Single-crystal-to-single-crystal 2+2 photodimerizations through absorption-tail irradiation


Material Information

Single-crystal-to-single-crystal 2+2 photodimerizations through absorption-tail irradiation
Physical Description:
ix, 204 leaves : ill. ; 29 cm.
Novak, Kathleen
Publication Date:


Subjects / Keywords:
Solid state chemistry   ( lcsh )
Photochemistry   ( lcsh )
Crystals   ( lcsh )
Chemistry thesis Ph. D   ( lcsh )
Dissertations, Academic -- Chemistry -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1993.
Includes bibliographical references (leaves 198-203).
Statement of Responsibility:
by Kathleen Novak.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001950901
oclc - 31200888
notis - AKC7443
System ID:

Full Text







This dissertation is dedicated to my family.

To my sister and brother,

Jeanne Annette Novak


Michael Anthony Novak

And to our parents,

Dolores Marlene Novak


James Joseph Novak

It was a family effort, as always.


I have looked forward to writing these pages of this dissertation--my thank you to

the many people contributing to this work--more than any other. It is certainly the part I

am most qualified to author.

If I was forced to choose only one person to thank here, it would be Professor Ken

Wagener, of course. I could write a dissertation on his skill as an advisor, his

commitment and generosity to his graduate students (as well as to any of the other

graduate students in the Chemistry Department and to the undergraduates in his courses),

and the joy and magic that he finds in the science of chemistry and reveals to those

students. Professor Wagener has created an exchange program with Professor Gerhard

Wegner from the Max-Planck-Institut fiir Polymerforschung (MPI-P) in Mainz,

Germany. My participation in this program was generously funded by the 3M Company

of St. Paul, Minnesota, and brought me to the MPI-P for two years during the course of

my doctoral work.

At our first meeting Professor Wegner suggested I irradiate the styrylpyrylium salt

crystals in their absorption tail to create single crystals of photoproduct This method

became the primary theme of the investigations, resulting in work that was eventually

presented at the ICCOSS XI conference in Jerusalem, 1993. I am deeply grateful for the

opportunity, made possible by Professor Wegner, to attend this conference. It was an

invaluable professional experience that has had a significant bearing on the development

of this dissertation. I also thank Professor Wegner for welcoming me to the MPI-P two

years ago and for all his support, encouragement and patience since then.

I smile at the thought of "acknowledging" anyone mentioned here, but especially

Dr. Volker Enkelmann. His ideas and labor are the heart of this dissertation. He was

working on the twenty-fifth full x-ray crystal structure associated with this research when

I left the Max-Planck-Institut. I wish I could describe how edifying and enlightening all

our discussions of this work were to me. But what measure exists for the gratitude an

apprehensive student possesses for a good and patient teacher?

Dr. Werner Kihler is the force behind the work in Chapter 6. He adapted

equipment he had been using in forced Raleigh scattering experiments and realized how

kinetic data could be obtained from thermal decay of the grating. He turned otherwise

tedious data collection monitoring into fascinating tutorials on holographic gratings.

Michael Steiert created many of the crystals used in Chapters 5 and 6, while

Bernhard Zimmer prepared the laser setup used for the styrylpyrylium triflate irradiation.

I owe their expertise and diligence an incalculable debt. Special thanks go to Herr

Zimmer, who spent days modifying and fine-tuning the laser setup used for the

irradiations reported in Chapter 3. Dr. Giinter Lieser's expertise with the

photomicroscope was also invaluable.

During my last days at MPI, Elke Muth slaved over the characterizations of many

of the compounds. Her crucial support is deeply appreciated. Carsten Fiilber generously

donated his time during the course of his own doctoral work to create the PV-Wave plots

found in Chapter 6. I thank Dr. Jeremy Titman for performing the C-13 solid state NMRs

as well as Walter Scholdei and Dr. Wolfgang Meyer for helpful discussions and advice.

I would also like to thank Lorraine Williams for her tireless efforts in arranging all

the paperwork associated with my travel and study overseas. If it were not for her and

Dr. Fabio Zuluaga, Dr. Dennis Smith and Dr. Will Vaughan taking care of my University

of Florida concerns, the trip to Germany would have been very difficult. Special thanks

go to Will Vaughan for the 1H-NMR in Figure 2-4. Finally, I would like to thank Dr.

James Konzelman, Dr. Jeffrey Linert and Professor James Deyrup for all their advice and

encouragement over the last four years. I am particularly grateful to Jim Konzelman for

his patient help with my first experiments and preparation for my seminars.


ACKNOW LEDGEM ENTS ................................................................... ......................... iii

ABSTRACT........................................................................................................................ vii


1 INTRODUCTION .....................................................................................................1

M motivation .................................................................................................................1
Preview ................................................................................................................ 4
Background ......................................................................................................... 5
[2+2] Photodimerizations of Styrylpyrylium Salts.........................................5
Topochemistry ........................................... ............... .............................. 9
Beyond Cinnamic Acids ........................................................................... 18
Appearance of the Product Phase: Homogeneous vs. Heterogeneous
M mechanisms ....................................................................................... 28
Summary of Results Presented in this Dissertation.........................................38

2 EXPERIM ENTAL ............................................................................................ 41

Instrumentation and Analysis................................. .............. ...........................41
General Information ................................................................................ 41
X-ray Structure Analysis ....................................................... ............... 41
Solid State C-13 M easurement ..................................................................42
Crystal Preparation .............................................................................................. 42
Syntheses ............................................................................ ...................... 45
Recrystallizations .......................................... .............. ..........................50
Irradiation................................................. ...................... ............................ 53
Photodimer Single-Crystal Preparation ........................................................53
M icrographs ............................................................................................54
Kinetic Studies........................................................................................ 55
Characterization ................................................................................................ 63
Elemental Analyses ............................................................................... 64
Infrared Analyses........................................................................ ............. 66
M ass Spectral Analyses ........................................ .................... ............ 71
NM R Analyses........................................................................................ 73

STYRYLPYRYLIUM TRIFLATE .................................................................84

Introduction ............................................... ...................................................... 84
Topochemical Reactions: A Type of Phase Transition .............................85
Homogeneous vs. Heterogeneous Product Formation ...............................90
Single-Crystal-to-Single-Crystal Topochemical Reactions: Phase
"Transitions" Involving Dimensionally Similar Lattices ......................94
Results .....................................................................................................................98
Crystal Modifications of a Styrylpyrylium Triflate and its
Photoproducts: The Overall Scheme..................................................98
[2+2] Photodimerization: The Forward Reaction.................................... 02
Thermal Cycloreversion: The Backward Reaction.....................................109
Substitutional Mixed Crystals: What Happens "In-Between".................. 115
Future Studies .................................................................................................125

4 CINNAM IC ACID: A REPRISE ................................................................... .... 127

Introduction .....................................................................................................127
Results .............................................................................................................128
t-Butyl Amine Salt of trans-Cinnamic Acid ............................................ 128
a-trans-Cinnamic Acid...................................................................... .... 130
Future Studies .................................................................................................. 140


Introduction..................................................................................................... 141
Crystal Packing ........................................................................................ 141
Kinetic Investigations on [2+2] Photodimerizations ................................ 142
Results ...................................................................................................................143
Crystal Structures.....................................................................................143
Kinetic Studies ............................................................................................157
Thermal Cycloreversion of Photodimers ........................................ ....165
A bis(Styrylpyrylium Triflate)........................... ...................................... 171
Future Studies .................................................................................................172


Introduction ......................................................................................... ............. 174
Background.............................................................................................. 174
The Setup ..............................................................................................178
Theory ......................................................................................... ........... 179
Results .............................................................................................................186
W writing the Grating.................................................................................. 186
Energy of Activation for Cyclobutane Cleavage...................................... 87
Energy of Activation for Photodimerization ........................................... 193
Future Studies .................................................................................................194

THE DRAWINGS OF M. C. ESCHER.................................................................196

REFERENCES............................................................................................................. 198

BIOGRAPHICAL SKETCH ........................................................................................204


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Kathleen Novak

December, 1993

Chairman: Kenneth B. Wagener
Major Department: Chemistry

A method for controlling the mechanism of product phase evolution in

topochemical reactions of photoactive molecular crystals is described that produces rare

single-crystal-to-single-crystal transformations. In the case of the [2+2]

photodimerizations of cinnamic acids and styrylpyrylium salts, a heterogeneous

mechanism, typical of solid state photodimerizations, has been the observed. In such

cases, the photoproduct is obtained as a polycrystalline powder and a full x-ray structure

analysis is not possible unless the product is recrystallized. However, a homogeneous

mechanism is induced when these photodimerizable monomer crystals are irradiated with

wavelengths in the long-wavelength tail of their electronic absorption spectra. This leads

to the formation of single crystals of photoproduct. Thus full x-ray structure analyses

may be done on the as-formed photoproduct and crystals of intermediate conversion.

Twenty-one crystal structures are reported in this dissertation: seven for the

styrylpyrylium triflate, its photodimer, and solid solutions of both and five analogous
structure analyses for the photodimerization of a-trans-cinnamic monomer crystals. In

both cases, the lattice evolves from the monomer to the product phase through a side-

group swivel. The analyses on the solid solutions are the first such crystal structures

reported on substitutional mixed crystals formed through a topochemical


The crystal modification obtained by homogeneously forming the photoproduct

may or may not be different from that obtained through heterogeneous conversion, as

determined by x-ray powder measurements. However, the recrystallized photoproduct

for styrylpyrylium photodimerization is vastly different from its as-dimerized

modification. The as-dimerized cyclobutane crystals thermally revert to single crystals of

monomer in the first documented single-crystal-to-single-crystal topochemical back


Two kinetic studies were performed: one was on the heterogeneous

photodimerization of an isomorphous series of styrylpyrylium salts, and the other

involved writing a thick phase grating into a single crystal. The growth and thermal

decay of the diffracted intensities provide data on the activation energies for the

photodimerization and thermal cycloreversion. This is the first use of holography to

study the kinetic behavior of a low molecular weight organic substance reacting in the

crystalline state.



The subject of this dissertation is single-crystal-to-single-crystal [2+2]

photodimerization reactions in organic solid state chemistry. In a recent review on

bimolecular photoreactions in crystals, Ramamurthy and Venkatesen write, "There is a

clear lack of understanding of the mechanism of solid state reactions. Present knowledge

exists only on the initial and final stages of the organic solid state reactions."' Until now,

crystallographic investigations have been limited because photoreactive crystals typically

disintegrate into polycrystalline powder during the reaction. This dissertation reports on

investigations of a general method for obtaining single-crystal-to-single-crystal

transformations of photoactive crystals and presents the first full x-ray structures of

crystals at an intermediate stage of a solid state photodimerization.

