Tautomerism by Hydrogen Transfer in Anils, Aci-Nitro and Related Compounds

685

Chapter 17

I

Tautomerism by Hydrogen Transfer in Anils, Aci-Nitro and Related Compounds E. Hadjoudis

ANIL TAUTOMERISM

1.1 Anil s o f SalicylaldehYdeS The reversible solid state photocoloration of anils of salicylaldehydes (1) was f i r s t observed by Senier and co-workers (refs. 1,2),who noted that of the ring-substituted anils, only a few were photochromic and that polymorphic modifications of the same anil were not necessarily all photochromic. These two observations revealed the topochemical effect on photochromism, but a firm state5 6 4 ~ C

H /

6 5 pR2

2

3

ment at that time was not made because of i l l defined experiments, especially with properly ring-substituted derivatives and variations of temperature. Cohen and co-workers (refs. 3-5) undertook a more systematic study of crystalline anils of salicylaldehydes and confirmed that many anils are dimorphic and that the two forms occasionally differ in color, yellow and orange-red. They noted also, in the photochromic anils, the existence of two temperature limits, the variation of size of the temperature interval (the "working range") and the importance of the ortho-OH group (refs. 3,4). I t was further found that "structural mimicry" (refs. 6,7) was operating, an effect that constitutes a particularly striking example of topochemical control. For example, the stable crystal form of N-salicylidene-4-chloroaniline is thermochromic, whereas the stable form of the corresponding bromo-derivative is photochromic, thus showing the absence of an apparent correlation of chemical properties with the electronic characteristics of the substituents (ref. 8). Cohen and co-workers classified the crystals of these compounds into two types on the basis of their spectroscopic properties (see Table 1). I t is to be stressed that this classification refers to the various compounds in given crystal structures. Thus, both compounds given as examples in Table 1 are dimorphic, with metastable and stable forms of different types; the classification given in the Table refers to the modifications which are the stable ones at room temperature and above.

686

TABLE 1 Classification of Crystalline N-Salicylideneanilines Type ~

Type B

Molecular structure

non-planar

planar

Effect of UV-light

reversible coloration~

no coloration~

no fluorescence

fluorescence

Effect of heat

no coloration

reversible coloration

Name

photochromic

thermochromic

ExampI e

RI=H ~ R2=2-cI

RI=H~ R2--4-Cl

In addition to the above "spectroscopic" approach, the authors carried out structural studies on a number of these compounds (refs. 9-12). From this information i t was concluded that there is a genelral distinction between the structures of crystals of types ~ and B: in the thermochromic crystals the molecules are planar and pack face-to-face with short intermolecular contacts, of the order of 3.3 ~, normal to the molecular planes~ in the photochromic crystals, the salicylaldimino part of the molecule is planar, but the aniline-ring lies 40 to 500 out of this plane, and the resulting structure is relatively open with no close face-to-face contacts between molecules. Further, in order to interpret the phenomena of photochromism and thermochromism, Cohen and co-workers proposed an intramolecular H-transfer mechanism as follows: there is a temperature-sensitive equilibrium in the crystal between the two tautomers (2) of the molecule, the one with the chelating hydrogen covalently bonded to the oxygen (the "OH-form") and the NH-form, with the hydrogen bonded to the nitrogen. The cis-keto NH-form absorbs at longer wavelengths~ raising the temperature increases the population of this form and thus causes a deepening of color.

H

enol ("OH-form"}

cis-keto ("NH-form")

trans-keto ("NH-form")

The intramolecular H-transfer can occur in either the ground or the excitedelectronic state. High energy is required for H-transfer in the ground electronic state of molecules in photochromic crystals because of their twisted con-

687

formation and., as a result, no absorption attributable to the NH-form is observed. H-Transfer can occur, however, in the excited electronic state, and the crystal structure is sufficiently open to permit a subsequent geometric isomerization leading to the colored trans-keto NH-form which is stabilized as a result of the rupture of the

intramolecular hydrogen bond. Thus in the c r y s t a l l i -

ne N-salicylideneanilines, photochromism and thermochromism are mutually exclusive properties (ref. 3). Fig. 1 shows the spectra of thin polycrystalline films of photochromic 2-chloro-N-sal i cyl ideneani Iine and thermochromic 5-chloro-N-sal icyl ideneani line under the conditions described. The spectra of the photochromic species are very

//',

,-,

o.so I-'II

I \

|~I ,'

~ o.,oi/i,

|\

(~1

",

-

'i

e I

".

,

i'. I 9 I.

-

-

".. 9

"k 500 A[nrnJ

600

400

500 k[nm]

!

"',

1:

'i

I

:,

I:

!1 "l

400

(b)

! . "

"

600

Fig. 1. (a) The absorption spectrum of a crystalline film of 2-chloro-N-salicylideneanilijqe before ( f u l l curve) and after (broken curve) irradiation. (temperature -131"; irradiation 20 min through Corning f i l t e r F 5874, 250-watt high pressure mercury arc.). (b) The absorption and fluorescence spectra of a crystalline film of the strongly therm~chromic 5-chloro-N-salicylideneaniline. 1 ( f u l l curve), absorption at-1~3 ; 2 (broken curve), absorption a t - 4 9 " ; 3 (dotted), fluorescence at-153 v on irradiation with 365 nm light. From Cohen and Schmidt, J. Phys. Chem. 66, 2442 (1962).

similar to thoseof the thermochromic species but show additional long wavelength absorption in the range 540-580 nm. This difference is thought to be due to the fact that in the case of the trans-keto form there is an additional n -~ n transition involving the lone pair of electrons on oxygen, not now involved in hydrogen bonding.

688

The fading rate of photocoloration in the solid showed a unimolecular decay process having an activation energy of the order of 25 kcal/mol. The decay rate was unaffected by

deuteration of the OH group~ indicating that the rate-de-

termining step in the fading reaction most l i k e l y corresponds to the trans-keto § cis-keto conversion (2) (ref. 12). The photochromic band appears also in rigid glasses (ref. 4); i t is very stable in this state (-150 to -175~ (-75~

but fades slowly in paraffin oil

I t was in this last case (paraffin o i l ) that erasure of the color occur-

red upon irradiation with visible light (ref. 4). Approximately the same maxima are obtained in the dark with polar solvents at room temperature, and presumably, similar quinoid structures are formed by H-transfer tautomerism as a result of both photochemical and solvent effects (ref. 4). As shown in Fig. l ( b ) , the fluorescence spectrum, as is the case for all the crystalline thermochromic anils, is appreciably Stokes shifted with respect to the absorption band of the anil species (ref. 5). This Stokes shift was interpreted in terms of proton transfer in the excited state (3) (ref. 13). eno hv

l

enol ~

~ ci s-keto w

-1

fluorescence

(3)

cis-keto

"OH-form"

"NH-form"

The isomerization process postulated in (2) was supported by flash photolysis experiments on N-salicylideneanilines in solution by Wettermark and co-workers (ref. 14), who demonstrated a photochemical trans-cis isomerization with an activation energy of 15 kcal/mol. They also found (ref. 15) a species with a lifetime in the millisecond region and an absorption maximum at 470 nm in r i gid glasses and polar solvents, corresponding to H-transfer photochromism as in the crystalline state, where also absorption maxima were observed in the same region. H-Transfer phototautomerism to the cis- and trans-quinoid structures

was

further supported by parallel studies on N-salicylideneanilines carried out by Becker and Richey (ref. 16), and Ottolenghi and McClure (ref. 17). These studies clarified many aspects of the operating mechanism. Low-temperature studies indicated that the colored keto-form

