Femtosecond UV/mid-IR study of photochromism of the spiropyran 1′,3′-dihydro-1′,3′,3′-trimethyl-6- nitrospiro[2H-1-benzopyran-2,2′-(2H)-indole] in solution

Article abstract of DOI:10.1016/S0009-2614(03)00949-7

The ring-opening reaction of 1′,3′- dihydro-1′,3′,3′- trimethyl-6-nitrospiro[2H-1-benzopyran-2,2′-(2H)-indole] is investigated in two solvents, by probing the evolution of the vibrational absorption spectrum with 130 fs time-resolution. Competition betwee

Full text of DOI:10.1016/S0009-2614(03)00949-7

Chemical Physics Letters 376 (2003) 214–219
www.elsevier.com/locate/cplett
Femtosecond UV/mid-IR study of photochromism
of the spiropyran 10,30-dihydro-10,30,30-trimethyl-6-nitrospiro-
[2H-1-benzopyran-2,20-(2H)-indole] in solution
a
Ann-Kathrin Holm , Matteo Rini , Erik T.J. Nibbering , Henk Fidder
b
b
a,
*
a
Department of Physical Chemistry, Uppsala University, P.O. Box 579, Uppsala S-751 23, Sweden
Max-Born-Institut f€ur Nichtlineare Optik und Kurzzeitspektroskopie, Max Born Strasse 2A, Berlin D-12489, Germany
b
Received 10 March 2003; in final form 25 April 2003
Published online: 1 July 2003
Abstract
The ring-opening reaction of 10 ,30 -dihydro-10 ,30 ,30 -trimethyl-6-nitrospiro[2H-1-benzopyran-2,20 -(2H)-indole] is in-
vestigated in two solvents, by probing the evolution of the vibrational absorption spectrum with 130 fs time-resolution.
Competition between internal conversion and photochemistry is found to depend on the solvent. The internal con-
version quantum yield is determined to be 0.63 in perdeuterated acetonitrile and 0.34 in tetrachloroethene. Based on
spectral features and biexponential kinetics, the formation of an additional merocyanine isomer in tetrachloroethene is
concluded.
Ó 2003 Elsevier Science B.V. All rights reserved.
1. Introduction
hydro-10 ,30 ,30 -trimethyl-6-nitrospiro[2H-1-benzopy-
ran-2,20 -(2H)-indole] (6-nitroBIPS) in two sol-
vents: tetrachloroethene (TCE) and perdeuterated
acetonitrile (ACN-d3).
Photochromism is the property that a substance
can undergo reversible colour changes by exposure
to light of specific wavelengths. Photochromism
has met with a longstanding scientific and indus-
trial interest [1–17], with visions of applying this
phenomenon in optical data storage [2], and op-
tically controlled molecular switching [3]. Here, we
investigate the spiropyran–merocyanine photo-
chemical ring opening/closure (Fig. 1) of 10 ,30 -di-
Scientifically this simple reaction already pre-
sents a rather complex puzzle. Not only is there the
issue whether singlet and triplet states both are
involved [4–7], but also the product can appear as
eight different cis–trans isomers [8–13]. On top of
that the merocyanine product can aggregate [14].
A wide range of experiments has been performed
on 6-nitroBIPS, including low temperature ab-
sorption [8], triplet sensitization [6], femto- [12,15]
and nanosecond [7] time-resolved UV–vis pump–
probe spectroscopy, and time-resolved resonance
* Corresponding author. Fax: +46-18-471-3654.
E-mail address: henk.fidder@fki.uu.se (H. Fidder).
0009-2614/03/$ - see front matter Ó 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0009-2614(03)00949-7

