Direct structural determination of conformations of photoswitchable molecules by laser desorption-electron diffraction

Article abstract of DOI:10.1002/anie.201003583

Fluorine in charge! The fluorine-iminium ion gauche effect has been exploited in the design of conformational probes for organocatalysis (see scheme). Stabilizing hyperconjugative [σ C- H→σ^*C-F] and/or electrostatic [N⁺ F ^δ-

Full text of DOI:10.1002/anie.201003583

Communications
DOI: 10.1002/anie.201003583
Conformational Analysis
Direct Structural Determination of Conformations of Photoswitchable
Molecules by Laser Desorption–Electron Diffraction**
Andreas Gahlmann, I-Ren Lee, and Ahmed H. Zewail*
The reversible isomerization between the closed spiropyran
(SP) and the open merocyanine (MC) forms of photoswitch-
able molecules, such as derivatives of 1,3,3-trimethylindoli-
nobenzospiropyran (BIPS), have attracted considerable
experimental and theoretical interest over the last decades.
Because light of different wavelength can initiate both the
forward and the reverse reaction in some BIPS derivatives,
the photochromic pair of isomers presents interesting oppor-
tunities in holographic data storage[1] and as molecular
switches to control material properties and biological func-
tion.[2]
The closed form consists of an indoline and a chromene
subunit joined together at the central spiro-carbon, with the
two fused ring systems forming two perpendicular planes.
Absorption of a UV photon leads to cleavage of the spiro-
ꢀ
carbon oxygen bond and to subsequent rearrangement of the
two subunits to form the extended open form(s) connected by
a central bridge segment consisting of three bonds, labeled
abg, as shown in Figure 1. Due to the larger conjugated p-
electron system, the open forms have a strong S0!S1
absorption maximum that is red-shifted relative to the
absorption of the closed form.
A large range of isomers is possible for the open forms,
because each of the three ethylenic bonds can exhibit either
the cis or the trans conformation. In addition, the involvement
of a triplet-state mechanism has been discussed extensively
for nitro-substituted BIPS derivatives.[3–9] Thus, a priori, 2·23 =
16 possible structures can exist for the open product(s).
Quantum chemical calculations have shown that the isomers
having the central bond in the trans configuration (designated
Figure 1. Structures of the photoswitchable closed and open cis–trans–
cis form of 6-nitro-BIPS. The open structures, formed after photo-
excitation, are labeled by the cis (C) or trans (T) configuration of the
indicated dihedral angels (abg) in the bridge segment.
TTT, CTT, TTC, CTC) are very similar in energy and are
more stable than the cis isomers.[10–12]
In solution, time-resolved spectroscopic studies have
concluded that the open photo-product is formed on the
picosecond time scale.[13–18] Even though a distribution to
different isomers has been proposed,[13,14,19,20] the number and
identity of species formed are still subject of discussion.[7,16]
Because the time scales of product formation are heavily
dependent on the nature of the solvent,[6,16,18] it is difficult to
deduce the molecular basis of the mechanism.[4–8,14,21] Using
NMR spectroscopy, the TTC isomer, in rapid equilibrium
with the TTT isomer, was identified as the thermally
populated species of 6,8-dinitro-BIPS, which exists in its
more stable open forms at room temperature;[22] similarly, the
TTC and CTC isomers have been detected in the related
compound spironaphtoxazine.[23] The TTT isomer of 6,8-
dinitro-BIPS[22] and the TTC isomer of 6-nitro-8-bromo-
BIPS[24] could be crystallized from acetone and water
solutions, respectively, and identified using X-ray crystallog-
raphy.
Here, we report our study of the molecule 6-nitro-BIPS
using laser desorption–electron diffraction in order to deter-
mine the nascent product structures upon photo-excitation.[25]
In this study, the diffraction is from isolated molecules made
into a plume with no perturbations from a solvent. While
ambiguities continue to exist regarding the competition
among reaction pathways, the various product yields, and
[*] A. Gahlmann, I-R. Lee, Prof. A. H. Zewail
Physical Biology Center for Ultrafast Science and Technology
Arthur Amos Noyes Laboratory of Chemical Physics
California Institute of Technology, Pasadena, CA 91125 (USA)
E-mail: zewail@caltech.edu
[**] This work was supported by the National Science Foundation and
the Air Force Office of Scientific Research in the Gordon and Betty
Moore Center of Physical Biology at Caltech.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003583.
6524
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 6524 –6527

