| ECCC-3 Contents Page | THEOCHEM Home Page | Elsevier Chemistry Home Page |
Model spirooxazines were studied via ab initio quantum-chemical
calculations. An anomeric interaction between nitrogen lone pair and antibonding
orbital of the Cspiro-O bond has been considered by Hartree-Fock
SCF level of theory with 3-21G basis set. Extent of the
interaction has been shown to depend on
planarity of the nitrogen atom, being maximized with its planarity growth.
Electronic properties and geometry parameters of model spirooxazines were
considered by limited CI under 3-21G basis set. Excitation lengthens the Cspiro-O
bond and causes a significant polarization of total wave function, which
results in increasing of the system dipole moment and transfering the
electronic density from the O2 atom and the phenyl fragment onto the N13=C14
bond in the case of spirooxazine 2 and onto the NO2 group
in the case of nitro-spirooxazine 3. Elongation of the C-O bond under
excitation is due to the
interaction in
spirooxazines. CI calculation of the excited singlet state of a model chromene
4 doesn't show elongation of this bond.
Spiropyrans (SP) and spirooxazines (SO) are typical organic compounds
exhibiting photochromism[1]. Upon ultraviolet irradiation
the Cspiro-O bond is broken to form the X-isomer which is then
being isomerized into a merocyanine form (Scheme 1).

An electronic structure and geometry of SP and structural mechanism of the photochromism have been extensively studied by various authors (see review [2]). Recently photochromism of SO has been of the increased interest. Compared to SP, they show better fatique resistance [3] which provides a possibility of their practical application. Thus, detailed study of a mechanism of the bond cleavage and isomerization of the open form of SO is an important task.
Spectroscopic studies of spiropyrans show that the both first excited singlet and triplet can be involved in the conversion mechanism [4,5]. In SP with the NO2-containing benzopyran moiety the bond rupture reaction occurs mainly via the triplet state, while in the NO2-free SP the photoreaction occurs via the singlet state.
This conversion has been extensively studied by various spectroscopic techniques [6,7] and different quantum-chemical calculations [8,9]. A femtosecond transient absorption spectroscopy study [6] of the spirooxazine revealed that the merocyanine is formed only from the first excited singlet.
Another interesting feature of SP and SO is an anomeric interaction between
the nitrogen lone pair and an antibonding orbital of the Cspiro-O
bond. There are many evidences [2,10] that this
interaction facilitates the ring opening reaction. The extent of this
interaction depends on the planarity of
the nitrogen atom and is maximum when nitrogen lone pair is collinear to the Cspiro-O
bond. This interaction results in shortening of the N-C bond and the
corresponding elongation of the Cspiro-O bond. For example, a
length of the C-O bond in six-membered oxygen-containing heterocycles is equal
to 1.41-1.43A, while in 3,3-diphenylpyrano[5,6-f]quinoline it is 1.458A
[11], and it is essentially elongated to 1.460-1.496A in
SP [2,10]. Thus, our objective was, using quantum-chemical
calculations, to check whether this experimentally found elongation is due to
the
interaction, and to study the
role of this interaction in the geometry of the excited state.
All calculations were carried out non-empirically by GAMESS
[12] program with 3-21G basis set on DEC Alpha 3000/600
computer. To reveal anomeric interaction, constraint geometry optimization of
1 has been used, with a value of the dihedral angle W being changed
step by step, while all the other coordinates were allowed to vary subject to
energy minimization. The dihedral angles W were used for controlling the N3
atom planarity. Optimized geometries of 1 are summarized in Tables 1
and 2 with constraint parameter W1(C1-N3-C4-C5) and W2(H7-N3-C4-C5),
respectively. To check out planarity of N3 in each table we used a sum of the
N3 atom valence angles,
(N2)={C1-N3-C4 + C4-N3-H7 +
C1-N3-H7}, and a dihedral angle Wpyr(C4-N3-C6-C1), which determines
a degree of the pyran cycle bending.
The ground state (S0), the first (S1) and the second
(S2) excited singlet states of 2, 3 and 4
were computed under Configuration Interaction (CI) with 3-21G basis set and
frozen core approximation applied. Active space was constructed from 3 double
occupied MO's, two independent MO's with alpha and beta spin, and 3 virtual
molecular orbitals. Only the first order excitations were considered. Thus we
had 1764 configuration state functions (CSF) included in each CI calculation.