The initial motivation for this study arose from the interest in developing systems

applicable to thermally reversible crosslinking in polymers;23 therefore, adduct forming

reactions, were sought which produced the original edducts on heating. The thermally

reversible solid state photodimerization4 of (E)-2,6-Di-tert-butyl-4-[2-(4-methoxyphenyl)-

ethenyl]pyrylium trifluoromethanesulfonate--hereafter referred to as the styrylpyrylium

triflate, 1-lc--shown in Figure 1-1 and Figure 3-6, falls into this category and was chosen

to be investigated. At that time, the photoreaction of this styrylpyrylium salt under typical

irradiation conditions--that is, irradiation corresponding to wavelengths in the range of its

absorption maximum--was known to proceed, as most [2+2] crystalline-state

photoreactions, with disintegration of the crystal into a polycrystalline powder due to a

heterogeneous mechanism of product phase evolution.

For example, it was known that the [2+2] photo-oligomerization of the diolefin

distyrylpyrazine (DSP) occurs with disintegration of the reactant single crystals into a

polycrystalline powder when those crystals were irradiated with wavelengths absorbed

strongly by the DSP monomer.5-9 However, it had also been proven by Braun6'7'10 that

intact single crystals of DSP oligomer could be obtained if the irradiation was effected with

wavelengths corresponding to the long-wavelength tail of the monomer

absorption.11,6'710 While crystal disintegration was ascribed to heterogeneous evolution

of the product phase within the crystal, homogeneous product formation was suggested as

the origin of the single-crystal-to-crystal transformation. Thus the mechanism of phase

transition could be controlled by proper choice of irradiation conditions.

The photodimerization of the styrylpyrylium triflate described in this dissertation

was found to show the same behavior as DSP; that is, intact single crystals of

photoproduct were obtained with tail-irradiation, even though crystal disintegration

occurred otherwise. Only one other single-crystal-to-single-crystal [2+2] photodimerization

had ever been demonstrated in the literature and no reversible single-crystal-to-single-

crystal transformations had ever been found.12.13 A recent reviewer of bimolecular

photoreactions in crystals writes, "Reversible photodimerizations in crystals are expected

to find wide application in molecular-scale devices. Successful application can be found

only for systems that are truly single-crystal-single-crystal transformations."13 Since

single-crystal-to-single-crystal transformations in general are quite rare and of significant

academic as well as technological interest, the work was redirected from the crosslinking to

carry out the following investigations and to answer the following questions.

For the styrylpyrylium triflate, crystallographic analysis of the crystals of the

monomer and as-dimerized dimer would verify that this reaction was topochemical and

show how much the lattice changed during the reaction. Viewing the photodimerizing

single crystals with an optical microscope would indicate whether phase boundaries did

emerge or if the reaction was single-phase. If the reaction was single-phase, could single

crystals of partial conversion be obtained and what could be learned about the mechanism

of the reaction from their analysis? For example, a recent challenge to Schmidt's

Topochemical Principle,1415 the foundation of much of organic solid state chemistry, had

appeared in the literature. What could be learned from this study to contribute to this

debate? Though several single-crystal-to-single-crystal reactions in the literature

existed,6'7,10,12.16-20 no examples of a single-crystal-to-single-crystal back reaction had

ever been demonstrated. The photodimerization of the styrylpyrylium triflate was known

to be thermally reversible, and since single crystals of the photodimer were now

obtainable, could single crystals of the thermally reversed monomer be achieved and how

would they compare to the original monomer? Furthermore, how would the crystal

structure of the recrystallized dimer compare to that of the as-dimerized dimer, and how

would their thermal behavior compare? How would the photoproduct obtained

heterogeneously compare to the product formed homogeneously?

To broaden the investigation to other styrylpyrylium salts, the synthesis of an

isomorphous series of salts through counter-anion replacement was planned. If this piece

of "crystal engineering" were possible,21-24 how would the structures and reactivities of

the members this series compare to one another? In addition to varying the counterion,

would it be possible to link two styrylpyrylium salts with either a methylene bridge or

through sharing an aromatic ring to create a photopolymerizable crystal that forms single

crystals of polymer?9,25-37

Most importantly, how general is the method of tail-irradiation for inducing single-

crystal-to-single-crystal reactions? Along these lines, it was planned to try to induce a

single-crystal-to-single-crystal transformation for the quintessential organic solid state

dimerization, that of a cinnamic acid derivative. Expecting that the single-crystal-to-single-

crystal photodimerization of cinnamic acid itself would not be possible, the synthesis of a

series of ammonium salts of several cinnamic acids was planned involving large

substituents on the nitrogen atom, with the idea of mimicking the styrylpyrylium salt


As originally suggested by Hesse and Hiinig,4 the thermally reversible

photochromic behavior of the styrylpyrylium triflate, 1-1c, might be applicable to

reversible information storage. A few examples of holographic recording systems based

on creating a spatial variation of the refractive index (a phase grating) within single crystals

of photochromic materials through photodimerization reactions exist in the literature.3841

These systems have been shown to be photochemically reversible but suffered eventually

from the typical crystal deterioration, as described above for DSP. Would it be possible to

write a holographic grating into a single crystal of a styrylpyrylium salt? If so, could this

grating be thermally erased and rewritten? What kinetic information about the

photodimerization and thermal cycloreversion could be extracted from holography?32,4243


Descriptions of the investigations and inquiries posed in the preceding four

paragraphs are fully addressed in chapters 3, 4, 5 and 6, respectively, and summarized at

the end this chapter. Besides outlining the initial work done on the photodimerization of

styrylpyrylium salts, the purpose of the balance of this chapter is to introduce certain

aspects of organic solid state chemistry through a brief review of the literature most

pertinent to issues discussed in this dissertation, primarily the evolution of the product

phase in topochemical reactions.

The Background section is divided into two parts. First, a synopsis is given of the

original work on the solid state photodimerization of styrylpyrylium salts that initiated the

work reported in this dissertation. In the second part, topochemistry is distinguished from

other types of organic solid state reactions. An outline of Schmidt's work on cinnamic

acids is presented as it pertains to the foundation of topochemistry, Schmidt's

topochemical principle, and some post-Schmidt conceptual refinements to the topochemical

principle are introduced. Specifically, the concepts "reaction cavity" and homogeneous vs.

heterogeneous mechanisms of product phase formation are introduced to explain single-

crystal-to-single-crystal transformations. Finally, a survey of selected examples of organic

solid state reactions is presented to compare and contrast various mechanisms of product

phase formation thus far reported in the literature. Separate introductions to chapters 3, 5

and 6 contain further background information from the literature specific to those chapters.

The field of organic solid state chemistry is widely acknowledged to be a largely

underdeveloped but rapidly growing science. Sophisticated and elegant investigations of a

number of fascinating issues, related to the topics of this dissertation but not having a

direct bearing, have been reported in the literature. Some of the most often-cited of these

reports and collections of papers have been summarized in the section The literature of

topochemistry (see p. 20).


f2+21 Photodimerizations of Styrvlpvrylium Salts

The synthesis of a series of nineteen a-styrylpyrylium salts has been reported, and

their photochemistry in the solid state and in solution, as well as the thermal behavior of

their photoproducts, has been explored by Hiinig and Hesse.4 This section of the

introduction to this dissertations briefly reviews their results. A summary is shown in

Table 1-1. The solid state photoactivity was tested by irradiating KBr pellets of each of the

salts. The preparative irradiations were done in hexane suspensions with wavelengths

strongly absorbed by the salts and thus no single-crystal-to-single-crystal transformations

were observed. The styrylpyrylium salts that are photoactive in the solid state were

shown to dimerize to form cyclobutane rings as illustrated in Figure 1-1.

Table 1-1. Previous work on the solid state photoactivity of some styrylpyrylium
salts and the solid state thermal behavior of their photodimers.4

salt substituent, X photoactivity dimer thermal behavior of dimer

1-1 a 4-NMe2
b 3,4,5-(OMe)3
c 4-OMe
d 4-Me
e 4-C1
f 4-H
g 4-CF3
h 4-N02
i 4-CN

1-2 a 4-NMe2
c 4-OMe
d 4-Me
f 4-H
h 4-N02
i 4-CN

1-3 c 4-OMe
d 4-Me
f 4-H
i 4-CN



53 %

1-4b 100% cycloreversion
1-4c 100% cycloreversion
1-4d 100% cycloreversion
1-4e modification/melting

1-4i modification/melting


CF3S03 t-Bu


x 4
BF4- Ph


S BF( Ph

Phenyl ring

PPhr- y h a Pyrylium ring
Pyrylium ringhv
Pyrylium ring


Figure 1-1. The [2+2] photodimerization of the styrylpyrylium salts in Table 1-1
occur to give head-to-tail (htt) dimers with an r-ctt (reference-cis,trans,trans) configuration.
See reference 44 for more details on the IUPAC rules for specifying the stereochemistry of

The preparative irradiations gave the yields indicated in Table 1-1. For 1-3d and
1-3f, edge filters (X>390 nm) were used to inhibit excitation of the product cyclobutane

back to monomer, but this did not improve the yield. This resulted in the conclusion that

the change in lattice parameters, that is, the lattice contraction, during the dimerization is

so severe that the unreacted molecules are shifted to unreactive orientations and distances.

The photostability of 1-la was suggested to be due to an unreactive crystal structure or the
strong delocalization of 7t-electrons in the exocyclic double bond. Through IR and NMR

analyses, the structure and E-configuration of the double bond in each of the salts was

verified. The stereochemical configuration of all the photodimers reported in this paper

were determined to be of the type [r-ctt htt],44 see Figure 1-1, by 1H-NMR analysis of the

cyclobutane proton signals in o-chlorophenol.

The compounds in series 1-1 form strongly fluorescent solutions in acetic acid

(series 1-2 and 1-3 are insoluble). Compounds 1-lc and 1-id photodimerize in this

solution to form only the cyclobutane isomer obtained from irradiation in their crystalline

state, [r-ctt htt dimer].44. The other possible dimers, [r-tct htt], [r-ctt hth], and [r-tct hth],

that would be expected from the reaction of an excited trans monomer with a ground state

trans monnomer, were not isolated, which indicates a pre-orientation of reacting partners in

solution. Concerning the solution photodimerization of 1-1c, it is stated, "Nach einiger Zeit

fallen orangefarbene Kristalle aus ... (64%) orangefarbene monokline Kristalle, Schmp.

