␯

^X₂₃ϩ␯

X

˜ ˜

together yield the result that the B-X intensity increases by a

621

27

X

788

867

995

1114

1092

743

859

973

˜ ˜

22

factor of 2 going from the T to G conformer while the B-A

intensity exactly goes to zero.

X

21

X

19

X

Of course we do not expect this simple model to be

completely correct and we also expect some variation in the

1039

1066

1204

1254

1282

1337

1365

1546

1947

2033

2132

2168

2251

2350

2399

2427

2603

3000

3089

3183

3223

3316

3409

17

X

18

^A₁₉ϩ␯

^A₁₇ϩ␯

^A₁₈ϩ␯

A

0

1209

1328

1306

˜ ˜

B-X and B-A intensities along the strong CO stretch pro-

˜ ˜

A

0

A

0

gression. Indeed for the alkoxies as a whole the B-A transi-

^X₁₈ϩ2␯

X

27

tion seems to slightly increase in intensity for higher CO

˜ ˜

1342

1378

1588

1990

2228

2184

^X₁₈ϩ␯

X

25

stretch quanta. It therefore seems reasonable to look for B-A

^X₁₈ϩ␯

X

23

CO stretch bands in the vicinity of the corresponding B-X

X

2␯

19

X

bands.

17

X

17

The triplet of bands ␯₁₉, ␯₁₈, and ␯ exhibit high ﬂuo-

18

2␯^A₁₉ϩ␯

2␯^A₁₇ϩ␯

2␯^A₁₈ϩ␯

A

0

2204

2442

2398

˜

rescence intensities in the X state, hence they are good can-

A

0

˜

A

0

didates for the assignment of their analog in the A state.

Indeed, the peaks appearing at 1204, 1254, and 1282 cm^Ϫ1

2␯^X₁₈ϩ2␯

X

27

2434

2470

2680

2985

3342

3276

2␯^X₁₈ϩ␯

X

A

25

could possibly be identiﬁed as ␯₁₉, ␯₁₇, and ␯₁₈, respec-

2␯^X₁₈ϩ␯

X

23

˜

tively. All three of these A state bands show a proﬁle of

intensities that appear decreased in a proportional way with

X

3␯

19

X

17

X

˜

respect to corresponding X state ones. From these lines one

would deduce an A-X separation of 214 cm^Ϫ1. Moreover, a

18

3␯^A₁₉ϩ␯

3␯^A₁₇ϩ␯

3␯^A₁₈ϩ␯

A

0

3199

3556

3490

˜ ˜

A

0

Ϫ1

A

0

˜ ˜

peak displaced from the B-X origin by 214 cm actually

appears in the DF spectrum albeit with a very low intensity.

Going to a higher-frequency region, we identiﬁed the ﬁrst

a

*

Unscaled harmonic frequencies computed at the B3LYP/6-31ϩG level of

theory ͑in cm^Ϫ1).

X

^bExperimentally observed vibrational bands ͑in cm^Ϫ1) relative to the vibra-

overtones of ␯₁₉, ␯₁₇, and ␯ peaking at the frequencies of

18

X

˜

1947, 2033, and 2132 cm^Ϫ1, respectively. The same pattern

tionless level of the X state ␯

.

0

˜

as before in the X state is consistently reproduced in the A

state, i.e., the peaks at 2168, 2251, and 2350 cm^Ϫ1may be

X

˜

combination bands but with two quanta of excitation of ␯

assigned as their corresponding A state bands. The second

18

overtones of the same modes appearing at 3000 (␯^X₁₉), 3089

(␯^X₁₇), and 3183 (␯₁^X₈) cm^Ϫ1exhibit ﬂuorescence intensity,

lead to the appearance of the following peaks in the DF

spectrum: 2399, 2427, and 2603 cm^Ϫ1. This repeated pattern

of observed frequencies in the energy region higher than

˜

with corresponding bands in the A state at 3223, 3316, and

2000 cm^Ϫ1of redshift strongly validates our previous assign-

3409 cm^Ϫ1

,

respectively. Obviously the poor Franck-

X

18

ments of the combination bands with ␯

.

˜

Condon factors for the A state make it impossible for the

majority of the vibrational modes to appear in the spectrum

One notes that unlike the T conformer, in the G con-

˜

former the A state origin ͑nor excited vibrations͒ does not

distinctly appear in the spectrum. The theoretical predictions

˜

save for the strongest modes identiﬁed in the X state. We

˜ ˜

consider the determination of the A-X separation of

˜

͑see Table III͒ indicate that the A state origin lies very close

214 cm^Ϫ1to be likely correct. However, the large number of

vibrational modes possibly active in the G conformer, makes

it impossible to deﬁnitely afﬁrm this value or rule out that

˜

to the X state origin. The experimental resolution could be

˜ ˜

insufﬁcient to resolve the B-A and B-X transitions. In this

case the experimental resolution places an upper limit on the

˜ ˜

the A-X separation is just too small to be resolved.

The largest discrepancies between the calculated har-

monic frequencies and the experimentally observed anhar-

Ϫ1

˜ ˜

A-X separation of Շ15 cm

.

˜

A second possibility would be that the A state origin

X

A

18

˜

monic frequencies lie, for the T conformer, with ␯₁₇, ␯

,

appears at a resolvable separation from that of the X state but

with considerably diminished intensity. Whereas in the T

A

X

and ␯ while for the G conformer with ␯ and ␯₁₇. We

17 24

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11789

FIG. 7. An expanded view of the DF spectra of the T₁G₂, T₁T₂, and G₁T₂

conformations of 1-butoxy radical. The top trace was obtained by pumping

band C ͑Fig. 2͒, the origin of the T₁G₂conformer, at 29 164 cm^Ϫ1, the

middle trace was obtained by pumping band B ͑Fig. 2͒, the origin of the

T₁T₂conformer, at 29 095 cm^Ϫ1while the bottom trace by pumping band A

͑Fig. 2͒, the origin of the G₁T₂conformer, at 28 649 cm^Ϫ1. The x axis

indicates ͑red͒ shift from the pump frequency. The intensities are not cor-

rected for the grating efﬁciency and the camera sensitivity curves. The ex-

FIG. 6. DF spectra of the T₁G₂, T₁T₂, and G₁T₂conformations of

1-butoxy radical. The top trace was obtained by pumping band C ͑Fig. 2͒,

the origin of the T₁G₂conformer, at 29 164 cm^Ϫ1, the middle trace was

obtained by pumping band B ͑Fig. 2͒, the origin of the T₁T₂conformer, at

29 095 cm^Ϫ1while the bottom trace by pumping band A ͑Fig. 2͒, the origin

of the G₁T₂conformer, at 28 649 cm^Ϫ1. The x axis indicates ͑red͒ shift

from the pump frequency. Assigned members of the predominant CO stretch

progression, ␯₂₃(T₁T₂), ␯₂₄(T₁G₂), and ␯₂₃(G₁T₂), are indicated by ticks

perimental resolution was 15 cm^Ϫ1

.

shows an expanded view of their DF spectra up to

Ϸ1600 cm^Ϫ1redshifted from the pumping frequency.