(2)

is stabilized by a ci s § trans iso-

merization immediately following the H-transfer reaction. Thus following light absorption a ~ - enol + nn~ -enol transition in the excited singlet state appears, probably,enhanced by the intramolecular H-bond of the enol form

689

which, by H-transfer forms the cis-keto ~ excited state, in competition with intersystem crossing to the 3nn -~enol t r i p l e t state. The cis__.__-keto~ converts f i n a l l y to the colored trans-keto form. The reaction is reversible, thus ground state isomerization converts the trans-keto form to the cis-keto form and then . . , . . , . . .

rapidly back to the starting enol form. The previously observed (ref. 3,16) yellow-green emission, which was assumed to be displaced fluorescence by intramolecular H-transfer in the excited singlet state, was shown to be the above t r i p l e t state (3nn) phosphorescence because of its h a l f - l i f e of about 75 ~s at 770K (ref. 17). Becker and Richey (ref. 16) provided spectral evidence for the cis-keto species in the photochromic reaction. They also showed that ( i ) the c is-keto spectrum obtained by warming the photolysis product at 77~ is essentially the same as that formed in the dark with polar solvents, and ( i i ) t h a t

photolysis of

the cis--keto form produces neither the trans-keto nor the enol structures, indicating that the excited states obtained by direc~c excitation (nn~) of the cis-keto form do not channel down to the same excited level produced by H-tran-

, . , . . , = _

sfer of the (nn~) enol. The tautomerism in the dark caused by polar solvents was also observed by Margerum and Sousa (ref. 18) and studied spectroscopicaly by Ledbetter (ref.19), while Dudek and Dudek (ref. 20), using proton magnetic resonance, measured the extent of the equilibrium and calculated an extinction coefficient in methanol for the quinoid intermediate, probably the cis___-keto form~of 1.46 x 104 at 434 nm. Earlier, Burr, Llewellyn, and Lothian (ref. 21) estimated the quantum yields of photocoloration for N-salicylideneaniline and N-salicylidene-m-tolui dine as 0.9 and 8, respectively. As pointed out later by Margerum and M i l l e r (ref. 22), these values are clearly two orders of magnitude too high because of the incorrect assumption that N-salicylideneaniline in ethanol is completely converted to the quinoid form. Andes and Manikowski (ref. 23) studied the photochromic characteristics of crystalline N-salicylideneaniline from the point of view of data storage applications and pointed out that several statements (refs. 24, 25) regarding rapid fatigue may be misleading, showing that photochromic side reactions do not occur in most cases except those produced by the irradiation of impure materials. Thus the s-form of crystalline N-salicylideneaniline showed excellent fatigue resistance when cycled between the yellow and red states by exposure to u l t r a v i o l e t and visible l i g h t for up to about 50,000 cycles,

therefore j u s t i f y i n g fur-

ther research for practical applications depending upon r e v e r s i b i l i t y . Concerning the isomerization during the course of the photoreaction, Rosenfeld, Ottolenghi and Meyer (ref. 26) noted that i t is legitimate to talk about a simple ci____ss§ trans isomerization only insofar as the C7 - N bond

(4)

is

690 concerned. The photochromic effect involves, in addition to proton transfer, ~

_, hv or

A

(4) ~ O " ' H

simultaneous rotation around both the C1- C7 and C7- N bonds. The structure, however, of the colored form of N-salicylideneanilines, to which the above quinoid structure was assigned, was considered by Russian workers (ref. 27) as a serious obstacle to the resolution of the nature of the photochemistry of anils. Nurmukhametov and co-workers (ref. 28) concluded, from an analysis of their spectral data, that the colored form of N-salicylideneaniline assumes a zwitterionic structure as a result of proton migration (process 2 in eq. 5), which proceeds in the lowest excited singlet state:

c\"-O, o_. . . . .

o-

M

c'.-O .§

M'"

M"

(5) I--

h~abs

i

.

M(SI ) L ~2 ..... $ !

M(So) I

-

MiSl)--~

.......

4

,

MiSo) ~

11

11

M (So)

h~fl [ uor 3

The zwitterionic molecule, however, was suggested to be unstable in the ground state and to rearrange to the original form (process 4). The fraction of I

the molecules in the S1 state before the emission of fluorescence undergo synanti isomerization (process 5) similar to the process discussed by Grellmann and Tauer (ref. 29). Reversion of the anti isomer from the zwitterionic structure to the original form (So), in the solid state, is impeded. Work at low temperatures, however, indicated the existence of various forms of N-salicylideneaniline with broken hydrogen bonds at the photochemical equilibrium (ref. 27). Ledbetter (ref. 30) also reached the conclusion that zwitterionic structures

do

exist and that they are determined by the solvent. Thus in

KBr, equilibria 1,2 and 3 are involved

(6)

9 In protonic solvents which are

not too acidic and in aprotic solvents having a high dielectric constant, equi-

691

librium 3 is solely involved. In these solvents all the infrared frequencies found in KBr are present except the v = ~H There are apparently no stabilic

<~

C/H

~C/H

~

H _~

(6) OH

~ H

+

~

OH / - - ~._...~

5.

C \+

zing forces as there must be in KBr to maintain the zwitterionic form. In strongly acidic solvents, equilibria involving the protonated species do occur, and the ~c = ~H band, in addition to the other bands, is observed in the infrared spectrum. Also, the ultraviolet spectrum indicates the existence of protonated species. Equilibria 3-6, with 4 predominating, must therefore be involved. The extent to which any species may be present in any equilibrium also depends on the substituents in the ring, as was also indicated in studies by Csaszar and co-workers (ref. 31). In connection with such equilibria Seliskar (ref. 32) noted that the major effect of the solvent medium on the neutral/dipolar ion-molecule equilibrium appears to be quite specific to proton-donating alcohol solvents. Nakagaki and co-workers (ref. 33)~however, using time-resolved spectroscopic techniques in the millisecond to picosecond range and Fourier transform IR spectroscopy, showed that the photochromic colored species has the keto-amine form. Their picosecond kinetic analysis led them to suggest the existence of an intermediate in the transformation of the cis___-keto amine to the photochromic species and therefore to propose the following mechanism (Fig. 2) for the photochromic phenomenon of N-salicylideneanilines. The excited singlet state E~ of the enol imine produced by photoexcitation of E results in formation of the photochromic colored species in the ground state, P, through the H-transfer and molecular rearrangement (cis-trans isomerization). The intermediate X* is an excited singlet state from which the photochromic colored species and the fluo-

692 4k

rescent state of the cis-keto amine, Kf, originate (ref. 26). X e ~

~.s"%=.j'='~.~

.. ,

9

hva

E

enol

ci__.~s-keto

trans-keto

Fig. 2. Schematic explanation of the photochromic species formation. From Nakagaki et. al. Bull. Chem. Soc. Japan 5__O, 1909 (1977).