185 190C, laut 1H-NMR [l-lc:dimer] (19:1)."4 It is unclear whether these mixed

crystals obtained from solution irradiation are of the same crystal modification as the dimer

formed from solid state irradiation or a second form, characteristic of the recrystallized

dimer. The latter is most likely the case, meaning the crystals were, based on the results

reported in this dissertation, not monoclinic, rather orthorhombic. Only in the case of 1-lh

was cis-trans isomerization observed when an acetic acid solution of the salt was

irradiated. However, it should be noted that an excited cis monomer reacting with a

ground state cis monomer would also give the [r-ctt htt dimer] found in the cases of 1-1c

and 1-id above.

The thermal behavior of the monomers and photodimers were analyzed by

differential scanning calorimetry (DSC). The only photodimers that undergo 100%

cycloreversion are 1-4b-d, summarized in Part I of Table 1-2. Because these systems

involve a shift to shorter wavelengths in the absorption spectrum of the photoproduct, they

are called inverse photochromic systems. Since they are thermally reversible, they may be

called thermochromic as well. Hiinig and Hesse4 suggested these systems might have

uses involving thermal information storage and erasure through irradiation. The large
differences in Xmax between the monomer and photodimer, 173 128 nm, was pointed

out as a favorable asset. The monomers 1-lb and 1-id also undergo thermal modifications

that are reversible, as are the melting transitions. The photoactivity of the modified

monomers 1-4b' and 1-4d' was not reported. However, the interconversion of 1-1c and

1-4c was cycled several times without noticeable side reaction. It is interesting to note that

the monomers 1-3c (photostable) and 1-3d (photoactive) also undergo modifications on

heating; however, they are irreversible transitions and such that 1-3d, at 176"C, is rendered

photostable. It was not reported if the photostability of 1-3c changes on heating. Of the

rest of the photodimers investigated, 1-4f and 1-4h simply melt on heating while 1-4e and

1-4i melt then recrystallize, as shown in Part II of Table 1-2. For 1-4e, it was determined

by 1H-NMR that at 170*C the cyclobutane had epimerized to a less strained isomer, for

example, the lower energy r-tct isomer.

The thermal stability of the 1-4c was tested in solution with the following results:

in o-chlorophenol and T=155'C, in 220 minutes, 78% of 1-4c had cycloreverted to 1-1c.

It was concluded that the crystal packing was vital for a fast back reaction at a defined

temperature and since both donar substituents and polar solvents favored the

cycloreversion, a zwitterionic intermediate was indicated.

Table 1-2. Thermal behavior of styrylpyrylium salts and photodimers reported by
Hiinig and Hesse.4

Part I. Photodimers undergoing cycloreversion reactions back to monomer.

photodimer monomer modification of monomer monomer melting

1-4b---> 1-1b(148C) ---> modification 1-1b'(164C) ---> 1-lb' melts (245'C)
1-4c---> 1-lc(101C) ---> 1-1c melts (209"C)
1-4d---> 1-1d(180'C) --->[modification 1-1d'(171'C)] ---> 1-d' melts (211'C)

Part II. Photodimers that recrystallize in a different modification after melting.

photodimer first melting modification of photodimer second melting
1-4e---> melts(157'C)---> crystallizes 1-4e'(165'C)---> 1-4e' melts (219'C)
1-4i---> melts(167'C)---> crystallizes 1-4i'(183'C)---> 1-4i' melts (240'C)


Topochemical vs. nontopochemical processes

The photodimerizations described above fall into the broad class of organic solid

state reactions known as topochemical reactions. G.M.J. Schmidt began his seminal work

on the photodimerization of cinnamic acids,45 which laid the foundation for the

topochemical principle in organic solid state chemistry, in this way:

The analysis of the mechanism of solid state reactions has for some
time now centered around the study of lattice imperfections ... This focus
on lattice irregularities has tended to ignore the major feature of the crystalline
state, namely three-dimensional regularity and its effect on reaction
mechanism, rate, and products. The work in progress here is intended to
analyse the role of lattice geometry and thereby to extend our understanding
of solid state chemistry.

This passage suggests two broad categories of reactions: those whose progress and

outcome are controlled by the three-dimensional regularity of the lattice of the reacting

phase, that is, topochemical reactions, and those that are not controlled by the lattice, but

rather trace their origin and/or propagation to disorder in the lattice, that is,

nontopochemical reactions. Since topochemical reactions are solid state reactions

controlled by the lattice geometry of the reacting species, if a reaction begins or continues

in a region of the crystal where the parent lattice is nonexistent, the reaction is said to

proceed nontopochemically. The parent lattice may become disordered by propagation of

defects in the crystal, which can be due to several causes, including by a reaction in the

ordered lattice or reaction at an incipient defect. Nucleation of the product phase within the

crystal may occur preferentially at defects in the parent lattice or may itself initiate

disordering. Whenever local melting occurs due to product "impurity" in the reactant

lattice, the reaction is no longer topochemical.

One type of crystal defect is a line defect or dislocation and may be caused by an

accumulation of vacancies within the lattice.46 The enhanced reactivity of solids at

dislocations has been attributed to a number of factors including the following.

Dislocations can act as energy traps. Furthermore, the relative stereochemical orientation

of neighboring molecules as well as the conformational freedom of molecules at a defect is

different compared to the regular lattice. Also, heterogeneous nucleation is easier "because

the stresses and strains set up during nucleation can be ameliorated more readily at a


Schmidt described a process that may cause a defect within an ordered lattice,

beginning with a polymerizable monomer stack of molecules A-D regularly spaced:47

Suppose that, on absorption of the activating radiation, the reactive
monomer A ... reacts with its neighbor B to form the reactive dimer AB*.
The covalent link formed between A and B pulls these two monomer units
together and, in so doing, increases the separation between B and its other
neighbor C. If this separation is still not too great to permit further reaction,
formation of the the trimer ABC will open a still wider gap between C and
D, so that eventually the orderly propagation of the polymer chain along the
stack will be arrested.

This explains why some topochemical reactions never reach quantitative

conversion, for example, 1-2f, 1-3d and 1-3f in Table 1-1. In other words, as yet

unreacted partners are pulled out of their reactive conformation by distortion of the lattice

from previous reactions. Reaction may cease at this point or the nontopochemical process

may proceed in the disrupted crystal structure. For example, disordered reactant around

the defect may continue to react by random translational movement to the center of

reactivity. As Wegner describes it, "In many cases reaction proceeds by diffusion of the

reactants to centers of reactivity or by nucleation of the new phase at certain centers of

disorder and, therefore, increased mobility under complete destruction of the parent


One phenomenon in solid state reactions is nucleation of the product phase at an

interface with the parent lattice where there is a coherent phase boundary. This may lead to

oriented growth of the product such that the product lattice and reactant lattice are related by

certain crystallographic equivalencies. One may contrast this with the idea that a daughter

crystal growing from a parent crystal, forms as it would from the melt and that no special

orientation exists. If the reaction proceeds through the bulk of the reactant, this type of

reaction is said to be topotactic. For example, in crystals of the cyclic monomer trioxane,

the molecules are aligned 4.3A apart along a certain crystallographic axis. The ring-

opening polymerization occurs to form crystalline polyoxymethylene with the chains

aligned along the direction of that same monomer axis in the original crystal.47 This

reaction, once thought to be topochemical, has been shown to be nontopochemical. 48

The photodimerization of 9-cvanoanthracene

The product for this photodimerization is not predicted by applying the

topochemical principle to the geometry of the regular crystal lattice, but it may be explained

by considering the geometry of well-defined defect sites within the crystal "where the

structure of the reaction product is predefined by the arrangement of the molecules around

the particular defect."11 As Thomas puts it, "[Defects] could, on topochemical grounds,

favour the 'wrong' product."46 The "wrong" product here refers to a non-topochemical

product. 9-Cyanoanthracene packs in a B-type structure with highly overlapping stacks

and an interplanar distance of 3.5A. The B packing--that is, a mirror symmetry relates the

closest monomers in a stack--should give head-to-head dimers, but head-to-tail dimers are

formed instead. By examining pairs of cleaved crystal surfaces, one of which had been

etched and one of which had been photodimerized, Thomas and coworkers were able to

correlate stacking faults in the crystal with centers of dimerization.49 The

photodimerization of 9-cyanoanthracene was shown to occur preferentially at these defects

in the anthracene crystal. The relative orientation of the monomers at these faults could

then be related to the stereochemistry of the anthracene dimer as shown in Figure 1-2.

However, it has been noted that it may not be the specific geometry that causes the

dimerization reaction to center at the dislocations. In correlating the etch pit surface with

the irradiated surface, the latter surface was examined for certain textures associated with

the 20% increase in crystal volume associated with the dimerization.46 Thus it may be that

these defects serve only to facilitate the crystallization of the photoproduct. Then how may

the head-to-tail dimer form? Later studies revealed the existence of a stress-induced

metastable phase within the regular parent lattice with a shorter head-to-head distance than

in the parent lattice.0 Other anthracenes also show such nontopochemical behavior.51'52

Ilt -.a J a ^

2P ,_ < < Bek

CS-gS- f^K- ~tow

Figure 1-2.46,49 This figure shows how head-to-tail anthracene photodimers might
form near a lattice defect. The closed arrows indicate the orientation of the substituent in
the 9-position on the anthracene. In the regular parent lattice the dimers would be oriented
so that nearest neighbors would have the substituents pointing in the same direction. But
here, one layer of anthracene molecules has slid relative to the parent lattice as indicated by
the dashed arrow. The open arrows indicate the stacks that would produce the photodimers
with subsitutents on opposite sides.

The concept of defect-controlled reactions is important when considering the final

outcome of phase transformations in organic solid state reactions, since phase

transformations, such as nucleation or recrystallization, are often initiated by crystal


Schmidt's topochemical principle

Lattice control. As described above, organic solid state reactions can be broadly

classified as topochemical or nontopochemical. The purpose of the preceding paragraphs

was to provide contrasting examples for the type of chemistry described in the balance of

this section, that is, topochemistry, in order to better illustrate this concept. In

topochemical reactions the chemical transformation is directed by the parent lattice

throughout the transformation. Thus there is a direct transition from the reactant to product

"without destruction of the crystal lattice and without formation of non-crystalline

intermediates.""1 Whereas in nontopochemical reactions there is random diffusion of

molecules to a center of reactivity, in the ideal topochemical reactions, the center of mass of

each molecule is stationary; therefore the molecules do not lose their location in the lattice.

For example, in the cases of topochemical polymerizations, the molecules are often

pictured as swiveling on their lattice positions to come within bonding distance of one

another. This notion was first described by Hirshfeld and Schmidt47 and first realized in

the laboratories of Wegner and coworkers for the polymerization of diacetylenes.16

This lattice-point swivel motion is shown in Figure 1-3a, a schematic illustration of

the polymerization of diacetylenes. However, this type of motion also applies to the

polymerization of distyrylpyrazine as shown in Figure 1-3b. The two remaining olefins in

the dimer pictured in the central stack of Figure 1-3b are now closer to their reacting

partners than they were in the unreacted monomer stack. Also, in dimerization reactions,

the molecule is seen to pucker. That is, the center moves in one direction while the edges

move in the opposite direction, and thus the center of mass is kept constant. Figure 1-3b

demonstrates how this is the case for [2+2] photopolymerizations as well.