Using our experience with the

T conformer of

˜ ˜

*

on the top horizontal bar while the indicates the line assigned to the A-X

origin transition. The intensities are not corrected for the grating efﬁciency

and the camera sensitivity curves. The experimental resolution was

1-propoxy, the vibrational interval of 1052 cm^Ϫ1measured

˜

from the X state origin may be associated with the CO

15 cm^Ϫ1

.

˜

stretch progression in the X state for the T₁T₂conformation

of 1-butoxy. Since the vibrational mode ␯ encompasses the

23

largest CO stretch character we assign this band and the

higher members of the progression to this mode. On the

grounds of matching well with the theoretically calculated

frequencies ͑see Table X͒, the following low-intensity, totally

need to recognize that while the B3LYP methodology is gen-

erally a cost-effective procedure for predicting reliable vibra-

tional frequencies, its results are obtained by use of the har-

monic approximation. Neither possible anharmonicity effects

˜

symmetric fundamentals were assigned in the X state: the

X

bands at 395 (␯₃^X₁), 950 (␯₂^X₇), and 1116 (␯₂^X₂) cm^Ϫ1involve

medium CCO deformation in their motion, and the band at

463 (␯^X₃₀) cm^Ϫ1large CCO deformation. The band at

͑probably ␯₁₇of T and G conformers͒ nor vibronic coupling

A

25

X

24

˜

interactions between the X and A states (␯ and 2␯ of the

T conformer͒ are incorporated in the vibrational frequency

X

976 cm^Ϫ1is identiﬁed as ␯₂₆, a mode with moderate CO

calculations. In addition, mechanisms like intensity borrow-

A

stretch and large CCO bend character. The band at 183 cm^Ϫ1

ing between energetically closely spaced bands ͑e.g., ␯

␯

,

19

A

X

₁₈, and ␯₁₇of the T conformer͒ of the same symmetry may

was assigned to ␯₃₄and is a backbone ﬂex mode. The vibra-

not be completely described by such calculations.

tion ␯₂₄involves a small amount of CO stretch character and

is assigned to the peak at 1017 cm^Ϫ1in the X state. Finally,

˜

the bands peaking at 1279 (␯^X₂₀), 1425 (␯₁^X₇), and 1437

(␯^X₁₆) cm^Ϫ1likely correspond to modes involving small CO

and small CCO deformation.

2. 1-Butoxy

For 1-butoxy, the DF spectra were recorded by pumping

bands A, B, and C shown in Fig. 2. Band A has been

identiﬁed⁶as the origin of the G₁T₂conformer, band B is the

origin of the T₁T₂conformer, and band C of T₁G₂. The DF

spectra of these species are shown in Fig. 6, while Fig. 7

Following the same arguments for the assignment of the

˜

A state origin in the T conformer of 1-propoxy ͑and within

the expectedly large interval of uncertainty of our theoretical

calculations quoted in Table III͒, we have identiﬁed the peak

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

11790

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Jin et al.

TABLE X. Assignment of the DF spectrum of the T₁T₂conformer of 1-butoxy radical. The vibrational

assignment is made for the emission pumping bands shown in Fig. 7.

Frequency

Assignment

Predicted^a

Experimental^b

˜

X

A

X

0

183

292

395

463

688

895

950

976

1017

1052

1116

1148

1191

1229

1255

1279

1319

1348

1400

1425

1437

1445

1493

1572

␯

0

X

185

34

A

0

X

391

467

31

X

(␯ and ␯^A₃₄ϩ␯^A₀)^c

477

683

759

30

A

␯

^A₃₁ϩ␯

^A₃₀ϩ␯

0

A

0

X

914

995

1061

1079

27

X

26

X

24

X

23

X

(␯ and ␯^X₂₇ϩ␯

)

X

34

c

1133 and 1099

1180

22

X

␯

^X₂₆ϩ␯

^X₂₄ϩ␯

^A₂₇ϩ␯

^A₂₆ϩ␯

34

X

1246

34

A

1206

1287

0

A

0

X

20

1282

^A₂₄ϩ␯

^A₂₃ϩ␯

A

0

1353

1371

1425

1391

A

0

^A₂₂ϩ␯

A

0

X

(␯ and ␯^A₂₇ϩ␯^A₃₄ϩ␯₀^A)^c

1369

1393

1413

17

X

␯

16

X

(␯ and ␯^A₂₆ϩ␯^A₃₄ϩ␯₀^A)^c

1472

1538

1574

15

A

0

␯

^A₂₄ϩ␯^A₃₄ϩ␯

^A₂₀ϩ␯

A

0

^aSee footnote a of Table VIII.

X

0

^bExperimentally observed vibrational bands ͑in cm^Ϫ1) relative to the vibrationless level of the X state ␯

.

˜

^cThe experimental bands may have dual assignments.

X

at 292 cm^Ϫ1as the vibrationless level of the A state. The

band redshifted from it by 1056 cm^Ϫ1͑at 1348 cm^Ϫ1) is

˜

CCO motion, respectively. Mode ␯ involves medium CCO

23

and small CO movements and reasonably correlates with a

peak at 1071 cm^Ϫ1and mode ␯ is characterized by a me-

X

22

correlated with the strong CO stretch character vibration ␯

,

23

˜

dium CO movement and corresponds to the peak appearing

at 1117 cm^Ϫ1. The bands centered at 1135 (␯₂^X₁), 1244

(␯^X₂₀), 1300 (␯^X₁₉), and 1338 (␯₁^X₈) are built upon vibrational

modes that include small amounts of CO stretch character in

their motion.

a mode that also leads a vibrational progression in the A

˜

state. By reference to our X state assignments, we make

˜

similar assignments in the A state, as detailed in Table X.

