Energy transfer in bichromophoric molecules: The effect of symmetry and donor/acceptor energy gap

Article abstract of DOI:10.1021/jp9831280

The dependence of the rate of singlet excitation transfer on the donor-acceptor energy gap was investigated in bichromophoric spiranes with symmetry-forbidden zero-order electronic coupling. The fluorescence measurements were performed in a supersonic jet in order to avoid collisional and inhomogeneous line broadening. Fluorescence excitation spectra and single-vibronic-level emission spectra of the model chromophores cyclopentaphenanthrene and 1,8-dimethylnaphthalene and the bichromophores spirofluorenephenanthrene and spirofluorenenaphthalene are presented and analyzed. Although the transition moments of the linked chromophores are rigorously perpendicular and the exchange coupling between the v′ = 0 states is computationally shown to be zero, all spiranes with energy gaps larger than ～1000 cm^-1 exhibited complete electronic energy transfer from all vibrational states of the electronically excited donor, including the undistorted v′ = 0 state. This behavior is explained in terms of vibronic coupling between the sparse states of the donor and the dense manifold (pseudocontinuum) of the acceptor states. The electronic energy transfer was sufficiently fast to result in measurable lifetime broadening of the donor absorption lines, from which the k_EET was estimated. The results demonstrate that the zero-order picture overestimates the degree of the molecular orbital symmetry control over electronic energy transfer and charge-transfer rates and that at sufficiently high driving forces the vibronically mediated "symmetry-forbidden" electronic energy transfer can be very rapid (～1 × 10¹² s^-1).

Full text of DOI:10.1021/jp9831280

10
J. Phys. Chem. A 1999, 103, 10-20
Energy Transfer in Bichromophoric Molecules: The Effect of Symmetry and Donor/
Acceptor Energy Gap
W. T. Yip† and Donald H. Levy*
Department of Chemistry and James Franck Institute, UniVersity of Chicago, Chicago, Illinois 60637
Renata Kobetic and Piotr Piotrowiak*
Department of Chemistry, Rutgers UniVersity, Newark, New Jersey 07102
ReceiVed: July 22, 1998; In Final Form: NoVember 3, 1998
The dependence of the rate of singlet excitation transfer on the donor-acceptor energy gap was investigated
in bichromophoric spiranes with symmetry-forbidden zero-order electronic coupling. The fluorescence mea-
surements were performed in a supersonic jet in order to avoid collisional and inhomogeneous line broadening.
Fluorescence excitation spectra and single-vibronic-level emission spectra of the model chromophores
cyclopentaphenanthrene and 1,8-dimethylnaphthalene and the bichromophores spirofluorenephenanthrene and
spirofluorenenaphthalene are presented and analyzed. Although the transition moments of the linked
chromophores are rigorously perpendicular and the exchange coupling between the V′ ) 0 states is
computationally shown to be zero, all spiranes with energy gaps larger than ∼1000 cm-1 exhibited complete
electronic energy transfer from all vibrational states of the electronically excited donor, including the undis-
torted V′ ) 0 state. This behavior is explained in terms of vibronic coupling between the sparse states of the
donor and the dense manifold (pseudocontinuum) of the acceptor states. The electronic energy transfer was
sufficiently fast to result in measurable lifetime broadening of the donor absorption lines, from which the
kEET was estimated. The results demonstrate that the zero-order picture overestimates the degree of the molecular
orbital symmetry control over electronic energy transfer and charge-transfer rates and that at sufficiently
high driving forces the vibronically mediated “symmetry-forbidden” electronic energy transfer can be very
rapid (∼1 × 1012 s-1).
Introduction
these strong molecular orbital symmetry effects were observed
only in nearly degenerate bichromophores (i.e., when the
The rate of electronic excitation transfer and charge-transfer
processes in linked donor-acceptor systems and in bichro-
mophoric molecules is determined by the electronic coupling
between donor and acceptor. The symmetry of the zero-order
localized molecular orbitals of the donor and the acceptor is an
important factor determining the magnitude of this coupling.
As a result, the mutual orientation of the donor and the acceptor
moieties and the symmetry of the relevant molecular orbitals
are frequently invoked when a particularly low or high transfer
rate is observed. While the general orientational and stereo-
chemical effects in electronic energy transfer (both dipole-
dipole and exchange coupled) and charge transfer (exchange
coupled) have been studied by several groups,1 detailed ex-
perimental investigations of systems in which the electronic
coupling is completely canceled by symmetry remain very rare.2
In recent years, we have initiated a molecular beam fluores-
cence study of a series of specially designed spirobifluorenes
(Figure 1a) in order to demonstrate that a total symmetry can-
cellation of electronic coupling (both the Coulomb and the
exchange couplings) between rigidly linked orthogonal chro-
mophores can indeed be achieved.3 It was shown that singlet
energy transfer in these systems is effectively blocked,4 despite
the close proximity of the donor and acceptor units, which are
separated only by a single sp3-hybridized carbon atom. However,
donor-acceptor energy gap is comparable with the spacing of
the low-frequency vibrational levels) at the bottom of the
respective vibrational manifolds (Figure 2a). When a sufficient
amount of excess vibrational energy (typically >200 cm-1) was
deposited in the S1 electronic state of the donor, singlet energy
transfer due to vibronic (Herzberg-Teller) coupling occurred
and was manifested by dual emission at the characteristic donor
and acceptor frequencies. It was concluded that two conditions
have to be satisfied in order for such zero-order, symmetry-
forbidden electronic energy transfer to become weakly allowed.
First, the overall symmetry of the donor-acceptor system has
to be broken. Second, a vibronic level of the acceptor nearly
degenerate with the appropriate level of the donor must be
available.
The second of the above requirements leads directly to the
topic of the present contribution. We investigate the influence
of the size of the energy gap between the V′ ) 0 levels of the
donor and the acceptor on the rate of singlet energy transfer in
systems where the symmetry of the zero-order molecular orbital
inhibits electronic interaction. To address this issue, several
newly synthesized spiranes with large energy gaps approaching
4000 cm-1 (Figure 1b) were studied. The members of the new
family of model compounds (Figure 3) (except for spirofluo-
renenaphthalene) have the same essential electronic symmetry
as the low energy gap compounds shown in Figure 1a. The
localized S0 T S1 transition dipoles of the chromophores are
† Present address: Department of Chemistry, University of Minnesota,
Minneapolis, MN 55455.
10.1021/jp9831280 CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/17/1998

Energy Transfer in Bichromophoric Molecules
J. Phys. Chem. A, Vol. 103, No. 1, 1999 11
Figure 1. Spiranes with orthogonal chromophores and S0 T S1 transition moments: (a) compounds with a small energy gap; (b) compounds with
a large energy gap.
vibrationally excited states of the donor in low energy gap
compounds. As the energy gap increases, the vibrational ground
state of the S1 electronic excited state of the donor, VD ) 0,
faces an increasing density of nearly degenerate vibrational states
of the S1 state of the acceptor (Figure 2b). The symmetry of
some of these states will inevitably be appropriate for vibronic
coupling and therefore excitation transfer. Thus, singlet energy
transfer and emission from both chromophores can be expected
when any vibronic state of the donor (including the zero-point
vibrational level) is excited. In the limiting case of a very large
energy gap (or “driving force” in the terminology of electron
and excitation transfer in condensed phases), the situation will
correspond to a coupling between a single well-defined donor
state and a continuum of acceptor states. This is the classical
setup of the Fermi “golden rule” describing nonradiative decay
processes. There are two consequences of this situation. First,
since intramolecular vibrational relaxation (IVR) proceeds much
faster to the dense manifold of the acceptor, total quenching of
the donor emission may occur. Second, absorption lines in the
excitation spectrum corresponding to the Franck-Condon
transitions of the donor will become broadened. Indeed, all large
energy gap bichromophores in this study (diazospirobifluorene,
dichlorospirobifluorene, spirofluorenephenanthrene, and spiro-
fluorenenaphthalene) exhibit complete singlet energy transfer
upon excitation of any donor state, and only acceptor fluores-
cence could be detected. Significant broadening of the donor
excitation spectrum was observed in all cases. In the absence
of an inhomogeneous contribution, there exists a direct link
between the line width, ∆E, and the lifetime of the donor
state, τ:
Figure 2. Schematic representation of the vibrational manifolds of
the electronically excited states of the donor and the acceptor: (a) when
the S1DONOR-S1ACCEPTOR energy gap is small and comparable with the
DONOR
spacing of the lowest vibrational levels; (b) when the S1
S1
-
ACCEPTOR
energy gap is much larger than the spacing between the
vibrational levels.