A.-K. Holm et al. / Chemical Physics Letters 376 (2003) 214–219
215
Fig. 1. Reaction scheme for 6-nitroBIPS.
Raman spectroscopy [9,10]. Nevertheless, the
number of investigations on the femto- and pico-
second timescale is still relatively limited.
BIPS in TCE and ACN-d3. The photochromic
reaction was initiated by 70 fs pulses at 330 nm (2
lJ/pulse), and the evolution of the vibrational
absorption spectrum was probed with 100 fs IR
pulses (ꢁ20 nJ/pulse), with the centre frequency
varying from 1260 to 1600 cmꢀ1. In each mea-
surement the detector covers 100–200 cmꢀ1, with a
spectral resolution of 4–7 cmꢀ1. More extensive
experimental details can be found in [16].
Recently, we started femtosecond UV/mid-IR
investigations on photochromism of spiropyrans,
focussing on the time evolution of the vibrational
absorption spectrum. In a first publication [16]
on the compound 10 ,30 -dihydro-10 ,30 ,30 -trimethyl-
spiro[2H-1-benzopyran-2,20 -(2H)-indole] (BIPS)
in TCE, we concluded that the photochemistry
quantum yield was maximally 0.1 due to a very
fast and efficient internal conversion process.
Moreover, appearance of the merocyanine product
seemed to involve a 28 ps time constant, much
slower than the 0.9 ps rise time concluded from
UV–vis experiments [12]. Here, we extend our in-
vestigations to 6-nitroBIPS. We find that internal
conversion competes with the photochemical
conversion pathways, although less efficient than
for BIPS. In addition, the efficiency of this process
appears to be solvent dependent. Data acquired
over a wider frequency range are presented and
comparisons are made to results from investiga-
tions on 6-nitroBIPS with other techniques [7–13].
3. Results
Fig. 2 shows the steady-state IR absorption
spectra of 6-nitroBIPS in ACN-d3 and TCE at
room temperature, corrected for solvent contri-
butions, in the range probed in our time-resolved
measurements. Electronic absorption spectra (not
shown) indicate that the compound is almost ex-
clusively in the closed spiropyran form in both
solutions. We estimate that at most 1% is in a
merocyanine form in ACN-d3 and 0.01% in TCE.
The IR spectra in Fig. 2 clearly illustrate that the
peak intensities of IR absorption bands can vary
quite considerably in different solvents. For in-
stance, the ratio of the peak intensities at ꢁ1342
cmꢀ1 divided by that at ꢁ1524 cmꢀ1 differs by 40%
for these two solvents. Solvent dependence of IR
2. Experimental
Tetrachloroethene (TCE) and 6-nitroBIPS were
purchased from Sigma–Aldrich. Perdeuterated
acetonitrile (ACN-d3; deuteration grade 99.8%)
was purchased from Deutero GmbH. The com-
pounds were used without further purification.
These solvents have limited IR absorption in the
investigated frequency range. Steady-state FT-IR
spectra of 6-nitroBIPS in TCE and ACN-d3 were
recorded with 1 cmꢀ1 resolution, using a Bio-Rad
FT-IR spectrometer. Time-resolved measurements
were performed on 20 mM solutions of 6-nitro-
Fig. 2. Steady-state IR absorption spectra for 6-nitroBIPS in
ACN-d3 (– –, 51 mM) and in TCE (——, 42 mM).

216
A.-K. Holm et al. / Chemical Physics Letters 376 (2003) 214–219
absorption intensities and frequencies is a known
effect [18].
Fig. 3 demonstrates overall transient differential
spectra created from our data for 6-nitroBIPS in
TCE (Fig. 3, upper panel) and ACN-d3 (Fig. 3,
lower panel), 2 ps after UV excitation, and at long
delay. For comparison with the bleaches at early
delays the steady-state IR absorption spectrum is
also plotted. The initial spectra resemble the stea-
dy-state IR spectra well, indicating removal of
ground state (closed) spiropyran species by the
femtosecond UV excitation pulse. Note further
that in TCE not a single sharp increased absorp-
tion feature is found at early times. In ACN-d3 the
initial spectrum shows a few positive features
(marked *), that we assign to vibrational hot
bands of the spiropyran species in the electronic
ground state formed after internal conversion. The
kinetics in ACN-d3 at 1264 cmꢀ1 depicted in
Fig. 4c, illustrate the vibrational cooling dynamics
for hot spiropyran [16].
Spectra at long delay times show for both sol-
vents (residual) bleaches as well as a number of
peaks related to photoproducts. First we point out
that in TCE the bleaches have recovered by ꢁ34%
Fig. 4. (a–b) Bleach recovery at two vibrational frequencies of
6-nitroBIPS (in the spiropyran form) in TCE (a) and in ACN-d3
(b). The numbers correspond to the central frequency associ-
ated with the diode pixels. The solid lines are fits corresponding
to decay times of 47 ps in TCE, and 14 ps in in ACN-d3. (c)
Kinetics related to the spiropyran vibrational hot band at 1264
cmꢀ1 in ACN-d3. The time scale at the bottom applies to both
panels band c.
in TCE after 150 ps, and ꢁ63% in ACN-d3 after 60
ps. Note also the strong lines at 1297 and 1519
cmꢀ1 (marked O) in TCE, which appear nearly
absent in ACN-d3. These lines we associate with
the formation of an additional merocyanine iso-
mer in TCE.
4. Discussion
Fig. 3. Transient IR-spectra between ꢁ1200 and ꢁ1650 cmꢀ1
,
4.1. Bleach recovery and internal conversion
for 6-nitroBIPS in TCE (upper panel) and in ACN-d3 (lower
panel), 2 ps after UV-excitation (thin solid lines), and at 150
and 60 ps, for TCE and ACN-d3, respectively (thick solid lines).
For comparison, the inverted steady-state IR spectra of 6-ni-
troBIPS in the respective solvents are also depicted (dotted
lines), roughly scaled to the bleaches at early times, with an
offset of )0.25 mOD for ACN-d3, and )0.5 mOD for TCE.
The kinetic behaviour related to the spiropyran
form of 6-nitroBIPS is illustrated in Fig. 4 at two
bleached vibrational absorption bands: at 1342
and 1612 cmꢀ1. These data show that both in TCE
(Fig. 4a) and in ACN-d3 (Fig. 4b) the bleaches