The fact that the photochromic phenomenon takes place when energy is available for further production of the trans-quinoid isomer is supported by the work of Laverty and Gardlund (ref. 34), who synthesized hydroxylated polyazomethines (7) in which the necessary isomerization step to photoproduct is inhibited, and as a result, these polymers are thermochromic. Therefore photo-

k> <>

N

CH=

H

N~C H

H2

CHz

(7)

OH

chromic polymers based on Schiff bases require structures that permit the needed movements for the production of the trans-keto colored form, as for instance the polypeptides containing photoisomerizable aza-aromatic chromophores studied f i r s t by Goodman and co-workers (ref. 35) and later by other investigators (ref. 36). For instance a polymer like that shown in 8 should be photochromic or thermochromic. Such a compound, to our knowledge, has not yet been prepared. fCO--CH-NH- ~ CH2

-CO-~H-NHI

O'<'C'o- CH:,. ~ N H~O

I t should be prepared and tested.

(8)

693

1.2 Heterocyclic Anils Hadjoudis, Moustakali-Mavridis and co-workers (ref. 37) extended the structural studies in three analogous series of heterocyclic anils (9). In the case of 5 4 R1

6

5 4 R

N 3

--"

5

6 N

az

4

t 6 5 Rz

4

saI icy I i dene-2aminopyri di nes

6

R1

Rz 5

saI icy I i dene-3ami nopyri dines

6

saI icy I i dene-4aminopyri di nes

(9)

salicylidene-2-aminopyridines, all the crystalline compounds examined were found to be thermochromic, i.e. photochromic properties as in the case of salicylideneanilines were not observed. Fig. 3 shows the thermochromic behavior of six compounds of this class. Fig. 4a shows the spectra of one of these compounds at various temperatures, and Fig. 4b gives the variation

with temperature of

the optical density of the maximum of the thermochromic band near 500nm, where there is almost no overlap of the bands of the two species. The energy difference between the colored and non colored form is found to be 2.17kcal/mole. All the thermochromic crystals, on irradiation with 365nm l i g h t , are strongly lumi-

Br).~

~o.sFk

N

I

IMe ."

I',

I

\

il\

\

',ix

K \ l 0"01",

.

300

,

400

.

,

500

',

", \

~

',,

,

300

400

500

300

400

500

~(nm) Fig. 3. The absorption spectra of crystalline films of the indicated compounds at room temperature (broken curves) and at liquid nitrogen temperature ( f u l l curves). From Hadjoudis et al. Isr. j . Chem., 18, 202 (1979). nescent, with the short wavelength cut-off at about 490-500nm. The fluorescence of these thermochromic compounds is, as in the case of N-salicylideneanilines (ref. 5), roughly a mirror image of the absorption of the ci___s-quinoid isomers. This generality was explained on the basis of their crystal and molecular structure. Thus the molecular packing of four compounds for which the crystal

694

structures were solved (ref. 38), is characteristic of that of planar molecules arranged in stacks along the shortest crystal axis with mean interplanar distance of 3.5 ~. The planarity is achieved because of the hetero-nitrogen of the

0.3

(a) ~'~~ 1.5 t~

~

(b)

0.2

2.0 Lll

80~

O.1

.17

<

l~\~Tt

- 25oc

~ \ ~

-2ooc

..o 1.0 ~ v - ~ ~ n

< 0.5 0.0 400

i

AI" ~,. ~

20oc

9 _o --0.1 .

".----~'~,~--=~ -o.31 500

Kcal/mole

oo

~

600 A,nm

9

,

2

,

m

3 l O-S/T

Fig. 4.(a) The effect of temperature on the absorption spectrum of thermochromic c r y s t a l l i n e 5-bromosalicylidene-2-aminopyridine. (b) log A against I/T from the absorption spectra of 5-bromosalicylidene-2-aminopyridine. From Hadjoudis et al. Isr. J. Chem., i__88,202 (1979).

pyridine ring. In the case of N-salicylideneanilines, there is steric hindrance due to the short distance of~, 2~ between the ortho-hydrogen exocyclic hydrogen

H7

H9

and the

when the molecule is planar (ref. 9). This repulsion

is relieved in the case of N-salicylidene-2- aminopyridines because the heteronitrogen atom is always at the cis position with respect to the

H7

hydrogen

atom (Fig. 5). The distance of about 2.5 ~ between these atoms corresponds to normal van der Waals contact (ref. 38). H7 2.5

9

H9

Fig. 5. Distances (X) for N-salicylidene-2-aminopyridine.

The thermochromic phenomenon was interpreted as due to a s h i f t in the tautomeric equilibrium (I0) as in the case of N-salicylideneanilines in which such

(10)

a tautomerization is in agreement with infrared studies (ref. 30).

695

The group of N-salicylidene-2-aminopyridines represents a good example of the concept of "crystal engineering" (refs. 39-41) according to which we can design molecules so as to guide their choice of crystal structure with desired chemical and physical properties. Thus, insertion of a nitrogen atom in the 2-position of the aniline ring of any anil molecule which is normally non-planar (photochromic crystal) yields a planar molecule (thermochromic crystal). Compare for instance: N-salicylideneani line (non-planar, photochromic) against N-sal icylidene-2-aminopyridine (planar, thermochromic). N-Salicylidene-3-aminopyridines are weakly thermochromic (ref. 39) in the sol i d state. The crystal structure analysis for the parent compound and for 5-methoxysalicylidene-3-aminopyridine shows (ref. 42) a rotation of the pyridine plane by 14.80. This deviation from planarity may be related to the weak thermochromic behavior (ref. 43). Among N-sal i cyl i dene-4-ami nopyri dines, photochromic and thermochromic compounds have been found and therefore non-planar and planar structures are expected (ref. 39). All the members of the heterocyclic anils (9) examined by Hadjoudis and coworkers have been found~to be photochromic in rigid glasses at spectroscopic dilution and the application of flash techniques has permitted the analysis of similar but transient phenomena in solution (ref. 37). Thus the transient absorption spectrum of N-salicylidene-2-aminopyridine shows a spectrum similar to that of the photoproduct in rigid glasses. The kinetic and spectral considerations of this compound indicated a quinoid photoproduct having an activation energy of 2.6 kcal/mole for the dark back reaction. Thus again, as in N-salicylideneanilines, when the factor of c r y s t a l l i n i t y is lost, as in rigid glasses and solutions, and the orientation of the molecules is random, all the molecules of these three classes appear to be photochromic. 1.3 Picosecond Flash Phgtol~sis The picosecond kinetics of intramolecular H-transfer in the lowest 1n-n* state of N-salicylideneaniline was investigated by Barbara and co-workers (ref. 44), who showed that the enol form of this molecule tautomerizes to a cis-keto form with a rate constant of 2 x 1011 s-1 following Franck-Condon excitation at

H ~C,,~ N "~ Ph

H.