Though the concept had been previously described, the first systematic

investigations that offered evidence of topochemical reactions in molecular crystals were

conducted by G.M.J. Schmidt and coworkers at the Weizmann Institute in Israel. These

investigations were reported in the Journal of the Chemical Society in a seminal

series53'45,54 that has since been republished in part with a selection of subsequent reports

and review articles by Schmidt and his collaborators.55 Jack Dunitz writes in a

retrospective on Schmidt's contributions to science,56 that up until the early 1960s, x-ray

crystallography had been used primarily to determine molecular geometry, for example,

interatomic distances and angles, or characteristic separations between nonbonded atoms.

But for most organic chemists, "a molecule in a crystal was a dead molecule" and organic

solid state reactions were largely regarded as a nuisance. Schmidt's genius was the

"synthesis of crystallography and chemistry."56 He looked beyond the individual

molecule to the way in which molecules approached one another in the crystal. He shifted

the focus from molecular geometry to molecular packing. Once the correlation between

crystal structure and solid state reactivity had been made, "Everything fell into place; order

had been created out of chaos."56

Figure 1-3a. Figure 1-3b.

Figure 1-3. Topochemical polymerization: stacks of difunctional monomers may
react in a regularly ordered crystal lattice, without the aid of defects, by rotating on their
lattice rotates. The rotation of one molecule brings its reactive ends into contact with two
others and so on in a zipper-like mechanism.
a) schematic diagram of diacetylene chain polymerization57
b) illustration of [2+2] step polymerization of DSP37

Cinnamic-acid investigations of G.M.J. Schmidt. The subject of Schmidt's most

heralded studies is the [2+2] photodimerizations of cinnamic acids and was initiated by a

dispute in the literature. In 1964, trans-cinnamic acid had been long known as dimorphic;
that is, it could exist in two types of crystalline forms, a and B, and that the a form gave

truxillic acid, the head-to-tail dimer. The argument centered over the stereochemistry of the

cyclobutane dimer formed from the B modification of trans-cinnamic acid which seemed to

give mixtures of both truxillic and truxinic acid, the head-to-head dimer.

In the first part of the series,45 Schmidt and Cohen state the thesis and three related

tenets supported by the studies reported in the following parts of the work and other

examples from the literature. This thesis, known as Schmidt's topochemical principle,

states that "reaction in the solid state occurs with a minimum amount of atomic or

molecular movement." The tenets are45

(1) chemically closely-related compounds show significant
differences in chemical behaviour in the solid state; (2) a given compound
reacts differently in the solid and dispersed phases; (3) polymorphic
modifications of a given compound show significant differences in chemical

Tenets (1) and (3) have found expression in the research reported in this dissertation: (1)

in chapter 4 in comparison of the photoactivities of various counteranion-substituted

styrylpyrylium salts and (3) in chapter 3 in comparison of the thermal behavior of

recrystallized vs. as-dimerized styrylpyrylium triflate photodimer, 1-4c.

In the second part of the series,54 Schmidt and coworkers resolve the controversy

surrounding the photochemistry of the B form of cinnamic acid by showing that there is a
phase transition from the B to the a around room temperature and that when the B

modification is irradiated at temperatures that exclude this transition, truxinic acid is

formed exclusively. Otherwise, irradiation of the B form gives a mixture of photodimers,
the truxillic acid arising from the a modification. These transformations are outlined in

Figure 1-4. (Almost 30 years later, the kinetics58'59 and mechanism of the B to a

transformation in trans-cinnamic acid continues to be investigated and has been shown to

be topotactic.60) In the second part of Schmidt's classic work, the preparation of crystal

modifications of thirty-six substituted cinnamic acids is also reported and their

photoproducts, if any, identified.

trans-cinnamic acid
monomr modificaion phaodnier
a- modification a-truxillic acid
cemrosymmetric dimer (1)

B-modification -h- B-truxinic acid
mirror dimer W(r

Figure 1-4. Crystalline modifications of trans-cinnamic acid and the
photoproducts arising from each. Note that these two forms may be distinguished
visually, the a modification being in the form of squarish platelets and the B modification
being in a longer, narrower lath shape.

The irradiation of each of the acids either produced only truxillic and/or truxinic-

type photodimers, or the crystal was photostable. Detailed experimental data and

references are given for the preparation and irradiation of the cinnamic acids with the "plea

that workers in solid state chemistry specify in sufficient detail the reacting crystalline

phase.. ."45 Once the identity and exclusivity of the photoproducts of dimerization had

been decisively established, Schmidt, in the third and culminating part of the series,53

related the molecular symmetry of the photoproducts with the crystallographic symmetry of

the reacting monomers. Crystallographic data on twenty-eight cinnamic acids are

presented in this part of the series.54 Analysis of their cell parameters compared with

crystal structures of five selected compounds revealed that among all the acids investigated,

three types of crystal modifications exist, a, B and 7, and that six of the acids are

dimorphic and one occurs in all three modifications. The five crystal structures presented

are one B type, p-chloro-trans-cinnamic acid; two y types, a-bromo-trans-cinnamic acid

and m-chloro-trans-cinnamic acid; and two a types, a-trans-cinnamic acid and trans-

crystal modification

Ph COOH ,COOH-=-Ph truxinic acid Ph
Ph -COOH.. COOH-c-Ph mirror symmetry H
Ph -COOH"COOH-c==-Ph

truxillic acid ph _P h

i centrosymmetric COOH


Ph COOHCOOH- =- Phphotostable
Ph COOH .* COOH-=--Ph

Figure 1-5. The crystal modifications of cinnamic acids have been identified with
their lattice symmetries and in turn with the stereochemical identity of their photoproducts
or with their photostability.61

cinnamide. Figure 1-5 shows that each type is built from a unit of two monomers,

hydrogen-bonded to one another and approximately in the same plane. These units form

parallel stacks that bring monomers within certain distances and relative orientations to one

another. In the B form the monomers in a stack are highly overlapped compared to the

other two modifications and the distance between reactive centers, which are related by

mirror symmetry, is approximately 4A. In the y form, the decreased overlap within the

stack results in a distance between nearest double bond contacts of 5A. In the a form, the

overlap offset is so great, that the closest monomer pairs in a reactive orientation are not in

the same stack, but in neighboring stacks. Thus, while neighbors within a stack are related

by mirror symmetry and are over 5.5A apart, the reacting pairs are only 4A apart and are

related to one another by a center of symmetry.

Thus it is easily seen that in the B lattice, the most easily formed dimer--the dimer

formed with the least atomic or molecular motion--is the head-to-head dimer. However, in
the a lattice the head-to-rail dimer is the dimer formed with the least lattice disturbance. In

all twenty-eight cases investigated by Schmidt, the lattice symmetry matched the

stereochemistry of the products or predicted the photostability of the lattice in the cases
where the y modification occurred.

Schmidt went on to investigate the photoreactions of heterocyclic analogs of

cinnamic acid,62-65 providing examples of the first tenet of the topochemical principle: that

chemically closely related compounds could show significantly different behavior in the

solid state. Thus, the lattice controls not only the reactivity of the compound

stereochemistry of any products, but in some cases, even the reaction type. For the first

time, it was demonstrated that the same crystalline phase could house two competing

reactions, in this case, dimerization and polymerization:

thienyl acrylic acid B -------> dimer

y -------> oligomer (ring participation)

furylacrylic acid B -------> dimer + oligomer (ring participation)

Spectroscopic evidence from the product oligomers indicate that the heteroaromatic rings


Beyond Cinnamic Acids

Schmidt's first work on the [2+2] photodimerization of cinnamic acids described

above allowed topochemical reactions to be identified as a class of organic solid state

reactions. Since then numerous other solid state reactions, both thermal and

photochemical, including unimolecular, bimolecular and polymeric, have also been shown

to be topochemical.9-16.66 The study of these has led to refinements of the original

principle and a deeper understanding of the mechanisms involved in topochemical

processes. Though there are many examples of thermallyinitiated solid state reactions in

crystals, most notably the thermal polymerization of diacetylenes and the thermochromism

of various classes of chromophores, it should be noted that photochemical transformations

dominate investigations into topochemical processes. Scores of examples of [2+2]

photodimerizations and polymerizations exist in the literature,66,13 and the most deeply
investigated solid state reaction is the y-radiation induced polymerization of diacetylenes.

Other chromophores investigated by Schmidt include butadiene and hexatriene

derivatives,67 p-quinones, chalcones,68-70 and the eight-center photodimerization of a 3-

keto-1,4-pentadiene derivatives.71 Schmidt and collaborators went on to investigate cis-

trans isomerization64 and photochromism in the solid state.7264

Reviewing solid state organic photochemistry,Schmidt outlined its progress as

having begun with the "phase of the topochemical principle," which was being followed

by the "phase of the locus of the reaction".73 On a microscopic scale, the locus refers to

the conformation of the photoreactants as it relates to the structure of the rigid matrix

surrounding them, that is, the "reaction cavity." On a macroscopic scale, the question of

the reaction locus involves the issues of crystal texture, for example, dislocations and

phase boundaries, on the outcome of the solid state reaction, as well as energy transfer

within a crystal. Thus the question, is the reaction defect-controlled or does it proceed in

the bulk of the crystal?

Schmidt called the third phase the "phase of crystal engineering."73 Along these

lines investigations into the photochemistry of solid solutions of photoactive species were

initiated.74'75 The first asymmetric syntheses through lattice-controlled reactions in chiral

crystals were also demonstrated.76-78 Also, "steering groups" have been investigated in

efforts to develop ways to control crystal packing, an intensely sought and highly prized

goal of solid state chemists. A text on various aspects of crystal engineering has recently

been published.79 Under the category of crystal engineering, one might include host-guest

topochemistry occurring in channel inclusion complexes such as the deoxycholid acid

example in Table 3-1, or matrix polymerization also represented in Table 3-1 with the

butadiene polymerization in the layered perovskite structure.