τ∆E ) p
(1)
Figure 3. Three-dimensional views of the model compounds (spiro-
fluorenephenanthrene (SFP) and spirofluorenenaphthalene (SFN). Note
the C2V symmetry of SFP and the Cs symmetry of SFN.
When the radiative lifetime of the donor is long (as it is in the
case of fluorene), τ is determined primarily by the rate of
nonradiative energy transfer to the acceptor, kEET ) 1/τ.
Therefore, by simply measuring the width of the donor line, it
is possible to establish the rate of electronic energy transfer.
Since the behavior of all large energy gap model compounds
was qualitatively identical, and because of space considerations,
only the spectroscopy of two representative systems, the
spirofluorenephenanthrene and the spirofluorenenaphthalene,
will be discussed in full detail.
mutually orthogonal, and the higher order multipole interactions,
as well as the exchange interaction, vanish by symmetry. We
studied the new model compounds under supersonic expansion
conditions to eliminate most fluctuations and sources of inho-
mogeneous line broadening that are present in condensed phases.
It can be expected that increasing the energy gap between the
donor and the acceptor may have a similar effect as populating

12 J. Phys. Chem. A, Vol. 103, No. 1, 1999
Yip et al.
The presented findings demonstrate that the zero-order
approach can overestimate the degree of the molecular or-
bital symmetry control over electronic excitation and charge-
transfer rates. Even at rather modest magnitudes of the
donor-acceptor energy gap (1500-4000 cm-1), the “symmetry-
forbidden” electronic energy transfer can occur on a picosecond
time scale, thanks to efficient vibronic coupling. The presence
of such strong vibronic interactions may explain why in all
previously published cases the differences between the rates of
the symmetry allowed and the symmetry forbidden exothermic
charge-transfer reactions are much smaller than would have
been expected from a zero-order localized molecular orbital
picture.2,5,6
Experimental Section
A. Synthesis. Spirofluorenephenanthrene (SFP). 4,5-Meth-
ylenecyclophenanthrene [4H-cyclopenta[def]phenanthrene] (Lan-
caster, 98%) was oxidized nearly quantitatively to the corre-
sponding ketone using the Jones reagent. The coupling of the
ketone with the 2-bromobiphenyl (Lancaster, 98%) Grignard
reagent and the subsequent ring closing were described previ-
ously.3 The overall yield of these steps was ∼80%. The product
was purified by recrystallization from a hexane/benzene mixture.
Spirofluorenenaphthalene (SFN). 1,8-Bis(bromomethyl)naph-
thalene was obtained in ∼70% by photoinitiated bromination
of 1,8-dimethylnaphthalene (Fluka, 97%). A stoichiometric
amount of a benzene/THF solution of the dibromide was added
dropwise to the methanolic slurry of fluorene (Aldrich, 99%)
with a slight excess of MeOLi (prepared in situ from lithium
wire and MeOH). The product (∼70% yield) was purified by
recrystallization from ethanol/hexane.
Figure 4. Fluorescence excitation spectra of (a) fluorene, (b) spiro-
fluorenephenanthrene, and (c) cyclopentaphenanthrene. There are two
spectral regions in (b) and the peak intensity in each region is only
relative to the most intense peak of the same region, i.e., the S1P and
S1F origin of cyclopentaphenanthrene at the low- and high-energy
region, respectively.
temporal profile of its fluorescence was captured by a Tektronix
2221 digital storage oscilloscope and the decay curve was
transferred to a 486DX2/33MHz PC via a GPIB interface. Each
decay curve was averaged over multiple laser pulses before the
data were transferred.
Cyclopentaphenanthrene and spirofluorenephenanthrene were
heated to 110 °C and 160-170 °C, respectively, and were
entrained in helium at backing pressures of 30 and 50 psig,
respectively. Dimethylnaphthalene was heated to 40 °C while
spirofluorenenaphthalene was heated to 160-180 °C in order
to obtain sufficient vapor pressure. To minimize the consumption
of dimethylnaphthalene and spirofluorenenaphthalene, a helium
backing pressure of only 10 psig was used. All sample molecules
were seeded in helium and were expanded into a vacuum
chamber through a 100 µm pinhole. With the continuous
supersonic jet, the typical working pressure inside the vacuum
chamber was kept below 25 mTorr. Cyclopentaphenanthrene,
spirofluorenephenanthrene, and spirofluorenenaphthalene were
prepared according to the method described in section A.
Dimethylnaphthalene (95%) was purchased from Aldrich and
was used as received.
Dichlorospirobifluorene (2Cl-SBF). 2,7-Dibromofluorene
(Lancaster, 97%) was converted to the 2,7-dichlorofluorene in
a nearly quantitative yield via the classical CuCl/CuCl2 halogen
exchange reaction. The purified product was oxidized quanti-
tatively to the 2,7-dichlorofluoren-9-one using the Jones reagent.
The coupling of the 2,7-dichlorofluoren-9-one with the 2-bro-
mobiphenyl Grignard reagent and the subsequent ring closing
were described previously.3 The overall yield of these steps was
∼80%. The product was purified by recrystallization.
1
All compounds were identified by NMR (300 MHz H and
75 MHz 13C), IR, and mass spectroscopy. The purity was
assessed by NMR and thin-layer chromatography (TLC). In the
case of dichlorospirobifluorene, the X-ray crystal structure was
obtained.