Angewandte
Chemie
the transient structures involved in solution, complete struc-
tural determination of all species involved in the absence of a
perturbing solvent sheds light on the intrinsic chemistry of the
molecule itself.
With electron diffraction we are able to determine both
ground-state and product structures for excitation of 6-nitro-
BIPS at 266 nm. For the ground-state structure, the exper-
imental and the theoretical molecular scattering function,
sM(s), together with the radial distribution, f(r), are shown in
Figure 2. Starting with a calculated structure of the ground
recorded at + 100 ns, is fit with a linear combination of
theoretically calculated sM(s) curves (see Experimental
Section for structure and temperature calculations). The
coefficients of this linear combination report directly on the
fractional abundance of a given species.
In the initial screening for product structures, we consid-
ered all possible open isomers (except the TCC structure, for
which no stationary point could be located)[11] as well as the
closed forms in their lowest singlet and triplet electronic
states (see Tables SI2 for calculated structure parameters and
frequencies). The quality of the fit is quantified in the c2 value
and the linear combinations are ranked in order of their
ability to reproduce the experimental data. We found that
fitting the data with less than two products is insufficient to
produce a good fit. If three or more products are included,
then ca. 70% of the product species tend to be in a closed
form while ca. 30% are in an open form. Additionally, the
linear combinations, which assign about equal abundance to
the vibrationally hot ground state and to the closed form in its
lowest triplet state, produce the best fits. Finally, the open
structures, for which the central bridge segment is in the CTC
or CTT conformation, are highly favored.
With the most plausible reaction products identified, we
attempted to refine the product structures (with their temper-
atures constrained at calculated values) in each linear
combination separately to further improve the quality of the
fit. Table SI3 shows the resulting ranking of linear combina-
tions after the attempt at structural refinement in each case.
Only three of the combinations allowed for the fitting of four
and five orthogonal parameters in each structure, while the
rest produced unphysical geometries after only fitting one or
two orthogonal parameters. The three combinations differ
only in the identity of the open form, while the fractions
remain similar. The c2 value of the combination including the
CTC structure represents the global minimum (best fit) on the
c2 hypersurface for the indicated number of orthogonal
parameters. We thus conclude that, based on the quality of the
fits, the following species are produced in the gas-phase:
SP(S0) at 1164 K, SP(T1) at 843 K, and MC CTC(S0) at
1132 K.
Using these product species, Figure 3 displays the fitted
DsM(s) curve with the experimental data at the + 100 ns time
delay and the corresponding Df(r) curve; also shown are the
theoretical curves, where the temperatures of all species are
held at the initial temperature of 510 K, thus isolating the
difference signal due to structural rearrangement only.
However, because electron diffraction is recording vibration-
ally-averaged structures and is therefore sensitive to temper-
ature, it is necessary to constrain the temperature of the
product species to calculated values, accounting for photon
absorption, when refining structural parameters (see Exper-
imental Section).
Figure 2. Diffraction results of the ground-state (GS) structure of 6-
nitro-BIPS. Shown are a) the molecular scattering function, sM(s), and
b) the radial distribution, f(r), for the experimental (circles) and
theoretical curves (line). Also shown is a weighted histogram of all the
internuclear distances measured in the diffraction experiment.
state, we refined seven orthogonal structural parameters, as
well as the internal temperature of the molecule (fitted value:
510 K), to achieve satisfactory agreement between the
experimental data and the theoretical model.[25] The refined
structural parameters are listed in Table SI1 in the Supporting
Information, together with values obtained from density
functional theory (DFT) calculations; they have discrepancies
of less than 0.056 ꢀ and 0.338 for bond lengths and angles,
respectively.
For studies of structural dynamics following UV excita-
tion, we recorded time-resolved diffraction patterns at
ꢀ100 ns and + 100 ns. To identify the structures present and
their abundance, the experimental frame-referenced DsM(s)
curve,[26] obtained by subtracting the reference diffraction
pattern recorded at ꢀ100 ns from the diffraction pattern
Regardless, the depletion of the first, second, and higher
order interatomic distances due to the ring opening and bond
twisting manifest itself in the more prominent negative peaks
in the Df(r) curve. The refined structural parameters are listed
in Tables SI4, together with values obtained from DFT
calculations. The successful identification of the product
structure among the many candidates demonstrates the
Angew. Chem. Int. Ed. 2010, 49, 6524 –6527
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6525