The literature of topochemistry

Most of the references cited above have been gathered into a collection of selected

papers by Schmidt and his collaborators, Solid State Photochemistry,55 that explores the

relationship between the geometrical properties of lattices of molecular crystals and the

photochemistry of those molecules. Besides reports of original research, it contains four

review articles, one on solid state polymerization,47 one on asymmetric synthesis78 and

two general reviews on solid state photochemistry,64 one focusing on dimerizations.73

Two other collections of papers centered involving the topic of solid state

photochemistry have also recently been produced: Photochemistry in Organized and

Constrained Media8s and Organic Solid State Chemistry.81 Pertinent to this work, the first

reference contains a review article on bimolecular photoreactions in crystals,13 a reprise of

a larger review article encompassing both unimolecular and bimolecular photochemical

reactions of organic crystals.66 Besides covering some of the prominent fields of research

in topochemistry (such as the photodimerization of benzylidenecyclopentanones,

intramolecular hydrogen abstraction in the solid state, "four-center" photopolymerization of

distyrylpyrazine, and host-guest topochemistry), the latter reference emphasizes the

interdisciplinary nature of organic solid state chemistry by including articles in the areas of

theoretical chemistry and solid state physics. Many short overviews of organic solid state

chemistry exist in texts devoted to solid state chemistry or to photochemistry, but a short,

comprehensive introduction to organic solid state chemistry is a review article by M. D.

Cohen and Bernhard S. Green,61 collaborators of Schmidt. The first single-crystal-to-

single-crystal transformation of an organic substance reported in the literature is the

polymerization of a diacetylene,16,82 which contributed much to the understanding of

product phase evolution in the solid state. Thus introductions to various issues involving

the mechanisms of solid state polymerizations by Gerhard Wegner are important to

mention here.6'7'11'48'57,83 Topochemical polymerizations are dominated by the two

examples in Figure 1-3: diacetylene photopolymerization84 and "four-center"

photopolymerizations. Hasegawa, the principal investigator of the latter type, has written

several reviews.934'36'37 Phonon spectroscopy has been used as a tool to investigate the

mechanism of product phase evolution in organic solid state reactions, most notably the

controversial DSP "four-center" reaction. A review by Prasad of investigations using this

technique may be found in Organic Solid State Chemistry, mentioned above.8

Polymerizations of unsaturated compounds by photocycloaddition reactions, primarily

[2+2] reactions, have been extensively reviewed by Dilling.35

In general, very few thorough kinetic studies exist for reactions in the solid state,

with the exception of diacetylene polymerization, easily the most well-investigated of

organic solid state reactions. Here Baughman has advanced a model that is based on the

idea that the kinetics of homogeneously polymerizing crystals (single phase) is dependent

on the strain induced within the crystal due to polymer-monomer lattice mismatch.85 In the

case of [2+2] photodimerizations, Hasegawa has proposed a reaction model for the

photoreaction of olefin crystals based on the temperature-dependent deviation of the two

olefin bonds from the optimal positions for the reaction.86 The only other method applied

to the kinetics of topochemical reactions in the literature is that refined by Hancock, Sharp

and Brindley for heterogeneous processes in barium and magnesium salts.87'88 Heyes has

used this method to explain the time vs. conversion curves of the [2+2] photodimerizations

of cinnamic acid, its o-methoxy derivative, the intramolecular [2+2] reactions of a

benzoquinone and cyclopentadiene adduct, the solid state racemization of a coboxalime
complex and the B to a phase transition of cinnamic acid.59 In this method, slopes from

In(ln(l-conversion)) vs. In (time) are used to identify reaction mechanisms such as

diffusion-control, nucleation and growth and phase-boundary control. Time vs.

conversion curves have been reported for the [2+2] photodimerization of azastilbenes and

coumarins.2'89 Quantum yield studies90 and a multiplicity study91 have been reported for

a few solid state intramolecular photoreactions.


Figure 1-6. The orbital overlap is defined as the distance T-T', which is the
distance between points 1.8 A, the van der Waals radius of carbon, along an axis passing
through the carbon atom, normal to the double bond plane.9

Geometric models

Since Schmidt's original formulation of the topochemical principle, hundreds of

examples of topochemical reactions have been investigated. Certain molecular crystals

expected to be reactive on the basis of their lattice geometry are not reactive, while many

expected to be stable are, in fact, reactive. These investigations have led to additional

efforts to explain the photoreactivity and selectivity of organic solid state reactions. An

informal rule for [2+2] photodimerizations is that the species must be within 4.2 A for a

reaction to be possible. Kearsley refines this parameter by suggesting a measure of the

orbital overlap between reacting orbitals as the distances between the apices of the reacting

atomic orbital lobes of carbon as shown in Figure 1-6.92 The overlap, SUM, is defined as

the sum of these two distances for a pair of double bonds and, along with the lateral

displacement of the double bonds, is calculated for almost 70 various photoactive and

photostable crystals. Kearsley also presents projections of carbon p-orbital contours

(based on electron density of H atomic orbitals) for nine compounds, an example of which

is shown in Figure 1-7 for benzyl benzylidenecyclopentanone (BBCP) and a napthalenone

derivative (NAPTH). It is clear why the BBCP is photoreactive but it was left to the

development of the concept of the "reaction cavity" by Cohen, described below, to explain

the photostability of NAPTH (see Table 1-3).

BBCP: reactive


S"X /= X-COOMe

NAPTH: unreactive

Figure 1-7. p-orbital contours for BBCP and NAPTH double bonds (enone). The
arc in the left-hand projection is a cross-section of a 4.2 A spherical shell originating from
the midpoint of the lower double bond, whereas the solid circle on the right is the
intersection of this shell with the plane of conjugation of the other double bond. The
dashed lines are the full van der Waal radii of the indicated portion of the molecule whereas
the partial circles represent a 1.75 A cross-section of the p-orbitals on the conjugated
carbon atoms.

Another method of elaborating on relative molecular orientations in a lattice

measures the angles swept out by a parallelogram constructed from the planes of two

neighboring double bonds, 82 and 03, shown in Figure 1-8.89 If the bonds are skewed

then 81 is considered. Though both of these methods, based on geometrical relationships

between potentially reacting pairs of molecules in the lattice, explain certain anomalies not

consistent with the simple 4.2 A rule, they fail to take into consideration the steric effect of

the neighboring molecules on the potential of the reaction to proceed as well as the overall

flexibility of the lattice. This has been addressed by Cohen with the concept of the reaction

cavity, as discussed below.

Figure 1-8. Angles considered to predict reactivity of double bonds in a crystal
lattice. Ideal values are 1=0", 02=90', 03=180".

A highly predictive geometrical analysis of a series of topochemical reactions was that

performed for the polymerization of diacetylenes.82 Figure 1-9 and Figure 1-10 show

how the diacetylene monomers are stacked according to repeat distance, d, and tilted with

respect to the axis of the monomer stack, (p, so that the minimum van der Waals distance,

02 fit.


Rv, is maintained. In order for the end carbons of the diacetylene unit to approach within

bonding distance, the monomer must rotate with the reacting carbons starting from a
separation distance of R, as determined by d and (p. The closed circles in Figure 1-10

represent reactive diacetylenes and the open circles, the unreactive monomer. Clearly, d
and (p must be such that R falls within 4 A or less for a structure to be reactive. The one

closed circle beyond the 4 A border is for the diacetylene DCH, (Cz-CH2-C=-C-C-C-

CH2-Cz, Cz=carbazolyl) discussed in chapter 3. Figure 1-10 emphasizes the first tenet of

Schmidt's topochemical principle as described by Enkelmann: "The reactivity is controlled

by the monomer packing and not by the chemical nature of the substituents."82

4.91 A

Figure 1-9. Geometrical parameters d and (p in the topochemical polymerization of
a diacetylene monomer stack.82


906 o 0 0 o


S6 8-

Figure 1-10. d vs p for various reactive (.) and unreactive (o) diacetylenes. Rv is
the van der Waals distance.82

Reaction cavity concept

Cohen reasoned that, if a reaction was controlled by the parent lattice, "at least up

to the transition state, the reaction complex maintains close contact with its environment in

the crystal."93'94 The contours of the inner boundaries of this environment form a space

of fixed size and shape in which the reaction may occur. This is the reaction cavity or

cage. Cohen restates the topochemical principle as this: distortion of the size and shape of

the reaction cavity due to the rotations and translations involved in bond formation and

breaking are minimized in a lattice-controlled reaction. Formation of voids within the

cavity are disfavored since they would lead to a large decrease in attractive forces that had

existed among the reacting molecules and their environment, while extrusions from the

cavity are disfavored due to an increase in repulsive forces that must arise. Thus, the

close-packed environment in a crystal resists large shape changes in the packed units. This

is illustrated in Figure 1-11.

transition state cavities



energetically\ /energentically
favorable / unfavorable

original cavity shape

Figure 1-11. The original reaction cavity undergoes minimum distortion in a
favorable reaction. However, creation of extrusions and voids is energetically

This concept has been applied to a number of examples of solid state reactions to

rationalize their specificity when two products are feasible and to explain the photostability

of favorably aligned molecules and the photoreactivity of unfavorably aligned


Computational investigations on the crystal structure of potentially photoactive

organic solids have been reported that take into consideration the influence of the

environment on the reactivity of molecules in crystals. One investigation calculated the

relative energies associated with the steric compression involved for hybridization of two

potentially reactive centers in a unimolecular H-abstraction to explain the specificity of the

product formation.95'96'66 Calculations involving lattice energies have also predicted

product specificity for dimerization reactions as well as photostability despite a favorable

geometric arrangement.95-97.66'13 Table 1-3 contrasts strictly geometrical parameters with

calculations considering reaction cavity principles for predictive value. Another type of


calculation involves computing packing density maps of crystal structures that allow the

identification of void space.98 These maps may explain why some molecules with

unfavorable geometric orientation, in contrast to above, nevertheless do react because of an

enhanced lattice flexibility.

Table 1-3. Calculations to predict topochemistry based on reaction environment
where simple geometrical considerations fail.


Unimolecular H-abstraction95,96'66
GEOMETRY C...H steric compression
C1...H 2.74 A 0 kcal / mole yes
C2...H 2.70 A 12 kcal / mole no

Olefins center to center rise in lattice
double bond ENERGY
7-chlorocoumarin97.66 distance
translational pairs 4.45 A 160 kcal / mole yes
centrosymmetric pairs 4.12 A 18,000 kcal / mole no


napthalenone derivative95,96.66,97
01=0', 02=82%, 03=154* 3.79 A 1500 kcal / mole no
7-methoxycoumarin97'66 3.83 A 160 kcal / mole yes
01=68', 02=110', 03=160'

Related to the idea of reaction cavity distortion is the concept of dynamical

preformation advanced by Craig and coworkers.13,66 It is well accepted that the molecular

geometry of the excited state may be radically different from that in the ground state, and

this must also be considered in the context of the reaction cavity principle.