While the T₁T₂conformer of 1-butoxy is a near-C_ssym-

metry structure, the geometry of the T₁G₂conformation is

characterized by signiﬁcant departure from the C_s-symmetry

˜

The assignment of the A state origin is based on the

consideration that the strong CO stretch progression appear-

˜

reference structure. The CO stretch progression in the X state

˜

ing in the X state should also appear in the A state. The band

is easily identiﬁable in Fig. 6. The vibrational 1-0 spacing of

the ﬁrst number of this relatively strong CO stretch progres-

at 1300 cm^Ϫ1is most likely the ﬁrst member of the CO

sion is 1050 cm^Ϫ1and we associated ␯ with strong CO

˜

24

stretch progression in the A state, and is likewise assigned to

stretch character mode with this band. ͑See Table 5 of the

Supplementary Material.͒ Using as a reference of the calcu-

lated frequencies, we make additional assignments of the

A

˜

␯

₂₄. This assignment identiﬁes the origin of the A state at

271 cm^Ϫ1above the X state. Other observed peaks at fre-

˜

quencies lower than 1600 cm^Ϫ1that may be assigned to the

˜

lower intensity bands in the X state. The weak peaks at 79

˜

A state are given in Table XI.

(␯^X₃₆) and 101 (␯^X₃₅) cm^Ϫ1are characterized as vibrations

that exhibit backbone ﬂex motions, and the peak at 208

(␯^X₃₄) cm^Ϫ1is characterized by a mode of CH₃torsion and

CCC bend. The bands peaking at 351 (␯^X₃₂) and 484

(␯^X₃₁) cm^Ϫ1are attributed to modes involving large CCO de-

formation in their motion, the peaks at 798 (␯^X₂₈), 886 (␯₂^X₇),

and 930 (␯^X₂₆) cm^Ϫ1correspond to modes containing the in-

ternal coordinates of medium-sized CCO and small-sized CO

movement, medium CCO deformation and medium CO/large

For the G₁T₂conformation, a pronounced vibrational

progression with a 1-0 interval of 1067 cm^Ϫ1is attributed to

˜

the CO stretch motion in the X state and is assigned to ␯

,

23

the character of which is predominantly the CO stretch

movement ͑see Table 6 of the Supplementary Material͒.

Based upon the match with the calculated frequencies, we

assigned the following low-intensity bands: 259 (␯^X₃₃) cm^Ϫ1

corresponds to a CH₃/CH₂torsion and CCC bend mode, 337

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11791

TABLE XI. Assignment of the DF spectrum of the T₁G₂conformer of 1-butoxy radical. The vibrational

assignment is made for the emission pumping bands shown in Fig. 7.

Frequency

Assignment

Predicted^a

Experimental^b

˜

X

A

X

0

79

101

208

271

351

367

484

621

756

798

886

930

1050

1071

1117

1135

1155

1201

1244

1300

1338

1342

1380

1397

1511

1567

1610

␯

0

X

97

117

208

36

X

35

X

34

A

␯

0

X

(␯ and ␯^A₃₆ϩ␯^A₀)^c

346

474

368

388

479

617

745

32

^A₃₅ϩ␯

A

␯

0

(␯ and ␯^A₃₄ϩ␯^A₀)^c

X

31

A

␯

^A₃₂ϩ␯

^A₃₁ϩ␯

0

A

0

X

869

887

957

1068

1078

1110

1192

28

X

27

X

26

(␯ and ␯^A₂₈ϩ␯^A₀)^c

X

24

1140

X

␯

23

X

22

X

21

^A₂₇ϩ␯

^A₂₆ϩ␯

A

1158

1228

0

A

0

X

20

1256

1307

1338

(␯ and ␯^A₂₄ϩ␯^A₀)^c

X

19

1339

X

18

␯

A

1349

1381

1463

1527

1578

1609

␯

^A₂₃ϩ␯

^A₂₂ϩ␯

^A₂₁ϩ␯

^A₂₀ϩ␯

^A₁₉ϩ␯

^A₁₈ϩ␯

0

A

0

A

0

A

0

A

0

A

0

^aSee footnote a of Table VIII.

X

0

^bExperimentally observed vibrational bands ͑in cm^Ϫ1) relative to the vibrationless level of the X state ␯

.

˜

^cThe experimental bands may have dual assignments.

A

(␯^X₃₂) cm^Ϫ1involves large CCO motion, 374 (␯₃^X₁) cm^Ϫ1is

characterized by medium-sized CCO motion, 510

(␯^X₃₀) cm^Ϫ1includes large contribution from the CCO inter-

nal coordinate, 850 (␯₂^X₈) cm^Ϫ1contains a small-sized CO

stretch character, 879 (␯^X₂₇) cm^Ϫ1is described by small CCO

and small CO participation, 946 (␯^X₂₆) cm^Ϫ1may be attrib-

uted to medium CCO and medium CO participation, 1033

(␯^X₂₅) cm^Ϫ1involves a small CO stretch activity in its mo-

tion, 1061 (␯^X₂₄) cm^Ϫ1includes the internal coordinate of a

medium character CO stretch motion, 1132 (␯^X₂₂) cm^Ϫ1re-

fers to a medium character CCO deformation, 1285

(␯₂^X₀) cm^Ϫ1and 1329 (␯₁^X₈) cm^Ϫ1correspond to a small con-

tribution of the CCO deformation internal coordinate in their

modes, and 1362 (␯₁^X₇) cm^Ϫ1includes a small CO stretch

character in its motion.

1196 cm^Ϫ1which we again associate with the mode ␯₂₃. By

˜

comparison with the X state assignments and having identi-

˜

ﬁed the A state origin in the T₁G₂conformer, it is a straight-

˜

forward task to complete our assignments in the A state.

Other observed peaks are assigned in Table XII.

3. 2-Propoxy

The DF spectra for the unique conformer of 2-propoxy

were recorded via two bands (A and B in Fig. 3͒, namely the

origin band and the ﬁrst member of the CO stretch progres-

sion. Figure 8 includes, in the top trace, the extended DF

spectrum recorded via the ﬁrst member of the CO progres-

˜

sion in the B state, and in the bottom trace an expanded view

of the region redshifted up to Ϸ1300 cm^Ϫ1from the pump-

ing frequency. The DF spectrum through the origin is almost

identical with the spectrum recorded by pumping one quan-

tum of CO stretch excitation but the latter has better signal/

noise and we focus our analysis on it.

˜

In the T₁T₂conformation the A state origin band ap-

pears relatively strong while it is somewhat diminished in the

G₁T₂conformer. In this sense the conformers of 1-butoxy

˜ ˜

show the same trend with respect to B-A origin intensity as

was suggested above for 1-propoxy. However, in G₁T₂

Our assignments for the nominally C_s2-propoxy are

necessarily built upon the same principles as for the primary

1-butoxy there is an observable band at 129 cm^Ϫ1which is

˜

likely the A state origin. It then follows that there is a CO

alkoxies. For the X state, we have identiﬁed a vibrational

˜

stretch progression in the A state starting with the 1-0 line at

progression starting at 932 cm^Ϫ1redshifted from the pump-

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

11792

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Jin et al.