Communications
In a recent gas-phase study, 6-nitro-BIPS was also excited
at 266 nm (S0!S3) and the results of the picosecond kinetics
were accounted for using a sequential mechanism involving
four distinguishable species. The authors detected a long-lived
(> 500 ps) photo-product emerging with a formation time of
12 ps, which they argued to represent an open form.[28]
However, as shown here, three long-lived product structures
in different states at 100 ns have been identified. Efforts to
determine whether these species are already present on the
picosecond time scale will be the subject of a future
contribution using ultrafast electron diffraction.[26]
In summary, we have determined the isolated-molecule
structure of photochromic 6-nitro-BIPS, as well as its major
photo-products after excitation at 266 nm. We were able to
identify three different nascent species and their abundances:
The closed form in the ground state (30%) and the lowest
triplet state (39%) and the open form in the CTC conforma-
tion (31%) in the ground state. This study from electron
diffraction combined with laser desorption elucidates the
importance of identifying and following all major product
structures, if the mechanism of this complex reaction in its
entirety is to be established.
Experimental Section
Figure 3. Diffraction results of the excited state(ES)-products struc-
Electron diffraction experiments were performed in our newly
constructed UED-4 apparatus.[25] Briefly, electron pulses (5 ꢁ 107
electrons/pulse, 60 keV, 1 kHz) were generated using the 266 nm
output from a ns-Nd:YAG laser. The camera distance (26.716 cm)
and the instrumental point spread function[27] (13.5 pixel FWHM)
were calibrated by fitting the measured diffraction data of CO2 gas to
its known structural parameters.[29]
tures of 6-nitro-BIPS, following UV excitation. Shown are a) frame-
referenced molecular scattering function, DsM(s; t=+100 ns,
t
ref =ꢀ100 ns), and b) the frame-referenced radial distribution, Df(r;
t=+100 ns, tref =ꢀ100 ns), for the experimental (circles) and theoret-
ical curves (line). Also shown in green are the difference curves, where
all species are held at the temperature of the ground state (510 K), to
highlight the contribution due to structural rearrangement.
The 6-nitro-BIPS sample powder (Sigma Aldrich, CAS#: 1498-
88-0) was ground with a mortar and pestle to reduce the particle size
and loaded into our desorption source. During the course of the
experiment, the fine powder was continuously transferred onto the
glassy carbon substrate by circular mechanical motion and subse-
quently desorbed into the gas-phase by a second ns-Nd:YAG laser.
Once vaporized, the sample was photo-excited at 266 nm using the
frequency-tripled output from a Ti:Sapphire femtosecond laser
system and subsequently interrogated by the electron pulses.
Structural analysis was conducted using home-built software, as
described previously.[26,30] Theoretically calculated curves were
obtained using the starting geometries and vibrational frequencies
from density functional theory calculations at the B3LYP/6-311G-
(d,p) level using the Gaussian 98 suite,[31] and we calculated the
internal temperature of the product species by considering the
absorbed photon energy, the vibrational frequencies, and the differ-
ence in electronic energy relative to the ground-state species.
Refinement of the (nonlinear) structural parameters was conducted
using the method of singular value decomposition.[32] Thus, the
orthogonal structural parameters that are fitted during refinement are
composed of linear combinations of the redundant internal coordi-
nates that are used to initially define the geometry of the molecules.
ability of electron diffraction to solve a complex (nm-scale)
molecular structure of low symmetry, even though the largest
internuclear distances are beyond the present instrumental
coherence length, which we estimate, given the different
parameters involved in these experiments, to be 6–7 ꢀ (88%
visibility).[27]
The structural dynamics reported here support the
involvement of a triplet-state mechanism in the overall
photochemistry of 6-nitro-BIPS, because we detect a closed
form in the T1 state. However, the open CTC structure is
detected in its electronic ground state and it is uncertain
whether this product was formed through that triplet-state
pathway or through a separate singlet-state pathway. Through
theoretical means, the CTC structure has been found to be the
first trans structure formed from the cis intermediate struc-
ture (CCC) after ring opening, due to the steric hindrance
between the two methyl groups on the tetrahedral indoline
carbon and the oxygen atom.[11–12] The quantum yield (30%)
of the internally hot ground state determined here is
comparable to the quantum yield (34%) found spectroscopi-
cally in trichloroethylene solution.[21] However, in our case,
this species could have been formed either through internal
conversion from the initially excited state, or through the
thermal reverse reaction of ring-closing.
Received: June 12, 2010
Published online: August 2, 2010
Keywords: conformational analysis ·
.
density functional calculations · electron diffraction ·
laser desorption · photochromism
6526 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 6524 –6527