Results
The synthesis of the small energy gap spiranes (dnh16-n
-
spirobifluorene, n ) 0, 1, 2, 8) and of the 4,5-diazospirobi-
fluorene (2N-SBF) was already described in detail elsewhere.3,7
B. Spectroscopy. Both fluorescence excitation and dispersed
fluorescence spectroscopy were used. Detailed descriptions of
the experimental setup may be found elsewhere.8 In brief,
fluorescence excitation spectra were obtained by exciting
samples entrained in a supersonic jet with the frequency-doubled
output of a Nd:YAG pumped dye laser. The total fluorescence
from the sample, as a function of the excitation frequency, was
collected by f/1 optics that were orthogonal to both the excitation
laser and the supersonic jet. To obtain a dispersed fluorescence
spectrum, the excitation frequency was tuned to a resonance
transition of the sample and the total fluorescence was dispersed
either by a 1.0 m Spex monochromator equipped with a
photomultiplier (RCA 8575) or a 0.275 m spectrograph (Acton
Research Corporation, SpectraPro 275) to which a linear array
detector (Princeton Applied Research, 1455B-700-HQ) was
attached. To measure the fluorescence lifetime of a sample, the
A. Cyclopentaphenanthrene. Figure 4c shows the fluores-
cence excitation spectrum of cyclopentaphenanthrene from
29200-34200 cm-1. The spectrum is constructed by connecting
seven excitation spectra together, and the relative intensities of
adjacent spectra are scaled by overlapping features. This scaling
method becomes inaccurate at high frequency since the peaks
beyond 31 000 cm-1 are sparse and weak. Comparison of peak
intensity is therefore not recommended. The spectral assignment
of cyclopentaphenanthrene is given in Table 1. There is no
previous report of the IR or Raman spectrum of cyclopen-
taphenanthrene to which we can refer, and our spectral assign-
ment is mostly based on the harmonic vibrational frequencies
computed from an AM1 optimized cyclopentaphenanthrene. In
general, AM1 provides a more reliable geometry prediction over
other semiempirical methods.9 As a result of anharmonic
corrections, quantum mechanical calculations systematically
overestimated harmonic vibrational frequencies. In the case of
AM1, a 10.4% error is expected.10

Energy Transfer in Bichromophoric Molecules
J. Phys. Chem. A, Vol. 103, No. 1, 1999 13
TABLE 1: Vibronic Transitions Assignment in the First
2000 cm-1 of the S1 r S0 Excitation Spectrum of
Cyclopentaphenanthrene
position/cm-1
shifta/cm-1
intensity
assignmentb
AM1c
29 370
29 723
29 780
29 844
29 877
29 893
29 944
30 004
30 043
30 369
30 379
30 414
30 468
30 520
30 535
30 540
30 553
30 558
30 571
30 609
30 677
30 684
30 718
30 722
30 734
30 749
30 755
30 775
30 812
30 839
30 873
30 881
30 887
30 982
31 616
0
353
410
474
507
523
574
634
673
vs
mw
s
ms
m
m
origin
a2‚a2
a1
a1
b1‚b1
b2
b2
b2
a1
360
376
474
508
522
570
595
659
m
s
ms
ms
ms
mw
mw
mw
ms
ms
w
999
474 + 523
b2
1009
1044
1098
1150
1165
1170
1183
1188
1201
1239
1307
1314
1348
1352
1364
1379
1385
1405
1442
1469
1503
1511
1517
1612
2246
1125
2 × 523
523 + 574
2 × 574
a1
1158
1164
b2
410 + 673
a1
b2
a1
634 + 673
b2
m
mw
m
1201
1230
1233
Figure 5. Single vibronic level emission spectra of (a) spirofluore-
nephenanthrene at 29 164 cm-1 (S1P) and (b) cyclopentaphenanthrene
at 29 370 cm-1 (S1). The spectral resolution is 20 cm-1 in both
spectra.
m
ms
mw
w
m
mw
m
1325
1340
1377
2 × 673
TABLE 2: Assignment of the SVL Emission of
Cyclopentaphenanthrene Excited at the S1 Origin
a1
353 + 1009
a1
position/cm-1 shifta/cm-1 intensity assignmentb
S1
S0
c
d
29 370
29 022
29 926
29 880
29 815
29 770
29 706
29 660
29 544
29 515
29 475
29 342
29 176
29 146
29 118
29 045
29 012
27 953
27 895
27 840
27 765
27 458
27 392
27 308
27 237
27 200
27 073
26 875
26 713
26 582
26 545
26 433
26 297
0
348
444
490
555
600
664
710
826
vs
origin
a2‚a2
a1
a1
b2
b2
a1
b2
a1
b2
s
410 + 999
a1
474 + 999
b2
vw
mw
mw
w
ms
ms
mw
ms
w
353
410
474
250
406
mw
mw
m
m
ms
s
1442
1482
1511
523, 574 540
b2
634
673
353 + 1165
a1
634 + 1612
1609
755e
812e
864e
710
830
mw
a Vibrational frequencies are blue-shifted with respect to the ori-
855
895
gin at 29 370 cm-1 b Assignment based on C2V symmetry. c S0 vibra-
.
vw
m
m
2 × 444
1028
1194
1224
1252
1325
1358
1417
1475
1530
1605
1912
1978
2062
2133
2170
2297
2495
2657
2788
2825
2937
3073
b2
1009
1165
1201
1239
1314
1379
1038
1200
tional frequencies calculated from AM1 optimized cyclopentaphenan-
threne.
a1
b2
a1
b2
w
vw
ms
mw
mw
s
vw
m
mw
w
mw
mw
vw
m
mw
mw
mw
w
mw
w
1244
1303
1365
1417
1441
1524
1608
There are some hot band transitions lying within 150 cm-1
to the red of the 29 370 cm-1 S1 origin, probably due to insuf-
ficient cooling of some twisting modes associated with the cyclo-
pentyl ring. Since the origin is the most intense feature, the
geometry of cyclopentaphenanthrene in the S1 state is similar
to that in the ground state. Most of the S1 vibronic activity is
a1
600 + 826
b2
710 + 826
a1
1503
1612
444 + 1475
490 + 1475
600 + 1475
664 + 1475
710 + 1475
826 + 1475
1028 + 1475
1194 + 1475
1325 + 1475
1358 + 1475
2 × 1475
1605 + 1475
confined within the first +2000 cm-1. Beyond 31 000 cm-1
,
the spectrum is dominated by weak and sharp transitions that
are sitting on top of a broad and slightly absorbing background.
At high vibronic energy, unfavorable Franck-Condon factors
can weaken the absorption cross section whereas the onset of
IVR can significantly broaden a transition and turn adjacent
transitions into a broad band. In addition to the totally symmetric
a1 vibrations, many b2 vibrations are observed. These b2 tran-
sitions probably gain their intensities by vibronic coupling to a
low-lying S2 state with B2 symmetry. This S2 state is most likely
located at ∼+5600 cm-1 in cyclopentaphenanthrene and resem-
bles the ∼+6000 cm-1 S2 state observed in phenanthrene.11-13
In Figure 5b, we show the emission spectrum of cyclopen-
taphenanthrene excited at the S1 origin. Most features can be
assigned to a1 and b2 fundamentals plus combination bands
containing the 1475 cm-1 b2 fundamental. The spectral assign-
ment of the emission spectrum is listed in Table 2. The strong
0-0 emission shows that the geometry of cyclopentaphenan-
threne is similar in S0 and S1, which is quite reasonable in view
of the rigidity of cyclopentaphenanthrene.
a Vibrational frequencies are red-shifted with respect to the origin
at 29 370 cm-1 b Assignment based on C2V symmetry. c Correlation
.
with S1 vibronic transitions. d Ground-state vibrational frequencies of
phenanthrene from ref 13. e S0 vibrational frequencies calculated from
AM1 optimized cyclopentaphenanthrene.
B. Spirofluorenephenanthrene. We compare the fluores-
cence excitation spectrum of spirofluorenephenanthrene in
Figure 4b with those of the two model compounds, fluorene
and cyclopentaphenanthrene, in parts a and c of Figure 4,
respectively. In Figure 4b, we focus on two spectral regions

14 J. Phys. Chem. A, Vol. 103, No. 1, 1999
Yip et al.