TABLE XII. Assignment of the DF spectrum of the G₁T₂conformer of 1-butoxy radical. The vibrational

assignment is made for the emission pumping bands shown in Fig. 7.

Frequency

Assignment

Predicted^a

Experimental^b

˜

X

A

X

0

129

259

337

374

387

466

510

640

850

879

946

979

1008

1033

1061

1067

1132

1162

1190

1196

1260

1285

1329

1362

1414

1460

1491

␯

0

A

0

X

253

341

426

33

X

32

X

31

^A₃₃ϩ␯

A

382

470

555

688

0

A

0

^A₃₂ϩ␯

X

(␯ and ␯^A₃₁ϩ␯^A₀)^c

559

30

A

0

␯

^A₃₀ϩ␯

X

855

873

960

28

X

27

X

26

^A₂₈ϩ␯

^A₂₇ϩ␯

A

0

984

1002

A

0

X

1030

1050

1089

1135

25

X

24

X

(␯ and ␯^A₂₆ϩ␯^A₀)^c

1089

23

X

␯

22

A

1159

1179

1218

1264

␯

^A₂₅ϩ␯

^A₂₄ϩ␯

^A₂₃ϩ␯

^A₂₂ϩ␯

0

A

0

A

0

A

0

X

1271

1337

1370

20

X

18

X

17

^A₂₀ϩ␯

^A₁₈ϩ␯

^A₁₇ϩ␯

A

0

1400

1466

1499

A

0

A

0

^aSee footnote a of Table VIII.

X

0

^bExperimentally observed vibrational bands ͑in cm^Ϫ1) relative to the vibrationless level of the X state ␯

.

˜

^cThe experimental bands may have dual assignments.

ing frequency ͑see Fig. 8͒ with ␯₁₉, a mode that bears sig-

niﬁcant CO stretch character ͑see Table VI͒. However, as

Table VI shows in 2-propoxy, unlike the primary alkoxy

radicals, there is a second mode ␯₁₄, which has comparable

CO stretch character. Referring to Table XIII we see that ␯

14

is assigned to a band at 1169 cm^Ϫ1. Clearly the 932 and

1169 cm^Ϫ1frequencies are much lower and higher, respec-

tively, than the CO stretch frequencies deduced above for the

primary alkoxies. We propose that when discussing CO

stretch motion in 2-propoxy, and comparing it with other

alkoxies, the average value, 1050 cm^Ϫ1, of these two fre-

quencies be considered. Based upon the match ͑see Table

XIII͒ with the calculated frequencies, we have assigned the

other bands shown in the spectrum Շ1200 cm^Ϫ1to the vi-

brational modes placed within the parentheses ͑in cm^Ϫ1):

254 (␯^X₂₆), 398 (␯₂^X₄), 483 (␯₂^X₃), 837 (␯^X₂₂), 1002 (␯^X₁₈),

1033 (␯^X₁₇), and 1125 (␯₁^X₅). The character of each of the

assigned modes in terms of its constituent internal coordi-

nates is shown in Table VI.

FIG. 8. DF spectra of 2-propoxy radical. The top trace was obtained by

pumping band B ͑Fig. 3͒, the frequency of the ﬁrst member of the CO

stretch progression, at 27 734 cm^Ϫ1. The bottom trace corresponds to an

expanded view of the DF data shown in the top trace. The x axis indicates

͑red͒ shift from the pump frequency. Assigned members of the predominant

˜

The assignment of the A state origin is somewhat chal-

˜

lenging. After the above fairly straightforward X state assign-

CO stretch progression ␯ are indicated by ticks on the top horizontal bar

19

ments, there are two relatively strong bands, in the low-

frequency region at 68 and 524 cm^Ϫ1, which are plausible

˜ ˜

*

while the indicates the line assigned to the A-X origin transition. The

intensities are not corrected for the grating efﬁciency and the camera sensi-

candidates. We favor the assignment of the B-X origin to the

tivity curves. The experimental resolution was 30–35 cm^Ϫ1

.

˜ ˜

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11793

TABLE XIII. Assignment of the DF spectrum pumped via the ﬁrst member of the CO stretch progression of the

excitation spectrum of 2-propoxy radical. The vibrational assignment is made for the emission pumping bands

shown in Fig. 8.

Frequency

Assignment

Predicted^a

Experimental^b

˜

X

A

X

0

68

254

398

483

524

837

892

932

␯

0

A

0

X

244

390

453

26

X

24

X

␯

23

A

521

877

␯

^A₂₃ϩ␯

0

X

22

809

984

␯

^A₂₂ϩ␯

A

0

X

19

␯

^A₁₉ϩ␯

A

0

1052

981

X

1054

1069

1002

1033

1054

1125

1169

1208

␯

18

X

17

A

0

1122

␯

^A₁₈ϩ␯

X

1207

1261

␯

15

X

14

(␯₁^A₄ϩ␯ and ␯₁^A₅ϩ␯^A₀)^c

A

0

1329 and 1275

^aSee footnote a of Table VIII.

X

0

^bExperimentally observed vibrational bands ͑in cm^Ϫ1) relative to the vibrationless level of the X state ␯

.

˜

^cThe experimental bands may have dual assignments.

68 cm^Ϫ1band for a couple of reasons. First there seems no

other explanation for a band at such a low frequency, as

Table VI indicates the lowest vibrational frequency should be

Ϸ200 cm^Ϫ1. In addition as Table III shows, both the calcu-

DF spectra from lines A and C and separately B and D are

nearly identical. Thus we believe that the LIF lines A and C

likely belong to a given conformer, which we call A, and

lines B and D belong to a second conformer, which we

call B.