Angewandte
Chemie
[21] A.-K. Holm, O. F. Mohammed, M. Rini, E. Mukhtar, E. T. J.
[1] G. Berkovic, V. Krongauz, V. Weiss, Chem. Rev. 2000, 100, 1741 –
1753.
[2] I. Willner, Acc. Chem. Res. 1997, 30, 347 – 356.
[3] T. Bercovici, E. Fischer, J. Am. Chem. Soc. 1964, 86, 5687 – 5688.
[4] D. A. Reeves, F. Wilkinson, J. Chem. Soc. Faraday Trans. 2 1973,
69, 1381 – 1390.
[5] S. A. Krysanov, M. V. Alfimov, Chem. Phys. Lett. 1982, 91, 77 –
80.
[6] C. Lenoble, R. S. Becker, J. Phys. Chem. 1986, 90, 62 – 65.
[7] T. Yuzawa, A. Shimojima, H. Takahashi, J. Mol. Struct. 1995, 352,
497 – 507.
Nibbering, H. Fidder, J. Phys. Chem. A 2005, 109, 8962 – 8968.
[22] J. Hobley, V. Malatesta, R. Millini, L. Montanari, W. O. N.
Parker, Phys. Chem. Chem. Phys. 1999, 1, 3259 – 3267.
[23] S. Delbaere, C. Bochu, N. Azaroual, G. Buntinx, G. Vermeersch,
J. Chem. Soc. Perkin Trans. 2 1997, 1499 – 1501.
[24] S. M. Aldoshin, L. O. Atovmyan, Mol. Cryst. Liq. Cryst. 1987,
149, 251 – 290.
[25] A. Gahlmann, S. T. Park, A. H. Zewail, J. Am. Chem. Soc. 2009,
131, 2806 – 2808.
[26] S. T. Park, J. S. Feenstra, A. H. Zewail, J. Chem. Phys. 2006, 124,
174707, and references therein.
[8] H. Gꢂrner, L. S. Atabekyan, A. K. Chibisov, Chem. Phys. Lett.
1996, 260, 59 – 64.
[27] A. Gahlmann, S. T. Park, A. H. Zewail, Phys. Chem. Chem.
Phys. 2008, 10, 2894 – 2909.
[9] H. Gꢂrner, Phys. Chem. Chem. Phys. 2001, 3, 416 – 423.
[10] Y. Futami, M. L. S. Chin, S. Kudoh, M. Takayanagi, M. Nakata,
Chem. Phys. Lett. 2003, 370, 460 – 468.
[11] G. Cottone, R. Noto, G. La Manna, Chem. Phys. Lett. 2004, 388,
218 – 222.
[12] Y. Sheng, J. Leszczynski, A. A. Garcia, R. Rosario, D. Gust, J.
Springer, J. Phys. Chem. B 2004, 108, 16233 – 16243.
[13] N. P. Ernsting, T. Arthen-Engeland, J. Phys. Chem. 1991, 95,
5502 – 5509.
[14] J. Z. Zhang, B. J. Schwartz, J. C. King, C. B. Harris, J. Am. Chem.
Soc. 1992, 114, 10921 – 10927.
[15] M. V. Alfimov, A. V. Balakin, S. P. Gromov, Y. V. Zaushitsyn,
O. A. Fedorova, N. I. Koroteev, A. V. Pakulev, A. Y. Resnyan-
skii, A. P. Shkurinov, Russ. J. Phys. Chem. 1999, 73, 1685 – 1694.
[16] A.-K. Holm, M. Rini, E. T. J. Nibbering, H. Fidder, Chem. Phys.
Lett. 2003, 376, 214 – 219.
[17] M. Rini, A.-K. Holm, E. T. J. Nibbering, H. Fidder, J. Am. Chem.
Soc. 2003, 125, 3028 – 3034.
[18] H. Fidder, M. Rini, E. T. J. Nibbering, J. Am. Chem. Soc. 2004,
126, 3789 – 3794.
[28] L. Poisson, K. D. Raffael, B. Soep, J.-M. Mestdagh, G. Buntinx, J.
Am. Chem. Soc. 2006, 128, 3169 – 3178.
[29] R. J. Mawhorter, M. Fink, B. T. Archer, J. Chem. Phys. 1983, 79,
170 – 174.
[30] S. T. Park, A. Gahlmann, Y. G. He, J. S. Feenstra, A. H. Zewail,
Angew. Chem. 2008, 120, 9638 – 9641; Angew. Chem. Int. Ed.
2008, 47, 9496 – 9499.
[31] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, V. G. Zakzrewski, J. A. Montegomery,
R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Milliam, A. D.
Daniels, K. N. Kudin, B. Mennuci, C. Pomelli, C. Adamo, S.
Clifford, J. Ochtersi, G. A. Patterson, P. Y. Ayala, Q. Cui, K.
Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B.
Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Kieth, M. A. Al-Laham, C. Y. Peng, A. Nanayak-
kara, G. Gonzales, M. Challacombe, P. M. W. Gill, B. G. Johnson,
W. C. Wong, M. W. Wong, J. L. Andres, M. Head-Gordon, E. S.
Replogle, J. A. Pople, Gaussian Inc., Pittsburgh, PA, 1998.
[32] W. H. Press, S. A. Teukolsky, W. T. Vetterling, B. P. Flannery,
Numerical Recipe in Fortran: The Art of Scientific Computing,
2nd ed., Cambridge University Press, Cambridge, 1992.
[19] R. Heiligman-Rim, Y. Hirshberg, E. Fischer, J. Phys. Chem.
1962, 66, 2465 – 2470.
[20] T. Yoshida, A. Morinaka, N. Funakoshi, J. Chem. Soc. Chem.
Commun. 1986, 437 – 438.
Angew. Chem. Int. Ed. 2010, 49, 6524 –6527
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6527