A.-K. Holm et al. / Chemical Physics Letters 376 (2003) 214–219
217
partially recover. This recovery has to be due to
regeneration of the spiropyran ground state, and
therefore indicates the occurrence of internal
conversion. The data demonstrate that the quan-
tum yield for internal conversion is solvent de-
pendent: about 34% in TCE, and 63% in ACN-d3
(error margin: 5%). Previous investigations had
not identified this apparently rather influential
decay channel. It may well be that the efficiency of
internal conversion is a key factor in determining
the overall photochemistry quantum yield. Given
the observed bleach recovery the maximum
quantum yield for photochemistry of 6-nitroBIPS
is 0.66 in TCE, and 0.37 in ACN-d3. (Absolute
[1,17] and relative [4,5] photochemistry quantum
yields have been reported for 6-nitroBIPS in var-
ious solvents.) The refill rate of the bleach, after
internal conversion and subsequent vibrational
cooling, is also different for these two solvents: 47
(ꢂ 10) ps in TCE, and 14 (ꢂ 2) ps in ACN-d3.
Takahashi and co-workers have concluded the
existence of five isomeric species [9] from time-re-
solved Raman spectra of 6-nitroBIPS, taken from
200 ns to 2 ms after UV excitation in three dif-
ferent solvents. These assignments were mainly
based on intensity variations with solvent of dif-
ferent Raman lines, and analogy to a similar study
on the parent compound BIPS [11]. Two species
were brought in connection with only a single
Raman line. As we demonstrated for spiropyran in
Fig. 2, IR absorption intensities of the same spe-
cies can already vary considerably from solvent to
solvent. Clearly, moderate intensity variations
alone are a poor basis for the identification of
different isomeric species. In a later investigation
on 6-nitroBIPS in cyclohexane, they followed the
time evolution of the Raman spectrum between
20 ns and 100 ls after UV excitation [10]. This
time only three transient species were invoked in
the interpretation. The species identified at earliest
times was ascribed to a merocyanine triplet state.
A convincing correlation between the decay of the
1409 cmꢀ1 Raman intensity (0.33 ls time con-
stant), and the increase of the 1525 cmꢀ1 Raman
intensity, led to assignment of the second species
as ground state merocyanine. On the microsecond
timescale a third species is identified as a dimeric
species. As none of these re-assignments require
the existence of different isomers, the previous
claim of five existing isomers should at least be
reduced to maximally three. Several other incon-
sistencies can be noted in their Raman assign-
ments. For example, the 20 ns data in [10] clearly
illustrate that assigning both the 1550 and 1523
cmꢀ1 lines to the same Ôinitial cisoid 1523 cmꢀ1
speciesÕ in [9] was incorrect.
4.2. Existence of different merocyanine isomers
Based on the spectra at long delays shown in
Fig. 3, combined with observed kinetics at different
frequencies, we identify with certainty product
bands in ACN-d3 around 1245, 1299, 1355, 1417
and 1551 cmꢀ1. In TCE bands are identified at the
similar frequencies 1251, 1297, 1351, 1413 cmꢀ1
and in the range 1537–1556 cmꢀ1. In addition, in
TCE we observe bands at 1263, 1318, 1439, 1463
and 1519 cmꢀ1. Note that, in general, the signals in
TCE are much stronger than in ACN-d3. There-
fore some of the weaker bands in TCE could be
below the detection threshold in ACN-d3. The
bands at 1297, 1413 and 1519 cmꢀ1 are much more
intense in TCE than in ACN-d3 (about an order of
magnitude or more). As the difference in internal
conversion efficiency can only justify a factor two
in photochemistry quantum yield, we conclude
that these three peaks constitute reliable evidence
for the existence of an additional different mero-
cyanine isomer in TCE, that is not present in
ACN-d3. Because all product bands in ACN-d3 are
also observed in TCE, the isomer that is formed in
ACN-d3 is likely to be present in TCE as well.
These conclusions are further supported by the
product formation kinetics discussed below.
About half of the observed resonance Raman
peaks correspond to lines we observe in IR ab-
sorption: for 6-nitroBIPS in ACN (after 2 ms, [9])
these are at 1358 and 1411 cmꢀ1. Data in TCE are
probably best compared to the 20 ns Raman data
in cyclohexane [10], that show related peaks at
1297, 1350, 1409 and 1552 cmꢀ1. Recall that Ra-
man and IR absorption yield complementary in-
formation.
Recently, density-functional-theory calculations
on the spiropyran form of 6-nitroBIPS and the
four most stable merocyanine isomers were