N~.~'Ph--I'~

c<-

.+l

I

~Phl"

IH'c"N'H I

:"I "OTONI

~

I

(11)

1

355 nm (11). They observed that the quinoid fluorescence at room temperature has a < 5ps rise time which is consistent with the above rate for tautomeric proton

696

transfer. At low temperatures, the fluorescence was found to have two components" a short-lived component which is formed within 5 ps of excitation, and a long-lived component which has the short-lived fluorescence state as a precursor. This behavior was interpreted as a very rapid proton transfer process occurring at all temperatures. The short-lived component of the fluorescence, blue shifted from the long-lived one, was tentatively assigned to vibrationally excited fluorescence. Lewis and Sandorfy (ref. 45), assuming that the photochemistry of N-benzylideneaniline (BA), in which isomerization results in an increase of the non-planarity of the aniline ring, and N-salicylideneaniline (SA) are closely related, proposed that the photoproduct of the latter has a similar structure. Thus like BA, the i n i t i a l enol form of SA exists in a configuration which is trans about

(12)

BA

SA

the C = N bond, and the aniline ring is somewhat twisted out of the molecular plane. I t should be noted that t_ran____ssand c_ i .__ssdescribe here the configuration about the C = N bond and not the relative positions of the hydrogen and oxygen atoms, as was the case in the basic proposals of Cohen and Schmidt (refs. 3-5). The photoproduct is thus determined to be a zwitterion, not an ortho-quinone, with a cis-configuration about the C = N bond, since hydrogen bonding between the C - 8 and ~ - H groups does not exist. The authors argue that the absorption of a proton causes an ultrafast proton transfer, as shown by Barbara and coworkers (ref. 44) above, which might be followed by a slower inversion at the nitrogen resulting in a c_ i __ssconfiguration about the C = N bond. To support this revised model, the authors offer a different interpretation of the infrared spectra presented earlier by Nekagaki and co-workers (ref. 33). In connection with the presence of protonated amine species, Ledbetter (ref.

cC]_|

(13)

o- .+ A

H#--<9 B

46), using resonance Raman spectra of pyridoxal 5'-phosphate and salicylaldehyde

697

Schiff bases of amino acids in water, showed the presence of the -C = NH+ bond (A). However, in less polar solvents, the tautomer of aryl Schiff bases exists as the quinoid resonance structure

(B). More recently, Lee and Kitaga-

wa (ref. 47), using the same technique, demonstrated protonated species in the case of N-salicylidenemethylamine in methanol (14). CH3

CH 3

I

I+

(14)

However, concerning N-salicylideneanilines, Grummt (ref. 48), studying V-shaped Hammett plots for the relaxation rates of the phototautomers of a number of Para-substituted compounds (15; R = MeO, Me, H, Cl, NO2, R1 = H; R = NO2, RI = OMe), interpreted them as a change in mechanism from O-protonation (donor R) to NH-deprotonation (acceptor R) as the rate-determining step; the observed acid and base catalysis strongly supports the ortho- quinoid structure of the phototautomer.

RIN--C--R

(15)

Higelin and Sixl (ref. 49) reexamined the reaction mechanism of the photochromism of N-salicylideneanilines in crystalline matrices, using dibenzyl and stilbene as host crystals, and rigid glasses. The authors, in contrast to older reports (refs. 3,12,21) which place the phenomenon of photochromism inside a range of temperatures, observed photochromism down to 1OK for the case of N-salicylideneaniline in a dibenzyl host crystal. Using, however, stilbene crystals as a host, they also observed a photochemical threshold at 180K. Their low-temperature emission spectra of the N-salicylideneaniline molecules

confirm older reports. Thus the large shift between absorption and emi-

ssion has been attributed to the H-transfer, OH---N O---HN,within the N-salicylideneaniline molecule, following photoexcitation of the enol configuration (refs.3,4,5s22), while the absence of a Stokes shift in the emission and absorption spectra of the photoproduct indicated the disruption of the original OH---N or O---HN hydrogen bond. This is in agreement with previous interpretations of the N-salicylideneaniline photoproduct structure, which has been assigned to the trans-keto configuration (refs. 3,4,5,18,20,33,45). Higelin and Sixl summarized in the energy-level scheme of Fig. 6 the essent i a l photochemical and thermal pathways of the forward and back reactions as

698

deduced from experiments on N-salicylideneaniline mixed crystals and from previous investigations in dilute solutions and rigid glasses. FORWARD PHOTOREACTION

~ E*

~

0A* QB ~

E(,(11 ~,,,,,,,,~QC*

J,

" rn (,3

o

E H

o

QB _

" E

QC

QA

_..17cH O~ H H ,,,,,to.e, O H ' N ~ '''ar~ C..,=C76-" QB

QA

QC B*

Eo.(2) ' ~ 2340 crn'l B

9400

OC* "E

",k QC

c r n -1

BACK REACTION Fig. 6. Energy-level scheme of the forward photoreaction and of the back reactions of N-salicylideneaniline in dibenzyl host crystals. From Higelin and Sixl, Chem. Phys. 7__77,391 (1983). Excitation of the enol (E) results in an extremely Stokes shifted emission of the type described by Weller (ref. 50), which is separated into a structured, high-energy tions QA~ §

A emission and a broad, low-energy B emission due to the transiand QB~ §

The fact that proton return from nitrogen to oxy-

gen in the final stable photocolored QC species is strongly hindered shows that an additional process, following immediately the proton transfer reaction which caused the Stokes shift, must be present to bring some stabilization or trapping of the NH configuration. Thus the hydrogen bond is cleaved by a distortion of the cis____-quinoid QB configuration about the C1 = C7 double bond in the excited

699

QA'statet and therefore the colored keto form is stabilized by a cis-trans isomerization process. QA ~is interpreted to be a distorted intermediate between QB and QC; I t may decay into QB and QC and is therefore suggested to be the precursor of both. That the final photoproduct has a trans-keto configuration in agreement with previous interpretations (refs. 3,4,5,17,20,33,43) is supported by the observed emission from the QC photoproduct, upon irradiation into its absorption bands, showing a complete disruption of the hydrogen bond. The luminescence lifetimes of the QA~and QB~states, 10 ps and 3 ns, respectively (ref. 33), are typical of fluorescence emission arising from x~excited singlet states. The conditions, however, under which the reaction proceeds directly or indirectly to the QC ground state remain unclarified, since the authors were unable to clearly separate the QC ~§ QC emission from the QB ~ § QB emission during the photoreaction after pulse excitation in the time-resolved spectra. The potential barrier B~in the photochemical pathway of the back reaction of salicylideneaniline in dibenzyl crystals is assumed to be identical to that of the forward reaction. On the basis of the determined barrier heights, a QB§ QC reaction is excluded in accordance with previous interpretations (refs. 17,33 ) 1.4 .Effect of Crystal Structure The problem of the effect of crystal structure on the photochromic properties of Schiff bases continues to be of interest. Thus Kawato and co-workers (ref. 51) prepared compounds with bulky substituents (see Table 2) and examined their photochromic behavior. The above workers found that tert-butyl substituents increase the stability of the photoproduct in cases where methylene X~ OH --N\~

TABLE 2 Anils with Bulky Substituents 1: X-H, Y-H

a: R-Ph

d'R= - ~

_~:X=H,Y=tert-butyl

b: R=CH2Ph

e-R:

3:X-tert-butyl, Y-tert-butyl

c:R-(CH2)2Ph

-~

groups are not present between the nitrogen atom and the aromatic ring (l_a,2_a, 3_a). The effect of the bulky substituent has been attributed to an increase in the open space for molecular movement in the crystal lattice. However compound 2~ is not photochromic, while compound 3_.eproduced a photoproduct 400 times more stable than that of unsubstituted la. In contrast, 3d is not photochromic, and this behavior was explained on the basis of X-ray crystal structures observed for N-salicylidene-2-aminopyridines by Moustakali-Mavridis and co-workers (ref.