In addition to the conformer assignment issue, the gen-

eral appearance of the 2-butoxy DF spectra is signiﬁcantly

different from the other species. Clearly there are fewer in-

dividual lines observed in the low-frequency, redshift region

and as the frequency increases, the lines become broader and

congestion becomes greater. There is also a tendency, as we

will see below, not to see a precise repetition of fundamental

frequencies in the combinations, nor a precise repetition be-

˜ ˜

indicate a value of Շ150 cm^Ϫ1, clearly inconsistent with an

lated and experimental trends for the A-X separation would

assignment to the 524 cm^Ϫ1. As Table XIII indicates there is

A

a good frequency match for the 524 cm^Ϫ1band to ␯₀^Aϩ␯

23

A

assuming the ␯ assignment of 68 cm^Ϫ1. The main problem

0

with the assignment is that there seems no particularly good

reason for the relatively strong appearance of this band, al-

˜ ˜

though a very small A-X separation may facilitate signiﬁcant

˜

vibronic perturbations between the A and X states with re-

˜ ˜

tween the A-X intervals. We believe all of these effects stem

sulting intensity anomalies. Using 68 cm^Ϫ1as the value for

from two factors: ͑i͒ the increasing size of the molecule with

additional vibrational degrees of freedom and ͑ii͒ a small

␯^A₀, the vibrational interval of the CO stretch progression

A

˜

formed in the A state ͑attributed to ␯₁₉) is, for the DF spec-

˜ ˜

A-X separation which combined with ͑i͒ engenders extensive

˜ ˜

904 cm^Ϫ1and for the spectrum pumped through its ﬁrst CO

trum pumped via the origin of the B-X LIF excitation,

˜

vibronic interactions between the A and X states.

As a consequence of the above we limit ourselves to a

somewhat less quantitative and less extensive analysis for

2-butoxy. Nonetheless, some quantitative information can be

obtained from the spectra and is summarized in Tables XIV

and XV. Using these Tables and Figs. 9 and 10 as a guide we

can make the following observations. For conformer A there

is a band at 125 cm^Ϫ1, much like the 68 cm^Ϫ1band in

2-propoxy. There is only one vibrational mode ͑see Tables

10–12 of the Supplementary Material͒, a backbone ﬂex for

which little intensity would be expected in the DF spectrum,

at anywhere near this low a frequency. Hence like 2-propoxy,

stretch member 913 cm^Ϫ1, making an average of 908 cm^Ϫ1

.

˜

Having the assignments to the X state as a guide, we make

˜

corresponding assignments to the A state as follows ͑in

A

cm^Ϫ1): 892 (␯^A₂₂), 981 (␯₁^A₉), 1054 (␯^A₁₈), and 1208 (␯₁₄and

A

␯

unresolved͒.

15

4. 2-Butoxy

There are some important differences between the

2-butoxy spectra and the other species that we have ob-

served. Like most of the other species there are multiple

conformers, denoted Gϩ, GϪ, and T ͑see Fig. 1͒ for a total

of 3. However, unlike the other species there are no rota-

tional analyses that assign a group of lines in the LIF spec-

trum to a given conformer. However, Fig. 10 shows that the

˜ ˜

B, a corresponding band appears at 55 cm^Ϫ1, which can be

we assign this band to the B-A origin. Turning to conformer

˜ ˜

identiﬁed as the B-A origin with even greater conﬁdence.

As one would expect for both conformers there is a rela-

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

11794

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Jin et al.

TABLE XIV. Assignment of the DF spectrum of the ‘‘B’’ conformer of 2-butoxy radical. The vibrational

assignment is made for the emission pumping bands shown in Fig. 10.

Frequency

Assignment

Experimental^a(B)

Experimental^a(D)

Average^b

Predicted^c

˜

X

A

X

0

58

509

929

992

1045

1105

1142

1188

0

53

509

929

993

1044

1104

1142

1187

0

55

509

929

992

1044

1104

1142

1187

␯

0

A

0

A

529

989

␯

^A₃₀ϩ␯

0

X

25

^A₂₅ϩ␯

A

0

1044

1101

1258

X

23

1046

1203

^A₂₃ϩ␯

A

0

X

20

^A₂₀ϩ␯

A

0

^aExperimentally observed vibrational bands ͑in cm^Ϫ1) relative to the excitation frequency of the vibrationless

X

0

˜

level of the X state ␯

.

^bThe average values between the observed vibrational bands shown under the second and third column of the

experimental results.

c

Unscaled frequencies computed at the B3LYP/6-31ϩG level of theory ͑in cm^Ϫ1). The vibrational levels of

*

˜

the A state were estimated by adding the average of the two experimentally observed vibrationless levels of the

A

˜

A state ͑denoted as ␯₀) to the corresponding theoretically predicted frequencies under the column of assign-

ment. The predicted frequencies were obtained for the T conformer which possibly correlates with bands B and

D. See text for details. The assignment was made on the grounds that the PES for the lowest two energy states

does not change considerably.

tively long CO stretch progression but only the ﬁrst member

of it has signiﬁcantly resolved structure. In conformer A, it

appears that there are four discernible lines ͑see Table XV͒

which we assign to the two bands with strongest CO stretch

2-butoxy spectra. Nonetheless there is one band around

500 cm^Ϫ1(554 cm^Ϫ1in conformer A and 509 cm^Ϫ1in con-

former B). These are somewhat similar to a corresponding

trend in this region for 2-propoxy, which we suggested was a

˜

˜ ˜

possible, but not particularly likely, B-A origin. The same

character for each the A and X states. Referring to Table XV

˜ ˜

must be said for the above bands in 2-butoxy. However, we

we note the A-X separations do not match those of the origin

A

region all that well, but we attribute this to the strong vi-

bronic interactions among the multiple levels in this region.

For conformer B, the pattern is even more complex in the

CO stretch region, but it is again at least consistent with that

expected.

feel the most likely assignment of this band is to ␯ for the

31

A

30

A conformer and ␯ for the B conformer, although like

2-propoxy the reason for the relatively strong intensity for a

band in this region is not obvious.

Before closing this section it is worth speculating a mo-

ment upon the correlation between the conformer labels A

and B assigned to various spectral bands and the actual con-

formers, Gϩ, GϪ, and T which have been the subject of

˜

To this point, except for the A state origin band, we have

ignored the region redshifted 0–900 cm^Ϫ1. As the ﬁgures

and tables show there is little activity in this region in our

TABLE XV. Assignment of the DF spectrum of the ‘‘A’’ conformer of 2-butoxy radical. The vibrational

assignment is made for the emission pumping bands shown in Fig. 10.

Frequency

Experimental^a(C)

Assignment

Experimental^a(A)

Average^b

Predicted^c

˜

X

A

X

0

129

484

555

916

1000

1136

1253

0

121

493

553

916

1000

1134

1252

0

125

488

554

916

1000

1135

1252

␯

A

0

^A₃₂ϩ␯

^A₃₁ϩ␯

A

490

556

0

A

0

X

936

1030

1194

26

X

(␯ and ␯^A₂₆ϩ␯^A₀)^d

1061

1155

1319

24

X

(␯ and ␯^A₂₄ϩ␯^A₀)^d

20

A

0

␯

^A₂₀ϩ␯

X

^aExperimentally observed vibrational bands ͑in cm^Ϫ1) relative to the vibrationless level of the X state ␯

.