TABLE 3: Vibronic Transitions Assignment in the First
1000 cm-1 of the S1N r S0 Excitation Spectrum of
Spirofluorenephenanthrene
TABLE 5: Assignment of the SVL Emission of
Spirofluorenephenanthrene Excited at the S1P Origin
c
position/cm-1
shifta/cm-1
intensity
assignmentb
S0
c
d
position/cm-1 shifta/cm-1 intensity assignmentb S1
S1
29 164
29 038
28 774
28 549
28 371
28 298
28 123
27 941
27 838
27 677
27 235
27 094
26 978
26 905
26 823
26 366
26 220
0
114
378
592
780
vs
w
w
ms
ms
mw
mw
mw
ms
s
w
m
vw
m
mw
m
origin
twist/bend
a1
b2
a1
b2
b2
29 164
29 203
29 322
29 398
29 432
29 576
29 591
29 633
29 734
29 831
29 849
29 898
30 006
0
39
vs
vw
w
mw
w
m
ms
mw
m
mw
w
vw
vw
origin
twist/bend
twist/bend
twist/bend
twist/bend
a1
twist/bend
a1
b2
b2
a1
43
141, 142
232
444
600
826
855
1028
158
234
268
412
427
469
570
667
685
734
842
247
854
410
1033
1215
1318
1475
1912
2056
2174
2248
2329
2770
2932
418
2 × 592
474
574
634
673
b2
1325
1475
b2
592 + 1318
592 + 1475
854 + 1318
780 + 1475
854 + 1475
1318 + 1475
2 × 1475
268 + 469
412 + 427
a Vibrational frequencies are blue-shifted with respect to the origin
at 29 164 cm-1 b Assignment based on C2V symmetry. c Correlation
.
m
with S1 vibronic transitions of cyclopentaphenanthrene. d Correlation
with S1 intramolecular modes of spirobifluorene.
a Vibrational frequencies are red-shifted with respect to the origin
at 29 164 cm-1 b Assignment based on C2V symmetry. c Correlation
.
TABLE 4: Vibronic Transitions Assignment in the First
1000 cm-1 of the S1F r S0 Excitation Spectrum of
Spirofluorenephenanthrene
with the S1 f S0 vibronic transitions of cyclopentaphenanthrene.
position/cm-1
shifta/cm-1
intensity
assignmentb
S1
c
33 124
33 270
33 274
33 422
33 528
33 673
33 678
33 763
33 824
33 852
33 913
33 929
33 977
34 002
34 109
34 121
0
146
150
298
404
549
554
639
700
728
789
805
853
878
985
997
vs
ms
s
w
ms
w
mw
m
mw
ms
w
vw
m
m
origin
?
a1
150
205
2 × 150
a1
395
146 + 404
150 + 404
b2
593
721
298 + 404
a1
150 + 639
2 × 404
a1
832
985
150 + 728
a1
?
ms
ms
a Vibrational frequencies are blue-shifted with respect to the origin
at 33 124 cm-1 b Assignment based on C2V symmetry. c Correlation
with S1 vibronic transitions of fluorene.
.
Figure 6. The first 1000 cm-1 fluorescence excitation spectrum of (a)
spirofluorenephenanthrene at 29 164 cm-1 (S1P), (b) cyclopentaphenan-
threne at 29 370 cm-1 (S1), (c) spirofluorenephenanthrene at 33 124
cm-1 (S1F), and (d) fluorene at 33 777 cm-1 (S1).
where the initial excitation is localized either at the low-energy
phenanthrene side or the high-energy fluorene side. We desig-
nate the first excited spirofluorenephenanthrene state localized
on the phenanthrene side as the S1P state and the first excited
spirofluorenephenanthrene state localized at the fluorene side
as the S1F state. The spectral assignment of spirofluorene-
phenanthrene is given in Tables 3 and 4. The S1P origin of
spirofluorenephenanthrene is located at 29 164 cm-1, a spectral
shift of -206 cm-1 relative to cyclopentaphenanthrene. The S1F
origin is at 33 124 cm-1, a spectral shift of -653 cm-1 relative
to fluorene.
In parts a and b of Figures 6, we compare the fluorescence
excitation spectra of spirofluorenephenanthrene and cyclopen-
taphenanthrene in the S1P region. The hot bands in cyclopen-
taphenanthrene have completely disappeared in spirofluore-
nephenanthrene. The attachment of a fluorene unit to cyclopenta-
phenanthrene creates low-frequency vibrational modes that can
help cool the higher frequency modes through collision-induced
IVR. The emission spectra of spirofluorenephenanthrene and
cyclopentaphenanthrene in Figure 5 are so similar that both of
them can be assigned to a1 and b2 fundamentals and combination
bands containing an active 1475 cm-1 mode. This indicates that
the energy level structure of cyclopentaphenanthrene remains
almost undisturbed in spirofluorenephenanthrene. The spectral
assignment of the S1P emission spectrum is given in Table 5.
We compare the fluorescence excitation spectra of fluorene
and spirofluorenephenanthrene in the S1F region in parts c and
d of Figure 6. The corresponding spectral assignment is given
in Table 4. The S1F excitation spectrum of spirofluorenephenan-
threne in the first 1000 cm-1 is dominated by a1 transitions,
with only a single transition assigned to b2 symmetry. The 205
cm-1 mode in fluorene is reduced to 150 cm-1 in spirofluore-
nephenanthrene, compared to 212 cm-1 in spirobifluorene.3 This
assignment is based on their similar propensities to form combi-
nation bands with other vibrational modes. Except for the 150
cm-1 mode, most vibrations in spirofluorenephenanthrene have
frequencies close to the corresponding vibrations of fluorene.
The emission spectra of spirofluorenephenanthrene excited
at different excess energies above the S1F origin are shown in
Figure 7. There is no detectable emission in the first -3000
cm-1 and the emission is red-shifted ∼4000 cm-1 from the S1F
origin. These spectra contain broad emission bands with the
first feature in every spectrum located at ∼29 164 cm-1, the
S1P origin of spirofluorenephenanthrene. For comparison, we

Energy Transfer in Bichromophoric Molecules
J. Phys. Chem. A, Vol. 103, No. 1, 1999 15
Figure 8. Fluorescence excitation spectrum of (a) fluorene, (b)
spirofluorenenaphthalene, and (c) 1,8-dimethylnaphthalene.
TABLE 6: Vibronic Transitions Assignment of the S1 r S0
Spectrum of 1,8-Dimethylnaphthalene
position/cm-1 shifta/cm-1 intensity assignmentb S1
S1
c
d
31 341
31 374
31 426
31 760
31 790
31 833
31 856
31 868
31 945
31 977
32 105
32 139
32 251
32 284
32 317
32 364
32 397
32 547
32 712
32 930
32 968
0
33
85
vs
s
mw
s
m
ms
ms
mw
s
ms
ms
w
origin
CH3-torsion
CH3-torsion
8h10
33
88
428
459
500
515
419
449
492
515
527
604
636
764
798
910
943
976
1023
1056
1206
1371
1589
1627
2191
2394
2611
2732
3043
419
484
33 + 419
901
Figure 7. Single vibronic level emission spectra of spirofluore-
nephenanthrene excited at (a) 33 124 cm-1 (S1F), (b) +150 cm-1, (c)
+404 cm-1, and (d) +728 cm-1. (e) Single vibronic level emission
spectrum of spirofluorenephenanthrene excited at 29 164 cm-1 (S1P).
The spectral resolution is 100 cm-1 in (a-d) and 20 cm-1 in (e). The
abscissa applies to (a-d) only; it does not apply to (e).
33 + 492
801
613
647
758
661
815
33 + 604
7h10
33 + 764
8h10910
mw
m
w
914
943
8h10810
reproduce the emission spectrum of the S1P origin in Figure 7e.
Except for the broadening, the S1F and S1P emission spectra in
Figure 7 are similar. Electronic energy transfer occurs in the
S1F state of spirofluorenephenanthrene and excitation to the
fluorene unit produces fluorescence exclusively from the phe-
nanthrene unit. The rate of electronic energy transfer is so fast
that the transfer of excitation energy is complete before the
initially excited fluorene unit has a chance to emit.
C. 1,8-Dimethylnaphthalene. The jet spectroscopy of di-
methylnaphthalene was first studied by Chakraborty and
Chowdhury.14 Our measurements agree fairly well with theirs;
however, we revise some of their spectral assignments. We
extend the measurement of dimethylnaphthalene well beyond
its S2 origin as shown in Figure 8c. The spectrum covers 5500
cm-1 continuously from 31200 to 36700 cm-1. The spectrum
is again constructed from several excitation spectra. The scaling
method works well in dimethylnaphthalene since there are many
sharp and moderately intense transitions across the 5500 cm-1
spectral region.