38). The rate constants of the thermal fading reaction of photochromers for the Schiff bases derived from benzylamine (l_b,2_b,3_b) were larger than those of the

700

other derivatives studied. Molecular models show that the flexible benzyl group is preferable to the phenyl group for molecular movement because of rotation around the C = N bond. Similar behavior was found for lc, whose rate constants were larger than those of lb and la. Thus i t was suggested that the photochrom mic process involves a simple C = N bond rotation or a C = N bond rotation with a change in hybridization at the nitrogen atom from sp2 to sp3. Hadjoudis and co-workers (refs. 52,53) in a continuation of previous efforts (refs. 37,43) to correlate the crystal structure with photochromism and/or thermochromism of the crystalline Schiff bases, prepared a number of compounds among which they hoped to find molecules clearly displaying both photochromic and thermochromic properties, as opposed to the exclusive photochromic or thermochromic behavior studied so far. The compounds prepared (see Table 3 below) were derivatives of N-salicylidene-2-thenylamine (ref. 52) and N-salicylidene2-benzylamine (ref. 53), since i t appeared that such behavior might result from salicylidene derivatives in which the amine is aliphatic or the amino group is insulated from the ring (ref. 13). TABLE 3 N-Salicylidene-2-thenylamines and N-Salicylidene-2-benzylamines R,

5

R2

6

Rz

4

N-CH

R,

~ N_CH2~

OH

R1

R2

Property

R1

R2

Property

H

H

Photochromi c

H

H

"

H

5-Br

"

Photochromic

H

5-Br

H

5-OCH 3

"

H

5-OCH3

"

3-OCH 3

II

H

3-OCH3

II

3-Br

5-Br

3-Cl

5-Cl

H

4-OCH 3

Thermochromic " Photo/Thermo-

chromic

3-Br

5-Br

3-Cl

5-Cl

H

4-OCH3

Thermochromi c " Photo/thermochromic

Among the compounds of Table 3, in which the amino group is insulated from the ring by the -CH2-grouping, photochromic and thermochromic examples have been observed and also a clear case (in each group) of a compound displaying both phenomena. Fig. 7 shows the dual behavior of N-(4-OCH3-salicylidene)-2-thenylamine in the crystalline state. The latter molecule is not planar due to the methylene group inserted in the bridge (ref. 54). The salicylaldimino moiety of

701

0.75 MeO~=N

- CHz ' ~ FILM

0.5

\

0.25

\

\ \ 2

"

0.0-

....

I

350

.

.

.

.

.

.

I ................

400

I. 450

. . . .

"

"

"

"---

- . - -

m

I 500

Wavelength,)~(nm) Fig. 7. Thermochromism and photochromism of a thin film of N-(4-OCH3-salicylidene)-2-thenylamine: 1 at 298K, 2 at 77K, 3 after 20 min of 365 nm light irradiation at 77K, 4 after staying in the dark overnight at 298K. From Hadjoudis, Vittorakis and Moustakali-Mavridis, Chemtronics, 1, 58 (1986).

the molecule is planar, however, thus allowing the formation of the intramolecular hydrogen bond. Due to the non-planarity of the molecule, the characterist i c packing of f l a t molecules with a 3.5 X distance between planes was not observed (Fig. 8). This structure does not preclude the hypothesis of c_ i ._%s + trans isomerization for photochromic behavior; i t is, however, dissimilar to the structures of the thermochromic N-salicylideneanilines (ref. 9) and N-salicylideneaminopyridines (ref. 38) determined so far.

Fig. 8. Stereoscopic view of N-(4-OCH3-salicylidene)-2-thenylamine. From Moustakali-Mavridis, Terzis and Hadjoudis, Acta Crystal., C43(1987) 1389-1391.

702

This class of compounds shows that the planarity or non-planarity of the molecule is not the only determining factor for thermochromic or photochromic behavior respectively and more structures are needed in order to c l a r i f y the extent of the structure effect on these properties. Thus concerning the prevailing mechanism(s) of photochromism and

thermo-

chromism of Schiff bases, a number of investigators confirm the basic proposals of Cohen and Schmidt in that they identify the cis-keto form as the species produced in the thermochromic process (as well as the second species formed in hydrogen-bonding solvents) and the trans-keto form as the species produced photochemi cal I y. The conclusions of later investigators, however, lead to the idea that i t is not necessary to invoke the presence of ortho-quinoid tautomers in order to explain the observed spectral changes in certain media e.g. protic solvents. 2

ACI-NITROPHOTOTAUTOMERISM Chichibabin and co-workers (ref. 55) studied the photochromism of c r y s t a l l i -

ne 2-(2',4-dinitrobenzyl)-pyridine (a-DNBP) and proposed a H-transfer from the methylene bridge to the nitrogen of the pyridine ring (16). Later Hardwick and co-workers (refs. 56,57) studied

solutions of a-DNBP and its isomer 4-(2,4-

dinitrobenzyl)-pyridine (y-DNBP) (17) which was also photochromic and therefore

L••CH2

02

"-""

"---

~--CH N~02 02 ~

(16)

(17)

02N~CH2CN NO2

led them to suggest an alternate mechanism in which the H-transfer is to the oxygen of the nitro group. According to this mechanism, the pyridine ring is not an essential structural feature for photochromic activity, and is therefore replaceable by other electrophilic groups. Thus experiments were accumulated indicating that this photochromic behavior is general for phenyl methanes, the structural requirement being a nitro group ortho

to at least one benzylic hy-

drogen (ref. 58). The results of the above authors are consistent with the photochemical production of an excited species in which hydrogen is transferred from the methylene carbon to the oxygen of the ortho- nitro group, producing a colored aci-quinoid structure in equilibrium with its anion. A general formulation for this photochemical transformation is therefore as follows (18): R1 may be H, C6H5, CH3, etc. R2 may be a substituent that increases the ionizing a b i l i t y of the central C - H bond without interfering with the light absorption

703

of the 2,4-dinitrophenyl moiety and which becomesconjugated with the quinoid structure of the aci--form (e.g. when R1 = CH3 and R2 = NO2).

R~ Rz'-C

R~ NO z

I

H 0//

-

~

R~

Rz--C

N 'W,

NO z ~

Rz--C

N

0

HO /

NO 2

+H

+

(18)

N N0

-0 /

Nitro-form, tet rahedraI ( co 1 o r l e s s )

'~ 0

Aci-form or its anion, cop 1 a n a r ( co 1 o red)

Sousa and Weinstein (ref. 59) noted that the para- nitro group is not required for the photochromic activity of this type of compound. Thus they found 2-(2-nitro-4-cyanobenzyl) pyridine to be photochromic. Wettermark (ref. 60) found, using flash photolysis, that short-lived colo red species are formed when aqueous solutions of ortho- nitrotoluene and of derivatives of ortho- nitrotoluene are exposed to ultraviolet light; 2,4-dinitrotoluene behaves similarly (ref. 61). I t has been proposed that the aci- nitro structure, HA, represents the acid form of the colored

CHz

CH 2

!.

~

CH 2

species obtained in

+

H -I-

(19)

'H"o~'N~" 0 (N)

HO/

~ (HA)

-0 (A')

aqueous solutions at low pH and that the conjugate anion, A-, constitutes the basic form of the colored species observed at high pH (19). As the col or formed on addition of base is likely due to the formation of the anion, this supports the assumption that the anion is one of the photochemically produced colored

species (refs. 61-63). The presence of this species can be explained by

H-abstraction by the excited nitro group to yield the unstable HA. Although the product has not been isolated, evidence for its existence was obtained by running the photolysis in D20 and observing incorporation of deuterium into the methyl group (ref. 64). More recent work by Sergeev and co-workers (ref. 65) in solution at low temperatures showed that photolysis of N gives the aci-form HA which, on further photolysis, led to another product(s). An analogous hydrogen abstraction, using ortho- nitrosotoluene, has been observed by Hadjoudis and Wettermark (ref. 66). Klemm and co-workers (ref. 67) investigated the photochromism of a-DNBP by nanosecond laser absorption spectroscopy and observed two colored transients, a short-lived species in the region 390-410 nm and a long-lived one in the region

704

510-580 nm in polar and nonpolar solvents. The overall reaction scheme, which is supported by absorption measurements at different pH-values, is as shown in (20).According to this scheme, the short-lived species is attributed to an aci- . . . . _ .