Because of the program limitations only partial geometry optimizations were
carried out, all the rest parameters were kept from full geometry optimizations
under single determinant Hartree-Fock SCF calculation with the same basis set.
We have studied the efficiency of the anomeric interaction between the nitrogen lone pair and the antibonding orbital of the Cspiro-O bond by varying the planarity of the N3 atom. We suppose that an increase of the nitrogen planarity causes a rise of its lone pair energy and a decrease of an energy gap between the n(N3) and anti-bonding(C1-O2) orbitals, making their interaction stronger. In case of the anomeric interaction this will be followed by shortening of C1-N3 and lengthening of C1-O2 bonds. Another behaviour of the length of these bonds depending the N3 planarity would disprove our supposition. To study the problem we used a model spirooxazine 1.


According to Wpyr and
(N2) values from
Tables 1 and 2, planarity of N3 increases from left to right. The C1-N3 and
C1-O2 bonds are shortened and elongated, respectively. These results confirm
that an increase of the N3 planarity leads to the increase of the anomeric
interaction.
| Parameter | W1=159.0° | W1=164.0° | W1=169.2°* | W1=174.0° |
|---|---|---|---|---|
| Wpyr(C4-N3-C6-C1),° | 155.1 | 158.5 | 162.5 | 166.8 |
(N3),° |
353.4 | 355.2 | 357.0 | 358.5 |
| C1-O2 | 1.4542 | 1.4554 | 1.4563 | 1.4566 |
| C1-N3 | 1.4457 | 1.4442 | 1.4430 | 1.4420 |
| Parameter | W2=14.0° | W2=8.7°* | W2=4.0° |
|---|---|---|---|
| Wpyr(C4-N3-C6-C1),° | 161.3 | 162.5 | 163.6 |
(N3),° |
354.0 | 357.0 | 358.3 |
| C1-O2 | 1.4560 | 1.4563 | 1.4566 |
| C1-N3 | 1.4453 | 1.4430 | 1.4414 |
It can be supposed that changes of the anomeric interaction we have
found are due rather to changes of the mutual orientation of the lone pair and
anti-bondind orbital of the C-O bond than to the planarity changes of the N3
atom. Actually, it is very difficult to determine theoretically orientation of
a normal vector of the nitrogen lone pair in the asymmetric environment to
check out this hypothesis. Nevertheless, we suppose this effect to be
insignificant. A simple example demonstrates how planarization raises an energy
of the lone pair. Table 3 shows results of semiempirical calculation of the NH3
molecule with C3V symmetry. In the planar structure of the NH3
molecule HOMO (nLP-N-H = 90°) is a pure nitrogen lone pair. The HOMO
energy has a linear dependence on the nitrogen atom planarity. This proves that
the N3 lone pair energy in 1 changes with changing of planarity of this
atom.
| nLP-N-H angle,°1 | Heat of formation, kcal/mol | Energy of HOMO, eV | S(N)2 | PZ(N)2 | S(H)2 |
|---|---|---|---|---|---|
| 120.0 | -1.02 | -11.439 | 0.42 | -0.83 | -0.21 |
| 115.0 | -5.91 | -10.911 | 0.38 | -0.87 | -0.19 |
| 109.93 | -7.29 | -10.418 | 0.32 | -0.91 | -0.16 |
| 105.0 | -6.51 | -10.041 | 0.25 | -0.94 | -0.12 |
| 100.0 | -4.91 | -9.769 | 0.17 | -0.97 | -0.08 |
| 95.0 | -3.56 | -9.615 | 0.09 | -0.99 | -0.04 |
| 90.0 | -3.05 | -9.566 | 0.00 | -1.00 | 0.00 |
1- nLP - normal vector of nitrogen lone pair;
2- atomic orbital
coefficients in HOMO;
3- equillibrium geometry.
Stucture of photochromic compounds has been extensively studied by X-ray diffraction experiments [2]. Less is known about the nature of their excited states.
In this section an influence of
interaction on the length of Cspiro-O bond in the excited state of
model spirooxazines
2 and 3 has been considered.
![]() |
![]() |
| 2 | 3 |
|---|---|
To establish the nature of the lowest excited singlet states and to reproduce their spins correctly CI was imposed. Results of these calculations are shown in Tables 4 and 5. To fit in our limited computer resources we had to use limited CI with 3-21G basis set. Thus the results presented are assumed to be qualitative.