˜

0

^bThe average values between the observed vibrational bands shown under the second and third column of the

experimental results.

^cSee footnote c of Table XIV. The predicted frequencies were obtained for the Gϩ conformer which possibly

correlates with bands A and C. See text for details.

^dThe experimental bands may have multiple assignments.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11795

FIG. 9. DF spectra of the A and B conformations of 2-butoxy radical. The

lowest trace was obtained by pumping band A ͑Fig. 3͒, the origin of the A

conformer at 26 760 cm^Ϫ1while the next to the lowest trace was obtained

FIG. 10. An expanded view of the DF spectra of the A and B conformations

of 2-butoxy radical. The lowest trace was obtained by pumping band A ͑Fig.

3͒, the origin of the A conformer at 26 760 cm^Ϫ1while the next to the

lowest trace was obtained by pumping band B ͑Fig. 3͒, the origin of the B

conformer, at 27 068 cm^Ϫ1. The trace obtained by pumping band C, at

27 321 cm^Ϫ1, corresponds to the transition to the ﬁrst quantum of the CO

by pumping band B ͑Fig. 3͒, the origin of the B conformer, at 27 068 cm^Ϫ1

.

The trace obtained by pumping band C, at 27 321 cm^Ϫ1, correspond to the

transition to the ﬁrst quantum of the CO stretch vibration of the A conformer

in the B state, while the trace through D, at 27 682 cm^Ϫ1, to the same type

˜

of excitation but for the B conformer. The x axis indicates ͑red͒ shift from

the pump frequency. Assigned members of the predominant CO stretch pro-

gression, ␯ ͑conformer B) and ␯ ͑conformer A), are indicated by ticks

stretch vibration of the A conformer in the B state, while the trace through

D, at 27682 cm^Ϫ1, to the same type of excitation but for the B conformer.

The x axis indicates ͑red͒ shift from the pump frequency. The intensities are

not corrected for the grating efﬁciency and the camera sensitivity curves.

23

24

˜ ˜

*

on the top horizontal bar while the indicates the line assigned to the A-X

origin transition. The intensities are not corrected for the grating efﬁciency

and the camera sensitivity curves. The experimental resolution was

The experimental resolution was 30–35 cm^Ϫ1

.

30–35 cm^Ϫ1

.

other is either the Gϩ or GϪ conformer, with the Gϩ con-

former more likely since it is computed to be the lowest

energy and for the primary alkoxies the lowest energy con-

former has always been observed.

our quantum chemistry calculations. The proper way of mak-

ing this correlation will be via the rotational analyses. How-

ever, the present studies at least provide some hints as to

what the correlation may be.

If one accepts the above reasoning, it remains to suggest

how A and B correlate to Gϩ and T. We argued earlier that

the T conformer seems structurally more similar in the local

O environment to 2-propoxy. Hence it may follow that the

It is useful to compare the C₁symmetry 2-butoxy to

2-propoxy which has C_ssymmetry. As we have previously

observed the ‘‘local’’ symmetry around the O chromophore

is the most important thing for determining the electronic

structure and spectrum. Referring to Fig. 1, it is reasonable to

argue that the local symmetry of T 2-butoxy is rather similar

to 2-propoxy with Gϩ and GϪ, while having similar struc-

tures among themselves, are both somewhat distinguished

from the T conformer. The qualitative observation is sup-

ported by the calculated vibrational frequencies of the three

conformers in Tables 10–12 of the Supplementary Material.

While the vibrational frequencies of all three conformers are

collectively rather similar, some distinctions appear. For ex-

ample all three conformers have two modes sharing the CO

stretch character, but the lower frequency ones are almost

identical ͑933 and 936 cm^Ϫ1) for the GϪ and Gϩ con-

former while it is 989 cm^Ϫ1for the T conformer.

55 cm^Ϫ1A-X separation of the B conformer, being very

˜ ˜

similar to the 68 cm^Ϫ1value for 2-propoxy, makes it more

likely correlated to the T conformer structure, leaving the A

band for the Gϩ conformer. While we believe the analysis

to be reasonable, we stress that deﬁnitive correlation must

await relational analyses of the A and B conformer bands.

IV. DISCUSSION

There are two primary kinds of quantitative information

that result from our DF studies of the alkoxy radicals,

˜ ˜

namely, the A-X separations and the vibrational frequencies

of the two states. We discuss these in turn and wrap up this

section with some observations on qualitative aspects of the

DF spectra.

It is important ﬁrst to comment on the reliability of the

˜ ˜

A-X separations from the DF data. The precision of these

values is Շ10 cm^Ϫ1but given the inherent spectral complex-

We note that there appear to be noticeable differences

between the spectra of the A and B conformers. Based upon

the above we expect one to be the T conformer while the

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

11796

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Jin et al.

˜

ity there is always the possibility of assignment error. For

a fairly large number of X ͑and A) state vibrational frequen-

cies for the alkoxies. Since modes involving the CO stretch

character are the strongest we have the most information for

them.

˜ ˜

example we now report a value of the A-X separation for

ethoxy although none was reported by Zhu, Kamal, and

Misra⁸in their analysis of the DF spectrum. We have reas-

˜

˜ ˜

Table XVI summarizes the experimentally determined X

signed to the B-A origin the line assigned by them to the

˜

and A state CO stretching frequencies of the alkoxy radicals

fundamental vibration ␯₁₈of the X state. We believe that this

˜

is very reasonable given the fact that this vibrational fre-

quency agreed least well of any they reported, when com-

pared to those of similar molecules. It also is a relatively

investigated thus far. It also includes the B state CO frequen-

cies. It can immediately be noticed that the CO stretching

˜

frequency of the X state is always larger than that of the B

state. The lowering of the frequency upon excitation is due to

the expansion of the CO bond upon the electronic transition,

as a consequence of exciting a bonding ␴ electron to a non-

bonding, orbital localized on the O atom. Comparing the CO

˜ ˜

strong transition, consistent with the B-X origin, but does not

show a progression as expected for a strong vibrational band.

Finally it is in good agreement with a value from a photo-

electron detachment study¹⁹and qualitatively with quantum

chemistry calculations.¹²

On the other hand, for 2-propoxy, the only other of these

alkoxy radicals studied by electron detachment,¹⁹there is a

marked discrepancy. However, if our value is correct, then it

˜

frequencies between the X and A states we notice an obvious

similarity, a fact that is not surprising owing to the theoreti-

cally predicted similarity of the corresponding PESs.