02N--'~CH2 ~ H~~ NO2

~ T O,N- - ~ C H O ~t N~QH ~ I01

H~

polar solvents @ (i.e. ethanol)

hv

NO2

..

F

"

--~,:,

J

~N~

~,~ "X'o_,

,~

N.

e,or" "~b_6

H6~

(20)

I ,•r OH

(i.e. ethanol) and nonpolar (i.e. n-heptane) solvents

02N'~~NCHN" ~e~" x O-H

nitro form (ref. 58) and the long-lived one to the azamerocyanine (refs. 68,69) form of ~-DNBP. Yokoyama and Kobayashi (ref. 70), using time-resolved resonance Raman spectroscopy observed several transient Raman bands due to the longlived species. Their assignment of the Raman bands, however, corresponds to either a quinoid form or an azamerocyanin form and is thus not conclusive. Sixl and Warta (ref. 71) investigated, by optical absorption spectroscopy in O0-

+

O-

N+

O-

~.§

o _

O" ~§

O-

o

o~

,H~

H "'CH 2.

.

.

.

H NH .

.

.

.

OH'"

Fig. 9. Molecular structures of photochromic DNBPmolecules according to the "CH2" crystal structure of DNBP. "CH2" is the state of lowest energy. The "OH" and "NH" species are photoproduct configurations with higher ground state energies. Due to the change in the hybridization of the central C-atom, the real "OH" and "NH" stereo-structure is expected to be different from the original "CH2" configuration. From Sixl and Warta, Chem. Phys. 94, 147 (1985).

the temperature range 10 < T < 330 K, the photochromism of DNBP single crystals

705

and, as in the early room-temperature work of Clark and Lothian (ref. 72) on DNBP single crystals, they observed the "NH" photoproduct, and in addition, the "OH" photoproduct in accordance with previous results in solution (refs. 67,68). Thus the above authors, showed that the molecules of the DNBP single crystals are tri-stable at low temperatures. Below 200 K the "OH" and "NH" photoproducts are stabilized by their internal reaction barriers. The molecular structures of the "CH2", "NH" and "OH" species in the DNBP single crystals are shown in Fig. 9 using the "CH2" geometry (ref. 73). Fig. 10 shows the absorption spectra on irradiation of DNBP crystals at 20OK. Wavelength

800 3.0-

.

600 i "'NH .

hinmJ

500 .... I .

.

2.52~ cl

400 i .

CH2""

"'OH" t::

1.5 1.o 0.5A=335

....

0-i_

~sGoo

nm

I

T= 200 l( j

2oGoo

25Goo

Energy E/hc [ c m " J

Fig. 10. Time-dependent absorption spectra of DNBP crystals at 200K upon photolysis at 420nm. At t - 0 the spectrum corresponds to the unirradiated original DNBP crystal of "CH2" configuration. During irradiation two photoproduct absorptions appear, corresponding to the "OH" form at 435nm and the "NH" configuration at 600nm. From Sixl and Warta, Chem. Phys. 94, 147 (1985). The different photo- and thermal reactions in DNBP single crystals can be represented as follows (21)" "oH" /

"7~

"CH2" ..

X'-

_

I

i i

',I

ktl

(21)

|hv 3

"NH" The corresponding energy levels and reaction pathways proposed by Sixl and Warta (ref. 71) are shown in the schematic diagram of Fig. 11. The energy levels given in wavenumbers correspond to the absorption energies of the "CH2" adduct and the "NH" and "OH" photoproducts. The excited state energy barriers (EcN, EOC, ENO) are deduced from the Arrhenius plots of the photoreaction rate constants and

706

those of the ground state configurations (ENc, EOC, EON) from the Arrhenius plots of the rate constants of the thermal reactions.

40 meV E'CN "'CH21"/ . . . . |

360 meV E'oc " ~ " ' * 1..CHZ.. 300 meV ~ .,~'l~'Tc~ ,

'8

f o;-

o

iO. eV "'0"" .,____tl "'CH~"

eV

eV~ ' / o e /li

"'NH'"

LL "'CH2."

Fig. 11. Energy ]eve] diagram of the DNBPsystem. The photoreactions are involved with excited state energy barriers; thermal reactions are involved with ground state energy barriers. From Sixl and Warta, Chem. Phys. 9__44,147 (1985).

The mechanism of Fig. 11. shows that the "NH" and "OH" photoproducts are generated from the photoexcited "CH2" adduct. Upon photoexcitation of the "NH" configuration the "OH" configuration is generated and upon photoexcitation of the "OH" configuration, the "CH2" adduct is recovered. The highest energy "OH" ground state configuration decays monoexponentially into the "NH" and back to the "CH " configurations 2 nally decays back to the "CH2II adduct. 3

The "NH" configuration f i -

OTHERH-TRANSFER PHOTOTAUTOMERISM

3. 1 Metal Dithizonates Irving and co-workers (ref. 74) and Webb and co-workers (ref. 75) reported independently that the mercury ( I I ) dithizone (diphenylthiocarbazone, H2Dz) complex is photochromic. Thus benzene or chloroform solutions of Hg(HDz)2 change on irradiation with visible light from their normal orange-yellow color to an intense royal-blue. The orange-yellow color returns slowly in the dark and these color changes can be repeated many times. Meriwether and co-workers (ref. 76) later examined this photochromic behavior in detail and found that photochromism was a general behavior of the heavy metal dithizonates. Further kinetic and infrared studies (ref. 77) showed that the photochromic mechanism involves a N4 to N2 H-transfer and a geometrical isomerization about the C - N bond as shown in (22). The i n i t i a l orange-yellow form has a strong absorption band around 500 nm

707

Ph %N4 --H

N3/

i Ph \

Ph\ \ H giN. .t~. . H

hv

\

/

s-'-c~ Ph ~ /N3 H --'N4 ~

Ph

"S_C_N/N-Ph

"~

\ ~ / N--N-H , /Hg I //N-Ph '

-

(22)

S=C-N

Ph

which, on irradiation, produces the blue photoproduct with a strong absorption at longer wavelengths, often around 600 nm. Fig. 12 illustrates the case of the mercury complex. 0.7

0.6 0.5

1

_

0.4 r

o 0.3

u~ J3 <

0.2 0.1 0.0 700

600 500 Wavelength (nm)

400

Fig. 12. Spectra showing the photochromism of mercury dithizonate Hg(KD?)p in benzene at 25v. 1, before and 2, after irradiation with visible l~ght. FI"o$ Meriwether, Breitner and Sloan, j . Am. Chem. Soc. 8_]_7,4441 (1965).