Appearance of the Product Phase: Homogeneous vs. Heterogeneous Mechanisms

The topochemical principle organizes the issues involving lattice packing, reactivity

and product identity in solid state reactions. However, the mechanism by which the

reactant phase transforms into the product phase lacked a cohesive explanation. The

concept of the shape of a reaction cavity in determining the topochemistry of a solid state

reaction applies to transition state geometries but may also inform the discussion on how

the product builds within the crystal. But up until the discovery of the first single-crystal-

to-single-crystal transformation of a molecular crystal, only speculation existed on the

course of the phase transformation from reactant to product. With the discovery of the

solid state topochemical reaction of diacetylene monomer crystals to yield single crystals of

polymer, new light was shed on the issue of phase transformations in the organic solid


Conceptual developments

Structural mimicry. Schmidt's original proposal on the appearance of the product

phase provides a rough outline that accommodates subsequent observations on the

mechanisms of phase transformations of topochemical reactions:53

It is reasonable to assume that the dimer goes into solid solution in the
lattice of the monomer and that as the dimer concentration rises the solubility
limit is exceeded and the new phase precipitated. However, since the
solubility of dimer in the monomer lattice and the stability of this solid
solution must vary from compound to compound ... no generalisations are
as yet possible.

Schmidt observed "a wide variety of effects," including that in X-ray powder diffraction

studies, powder lines of the product phase appear over a wide range of product

concentrations for various substances. Schmidt also observed that some of the 1 acids

"crystallise during the course of reaction in a metastable form not identical with the

modification obtained by crystallisation from solvent." As for single-crystal-to-single-

crystal transformations, Schmidt proposed that53

the absence of orientation [of the product phase] is due to too large a
difference between the crystal structures of the monomers and dimers and
that, where two structures are sufficiently closely related... direct single-
crystal transitions are in fact possible.

Ten years later when introducing the concept of the reaction cavity in organic solid

state reactions, Cohen refined the proposal that organic solid state reactions begin with the

formation of a solid solution of reactant and product with the idea that due to the

geometrical constraints of the cavity, the product molecules are not only a substitutional

solute in the parent crystal but that the conformation of the product molecules was that

"which best fits them to solid-solubility in the reactant crystal."'5394,93 Therefore, solid

solubility was dependent on conformational mimicry. Moreover, he stated that this will be

most readily observed when the product molecule is "intrinsically readily deformable."

Along these lines, the remarkable case of the single-crystal-to-single-crystal conversion of

diacetylene monomer crystals to polymer crystals was informative. For this case, Cohen

notes that "complete conversion can be achieved without phase change since the polymer

formed is crystallographically isomorphous with the starting monomer."94,93

The role of van der Waals contacts among side groups. Van der Waals contacts

among bulky side groups determine the packing of a monomer in the solid state. The

diacetylene examples illustrate that these groups stabilize the packing of the reacting

monomer throughout the transformation to photoproduct. Since the side groups do not

react, their final conformation in the daughter lattice is similar to that in the parent lattice.

In this way, the side groups control the production of a structure isomorphous with the

incipient structure. Figure 1-12 demonstrates this concept by superimposing projections of

monomer and polymer crystal structures of bis(p-toluene sulfonate of dodeca-5,7-diyne-

1,12-diol) PTS-12. The side groups move very little in the transformation from reactant to

product. It was noted that although these ideas were true for a type of polymerization

reaction, for reactions producing low-molecular weight products such as

photodimerizations, "lattice control of the conformation of the product is less clearly

established.'94,93 A single-crystal-to-single-crystal photodimerization has also now been

achieved. The full crystal structure of the photoproduct superimposed on that of the

reactant is also seen in Figure 1-12.12

Solid solutions. The terms "homogeneous" and "heterogeneous" were established

to describe the two types of mechanisms by which topochemical polymerizations occurred

and illustrates them as shown in Figure 1-13.83'1148,57The figure on the right illustrates

polymer-formation proceeding homogeneously inside the monomer crystal starting at

randomly distributed points where polymerization has been initiated. The result is that the

product is dispersed in the undisturbed monomer matrix so that a solid solution of product

and reactant is obtained. Thus coherence between all parts of the crystal, and thus the

single-crystal texture, is retained. This is believed to be the mechanism by which several

diacetylenes polymerize to give macroscopic single crystals of extended chain polymers.

In fact, the polymerization of a diacetylene monomer is the first documented example of a

topochemical reaction producing single crystals of product

Work done by Braun, et al, on the photopolymerization of distyrylpyrazine

(DSP)6,7.10 revealed that its mechanism of polymerization--homogeneous or

heterogeneous--could be controlled with irradiation conditions. When the UV-VIS

photoactive DSP crystals are irradiated with light of wavelengths for which the

chromophore has a large extinction coefficient, most of the light is absorbed at the crystal

surface, according to Beer's Law. Thus product formation is isolated at the crystal surface

and the interior of the crystal, which receives little intensity, remains unreacted. The

product eventually reaches its limit of solubility at the crystal surface and precipitates out

It is the dimensional mismatch between the parent phase and new phase which cause

incoherent phase boundaries and disintegration of the crystal into polycrystalline

powder.53,100,11,34,36 Most solid state photoreactions proceed by such a heterogeneous

mechanism, illustrated in Figure 1-13 on the left.

However, when DSP crystals are irradiated with wavelengths of light for which the

chromophore has a small extinction coefficient, for example, in the absorption tail of the

monomer, then the light is absorbed relatively evenly throughout the bulk of the crystal and

thus product formation is also randomly distributed throughout the crystal, just as in the

diacetylene polymerization. This random product formation results in a single-phase

transformation from DSP monomer to oligomer and thus the single-crystal texture of the

monomer is preserved.

Conformation of side-groups in solid solutions and structural mimicry. This

dissertation provides the first glimpse of the conformational relationships between reactant

and product molecules in a crystal in the process of undergoing a topochemical reaction.

This is done through the first x-ray crystal structure analysis of a solid solution of reactant

and its photoproduct formed through a topochemical reaction.

Here the monomer and dimer structures--for the photodimerization of the

styrylpyrylium triflate 1-lc and a-trans-cinnamic acid--though having different

equilibrium conformations, are seen to be constrained to the same lattice positions to form

a solid solution. The dimer mimics the monomer at low conversions, whereas the

monomer appears to mimic the dimer at high conversions. Because the degree of

conversion is kept uniform over the entire crystal and over the course of the reaction

through, a single-crystal-to-single-crystal transformation is observed. Throughout the

dimensions of the macroscopic single crystal, the monomer and dimer are constrained to

resemble one another through the formation of a solid solution. This is achieved through

random product formation throughout the crystal via absorption-tail irradiation. Thus the

monomer and dimer remain in a single phase throughout the conversion.

However, the photodimerization of BBCP in the solid state is activated with UV-

VIS light and occurs with retention of the single-crystal character without specific attention

to the irradiation conditions. Assuming the product formation is caused with strongly

absorbed wavelengths and is therefore heterogeneous, why is crystal disintegration not

observed? In this case, the equillibrium conformation of the monomer and the as-

dimerized conformation of the dimer are so similar, that is, they mimic one another so

strongly, that no phase separation occurs apparently due to an insufficient difference in

their lattice dimensions. On the other hand, in the case of DSP, the styrylpyrylium triflate

1-1c and cinnamic acid, the equillibrium conformation of the monomer and the as-

dimerized conformation of the dimer are different enough that they require solid solutions

to bridge the gap from reactant to product single crystal. In these solid solutions monomer

and dimer mimic one another in intermediate conformations that are dependent on the

degree of conversion.

Examples of homogeneous and heterogeneous mechanisms

Figure 1-14 summarizes some of the courses documented for the emergence of the

product phase in solid state reactions. First, as described originally by Schmidt, product

accumulates in the parent crystal in solid solution with the reactant.53,64,94.93 This is path

Ia, in Figure 1-14. As Schmidt and Cohen suggest, this produces a substitutional mixed
crystal of reactant and product, R P Two possible fates await this solid solution as the

reaction proceeds and the product concentration increases: continuation of a single-phase

transformation, paths I, which are extremely rare, or phase separation, paths II.

Under the category of single-phase transformations, consider the case where the product

remains in the solid solution with the reactant until quantitative conversion is reached

following the ideal homogeneous mechanism first realized in the laboratories of Wegner

and coworkers for the polymerization of the diacetylene 1,6-bis(p-toluene sulfonate) of

2,4-hexadiyne diol (PTS) and the oligomerization of DSP.16'99'6'7 Ia. A characteristic of

the PTS polymerization is the smooth and continuous change in cell parameters from

monomer to polymer phase. This is a result of the solid solution of monomer and

polymer. The lattice changes to accommodate the gradual increase in concentration of the

polymer and decrease in monomer concentration. Other single-crystal-to-single-crystal

transformations that have been documented to occur with a smooth, continuous change in

lattice parameters are the BBCP photodimerization12 and the racemization of a coboxalime

complex.18 Cell parameter vs. conversion plots for PTS polymerization and BBCP

photodimerization are shown in Figure 1-15.

Before quantitative conversion is reached, however, at some critical concentration,

a single-crystal-to-single-crystal structural phase transition may take place, Ib, after which

the homogeneous topochemical reaction continues, Ia. This phenomenon has been

Figure 1-12. Superimposed crystal structures of monomer and photoproduct for
the single-crystal-to-single-crystal transformations of BBCP photodimerization (top) and
PTS polymerization (bottom).

-- -. -_ -', ,- _
'-->S^ ^SSS y ; i ; 31 f
h | ;h
ii- i i SS S; ;1^
;;-:;:::=: === =:;;==

Figure 1-13. A heterogeneous mechanism of topochemical polymerization is
pictured on the left, a homogeneous mechanism on the right. In the heterogeneous
mechanism, nucleii of the product phase are formed, whereas in the homogeneous case, the
product is formed randomly and a solid solution of monomer and polymer is
maintained." 183,99


documented for the polymerization of two diacetylene monomers, DCH (Cz-CH2-C=C-

C=C-CH2-Cz, Cz=carbazolyl) and PTS-12.82

The category of phase separation is more complex. Any type of product phase

separation, whether this second phase emerges as crystalline or amorphous, has the

potential to destroy the parent crystal and lead the reaction into a non-topochemical mode

by introducing "defects" or "impurities" into the parent crystal and through this, destroying

it Then the reaction proceeds as if in a melt, that is, by random diffusion of reactants.

However, assuming the parent crystal is not destroyed, the product phase may

nucleate in a certain crystalline modification, and proceed with coherent contact between

the parent crystal phase and product phase. Or the phase boundaries may be incoherent

due to a mismatch in the dimensions of parent and product phases. Build-up of product

such that phase separation occurs across incoherent phase boundaries is the dominant

mechanism of product formation in topochemical reactions documented thus far. This is

path IIa in Figure 1-14 and is the established mechanism for most topochemical reactions

based on [2+2] photodimerizations which are induced by strongly absorbed irradiation and

occur to yield the product as a polycrystalline powder.