˜

The X state CO stretch frequencies of the T conformer

is fairly easy to understand an error in the electron detach-

of 1-propoxy, and T₁T₂and T₁G₂conformations of

1-butoxy are almost identical and very similar to methoxy.

The G conformer of 1-propoxy shows a similar CO fre-

quency with the G₁T₂conformer of 1-butoxy as well, but

higher than the values for the other radicals. It seems that

there is a consistent trend within the conformers of the pri-

Ϫ1

˜ ˜

ment work. It is unlikely that an A-X separation of 68 cm

would have been resolved in those studies. Therefore, an

average Ϸ1225 cm^Ϫ1of two strong peaks was previously

˜

assigned as the A state origin. We believe that transition

corresponds to the CO stretch fundamental. In the report on

the photodetachment work, the authors pointed out that their

˜

mary alkoxies that involves a larger X state CO stretching

frequency for the conformers that have a G local electronic

environment around oxygen.

We conclude our discussion with a few words about the

general appearance of the DF spectra. As expected as the

˜

A state assignment in 2-propoxy did not have the support of

polarization data as did their assignment for ethoxy. We have

pointed out the possibility of assigning our observed line at

524 cm^Ϫ1to the B-A origin, but if the A-X separation were

this large it would be hard to understand while it was not

observed in the photodetachment studies and it would be

rather inconsistent with the quantum chemistry calculations,

which as noted earlier are admittedly not very accurate for

˜ ˜

˜

energy in the X ͑or A) state increases the spectra become

more congested and eventually discrete lines are not ob-

served. It is also expected that this phenomenon occurs at

progressively lower energy as the size of the radical in-

creases.

˜ ˜

the A-X separation. Finally we pointed out that for 2-butoxy

However, what is not particularly expected is signiﬁ-

cantly different behavior for different isomers and conform-

ers which is most clearly indicated for propoxy in Figs. 5 and

8. Striking is the degree of difference for the C_sconformers

of 1- and 2-propoxy. Even more interesting is how different

the spectra of the T and G conformers of 1-propoxy appear.

Consider also the vast difference in the congestion in the T

again it is possible to assign a transition, giving a much

larger A-X separation (տ500 cm^Ϫ1) but the proponderance

˜ ˜

of the evidence favors the Շ130 cm^Ϫ1values.

With these comments we now turn to the discussion of

˜ ˜

the values of the A-X separation in Table III. Several trends

are apparent. For a given species the conformer with the C_s

˜ ˜

plane always has the largest A-X separation and it appears

˜

and G 1-propoxy spectra for the X and A states at nearly the

same energy. These results indicate the vibronic coupling

must be rather isomer and conformer speciﬁc with possible

implications for internal vibrational redistribution and dy-

namics in these species. We are presently developing a pic-

the stronger the deviation from symmetry near the O atom

chromophore, the more pronounced the decrease in splitting.

For a given 1-alkoxy conformer, T₁or G₁near the oxygen

˜ ˜

atom, increasing the alkyl chain decreases the A-X separa-

tion. Finally in going from a primary to a secondary alkoxy

˜ ˜

˜

ture of the coupling in the X, A, and B states and plan to

˜ ˜

radical isomer, the A-X separation decreases signiﬁcantly.

publish these results in the future.

These experimental results are quite valuable, since the

˜

energy of the excited A state is quite important with respect

to its participation in thermal reaction chemistry. Moreover,

as we discussed earlier, the most advanced quantum chemis-

try calculations today can only give a qualitative idea at best

as to its value.

The ability of calculations to predict vibrational frequen-

cies is of course much better and we have strongly relied

upon them for our vibrational assignments. However experi-

ment clearly provides the ‘‘gold standard’’ for the frequen-

cies as well. As we have shown the DF experiments measure

V. CONCLUSION

DF spectra of the alkoxy radicals have been observed in

a free jet expansion environment. Analysis of the spectra

yields the energy separation between the vibrationless levels

˜

of the X and A states for most of the isomers and conformers

of C_nH_2nϩ1O for nϭ3 and 4. These studies also provide

˜

considerable insight into the X and A state vibrational struc-

ture. The CO stretch progression dominates for all DF spec-

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

J. Chem. Phys., Vol. 121, No. 23, 15 December 2004

Alkoxy radicals

11797

TABLE XVI. Summary ͑in cm^Ϫ1) of experimentally determined X, A and B state vibrational fundamental frequencies of the CO stretch mode of the primary

and secondary alkoxy radicals. The table summarizes the results of this work unless noted otherwise. The reported standard deviation of ␻_ex_eis determined

by the ﬁt and it does not include the uncertainties of ␻₀.

˜ ˜

˜

X

˜

A

˜

B

Alkoxies

Conformers

␻

␻_ex_e

␻

␻_ex_e

␻

e

␻_ex_e

0

e

0

e

0

a

c

a

c

b

c

b

c

Methoxy

ethoxy

1-propoxy

1047͑2͒

1057͑3͒

1078͑2͒

7.0͑7͒

6.2͑4͒

662.4͑5͒

603^c

667.4͑1͒

609.4͑3͒

2.56͑3͒

2.96͑5͒

1068 ^c

991^c

T

G

T₁T₂

T₁G₂

G₁T₂

1050͑10͒

1066͑10͒

1052͑10͒

1051͑10͒

1067͑10͒

1055.8͑5͒

1072͑3͒

1055͑1͒

1058͑2͒

1093͑15͒

2.8͑1͒

2.6͑7͒

2.0͑4͒

4.7͑4͒

10͑4͒

1011͑10͒

1068͑10͒

1056͑10͒

1029͑10͒

1067͑10͒

1026͑10͒

1049͑10͒

1010͑10͒

1010.3͑5͒

1080͑4͒

1063͑9͒

1031͑1͒

1092͑16͒

Ϫ0.3͑1͒

3.9͑8͒

6͑2͒

0.4͑2͒

10͑4͒

676^d

596^d

1-butoxy

672^e

f

g

2-propoxy

2-butoxy

1050͑10͒

1044͑10͒

1000͑10͒

569^g

611ⁱ

559ⁱ

574.6͑8͒

2.7͑1͒

h

B

A

h

^aTaken from Refs. 20 and 21.

b

˜

Data taken from Ref. 22. The least squares ﬁt of the LIF data was done for ␯ϭ1–4 of the CO stretch progression in the A state. The ﬁrst excited state of

˜

methoxy is designated as A, but owing to its electronically similar structure to the B state of the rest of the reported alkoxy radicals, we report its

˜

spectroscopic characteristics under the column for the B state.

c

˜

See Ref. 8. We have reﬁtted the data of Table II for the CO stretch progressions in the X and B states and include the errors of ␻ and ␻_ex_ein parenthesis.

e

˜

Based on our assignment of the A state origin, we determined ␻ of the CO stretch progression in the A state.