218
A.-K. Holm et al. / Chemical Physics Letters 376 (2003) 214–219
presented, and compared to IR absorption spectra
in an argon matrix at 8 K [13]. Reasonable
agreement was obtained between experiment and
calculations, although insufficient to unequivocally
identify the isomers. More specific, we point at the
intense experimental line at 1529 cmꢀ1, which
corresponds to a calculated line at 1551 cmꢀ1. This
experimental 1529 cmꢀ1 line could be associated
with the extra isomer in our TCE data, whereas
the calculated 1551 cmꢀ1 line agrees with both
isomers observed by us.
Altogether we are convinced that existing ex-
perimental results provide insufficient evidence for
claiming formation of five isomers. It seems
therefore prudent to limit our conclusion to strong
indications for existence of a second isomeric
species in TCE, not formed in ACN-d3.
4.3. Product formation kinetics
Fig. 5. (a) Merocyanine product formation kinetics for 6-ni-
troBIPS in TCE at 1551 (Ã) and 1556 cmꢀ1 (j). (b) Product
formation kinetics for 6-nitroBIPS in ACN-d3 at 1418 (d) and
1551 cmꢀ1 (s).
Kinetics at selected wavelengths is depicted in
Fig. 5. In Fig. 5bthe product formation kinetics
in ACN-d3 are shown at 1418 (d) and 1551 cmꢀ1
(s). Single exponential fits of these two traces
indicate a single exponential formation time of
9.5 ꢂ 1.5 ps. No further change in product ab-
sorption was observed between 30 and 100 ps.
However, there seems to be some variation in rise
time over the bands; e.g., at 1413 cmꢀ1 we obtain
7.1 ps. This behaviour is observed even more
clearly in TCE, where the overall product for-
mation in TCE is considerably slower. As an il-
lustration Fig. 5a shows the product related
signals at 1551 (Ã) and 1556 cmꢀ1 (j). Clearly
the signal rises much faster at 1551 cmꢀ1 during
the first 50 ps. Single exponential fits at different
product bands produce rise times in the range of
35 to 60 ps for 6-nitroBIPS in TCE. As indicated
by the data in Fig. 5b the kinetics require at least
a biexponential fitting approach. For the stron-
gest band (around 1413 cmꢀ1) this procedure re-
sulted in a faster component increasing from
about 5 ps at 1397 cmꢀ1 to 28 ps at 1413 and
1418 cmꢀ1, and a second component typically
>100 ps. Because the data are only up to 150 ps,
the time constant of the slow component cannot
be extracted reliably. The most likely source for
the observed frequency dependence of the faster
component is vibrational cooling dynamics of the
merocyanine products.
Absorption change studies on 6-nitroBIPS in
hexane and acetonitrile by Lenoble and Becker [7],
with 1 ns time resolution, indicated that in aceto-
nitrile the rise of the 570 nm product related band
was complete within 1 ns. Our measurements
lower the upper limit for the product formation
time in (perdeuterated) acetonitrile to about 9.5 ps.
The fast component in TCE (ꢁ5–28 ps), and the
time constant in ACN-d3 (ꢁ7–10 ps), are in line
with a 20 ps time constant found by Ernsting and
Arthen-Engeland [12] in a femtosecond UV/white-
light investigation of 6-nitroBIPS in n-pentane.
Based on solvent polarity one may suspect that
the kinetic behaviour in hexane, n-pentane, and te-
trachloroethene should be rather similar. This
seems true for our measurements compared to
those of Ernsting and Arthen-Engeland [12]. How-
ever, in hexane the fastest rise time identified by
Lenoble and Becker was 4 ns at 430 nm [7].
Whether the slow >100 ps component in TCE can
be reconciled with their nanosecond result in
hexane remains an open question.