Photochromism is probably an inherent property of the l igand since i t occurs in the presence of a variety of metals. The central metal atom determines the photochemical s t a b i l i t y , rate of the return reaction, and in some cases, the colors of the complexes. 3. 20_rtho-Alkyl Aromatic Imines Toshima and co-workers (ref. 78) found that o_rtho__- alkyl aromatic imines isomerize photochemically to enamide derivatives by a H-transfer (23). Thus at 77K, in an anhydrous propan-2-ol-methanol (1"1) glass, N-acetyl-o-methyldiphenylmethaneimine (la) on irradiation develops a new absorption band (Fig. 13) owing, most probably, to the formation of its quinoid form 2a with absorption maxima

708

around 405 and 430 nm.

/Ph ~

P

h

Ac

hv

(la) R=H (2a) R=H -Ac

A

R~CH 2 - -

(23)

(lb) R=Ph (2b) R:Ph

This new band (2 in Fig. 13) is stable at low temperature and reverts to the original spectrum at room temperature.

8 <

,

I

350

.,,

I

I

400 450 Wavelength (nm)

Fig. 13. The electronic spectrum of N-acetyl-o-methyldiphenylmethaneimine (la) in propan-2-ol-methanol ( I ' I ) (ca. O.IM) at 77 K. 1, before i r r a d i a t i o n ; 2, after i r r a d i a t i o n for 5 min.; From Toshima, Saeki and H i r a i , Chem. Commun., 1424 (1971). The maxima of

2b (430 and 480 nm) are red shifted in comparison with those

of 2a. The ~max values of these photochemical isomers are comparable with those reported for o-alkylbenzophenones(refs. 79-81). On i r r a d i a t i o n of a solution of la at room temperature, no change was observed, but this is most probably due to a very fast back reaction. Thus Hadjoudis and Hayon (ref. 82), using flash photolysis, observed similar quinoid transient species at room temperature in heptane solutions of o-methylbenzylideneaniline. 3. 30rtho-Nitrobenzylidene-acyl - Hydrazides Ellam and co-workers (ref. 83) investigated the effect of extending the conjugation of 4-(2,4-dinitrobenzyl)-pyridine (refs. 56,57) and found that solid o--nitrobenzylidene-pyridine-4-carboxylic acid hydrazide (A) is photochromic. Thus when this pale-yellow compound is irradiated with uv-light, there is a dra-

709

matic buildup

of a reddish coloration. The process may be reversed by heating

the material gently. The color change reverses exponentially with a h a l f - l i f e of 52 min

at 80~

Determination of the rate constants at several temperatures

gave an activation energy of 10 Kcal/mole for the process. The suggested mechanism (24) involves a hydrogen shift via an excited state resulting in a change in the double bond sequence of the compound. Fig. 14 shows the reflectance spec-

oN.N-CH

tra of this solid compound before and after irradiation with uv-light. The au-

0 H ~N ~i----C~ U

O N

-.-=-..=.o~

OzN. "N"

hv

Yellow form

-N

|

Excited state

.

o! N"OH

o

Red f o r m ( 2 4 )

==

thors suggested that the high concentrations of the crystalline state

are

essential to the process.

1 E 1.0" oo _Q =..

8 0.5.

I, 400

,

! 500 Wavelength

! , 600

,

(nm)

Fig. 14. Reflectance spectra of solid o-nitrobenzylidene-pyridine-4-carboxylic acid hydrazide; 1, before irradiation;-2, after 95 min of irradiation with uvlight. From Ellam et al. Chemistry and Industry, (1974) 77.

Hadjoudis and co-workers (ref. 84), however, detected by flash photolysis experiments, in PMMArigid glasses at room temperature a colored transient species with a spectrum which is similar to that of a polycrystalline film and comparable to the above reflectance spectrum 2. Therefore the photoreaction takes place in other states as well, e.g. in rigid glasses and at relatively low concentrations (:10-4M), but faster techniques as the flash photolysis for its detection.

are needed

710

ACKNOWLEDGMENT I would like to express my appreciation to prof. J.R. Scheffer, for reading the manuscript and for making valuable comments and suggestions which contributed to its improvement. REFERENCES I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

A. Senier and F,G. Shepheard, J. Chem. Soc., 95(1909) 1943. A. Senier, F.G. Shepheard and R. Clarke, J. Chem. Soc., 101(1912) 1952. M.D. Cohen and G.M.J. Schmidt, J. Chem. Phys., 66(1962) 2442. M.D. Cohen, Y. Hirshberg and G.M.J. Schmidt, J. Chem. Soc., (1964) 2051. M.D. Cohen and S. Flavian, J. Chem. Soc. B, (1967) 334. L. Sacconi~ M. Ciampolini and G.P. Speroni, J. Amer. Chem. Soc., 87(1965) 3102. A. Chakravorty, Inorg. Chem., 4(1965) 128. M.D. Cohen, J. Chem. Soc.(B), (1968)373 , M.D. Cohen and B.S. Green, n, Chem. Britain, 9(1973) 490. Rz=H, R2=2-CI by J. Bregman, L. Leiserowitz and K. Osaki, J. Chem. Soc., (1964)2086-2100; R~= 5-Cl, R2=H by j . Bregman, L. Leiserowitz and G.M.J. Schmidt, J. Chem. Soc., (1964)2068. Rz=H, R2=4-CI by J. Bregman, E. Mond and G.M.J. Schmidt, unpublished results. Rz=H, R2=4-Br by G.M.J. Schmidt, unpublished results; see also M.D. Cohen, Y. Hirshberg and G.M.J. Schmidt in: D. Hadzi (Ed.), Hydrogen Bonding, Pergamon Press, London, 1959, p. 293. M.D. Cohen, G.M.J. Schmidt and S. Flavian, J. Chem. Soc., (1964)2041. G.M.j. Schmidt, The photochemistry of the solid state, in" Reactivity of the Photoexcited Organic Molecule, Interscience, London, 1967, pp. 227-284. D.G. Anderson and G. Wettermark, J. Am. Chem. Soc., 87(1965) 1433. G. Wettermark and L. Dogliotti, j . Chem. Phys., 40(1964) 1486 9 R.S. Becker and W.F. Richey, J. Am. Chem. Soc., 89(1967) 1298. M. Ottolenghi and D.S. McClure, J. Chem. Phys., 46(1967) 4613. J.D. Margerum and J.A. Sousa, Appl. Spectrosc., 19(1965) 91. J.W. Ledbetter, Jr., J. Phys. Chem., 70( 1966)2245. G.O. Dudek and E.P. Dudek, J. Am. Chem. Soc., 88(1966) 2407-2411. A.A. Burr, E.J. Llewellyn and G.F. Lothian, Trans. Faraday Soc., 60(1964) 217. J.D. Margerum and L.J. Miller in G.H. Brown, Ed., Photochromism, WileyInterscience, New York, 1971, pp. 557. R.V. Andes and D.M. Manikowski, Appl. Opt., 7(1968) 1179. G.H. Brown and W.G. Shaw, Rev. Pure Appl. Chem., 11(i961) 1. R.E. Exelby and R. Grinter, Chem. Rev., 65(1965) 247. T. Rosenfeld, M. Ottolenghi and A.Y. Meyer, Mo]. Photochem., 5(1973) 39. A.P. Simonova, R.N. Nurmukhametov and A.L. Prokhoda, Dokl. Phys. Chem., 230 (1976) 936. R.N. Nurmukhametov, O.I. Betin and D.N. Shigorin, Dokl. Phys. Chem,, 230 (1976) 828. K.H. Grellmannand E. Tauer, Tetrahedron Lett., (1974) 3707. J.W. Ledbetter, Jr., J. Phys. Chem., 81(1977) 54. j . Csaszar, J. Balog and A. Makary, Acta Chim. (Budapest) (1978) 473. C.J. Seliskar, J. Phys. Chem., 81(1977) 1331. R. Nakagaki, T. Kobayashi, J. Nakamura and S. Nagakura, Bull. Chem. Soc., Jpn., 50(1977) 1909. J.J. Laverty and Z.G. Gardlund, Polymer Letters, 7(1969) 161. M. Goodman and A. Kossoy, J. Am. Chem. Soc., 88(1966) 5010 ; M. Goodman and M.L. Falxa, Ibid, 89 (1967) 3863. A. Ueno, J. Anzai, T. Osa and Y. Kadoma, Bull. Chem. Soc. Jpn., 52(1979) 549; O. Pieroni, J.L. Houben, A. Fissi and P. Costantino,