Two first excited singlet states of 2 and 3 were computed. We assume that limited CI may not correctly reproduce a sequence of the excited states, which doesn't allow us to discuss a difference between the first and second excited states. So we are focusing on determining a photochemically active excited singlet state (PAES) that leads to the opened merocyanine form. To determine it we relied on assumption that excitation should weaken the Cspiro-O bond which will be broken later in the way of transforming the molecule into a merocyanine form. The states in the Tables 4 and 5, marked by asterisk, confirm this criterion. Nevertheless we list both excited singlet states of 2 and 3 in the Tables 4 and 5, respectively, to let a reader make his own choice.
| CSF(S0) | Coeff.(S0) | Popul.(S0) | CSF(S1) | Coeff.(S1) | Popul.(S1) | CSF(S2*) | Coeff.(S2*) | Popul.(S2*) | |
|---|---|---|---|---|---|---|---|---|---|
| 15 | 0.9694 | 22220000 | 12 | 0.7651 | 12221000 | 5 | 0.8859 | 22121000 | |
| 365 | -0.1076 | 22200020 | 14 | 0.4136 | 22211000 | 9 | -0.1889 | 21221000 | |
| 1310 | -0.0845 | 20220200 | 53 | 0.3600 | 12220001 | 522 | -0.2230 | 21220100 |
| Total Energy, hartree | S0 = -641.56420991 | S1 = -641.34348432 | S2* = -641.32799545 |
|---|---|---|---|
| Dipole moment | |||
| X | -1.0405 | -2.0806 | -0.8136 |
| Y | -0.9557 | -1.5909 | -0.6087 |
| Z | -0.4517 | 0.0654 | -3.4873 |
| Total | 1.4833 | 2.6200 | 3.6323 |
*expected photochemically active excited singlet state
| CSF(S0) | Coeff.(S0) | Popul.(S0) | CSF(S1) | Coeff.(S1) | Popul.(S1) | CSF(S2*) | Coeff.(S2*) | Popul.(S2*) |
|---|---|---|---|---|---|---|---|---|
| 15 | 0.9735 | 22220000 | 5 | 0.8976 | 22121000 | 9 | 0.6251 | 21221000 |
| 105 | -0.1253 | 22200002 | 494 | -0.1521 | 22021100 | 125 | 0.1667 | 22120010 |
| 1310 | -0.0710 | 20220200 | 522 | -0.2690 | 21220100 | 510 | 0.6121 | 22120100 |
| Total Energy, hartree | S0 = -805.03486598 | S1* = -804.80993183 | S2 = -804.80125624 |
|---|---|---|---|
| Dipole moment | |||
| X | -6.4958 | -13.4786 | -9.1303 |
| Y | -2.8244 | -5.4002 | -3.6994 |
| Z | 1.7427 | 5.8062 | 2.2941 |
| Total | 7.2945 | 15.6380 | 10.1149 |
*expected photochemically active excited singlet state
From the total 1764 considered CSF's only limited number of the configuration states have large weighting factors. Tables 4 and 5 list only first three CSF's with the greatest coefficients. It is well known from spectroscopic experiments [14] that in spiropyrans electronic density transfers from an oxygen atom onto vacant molecular orbitals under photoexcitation and formation of PAES. Assuming close similarity in behaviour of spiropyrans and spirooxazines we should examine those MO's of ground states of 2 and 3 where the oxygen atom has significant atomic orbital coefficients. From our calculations HOMO doesn't have considerable coefficients on the oxygen atom, all such orbitals lie lower, they are HOMO-1, HOMO-3, for example, and so on. This explaines why HOMO doesn't play a significant role in the formation of the PAES and testifies for a multiconfigurational nature of the PAES, because excitation proceeds from deep MO's of 2 and 3. Population of those MO's included in CI are given in the Tables 4 and 5; they show that HOMO-1 (CSF 5) plays the most important role in formation of the both S2 of 2 and S1 of 3.