A.-K. Holm et al. / Chemical Physics Letters 376 (2003) 214–219
219
ganic Chemistry, vol. 40, Elsevier, Amsterdam, 1990,
Chapter 8 and 23, and references therein.
The biexponentiality of the product formation
in TCE, versus single exponential kinetics in ACN-
d3, clearly corroborates the conclusion drawn
above on formation of a second isomeric species in
TCE. This biexponentiality also implies that these
species cannot be formed in direct competition
from a mutual precursor state, since this would
lead to the same formation constant for both. Note
that published conclusions on different isomers was
based on data obtained 20 ns or more after UV
irradiation of a sample. These data therefore con-
stitute the first time-resolved evidence that a second
isomer already appears within 100 ps.
[3] B.L. Feringa (Ed.), Molecular Switches, Wiley-VCH,
Weinheim, Germany, 2001.
€
[4] A.K. Chibisov, H. Gorner, J. Phys. Chem. A 101 (1997)
4305.
€
[5] H. Gorner, Chem. Phys. Lett. 282 (1998) 381.
[6] D.A. Reeves, F. Wilkinson, J. Chem. Soc. Faraday Trans.
2 (69) (1973) 1381.
[7] C. Lenoble, R.S. Becker, J. Phys. Chem. 90 (1986)
62.
[8] T. Bercovici, R. Heiligman-Rim, E. Fischer, Mol. Photo-
chem. 1 (1969) 23.
[9] H. Takahashi, H. Murakawa, Y. Sakaino, T. Ohzeki, J.
Abe, O. Yamada, J. Photochem. Photobiol. A 45 (1988)
233.
[10] T. Yuzawa, A. Shimojima, H. Takahashi, J. Mol. Struct.
352/353 (1995) 497.
Acknowledgements
[11] H. Takahashi, K. Yoda, H. Isaka, T. Ohzeki, Y. Sakaino,
Chem. Phys. Lett. 140 (1987) 90.
H.F. gratefully acknowledges financial support
from the Swedish Research Council (Vetens-
[12] N.P. Ernsting, T. Arthen-Engeland, J. Phys. Chem. 95
(1991) 5502.
[13] Y. Futami, M.L.S. Chin, S. Kudoh, M. Takayanagi, M.
Nakata, Chem. Phys. Lett. 370 (2003) 460.
[14] H. Eckhardt, A. Bose, V.A. Krongauz, Polymer 28 (1987)
1959.
ꢀ
kapsradet). This work benefitted from financial
travel support through LIMANS (Project
MBI000237), and instrument support by the
Deutsche Forschungsgemeinschaft (Project DFG
NI 492/2-2).
[15] M.V. Alfimov, A.V. Balakin, S.P. Gromov, Y.V. Zaushit-
syn, A.Y. Resnyanskii, A.P. Shkurinov, Russ. J. Phys.
Chem. 73 (1999) 1685.
[16] M. Rini, A.-K. Holm, E.T.J. Nibbering, H. Fidder, J. Am.
Chem. Soc. 125 (2003) 3028.
References
[17] A.S. Kholmanskii, K.M. Dyumaev, Russ. Chem. Rev. 56
(1987) 136.
[1] R.C. Bertelson, in: G.H. Brown (Ed.), Photochromism,
Wiley, New York, 1971, p. 45-ff, and references therein.
[18] C. Reichardt, Solvents and Solvent Effects in Organic
Chemistry, third ed., Wiley-VCH, Weinheim, Germany,
2003, Chapter 6 and references therein.
€
[2] R. Guglielmetti, in: H. Durr, H. Bouas-Laurent (Eds.),
Photochromism: Molecules and Systems, Studies in Or-