711

J. Am. Chem. Soc., 102(1980) 5913. 37 E. Hadjoudis, I. Moustakali-Mavridis and J. Xexakis, Isr. J. Chem., 18(1979) 202. 38 I. Moustakali-Mavridis, E. Hadjoudis and A. Mavridis, Acta Crystal., B34(1978) 3709. 39 E. Hadjoudis, M. Vitorakis and I. Moustakali-Mavridis, Mol. Cryst. Liq. Cryst., 137(1986) 1. 40 M.D. Cohen, Recent research in topochemistry at the Weizmann Institute of Science, in" D. Ginsburg (Ed.),Solid State Photochemistry, Verlag chemie, Weinheim, 1976, pp. 233-254. 41 B.S. Green, R. Arad-Yellin and M.D. Cohen, Stereochemistry and Organic Solid-State Reactions in: E l i e l , Wilen and Allinger (Eds.), Topics in Stereochemistry, V16, Interscience, 1986, p. 170. 42 I. Moustakali-Mavridis, E. Hadjoudis and A. Mavridis, Acta Crystal., B36 (1980) 1126. 43 E. Hadjoudis, J. Photochem., 17(1981) 355. 44 P.F. Barbara, P.M. Rentzepis and L.E. Brus, J. Am. Chem. Soc.,102(1980) 2786. 45 J.W. Lewis and C. Sandorfy, Can. J. Chem., 60(1982) 1720. 46 J.W. Ledbetter, J. Phys. Chem., 86(1982) 2449. 47 H. Lee and T. Kitagawa, Bull. Chem. Soc. Jpn., 59(1986) 2897. 48 U.W. Grummt, Journal f. prakt. Chemie, 327(1985) 220. 49 D. Higelin and H. Sixl, Chem. Phys., 77 (1983) 391. 50 A. Weller, Z. Elektrochem.,60(1956) 1144; A. Weller in G. Porter; Ed., Progress in Reaction Kinetics, V1, Pergamon, London 1961, p.188. 51 T. Kawato, H. Koyama, H. Kanotomi and M. Isshiki, J. Photochem., 28(1985) 103. 52 E. Hadjoudis, M. Vittorakisand I. Moustakali-Mavridis, Chemtronics, 1(1986) 58. 53 E. Hadjoudis, M. Vittorakis and I. Moustakali-Mavridis, Tetrahedron, 43 (1987) 1345. 54 I. Moustakali-Mavridis, A. Terzis and E. Hadjoudis, Acta Crystal., C43 (1987) 1389. 55 A.E. Chichibabin, B.M. Kuindzhi and S.W. Benewolenskaja, Ber., 58(1925) 1580. 56 R. Hardwick, H.S. Mosher and P. Passailaigue, Trans, Faraday Soc., 56(1960) 44. 57 H.S. Mosher, C. Souers and R. Hardwick, J. Chem. Phys., 32(1960) 1883. 58 J.D. Margerum, L.J. Miller, E. Saito, M.S. Brown, H.S. Mosher and R. Hardwick, J. Phys. Chem., 66(1962) 2434, G.Wettermark, J. Am. Chem. Soc., 84(1962) 3658. 59 J.A. Sousa and J. Weinstein, J. Org. Chem., 27(1962) 3155. 60 G. Wettermark, Nature, 194(1962) 677. 61 G. Wettermark and R. Ricci, J. Chem. Phys., 39(1963) 1218. 62 G. Wettermark, J. Phys. Chem., 66(1962) 2560. 63 G. Wettermark, E. Black and L. Dogliotti, Photochem. and Photobiol., 4(1965) 229. 64 H. Morrison and B.H. Migdalof, J. Org. Chem., 30(1965) 3996. 65 A.M. Sergeev, R.N. Nurmukhametov and R.N. Barov, Khim. Fiz., 8(1982) 1096 (Russ); CA: 101:129967 d. 66 E. Hadjoudis, A. Tsoka and G. Wettermark, J. Photochem., 8(1978) 233. 67 E. Klemm, D. Klemm, A. Graness and J. Kleinschmidt, Chem. Phys. Letters, 55(1978) 113. 68 E. Klemm, D. Klemm, A. Graness and J. Kleinschmidt, Chem. Phys. Letters, 55(1978) 503. 69 D. Klemm, E. Klemm, A. Graness and J. Kleinschmidt, Z. phys. Chemie, Leipzig, 260(1979) 555. 70 K. Yokoyama and T. Kobayashi, Chem. Phys. Letters, 85(1982) 175. 71 H. Sixl and R. Warta, Chem. Phys., 94(1985) 147. 72 W.C. Clark and G.F. Lothian, Trans. Faraday Soc., 54(1958) 1790.

712

73 K. Seff and K.N. Trueblood, Acta Cryst., B24(1968) 1406. 74 H. Irving, G. Andrew and E.J. Risdon, J. Chem. Soc., (1949) 541-547. 75 J.L.A. Webb, I.S. Bhatia, A.H. Corwin and A.G. Sharp, J. Am. Chem. Soc., 72(1950) 91. 76 L.S. Meriwether, E.C. Breitner and C.L. Sloan, J. Am. Chem. Soc., 87(1965) 4441. 77 L.S. Meriwether, E.C. Breitner and N.B. Colthup, J. Am. Chem. Soc.,87(1965) 4448. 78 N. Toshima, M. Saeki and H. Hirai, Chem. Commun., (1971) 1424. 79 T. Okada, M. Kawanisi, H. Nozaki, N. Toshima and H. Hizai, Tetrahedron Lett., (1969) 927. 80 A. Padwa, W. Bergmark and D. Pashayan, J. Am. Chem. Soc., 91(1969) 2653. 81 B. Fraser-Reid, A. McLean and E.N. Usherwood, Canad. J. Chem., 47(1969) 4511 9 82 E. Hadjoudis and E. Hayon, J. Phys. Chem., 74(1970) 3184. 83 R.~,I. Ellam, P.B. East, A. Kelly, R.M. Khan, J.B. Lee and D.C. Lindsey, Chemistry and Industry, (1974) 74. 84 E. Hadjoudis, I. Argyroglou and A. Tsoka, Chim. Chronika, 7(1978) 203. See Additional Literature (1989 - 2001): Anils, A107 Literature on Group Transfer Photochromism of Quinones, A 111 Literature Survey for Photochromism based on Electron Transfer of Bipyddinium-salts (Viologenes), A115