Generally, single configuration excited singlet states is constructed only from HOMO-LUMO excitations, but as follows from our calculations the studied systems (2 and 3) and probably spirooxazines at all are exclusions from this very simple rule. In other words one can state an important role of electronic correlation for correct prediction of electronic properties and geometries of spirooxazines.
| Bond | S0(2) | S1(2) | S2(2) | S0(3) | S1(3) | S2(3) | |
|---|---|---|---|---|---|---|---|
| C1-O2 | 1.4530 | 1.4463 | 1.4603 | 1.4629 | 1.4660 | 1.4628 | |
| C1-N3 | 1.4519 | 1.4514 | 1.4556 | 1.4435 | 1.4413 | 1.4432 | |
| C1-C14 | 1.5075 | 1.4960 | 1.4894 | 1.5075 | 1.5070 | 1.5000 | |
| O2-C7 | 1.3746 | 1.3979 | 1.3604 | 1.3595 | 1.3565 | 1.3544 | |
| N3-C4 | 1.4131 | 1.4197 | 1.4167 | 1.4152 | 1.4180 | 1.4176 | |
| C4=C5 | 1.3228 | 1.3249 | 1.3241 | 1.3231 | 1.3241 | 1.3242 | |
| C5-C6 | 1.5217 | 1.5214 | 1.5196 | 1.5226 | 1.5240 | 1.5240 | |
| C7-C12 | 1.3834 | 1.3995 | 1.4192 | 1.3870 | 1.4186 | 1.4001 | |
| C9-C10 | 1.3886 | 1.4047 | 1.4244 | 1.3832 | 1.4148 | 1.3964 | |
| C12-N13 | 1.4189 | 1.4007 | 1.3800 | 1.4153 | 1.3914 | 1.3962 | |
| N13=C14 | 1.2608 | 1.2814 | 1.2889 | 1.2607 | 1.2795 | 1.2743 |
Optimized geometries of 2 and 3 are given in Table
6. Excitation results in significant geometry changes. As will be shown later
this is due to essential polarisation of the total wave function and charge
redistribution. The most important geometry changes are related to the C1, O2,
N13, C14 atoms and phenyl fragment.
The C1-O2 bond is shortened in S1 and elongated in S2 state of 2 (Table 6). We suppose the S2 state is the PAES for 2 where cleavage of the C1-O2 bond takes place. We assume that the more precise CI can reorder the states and bring this S2 to be the first excited singlet state for 2.
It is interesting to note that excitation lenghtens N13=C14 bond both in 2 and 3, populating its vacant antibonding orbital. Later we will show that electronic density is transfering on this bond in the excited state. Elongation of the C1-O2 and N13=C14 bonds and shortening of the C1-N3 one take place in S1 state of 3 and here this effect is more evident than that of S2.
Significant elongation of the C7-C12 and C9-C10 bonds in S1 of 3 and S2 of 2 shows an important role of the phenyl fragment in polarisation of total wave function under excitation. The maximum elongation of these bonds takes place in S1 of 3 and S2 of 2 again, that confirms special nature of these states.
| Atoms | S0(2) | S1(2) | S2(2) | S0(3) | S1(3) | S2(3) |
|---|---|---|---|---|---|---|
| O2 | -0.75 | -0.69 | -0.65 | -0.75 | -0.68 | -0.66 |
| N13 | -0.65 | -0.52 | -0.76 | -0.65 | -0.58 | -0.66 |
| C14 | 0.18 | -0.14 | -0.01 | 0.18 | 0.20 | 0.09 |
| C1 | 0.42 | 0.48 | 0.42 | 0.40 | 0.40 | 0.41 |
| N3 | -0.79 | -0.71 | -0.78 | -0.81 | -0.80 | -0.80 |
| C5 | -0.34 | -0.30 | -0.34 | -0.32 | -0.31 | -0.31 |
| N(NO2) | - | - | - | 0.17 | -0.02 | 0.09 |
| O(NO2) | - | - | - | -0.39 | -0.50 | -0.43 |
| C7 | 0.42 | 0.35 | 0.51 | 0.46 | 0.41 | 0.44 |
| C8 | -0.25 | -0.23 | -0.24 | -0.25 | -0.23 | -0.16 |
| C9 | -0.24 | -0.30 | -0.34 | -0.18 | -0.13 | -0.31 |
| C10 | -0.26 | -0.23 | -0.07 | 0.26 | 0.36 | 0.41 |
| C11 | -0.21 | -0.26 | -0.28 | -0.16 | -0.22 | -0.13 |
| C12 | 0.19 | 0.22 | 0.24 | 0.19 | 0.33 | 0.18 |
| O2* | 0 | 0.06 | 0.10 | 0 | 0.07 | 0.08 |
| N13=C14* | 0 | -0.19 | -0.29 | 0 | 0.09 | -0.09 |
| C1,N3,C5* | 0 | 0.19 | 0.01 | 0 | 0.02 | 0.03 |
| NO2* | - | - | - | 0 | -0.42 | -0.16 |
| Ph* | 0 | -0.10 | 0.17 | 0 | 0.21 | 0.12 |
Electronic densities and charge redistribution in S0, S1 and S2 of 2 and 3 are shown in Table 7. They demonstrate that electronic density leaves the O2 atom in S2(2) and S1(3). The N13=C14 bond is a single acceptor of electronic density in S2(2). All the density transfers onto NO2 group in S1(3), since this group is a very strong acceptor. The N3 atom doesn't change its electronic population significantly in both molecules. As a non trivial result we have found that phenyl group looses the density under excitation in greater extent than O2 atom in S2(2) and S1(3). It is possible because phenyl fragment has six carbon atoms versus one oxygen atom.