The most intense peak is the S1 origin, and low-frequency
vibrations are obvious within the first 100 cm-1 of the S1 origin,
probably due to torsional motions associated with the two methyl
groups. Like naphthalene, the excitation spectrum of dimeth-
ylnaphthalene is dominated by a1 and b2 modes. Most vibronic
activity is confined to the first 1500 cm-1 with only a few bands
extending to the S2 region. On the basis of the excitation
spectrum, the relatively strong transition at 34 384 cm-1 (+3040
cm-1 from the S1 origin) is likely to be the S2 origin. This S2-
S1 gap is +3900 cm-1 in naphthalene.15,16 Possibly the S2 origin
970
33 + 943
802
ms
mw
mw
m
m
m
mw
mw
mw
mw
ms
m
ms
m
ms
ms
ms
1021
33 + 1023
802
1204
1370
1590 1579e
7h10810
8h10610
8h10820
8h10810610
33 532
33 735
33 952
34 073
34 384
6h10410
2603e
2773e
3046e
3076e
3111e
7h10810410
4h10920810
7h10930810
34 417
3076
34 459
34 599
34 658
34 688
34 717
3118
3258
3317
3347
3376
a Vibrational frequencies are blue-shifted with respect to the origin
at 31 341 cm-1 b Convention based on ref 28. c Measurements made
.
in ref 14. d Measurements made in ref 12. e Measurements made in
ref 15.
of dimethylnaphthalene lies between 34200 and 34800 cm-1
.
The humps that begin at 34 700 cm-1 resemble those observed
in the S2 region of naphthalene. The spectral assignment of
dimethylnaphthalene is given in Table 6.
In naphthalene, S0 T S1 is a symmetry-allowed transition,
yet this transition has a very small oscillator strength to the

16 J. Phys. Chem. A, Vol. 103, No. 1, 1999
Yip et al.
TABLE 7: Assignment of the SVL Emission of
1,8-Dimethylnaphthalene Excited at the S1 Origin
c
position/cm-1
shifta/cm-1
intensity
assignmentb
S1
31 341
31 249
31 154
31 054
30 864
30 781
30 723
30 549
30 303
30 226
30 170
29 991
29 898
29 725
29 508
29 449
29 371
28 954
28 878
28 827
28 648
28 555
28 384
0
92
187
287
477
560
618
792
1038
1115
1171
1350
1443
1616
1833
1892
1970
2387
2463
2514
2693
2786
2967
vs
origin
CH3-torsion
0
85
ms
vw
vw
ms
mw
m
2 × 92
?
8h01
419
492
604
764
910
801
7h01
w
m
8h01901
920
mw
mw
s
ms
mw
mw
mw
mw
m
mw
mw
ms
m
901801
510
92 + 1350
287 + 1350
8h01501
901501
Figure 9. Single vibronic level emission spectrum of 1,8-dimethyl-
801501
naphthalene excited at 31 341 cm-1 (S1). The spectral resolution is
8h01901501
901501
40 cm-1
.
901801501
520
ground state. This is because of an accidental cancellation of
the S1 transition moment due to configuration interactions of
low-lying excited states.15,16 The S2 origin in naphthalene is ∼20
times stronger than the S1 origin, leaving the bulk of the S1
spectrum borrowing intensity from the S2 state via vibronic
coupling. This delicate cancellation of the transition moment is
vulnerable to the breaking of symmetry induced by chemical
substitutions. The S1 origin of dimethylnaphthalene is almost
an order of magnitude more intense than the S2 origin. This
same effect has also been reported in various mono- and
disubstituted naphthalene derivatives.17-19
92 + 2693
287 + 2693
mw
a Vibrational frequencies are red-shifted with respect to the origin
at 31 341 cm-1 b Convention based on ref 28. c Correlation with S1
vibronic transitions of 1,8-dimethylnaphthalene.
.
The emission spectrum of dimethylnaphthalene is shown in
Figure 9. The emission spectrum is dominated by a1 and b2
fundamentals and combinations containing an active 1350 cm-1
a1 mode. The spectral assignment is given in Table 7.
D. Spirofluorenenaphthalene. Figure 8b shows the fluores-
cence excitation spectrum of spirofluorenenaphthalene. The
spectrum contains two parts with very different intensity
distributions. The low-energy part (below 33 500 cm-1) with a
weaker transition moment is naphthalene-like. The high-energy
part (33 500 cm-1 and above) with a much stronger transition
moment has fluorene-like absorptions sitting on top of a broad
background that begins near 33 800 cm-1. The excitation
spectrum of spirofluorenenaphthalene looks very much like a
superposition of a dimethylnaphthalene and a fluorene spectrum
with some spectral shifts. We assign the peaks at 31 325 and
33 512 cm-1 as the S1N and S1F origins of spirofluorenenaph-
thalene with the initial excitation localized at the naphthalene
and fluorene unit, respectively. The broad background that
begins at 33 800 cm-1 is most likely the S2 origin belonging to
the naphthalene unit. We assign this as the S2N state of
spirofluorenenaphthalene.
Figure 10. The first 2000 cm-1 fluorescence excitation spectrum of
(a) spirofluorenenaphthalene at 31 325 cm-1 (S1N) and (b) 1,8-
dimethylnaphthalene at 31 341 cm-1 (S1). The first 1000 cm-1
fluorescence excitation spectrum of (c) spirofluorenenaphthalene at
33 512 cm-1 (S1F) and (d) fluorene at 33 777 cm-1 (S1).
spectrum. This is the reason for the difference between the two
spectra in Figure 10. In general, Figure 10a looks more like the
excitation spectrum of acenaphthalene in which the two methyl
groups are directly linked to each other.14 In this study, the
naphthalene unit in spirofluorenenaphthalene serves as an inter-
mediate case between the rigid acenaphthalene and the relatively
free dimethylnaphthalene. The S1N state spectral assignment is
listed in Table 8. The excitation spectrum of the first 1100 cm-1
into the S1F region of spirofluorenenaphthalene is compared with
the S1 spectrum of fluorene in parts c and d of Figure 10, and
the complete spectral assignment is presented in Table 9.
Excitation to the S1F state leaves the bichromophore locally
excited on the fluorene side only. Figure 11 shows a series of
emission spectra at different excess energies above the S1F
A comparison between spirofluorenenaphthalene and dimeth-
ylnaphthalene at S1N is shown in parts a and b of Figure 10.
While dimethylnaphthalene shows some torsional activity of
the two methyl groups, spirofluorenenaphthalene exhibits none
since the two methyl groups are linked together via a carbon
atom in spirofluorenenaphthalene. In dimethylnaphthalene, tor-
sional motions of the methyl groups are free and are decoupled
from the ring modes. In spirofluorenenaphthalene, these same
torsional motions are arrested by coupling to the ring modes.