It must be emphasized that there is an intimate relationship between paths Ia and IIa

in Figure 1-14 for topochemical reactions induced with light in the ultraviolet-visible

regions of the spectrum. For example, it is possible to switch from the mechanism of path

Ia to the mechanism of path Ia with proper choice of irradiation conditions as described

for the oligomerization of DSP and in the studies presented in this dissertation in chapters 3

and 4. How does changing the mechanism change the product identity? To put it another

way, how do the as-dimerized products from a homogeneous vs. heterogeneous

mechanism compare? Up until now, no comparisons have been possible for topochemical

photodimerizations. For the cases reported in this dissertation, however, it has been

established that the photodimer of the styrylpyrylium triflate, 1-4c, has the same metastable

modification whether produced through a homogeneous or heterogeneous mechanism.

I O R, P

I. single-phase II. phase-separation

Ib. la.

single-crystal-to-crystal nuc ion
structuralphase transitior proct p
topochemicalformation of O product phase through solid R< P
0 + r
Product Phase
PTS polymerization I
DCH and PTS-2 DSP oligomerization
by tail irradiation amorphous--
A rpy zation BBCP photodimerization nonpohemical
Acridizminium Cobaloxime complex
salt photodimerization racemization

crystalline crystal


incoherent coherent ic.
boundaries boundaries

"4-center" polymerizations Ni coordination complex
Cinnamic acid polymerization
photodimerizations Trioxane polymerization -
Most other topochemical
reactions I Ib. amorphous or crystalline
I product phase produced
p an example of this type
A would be the
of m-PDA Me

Figure 1-14. A summary of some of the ways in which a product phase may
emerge in a solid state reaction. The main mechanisms are homogeneous, Ia, and
heterogeneous, IIa.


SE-- 2915- ..91S

Figure 1-15. Plots of cell parameters vs. conversion for the single-crystal-to-
single-crystal transformations of BBCP to photodimer (top)12 and the diacetylene
monomer, PTS to photopolymer (bottom).82

However, the styrylpyrylium photodimer, 1-4c, recrystallized from ethanol has a different

modification. For cinnamic acid, the homogeneously vs. heterogeneously produced

photodimers are different as determined X-ray powder diffraction measurements. The

crystal structure of the recrystallized cinnamic acid photodimer, that is, truxillic acid, is in


If the product phase emerges as crystalline or amorphous, but in coherent contact

with the parent crystal, then a topotactic reaction is possible, path lib. Figure 1-14 shows

that although topochemical control may often lead naturally to topotaxy, the phenomena of

topotaxy can occur through an amorphous intermediate and thus have nothing to do with

topochemistry. Examples of topotactic reactions in the literature are those reported for

trioxane'01 and Ni coordination complex polymerizations.102

An example of a nontopochemical reaction, path IIc, in which an amorphous

polymer is produced from crystalline monomer is that of [2+2] photopolymerization of the

meta-substituted methyl ester of phenylenediacrylic acid, m-PDA Me. For this reaction,

Hasegawa reports,34

At the initial stage of photoirradiation, the molecules...react topochemically to
form the corresponding dimer having a cyclobutane ring with a mirror symmetry.
Then, the subsequent reaction between the dimer and its neighbor in the destroyed
crystal lattice results in an amorphous oligomer exhibiting more than two kinds of
cyclobutane structures.

Figure 1-14 organizes the two principle paths established for the mechanisms of

product phase formation in topochemical reactions. The homogeneous mechanism is a

single-phase transformation, whereas, the heterogeneous mechanism involves phase

separation. In some cases of topochemical photoreactions, these mechanisms can be

controlled with irradiation conditions. A few other solid state processes, bearing on the

emergence of the product phase, have also been inserted into the scheme as they relate to

these two main mechanisms.

Summary of Results Presented in this Dissertation

In chapter 3, seven crystal structures are reported: the styrylpyrylium triflate

monomer, 1-1c, at a higher and lower temperature, the styrylpyrylium triflate photodimer,

1-4c, the monomer crystals obtained when the photodimer is thermally treated, the

recrystallized photodimer, and two single crystals of monomer and dimer formed by the

partial reaction of monomer through a single-crystal-to-single-crystal photodimerization.

The crystal structures of monomer and as-dimerized dimer confirmed topochemical control

of the reaction but also revealed large changes in the side-group conformations.

Consequently, the crystal structures of the monomer-dimer solid solutions formed from the

irradiation of a monomer crystal were of keen interest. They revealed that the side groups

occupied the same position in the lattice, whether they were attached to a product or

reactant molecule. This results in bond angle deformation involving the ring sp2 carbon

linking the double bond or cyclobutane to the rest of the side-ring. The two crystal

structures done on crystals at intermediate conversions catch the side groups in two stages

of a smooth rotation from monomer to product structure. Photos of the crystals in the

process of photodimerizing and thermally reversing show no second phase emerging in the

crystals during either process. Furthermore, the crystal structure of the monomer crystal

obtained from thermal reversion was identical to that of the original and provided the first

documentation of a single-crystal-to-single-crystal back reaction. The crystal structure of

the recrystallized dimer proved that the as-dimerized and recrystallized dimer were in

different modifications. The recrystallized dimer single crystals were resistant to thermal

cycloreversion under the same conditions that caused this reaction in the as-dimerized

crystals. Calculated and measured powder diffractograms of the homogeneously and

heterogeneously-formed photodimers indicate that these products are identical.
Six full crystal structures are reported in chapter 4. Five are of a-trans-cinnamic

acid and photodimer single crystals at 0%, 28%, 40%, 67% and 100% conversion. A

crystal structure of an ammonium salt of cinnamic acid was also performed. The ability to

achieve a single-crystal-to-single-crystal phototransformation on cinnamic acid indicates

that the method of tail irradiation may have some general applicability to topochemical

photoreactions. Again, the side groups in this photodimerization occupy the same location

in the lattice regardless of whether they are attached to an olefin or cyclobutane. In both

the styrylpyrylium triflate and cinnamic acid, the phenyl ring rotates approximately 20"

from the reactant to product structure. The same type of bond angle deformation is seen in

both cases as well. The main difference in the two cases comes from X-ray powder

diffraction measurements on the heterogeneously formed dimer. In the case of cinnamic

acid, the X-ray powder diffraction measurements for the heterogeneously produced dimer

do not match that calculated from the crystal structure of the homogeneously-produced

photodimer single crystal.
Eight full crystal structures are reported in chapter 5: the BF4-, ReO4-, C104-,

SnCl5-, and AuC12- salts of the styrylpyrylium monomer, discussed in chapter 3, as well

as the photodimers of the first two, plus the bis(styrylpyrylium triflate) created by linking

two styrylpyrylium triflate chromophores, 1-lc, by replacing the methyls of their methoxy

groups with a four carbon methylene bridge. The first three salts are isomorphous with

one another, the lattice adjusting slightly from the smaller BF4" and C104- counterions to

accommodate the larger ReO4'. Time vs. conversion curves showed the perrhenate salt to

have a slower rate once higher conversions were reached. The method for obtaining

single-crystal-to-single-crystal conversions by tail-irradiation succeeded with these salts as

well and the crystal structures of the photodimers allows comparison of the counterion

conformation from the reactant to product structure. The other salts, except for the tin

counterion, were photoinactive.

In chapter 6, evidence is presented of a holographic grating written into a

styrylpyrylium triflate, 1-lc, single crystal by two interfering laser beams. Besides a

photograph of the grating, diffraction efficiency vs. time curves for the writing and thermal

erasing of the grating were obtained. Once the sin2 dependence of the grating growth was

established, the diffraction efficiency vs. time curves could be related to rate of product

formation. Then gratings were written and erased at various temperatures to obtain rate

constants at various temperatures which were used to construct Arrhenius plots which

yielded energies of activation for crystalline-state photodimerization and thermal

cycloreversion of the styrylpyrylium salt, 1-lc, and its photodimer, 1-4c.

Preparative work for the styrylpyrylium triflate, 1-1c, was begun at the University

of Florida (UF), Gainesville, Florida. All of the absorption-tail irradiations and

crystallographic analyses were performed at the Max-Planck-Institut flir Polymerforschung

(MPI-P), Mainz, Germany. The work on the bis(stryrylpyrylium triflates) was begun at

UF with the synthesis of compounds 2-25, 2-29 and 2-33. The balance of the

bis(styrylpyrylium salts) were synthesized at MPI-P. All of the work reported in Chapters

4 and 6 was done at MPI-P. All of the mass spectral, nuclear magnetic resonance, and

infrared analyses were performed at MPI-P.


Instrumentation and Analysis

General Information

All reagents were purchased from Aldrich and Fluka chemical Co. and used without

further purification. Melting points were determined from DSC traces done on a Mettler

DSC 30 at a heating rate of 5'C / minute. Routine NMR spectra were recorded on a Bruker

AC 300 spectrometer (1H: 300 MHz; 13C: 75.5 MHz) or a Varian Gemini 200 Fourier

NMR spectrometer. The residual 1H peak of the deuterated solvent is used as an internal
standard (CHC13: 1H, 8 = 7.24; 13C, 8 = 77.00). Elemental analyses were carried out by

the Analytische Laboratorien in Engelskirchen, Germany. Infrared measurements were

performed with KBr pellets on a Nicolet 730 FT-IR Spectrometer. The photographs in

Figures 3-10, 11, 16 and 6-6 were taken on a Zeiss Photomicroscope III with a high

pressure halogen lamp. Mass spectra were obtained using a Varian MAT 7A (70 eV).

X-ray Structure Analysis

All of the x-ray structure analyses were performed by Dr. habil. Volker Enkelmann,

Privat Dozent, at the MPI-P, with an Enraf-Nonius CAD-4 four circle diffractometer with

graphite monochromated Cu-Ka radiation. Lattice parameters were obtained by least-

squares fits to the setting angles of 25 reflections with 0>200. The intensity data collection

was performed by 0-20 scans. The intensities of three control reflections were measured at

regular intervals to check the stability of the crystals. The raw data were corrected for

Lorentz and polarization effects. The structures were solved by direct methods and refined

by full matrix least squares analyses with anisotropic temperature factors for all atoms

except H. Positions of the H atoms were calculated using the known molecular geometry

and refined in the riding mode with fixed isotropic temperature factors. Empirical

absorption corrections were applied to the data and unit weights were used throughout the


X-ray powder diffractograms were recorded using a Philips PW1820 powder
diffractometer with Ni filtered Cu-Ka radiation and a secondary graphite monochromator.