From the standpoint of polarization it is interesting to consider the value of dipole moments in excited states (Tables 1 and 2). As one can see excitation increases a value of total dipole moment of the molecules. The maximum dipole moment is for S2(2) and S1(3) which we assume to be the PAES. The value of dipole moment of S1(3) is probably overestimated, although it should be greater than that of S2(2).
As it was above mentioned the length of C-O bond in spiropyrans and spirooxazines is elongated in their ground state due to anomeric interaction. A model chromene 4 doesn't have such interaction.

In this context it would be interesting to compare the length of the C-O bond in the excited state of 2, 3 and 4. Results of CI calculation of 4 are summarized in Tables 8. Two first excited states of 4 were computed.
| CSF(S0) | Coeff.(S0) | Popul.(S0) | CSF(S1) | Coeff.(S1) | Popul.(S1) | CSF(S2) | Coeff.(S2*) | Popul.(S2) | |
|---|---|---|---|---|---|---|---|---|---|
| 15 | 0.966 | 22220000 | 9 | 0.863 | 21221000 | 5 | -0.362 | 22121000 | |
| 140 | 0.073 | 12211010 | 137 | 0.395 | 21220010 | 14 | 0.784 | 22211000 | |
| 1284 | -0.085 | 22020200 | 522 | 0.122 | 21220100 | 530 | -0.270 | 22210100 |
| Total Energy, hartree | S0 = -433.91098660 | S1 = -433.69110068 | S2 = -433.68942629 |
|---|---|---|---|
| Dipole moment | |||
| X | -0.451 | -807 | 1.842 |
| Y | 0.875 | 0.477 | 1.917 |
| Z | 0.000 | 0.000 | 0.000 |
| Total | 0.984 | 0.938 | 2.659 |
| Bond length | |||
| C1-O2 | 1.4469 | 1.4412 | 1.4461 |
| C=N | 1.2619 | 1.2810 | 1.2883 |
| Charge transfer by fragment | |||
| O2 | 0 | 0.084 | 0.029 |
| C=N | 0 | -0.075 | -0.230 |
| Ph | 0 | -0.061 | 0.087 |
The length of C1-O2 bond in ground state of 2, 3 and 4
is 1.4530, 1.4629 and 1.4469, respectively (Tables 6,8). Thus the role of
anomeric interaction in elongation of the C-O bond in the ground state of 2
and 3 is undoubted. In the excited states S2(2) and
S1(3) this bond is elongated, compared to the ground state.
Shortening of the C-O bond have been found in both excited states S1
and S2 of 4 in contrast to spirooxazines, which confirms an
assumption about important role of
interaction for the value of the C-O bond in an excited state.
Model spirooxazines were studied via ab initio quantum-chemical calculations. These results demonstrate that anomeric interaction between nitrogen lone pair and antibonding orbital of Cspiro-O bond plays important role in elongation of the C-O bond both in the ground state and the excited one. It is very important to use CI for correct reproducing of an excited singlet state of spirooxazines because of its multiconfigurational nature. Excitation results in increasing of a total dipole moment of the system. Upon excitation electronic density transfers from O7 atom and phenil fragment onto N13=C14 bond in the case of spirooxazine 2 and onto NO2 group in the case of nitro-spirooxazine 3.
The authors are grateful to Russian Foundation for Basic Research, grant number 96-03-32046, and to International Science Foundation, grant number IFN000 for support of these investigations.