As a consequence, these low-frequency torsional modes mix
with the ring modes and produce a fairly complicated excitation

Energy Transfer in Bichromophoric Molecules
J. Phys. Chem. A, Vol. 103, No. 1, 1999 17
TABLE 8: Vibronic Transitions Assignment in the First
2000 cm-1 of the S1N r S0 Excitation Spectrum of
Spirofluorenenaphthalene
TABLE 9: Vibronic Transitions Assignment in the First
1500 cm-1 of the S1F r S0 Excitation Spectrum of
Spirofluorenenaphthalene
c
position/cm-1 shifta/cm-1 intensity assignmentb S1
AM1d
position/ shifta/
c
d
e
cm-1
cm-1 intensity assignmentb S1
S1
S1
31 325
31 704
31 730
31 742
31 751
31 790
31 810
31 817
31 880
31 892
31 939
31 944
31 994
32 070
32 082
32 108
32 121
32 129
32 193
32 218
32 270
32 294
32 298
32 309
32 318
32 355
32 370
32 403
32 408
32 466
32 736
32 754
32 762
32 771
32 781
32 913
0
379
405
417
426
465
485
492
555
567
614
619
669
745
757
783
796
804
868
893
945
969
973
984
993
1030
1045
1078
1083
1141
1411
1429
1437
1446
1456
1588
vs
m
ms
vs
w
ms
w
w
m
m
m
m
mw
mw
m
origin
2 × 189
b2
356 2 × 186
399 406
413 459
426 439
424 467
482
33 512
33 664
33 670
33 681
33 716
33 832
33 850
33 870
33 886
33 912
34 051
34 080
34 169
34 200
34 220
34 250
34 365
34 497
34 510
34 665
34 725
34 832
34 960
34 971
0
152
158
169
204
320
338
358
374
400
539
568
657
688
708
738
853
985
998
1153
1213
1320
1448
1459
vs
origin
? b1
twist/bend
a1
mw
mw
s
150
205
146
150
119
b2(8h1)
0
210
222
a1
a1(91)
ms
mw
m
mw
mw
ms
w
? b1
0
152 + 169
a1
b2
a1
2 × 169
504
561
152 + 204
169 + 204
3 × 189
a1
b2
395
537
404
639
399
542
a1
603 560
608 639
684
a1(81)
0
m
mw
s
169 + 400
a1
b2
593
601
683
b2(7h1)
738
4 ×0189
379 + 405
379 + 417
379 + 426
a1/b2
? b1
mw
ms
m
ms
w
169 + 539
m
775 766
793 810
a1
721
832
985
728
853
985
997
725
834
985
977
ms
vw
w
a1
a1
b2
868/871
918
a1(71)
w
5 ×0189
417 + 555
405 + 567
417 + 567
426 + 567
950
970
978
987
m
169 + 985
mw
mw
mw
mw
mw
m
w
m
m
mw
m
ms
m
mw
mw
s
ms
ms
ms
m
a1
a1
a1
a1
1231 1227
1321
1457
1532
997
a Vibrational frequencies are blue-shifted with respect to the origin
465 + 567 1032
379 + 669
at 33 512 cm-1 b Assignment based on C2V symmetry. c Correlation
.
with S1 vibronic transitions of fluorene. d Correlation with S1F vibronic
transitions of spirofluorenephenanthrene. e Fluorene vibronic measure-
ments made in ref 28.
465 + 614 1041
465 + 619 1059
b2
1185 1143
465 + 945 1421
669 + 757 1428
405 + 1030 1439
417 + 1030 1469
669 + 783 1476
levels of the bichromophore will allow vibrationally induced
electronic energy transfer. In d8h8-spirobifluorene, the energy
gap between the two chromophores was 106 cm-1, and at low
vibrational excitation, efficient energy transfer depended on acci-
dental coincidences between donor and acceptor vibronic levels.3
The energy gap is 3960 cm-1 in spirofluorenephenanthrene.
When the S1F state is initially excited, those S1P vibronic levels
that are isoenergetic to the S1F state are left unexcited due to
poor Franck-Condon factors. This gives a fluorene-like excita-
tion spectrum. As in the case for spirobifluorene, vibronic
coupling in spirofluorenephenanthrene allows energy transfer
from fluorene to phenanthrene. This vibrationally induced
electronic energy transfer is induced by nontotally symmetric
vibrations that break the symmetry of the bichromophore (e.g.,
interchromophoric scissoring, wagging, and bending modes).
With the 3960 cm-1 energy gap in spirofluorenephenanthrene,
the density of states of the acceptor is high enough that energy
transfer no longer depends on accidental coincidences even at
the zero-point energy level of the S1F state. Our observation
that only phenanthrene-like fluorescence is observed when any
fluorene-like level is excited indicates that energy transfer is
fast compared to the radiative decay rate of fluorene.
The efficient energy transfer in spirofluorenephenanthrene
also lengthens the fluorescence lifetime over that of fluorene.
We measured the fluorescence lifetime of fluorene to be 16.7
( 0.8 ns, that of cyclopentaphenanthrene to be 61.3 ( 5.1 ns,
and that of spirofluorenephenanthrene excited at the S1F origin
to be 49.0 ( 0.9 ns. The fluorescence lifetime of phenanthrene
is 78 ns.11
The electronic energy transfer reported in this paper is in the
statistical limit of radiationless transitions where a single donor
state is coupled to numerous isoenergetic acceptor states.23-25
This leads to irreversible electronic energy transfer after
excitation of the donor and spectral broadening in the absorption
b2(8h161)
1581
0
0
a Vibrational frequencies are blue-shifted with respect to the origin
at 31 325 cm-1 b Assignment based on C2V symmetry. Correlation to
.
S1 vibronic transitions of 1,8-dimethylnaphthalene is enclosed in
parentheses. c Measurements made in ref 14. d S0 vibrational frequencies
calculated from AM1 optimized 2,3-dihydrophenalene.
origin. In all cases, the emission is naphthalene-like. Also shown
in Figure 11f is the emission spectrum of dimethylnaphthalene
from its S1 origin. The emission spectra show essentially
complete energy transfer.
Discussion
A. Cyclopentaphenanthrene/Spirofluorenephenanthrene
System. In electronic energy transfer in spirofluorenephenan-
threne, fluorene is the donor and phenanthrene is the acceptor.
As in the previously studied case of spirobifluorene, the planes
of the two chromophores are perpendicular and this places
symmetry restrictions on the mechanism responsible for energy
transfer.3 The S1 T S0 transition moment of fluorene is polarized
along the long axis, whereas the S1 T S0 and S2 T S0 transition
moments of phenanthrene are both in-plane and are polarized
along the short and long axes, respectively.11 Because the
transition moments of fluorene and phenanthrene are perpen-
dicular, energy transfer via dipolar coupling is forbidden as long
as the symmetry is maintained.20,21 Because the π systems of
the two rings are orthogonal, exchange-coupled energy transfer
is also forbidden in the symmetric molecule.22
In the case of spirobifluorene, coupling between the initially
excited donor vibronic level and nontotally symmetric vibronic

18 J. Phys. Chem. A, Vol. 103, No. 1, 1999
Yip et al.
Figure 12. The fluorescence excitation profile of (a) the S1F origin of
spirofluorenephenanthrene and (b) the S1 origin of fluorene.
of spirofluorenephenanthrene to the S1 origin of fluorene. The
fwhm of fluorene is 1.83 cm-1, which is presumably due to the
heterogeneious broadening. The fwhm of spirofluorenephenan-
threne is 6.43 cm-1. By numerically evaluating the width of
the Voight line shape,27 we extract a homogeneous line width
of 5.88 cm-1 or kEET ) 1.11 × 1012 s-1 (τEET ) 0.90 ps), which
is 4 orders of magnitude shorter than the fluorescence lifetime
of fluorene.
B. Dimethylnaphthalene/Spirofluorenenaphthalene Sys-
tem. In electronic energy transfer in spirofluorenenaphthalene,
fluorene is the donor and naphthalene is the acceptor. Since
the two intervening methylene carbons on the naphthalene side
are sp3 hybridized, the symmetry of spirofluorenenaphthalene
is different from that of spirobifluorene and spirofluore-
nephenanthrene. The naphthalene plane bends away from the
short axis of fluorene and makes the C2V axes of fluorene and
dimethylnaphthalene incline at a tetrahedral angle of 109.5°.