0

^dTaken from Ref. 5.

^eTaken from Ref. 23.

X

A

19

^fThe frequency of 1050 cm^Ϫ1(X state͒ is the average of 932 (␯₁₉) and 1169 (␯₁₄) cm^Ϫ1. The frequency of 1026 cm^Ϫ1(A state͒ is the average of 913 (␯

)

˜

and 1140 (␯^A₁₄) cm^Ϫ1. All of these values were determined from the DF spectrum that corresponds to emission from the ␯ϭ1 CO stretch vibrational level of

˜

the B state.

g

˜

Data taken from Ref. 3. The least squares ﬁt of the LIF data was done for ␯ϭ1–6 of the CO stretch progression in the B state.

^hThe unresolved structure of the 2-butoxy DF spectra above 1500 cm^Ϫ1allowed only for the determination of the fundamental frequencies with any precision.

ⁱTaken from Ref. 6.

˜

B. Moore, 58th OSU International Symposium on Molecular Spectroscopy

͑The Ohio State University, Columbus, OH, 2003͒.

tra providing the ␯_COstretching frequency for the X state and

˜

in most cases for the A state as well as numerous other

⁸X. Zhu, M. M. Kamal, and P. Misra, Pure Appl. Opt. 5, 1021 ͑1996͒.

⁹A. H. Blatt, Organic Synthesis, Collective Volume ͑Wiley, New York,

1943͒.

vibrations. Quantum chemical calculations have been per-

formed and compared with the experimental observations.

Generally speaking, the experimental results agree well with

the theoretical calculations. Observations are made concern-

ing the degree of congestion of the DF spectra as the radical

size increases.

¹⁰M. J. Frisch, G. W. Trucks, H. B. Schlegel et al., GAUSSIAN 98, Revision

A.7 ͑Gaussian, Inc., Pittsburgh, PA, 1998͒.

¹¹J. F. Stanton, J. Gauss, W. J. Lauderdale, J. D. Watts, and R. J. Bartlett,

The package also contains modiﬁed versions of the MOLECULE Gaussian

¨

integral program of J. Almlof and P. R. Taylor, the ABACUS integral deriva-

tive program written by T. U. Helgaker, H. J. As. Jensen, P. Jørgensen, and

P. R. Taylor, and the PROPS property evaluation integral code of P. R.

Taylor.

ACKNOWLEDGMENTS

¹²G. Tarczay, S. Gopalakrishnan, and T. A. Miller, J. Mol. Spectrosc. 220,

276 ͑2003͒.

The authors gratefully acknowledge useful discussions

with Lily Zu. The authors are pleased to acknowledge the

support of parts of this research by the National Science

Foundation via Grant No. 0211281, and other parts by the

Chemical Sciences, Geosciences, and Biosciences Division,

Ofﬁce of Basic Energy Sciences, Ofﬁce of Science, U.S.

Department of Energy, via Grant No. DE-FG02-01ER15172.

The authors also acknowledge a grant of computer time from

the Ohio Supercomputer Center.

¹³See EPAPS Document No. E-JCPSA6-121-010446 for the calculated vi-

brational frequencies. A direct link to this document may be found in the

online article’s HTML reference section. The document may also be

reached via the EPAPS homepage ͑http://www.aip.org/pubservs/

epaps.html͒ or from ftp.aip.org in the directory /epaps/. See the EPAPS

homepage for more information.

¹⁴R. S. Mulliken, J. Chem. Phys. 23, 1997 ͑1953͒.

15

´

A. G. Csaszar, W. D. Allen, Y. Yamaguchi, and H. F. Schaefer, in Compu-

tational Molecular Spectroscopy, edited by J. Jensen and P. R. Bunker

͑Wiley, New York, 2000͒, Chap. 2.

¹⁶E. R. Davidson and W. T. Borden, J. Phys. Chem. 87, 4783 ͑1983͒.

¹⁷J. F. Stanton, Chem. Phys. Lett. 237, 20 ͑1995͒.

¹⁸J. F. Stanton, J. Chem. Phys. 115, 10382 ͑2001͒.

¹⁹T. M. Ramond, G. E. Davico, R. L. Schwartz, and W. C. Lineberger, J.

Chem. Phys. 112, 1158 ͑2000͒.

¹M. E. Jenkin and G. D. Hayman, Atmos. Environ. 33, 1275 ͑1999͒.

²R. Atkinson, Int. J. Chem. Kinet. 29, 99 ͑1997͒.

³C. C. Carter, J. R. Atwell, S. Gopalakrishnan, and T. A. Miller, J. Phys.

Chem. A 104, 9165 ͑2000͒.

⁴C. C. Carter, S. Gopalakrishnan, J. R. Atwell, and T. A. Miller, J. Phys.

Chem. A 105, 2925 ͑2001͒.

²⁰Y.-Y. Lee, G.-H. Wann, and Y.-P. Lee, J. Chem. Phys. 99, 9465 ͑1993͒.

²¹S. C. Foster, P. Misra, T.-Y. D. Lin, C. P. Damo, C. C. Carter, and T. A.

Miller, J. Phys. Chem. 92, 5914 ͑1988͒.

⁵S. Gopalakrishnan, C. C. Carter, L. Zu, V. Stakhursky, G. Tarczay, and T.

A. Miller, J. Chem. Phys. 118, 4954 ͑2003͒.

²²D. E. Powers, M. B. Pushkarsky, and T. A. Miller, J. Chem. Phys. 106,

6863 ͑1997͒.

⁶S. Gopalakrishnan, L. Zu, and T. A. Miller, J. Phys. Chem. A 107, 5189

͑2003͒.

²³S. Gopalakrishnan, L. Zu, and T. A. Miller, Chem. Phys. Lett. 380, 749

͑2003͒.

⁷X. Liu, V. Stakhursky, E. D. Olmon, V. A. Lozovsky, T. A. Miller, and C.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:

128.240.225.44 On: Mon, 22 Dec 2014 01:12:56

Article Doi

Dispersed fluorescence spectroscopy of primary and secondary alkoxy radicals

DOI: 10.1063/1.1814104

Source and publish data:

Authors:

Article abstract of DOI:10.1063/1.1814104

Full text of DOI:10.1063/1.1814104

Products guided by the article

R&D Labs maybe for 16499-18-6

Relevant to this article

Hot Product