Thus, only the long molecular axis of dimethylnaphthalene is
perpendicular to the S1 r S0 transition moment of fluorene.
The S1 r S0 and S2 r S0 transition moments in naphthalene
are respectively long and short axes polarized,16 and therefore,
dipolar coupling from fluorene is forbidden to the S1 but allowed
to the S2 state of naphthalene if the symmetry of the molecule
is maintained.20,21 Energy transfer via exchange couplings is
also forbidden in the symmetric molecule because the π systems
of the two rings are orthogonal.22
Figure 11. Single vibronic level emission spectra of spirofluo-
renenaphthalene excited at (a) 33 512 cm-1 (S1F), (b) +168 cm-1, (c)
+204 cm-1, (d) +399 cm-1, and (e) +688 cm-1. (f) Single vibronic
level emission spectrum of 1,8-dimethylnaphthalene excited at 31 341
cm-1 (S1N). The sharp peaks at high energy in (a-e) are scattered light
from the excitation laser. The spectral resolution is 40 cm-1 in all
spectra.
spectrum due to a shorter excited-state lifetime. Each donor
absorption peak will have a Lorentzian line shape with a width
that is determined by the rate of electronic energy transfer. If
the observed line width (fwhm) ∆DA is produced by a
homogeneous Lorentzian line width (fwhm), Γ, and a Gaussian
heterogeneous line width (fwhm), ∆D, the observed line shape
is a convolution of a Gaussian with a Lorentzian known as a
Voight profile.
exp[-(x - y)2/η2]
∞
1
I(x) )
dy
(2)
∫
In spirofluorenenaphthalene, the S1N state is at 31 325 cm-1
,
x
πη
1 + y2
-∞
the S1F state is at 33 512 cm-1 (energy gap ) 2187 cm-1), and
the S2N state is probably at 33 800 cm-1. Since the S1F origin
of spirofluorenenaphthalene lies roughly 300 cm-1 below the
In this formula, η is ∆D/Γ, the ratio of the Gaussian to
Lorentzian line widths. A formula giving the homogeneous line
width (fwhm), Γ, of a vibronic peak broadened by radiation-
less transitions in terms of the observed line width of the peak,
S2N origin, electronic energy transfer from fluorene to dimeth-
ylnaphthalene must be vibrationally mediated for the S1F origin
and low-lying vibronic states but might be enhanced by direct
dipolar coupling when higher vibronic levels of the donor are
excited.
Table 10 gives the rate of electronic energy transfer in
spirofluorenephenanthrene and spirofluorenenaphthalene at dif-
ferent vibronic levels inferred from the spectral broadening using
eq 3. In neither molecule is there any dramatic dependence of
the electronic energy transfer rate on vibronic excitation. If there
is any increase in rate in spirofluorenenaphthalene above the
S2N origin, the effect is small. Presumably, the density of states
of the S2N state is low, and direct dipolar coupling with the S1F
state depends on accidental coincidences.3
∆
DA, and the observed line width in the absence of the
radiationless transition, ∆D, was given by Amirav et al.26 This
model assumed that in the absence of radiationless transitions,
the line shape was Gaussian and the width was produced by
heterogeneous broadening, either unresolved rotational structure
or laser bandwidth. The Amirav et al. model corrects for this
heterogeneous broadening. For bichromophores, the correction
is given by
2
Γ ) [∆DA2 - [∆D /(2 ln 2)]}/∆DA
(3)
In Figure 12, we compare the absorption profile of the S1F origin

Energy Transfer in Bichromophoric Molecules
J. Phys. Chem. A, Vol. 103, No. 1, 1999 19
TABLE 10: Rate of Electronic Energy Transfer in
Spirofluorenephenanthrene (SFP) and
Spirofluorenenaphthalene (SFN) at Different Vibronic
Energy Levels
for the model compound, dimethylnaphthalene. Both the
magnitude and energy dependence of these lifetimes are different
from those of spirofluorenenaphthalene. This is not surprising
considering the different vibrational frequencies in the two
molecules. The very low frequency modes associated with the
nearly free internal rotation of the methyl groups in dimethyl-
naphthalene are shifted to much higher frequency when the
methyl groups are bound to fluorene in spirofluorenenaphtha-
lene. This would have a significant effect on the density of states
and on the rate of radiationless decay.
molecule excess energy/cm-1
∆
DA/cm-1 Γa/cm-1 kEET/1012 s-1
SFP
0
+150
+404
+728
+985
+997
0
+169
+204
+400
+688
+853
+985
6.43
3.85
3.84
6.31
4.54
5.45
3.37
2.77
3.69
6.08
4.32
7.04
4.55
6.05
3.22
3.21
5.93
4.01
5.00
2.65
1.90
3.03
5.68
3.76
6.69
4.02
1.14
0.61
0.60
1.12
0.76
0.94
0.50
0.36
0.57
1.07
0.71
1.26
0.76
SFN
C. 2,7-Dichlorospirobifluorene and 4,5-Diazospirobifluo-
rene. The two remaining large energy gap spiranes, dichlo-
rospirobifluorene and diazospirobifluorene, exhibited the same
qualitative behavior as spirofluorenephenanthrene and spiro-
fluorenenaphthalene. Specifically, in the case of dichlorospiro-
bifluorene, only the characteristic dichlorofluorene (acceptor)
emission was observed upon excitation of any vibronic transition
of the donor, proving that a complete singlet excitation transfer
occurred regardless of which vibrational state of the donor was
initially populated.
a
∆
) 1.83 cm-1 is assumed.
D
The fluorescence of the monomeric diazofluorene was not
sufficiently strong to be detected in the molecular beam
experiments. From the weak emission in solution, we estimate
that the S1 state of diazofluorene is approximately 1100 cm-1
lower in energy than the S1 state of fluorene. Attempts to excite
bichromophoric diazospirobifluorene in the spectral region of
the fluorene absorption also did not yield measurable emission
in the molecular beam experiment. It was concluded that the
donor fluorescence is quenched via rapid singlet excitation
transfer from the S1 state of fluorene to the nonemissive S1 state
of diazofluorene.
Conclusions
Electronic energy transfer has been observed in spirofluore-
nephenanthrene and spirofluorenenaphthalene. In these two
molecules, emission exclusively from phenanthrene and naph-
thalene was observed upon exciting the fluorene origin (S1F).
The special symmetry of these two molecules prohibits energy
transfer by either dipolar or exchange couplings if the symmetry
of the molecules is maintained. The occurrence of energy
transfer in spirofluorenephenanthrene and spirofluorenenaph-
thalene is explained by vibronic coupling via nontotally sym-
metric vibrations. Nontotally symmetric vibrations break the
symmetry of the molecule and reactivate both dipolar and
exchange couplings that promote electronic energy transfer. The
observed spectral broadening indicates that the rate of energy
transfer is substantially faster than the radiative decay rate of
fluorene, consistent with the fact that only acceptor like emission
was observed. The very large energy gap between the donor
and the acceptor also suggests that radiationless relaxations in
these two molecules occur in the statistical limit. Thus, electronic
energy transfer is rapid and irreversible. This is in marked
contrast to the behavior observed in spirobifluorene, where the
rate of energy transfer strongly depends on accidental coinci-
dences of energy levels between the donor and the acceptor.
Figure 13. Fluorescence lifetime of spirofluorenenaphthalene at
different excess vibrational energies above the S1N origin. The three
filled circles are the fluorescence lifetimes of 1,8-dimethylnaphthalene
measured at the S1 origin and two other levels above the S2 origin.
As in the case of spirofluorenephenanthrene, electronic energy
transfer in spirofluorenenaphthalene produces a lengthening of
the fluorescence lifetime from that of fluorene (16.7 ( 0.8 ns)
to that of dimethylnaphthalene (198 ( 16.4 ns). The variation
in the lifetime of spirofluorenenaphthalene with excess vibra-
tional energy is plotted in Figure 13. The fluorescence lifetime
of spirofluorenenaphthalene at the S1N origin is 151 ( 8 ns.
The lifetime drops to 117 ( 8 ns at the S1F origin and it keeps
decreasing slowly until it levels off at even higher vibrational
energy. Variations in the fluorescence lifetime have been
observed in substituted naphthalenes. These variations reflect
the changes in the rate of radiationless decay that leads to
intersystem crossing to the low-lying T1 state. In general, the
rate of radiationless decay increases with the density of states
of the T1 state.19 In Figure 13, we have also plotted three points
Acknowledgment. The work in the laboratory of D.H.L.
was supported by the National Science Foundation under Grant
CHE-9319958. The work in the P.P. laboratory was supported
by the Office of Basic Energy Sciences, U.S. Department of
Energy, through Grant DE-FG02-97ER14756. The authors thank
Mr. Gene A. Plaisance for his contribution to the synthesis of
one of the model compounds.

20 J. Phys. Chem. A, Vol. 103, No. 1, 1999
Yip et al.
(7) Piotrowiak, P.; Kobetic, R.; Schatz, T.; Tapper, S. Bull. Pol. Acad.
Sci. 1994, 42, 445.
(8) Sharfin, W.; Johnson, K. E.; Wharton, L.; Levy, D. H. J. Chem.
Phys. 1979, 71, 1292.
(9) Mu¨ller, M.; Hohlneicher, G. J. Am. Chem. Soc. 1990, 112, 1237.
(10) Healy, E. F.; Holder, A. J. Mol. Struct. 1993, 281, 141.
(11) Ohta, N.; Baba, H. Mol. Phys. 1986, 59, 921.
(12) Warren, J. A.; Hayes, J. M.; Small, G. J. Chem. Phys. 1986, 102,
323.
(13) Schettino, V.; Neto, N.; Califano, S. J. Chem. Phys. 1966, 44, 2724.
(14) Chakraborty, T.; Chowdhury, M. Chem. Phys. 1992, 159, 439.
(15) Beck, S. M.; Powers, D. E.; Hopkins, J. B.; Smalley, R. E. J. Chem.
Phys. 1980, 73, 2019.
(16) Behlen, F. M.; McDonald, D. B.; Sethuraman, V.; Rice, S. A. J.
Chem. Phys. 1981, 75, 5685.
(17) Warren, J. A.; Hayes, J. M.; Small, G. J. J. Chem. Phys. 1984, 80,
1786.
(18) Ichimura, T.; Auty, A. R.; Jones, A. C.; Phillips, D. J. Spectrosc.
Soc. Jpn. 1985, 34, 83.
(19) Jacobson, B. A.; Guest, J. A.; Novak, F. A.; Rice, S. A. J. Chem.
Phys. 1987, 87, 269.
(20) Fo¨rster, T. Discuss. Faraday Soc. 1959, 27, 7.
(21) Fo¨rster, T. Ann. Physik. 1948, 2, 55.
References and Notes
(1) (a) Speiser, S. Chem. ReV. 1996, 96, 1953 and references therein.
(b) Newton, M. D. Chem. ReV. 1991, 91, 767 and references therein. (c)
Chattoraj, M.; Bal, B.; Closs, G. L.; Levy, D. H. J. Phys. Chem. 1991, 95,
9666. (d) Chattoraj, M.; Bal, B.; Shi, Y.; Closs, G. L.; Levy, D. H. J. Phys.
Chem. 1993, 97, 13046. (e) Chattoraj, M.; Chung, D. D.; Bal, B.; Closs, G.
L.; Levy, D. H. J. Phys. Chem. 1994, 98, 3361. (f) Closs, G. L.; Piotrowiak,
P.; Miller, J. R. In Photochemical Energy ConVersion; Norris, J., R., Jr.,
Meisel, D., Ed.; Elsevier: New York, 1989. (g) Maki, A. H.; Weers, J. G.;
Hilinski, E. F.; Milton, S. V.; Rentzepis, P. M. J. Chem. Phys. 1984, 59,
2288.
(2) (a) Zeng, Y.; Zimmt, M. B. J. Am. Chem. Soc. 1991, 113, 5107.
(b) Zeng, Y.; Zimmt, M. B. J. Phys. Chem. 1992, 96, 8395. (c) Oliver, A.
M.; Paddon-Row, M. N.; Kroon, J.; Verhoeven, J. W. Chem. Phys. Lett.
1992, 191, 371. (d) Kumar, K.; Lin, Z.; Waldeck, D. H.; Zimmt, M. B. J.
Am. Chem. Soc. 1996, 118, 243.
(3) van Dantzig, N. A.; Levy, D. H.; Vigo, C.; Piotrowiak, P. J. Chem.
Phys. 1995, 103, 4894.
(4) No exciton splitting and no acceptor fluorescence were detected,
indicating that the electronic coupling was smaller than the spectral
resolution of the experiment (∼0.05 cm-1) and that the singlet excitation
transfer, if present, was several orders of magnitude slower than the decay
of the donor S1 state, τ > 20 ns.
(5) (a) Zasukha, V. A. Chem. Phys. 1988, 128, 289. (b) Glauser, W.
A.; Raber, D. J.; Stevens, B. J. Phys. Chem. 1991, 95, 1976.
(6) It is a common practice in the field of excitation and electron-
transfer studies to use the first-order perturbation theory methods to estimate
the electronic coupling between the donor and the acceptor. In the simplest
and most frequently used approach, the off-diagonal electronic matrix
elements, VDA, are obtained from the splittings (i.e., the shifts of the diagonal
elements) calculated at the ab initio or semiempirical level for a degenerate
donor-acceptor pair and transferred to the nondegenerate systems under
study. A better solution involves a calculation on a realistic donor-acceptor
pair and direct extraction of the off-diagonal elements. Since both approaches
by default calculate the coupling between the V ) 0 levels of the donor-
acceptor pair, neither of them is capable of accounting for the presence of
strong vibronic coupling. Therefore, in cases with complete symmetry
cancellations of the Coulomb and exchange integrals involving the localized
electronic wave functions of the donor and the acceptor, such calculations
do not yield a useful estimate of the magnitude of the electronic interaction
between the two chromophores.
(22) Dexter, D. L. J. Chem. Phys. 1953, 21, 836.
(23) Bixon, M.; Jortner, J. J. Chem. Phys. 1968, 48, 715.
(24) Freed, K. F.; Jortner, J. J. Chem. Phys. 1969, 50, 2916.
(25) Freed, K. F. J. Chem. Phys. 1970, 52, 1345.
(26) Amirav, A.; Sonnenschein, M.; Jortner, J. J. Chem. Phys. 1984,
88, 5593.
(27) There is extensive literature on the evaluation of the Voight line
shape (see the following: Armstrong, B. H. J. Quant. Spectrosc. Radiat.
Transfer 1967, 7, 61 and references contained therein. For the following
formula for evaluating the Voight profile in the limit where the homogeneous
line width is greater than the inhomogeneous line width, see:
Γ ) [∆DA2 - [∆D2/(2 ln 2)]}/∆DA
Amirav, A.; Sonnenschein, M.; Jortner, J. J. Phys. Chem. 1984, 88, 5593.
(4)
Use of this formula gives a value of Γ ) 6.05 cm-1
.
(28) Bree, A.; Zwarich, R. J. Chem. Phys. 1969, 51, 903. (b) Bree, A.;
Zwarich, R. J. Chem. Phys. 1969, 51, 912.