Further details of the crystal structure analyses for the crystal structures reported in

chapter 3 are available upon request as described in Angewandte Chemie.118 Those

reported in chapter 4 are available from the Journal of the American Chemical Society.117

Solid State C-13 Measurement

The solid state C-13 NMR measurements referred to in chapters 3 and 4 were

performed by Dr. Jeremy Tittman at the MPI-P on a Bruker MSL 300 at 75.47IR

measurements were performed on a Bruker MSL 300 at 75.47 MHz. The MAS (Magic

Angle Spinning) speed was 3 kHz, and the Cross Polarization time was 2 msec. A Total

Suppression of Spinning Side Bands (TOSS) pulse sequence was used. The signal

acquisition was done over 35 msec., the 90" pulse length was 4.5 Wmsec., and the spectral

width was 29,411 Hz. 2048 Data points were collected.

Crystal Preparation

Tables 2-1 2-6 list the series of compounds which were synthesized as part of this

work. Reference numbers are given to each compound. A series of ammonium salts from

3 types of cinnamic acids were prepared and are listed in Table 2-6. Two classes of

styrylpyrylium salts were synthesized: simple p-methoxy-substituted styrylpyrylium salts,

shown in Table 2-3, and diolefins involving the styrylpyryium group. The diolefins fall

into two categories: Hasegawa-type diolefins, shown in Table 2-4, which are fully

conjugated with one another through sharing a styryl moiety and diolefins linked together

with a methylene bridge, shown in Table 2-5.

Pyrylium salts were required for the syntheses of all classes. These are shown with

their numbering scheme for the dissertation in Table 2-1 below. Methylene bridged

dialdehydes, shown in Table 2-2, were synthesized in order to prepare the methylene-

bridged diolefins.

Table 2-1. 2,6-Di-tert-butyl-4-methylpyrylium salts

counterion. X: compound
BF4- 2-1
ReO4- 2-2
C104- 2-3
SnCl6= 2-4 t-Bu
AuCl4- 2-5
SbF6- 2-6 -

Table 2-2. 4,4'-(Alkylenedioxy)dibenzaldehydes

n compound
2 2-7
3 2-8
4 2-9
6 2-10
10 2-11 0 o( cr2
Table 2-3. (E)-2,6-
Di-tert-butyl-4-[2-(4-methoxyphenyl)-ethenyl]pyrylium salts and their photodimers, 4,4'-c-
2,t-4-Bis(4-methoxyphenyl)-r-1 ,t-3-cyclobutandiyl]bis(2,6-di-rert-butylpyrylium-bis salts

counterion. X- compound
BF4- 2-12a
photodimer 2-12b
ReO4- 2-13a
photodimer 2-13b
C104- 2-14a t-Bu
photodimer 2-14b
SnCl5- 2-15a
photodimer 2-15b MeO t-
AuCl2- 2-16 t
SbF6- 2-17

Table 2-4. p-Bis[(E)-2,6-di-tert-butyl-4-(2-ethenyl)-pyrylium]benzene salts

counterion, X-




t-Bu t-Bu

t-Bu '-Bu

Table 2-5. a,(o-(Alkylenedioxy)bis[(E)-2,6-Di-tert-butyl-4-[2-(4-phenyl)-
ethenyl]pyrylium] salts
counterion, X' n=2 n=3 n=4 n=6 n=10







t-Bu t-Bu

\ 0 ( CH2) n / 0
t-Bu ( cn t-Bu

Table 2-6. Ammonium salts of cinnamic acids


cinnamic acid portion

a-trans-cinnamate p-methyl-a-trans-cinnamate p-methoxy-a-trans-cinnamate






Pyrylium salt syntheses (Table 2-1)
The following syntheses are adapted from syntheses reported by Balaban and
Nenitzescu103 and Dimroth.104
Pyrvlium salt 2-4 (SnClf2-) 20.24 g (0.16 mol) pivaloyl chloride was stirred with

21.92 g SnC14 (0.084 mol) under Argon. 7.77g (0.084 mol) t-Butyl chloride was added

dropwise to that mixture. The mixture was stirred for 16 hours resulting in an orange,

viscous reaction mixture. This mixture was poured into 300ml of distilled water where it

forms a partially solidified oil. Ether (200ml) was added to this heterogeneous mixture to

dissolve the solidified oil, then the ether layer removed and the water layer further extracted

with ether until the ether layer remains clear (3 times with 200ml of ether each time). Water

(150ml) was removed and the water layer cooled to give 20g of a white solid. The solid

was filtered and the filtrate (mother liquor) used in the preparation of additional pyrylium

salts described below. The white solid was dried to give 20g of white crystals which were

recrystallized in water. Proton and carbon NMR of this substance indicate a pyrylium

cation. Elemental analysis revealed that the counteranion contained tin and was mostly
SnCl62-, though SnCl5- was present as well.

Pyrylium salts 2-1 2-3. 2-5 and 2-6. Aqueous solutions of each of the following

acids were added to a portion of the mother liquor from the tin-containing pyrylium salt
synthesis to give pyrylium salts with the corresponding counterion: HBF4, HC104,

HReO4, HSbF6. White precipitates are seen immediately in the case of C104-, ReO4- and

SbF6- pyrylium salts. The BF4- pyrylium salt crystallizes slowly out of solution as

colorless needles. HAuC14 was added to a portion of the mother liquor from the tin-

containing pyrylium salt synthesis; the gold-containing pyrylium salt emerges from the

solution as orange crystals. Proton and carbon NMR of this substance indicate a pyrylium
cation. Elemental analysis revealed that the counteranion was mostly AuCl4-, however,

AuCl2- was present as well. Each of these pyrylium salts may be recrystallized from


Miscellaneous pyrylium salts. Attempts were made to obtain other pyrylium salts

of the type found in Table 2-1 as described below.

For the counterion PF6', an aqueous solution of the acid HPF6 was added to the

mother liquor from the synthesis of the 2-4 to yield a white precipitate. Elemental analysis
indicated that a pyrylium salt of the type found in Table 2-1with a PF6- counterion had

precipitated out, but this compound could not be converted to a styrylpyrylium salt with a

phosphorous-containing counterion.
For the counterions PtCl6', IrCl6-, and OsCl6-, aqueous solutions of the acids,

H2PtCl6, H2IrCl6, H20sCl6, were added to the mother liquor from the synthesis of the

tin-containing pyrylium salts. However, no precipitates were observed.
For the counterion CH3(CH2)nSO3-, the synthesis of the pyrylium salt with a

triflate counterion105 was repeated using CH3SO3H instead of triflic acid in an attempt to

synthesize a pyrylium methanesulfonate; however, only starting materials could be

recovered. Addition of methane sulfonic acid to an aqueous solution of the pyrylium tin

hexachloride yielded no precipitate. An equimolar solution of the styrylpyrylium triflate, 1-

Ic, in methane sulfonic acid was subjected to vacuum distillation to remove the triflic acid;

however, under these conditions the pyrylium cation decomposed. Further attempts to

obtain this compound via ion-exchange columns from Fluka's Amberlyst resins also failed.

Methvlene-bridged dialdehvdes (Table 2-2)

The series of dialdehydes in Table 2-2, as well as dialdehydes bridged with 7 and

11 methylene units, were prepared as follows. A solution of sodium ethanolate was

freshly prepared by dissolving 5.75g Na in 150ml anhydrous ethanol from a freshly

opened bottle. The residual moisture from a 3-necked round bottom flask fitted with a

reflux condenser was driven off with a heat gun. The apparatus was flushed with argon

via a connection at the top of the condenser and freshly cut Na was added to the round

bottom flask. Dry ethanol (150ml) was quickly added with a funnel through one of the

necks and the apparatus resealed and allowed to stand until all the sodium had dissolved.

p-Hydroxy benzaldehyde was dissolved under argon in dry ethanol and quickly pipetted

into the sodium ethanolate solution along with the dialkylbromide. The reaction mixture

was stirred under reflux for 4 hours. After 1 hour a precipitate emerged.

The reaction mixture was then cooled and the ethanol removed with a rotovap.

100ml of 5% NaOH solution was added to the reaction residue and the dialdehyde extracted

with diethylether. This organic layer was washed repeatedly with fresh portions of water

until the water layer tested neutral for pH. The ether layer was then dried for 15 minutes

over calcium chloride, filtered and the ether layer removed with a rotovap to yield a dull

white flaky residue in the case of n=2, 3, 6 and 10. n=4 gave shiny, translucent flakes.

Raw yields in each of these cases were around 100%. The yields reported are the yields

obtained after recrystallization. Very low raw yields in the case of n=7 and n=ll were

obtained probably due to an insufficient amount of sodium ethanolate. White powders

were obtained in those cases. They were not recrystallized or used further.

Table 2-7. Synthesis data for methylene-bridged dialdehydes (Table 2-2)

p-hydroxy dialkylbromide sodium ethanolate yield
n=2 30g(250mmoles) 9.4g(50mmoles) 150ml x 1.7M 2.0g 15%
n=3 15.4g(124mmoles) 10.0g(50mmoles) 150ml x 0.83M 6.3g 16%
n-4 10.7g(50mmoles) 150ml x 1.7M 4.8g 27%
n=6 12.1g(50mmoles) 150ml x 1.7M 4.0g 25%
n=7 49g(400mmoles) 8.6g(33mmoles) 50ml x 0.83M 50mg 4%
n=10 30g(250mmoles) 14.9g(50mmoles) 150ml x 1.7M 4.5g 24%
n=ll 49g(400mmoles) 10.5g(33mmoles) 50ml x 0.83 50mg 4%

For n=2, 4, 6 and 10 dialdehydes were recrystallized in ethanol, n=3 in methylene
dichloride, and used in Knoevenagel condensations with the ReO4-, BF4- and C104-

pyrylium salts as well as the triflate pyrylium in the cases of n=3 and n=4.

Styrylpvrylium salt syntheses (Tables 2-3 2-5)

p-Methoxv-substituted styrvlpvrylium salts (Table 2-3) Each of the styrylpyrylium

salts in Table 2-3, was synthesized by Knoevenagel condensation of the corresponding

pyrylium salt described above (for 1-lc, the synthesis of the pyrylium salt was carried out
as described previously in the literature105) with p-anisaldehyde via a procedure directly

analogous with that reported by Hiinig for the p-methoxy substituted styrylpyrylium

triflate, 1-1c. Each of the pyrylium salts was refluxed for 16 hours in glacial acetic acid

(5ml/lg pyrylium salt) with 3 molar equivalents of p-anisaldehyde (MW=136,

density=1.119 g/ml). The reaction mixture is then poured into an equivalent volume of

ether to yield the photoactive crystals. The crystals are sucked dry and washed repeatedly

with ether to remove the acetic acid. Typical reaction data is given below. These syntheses

were repeated as often as further product was required. The yields were in the range 70-


Table 2-8. Synthesis data for p-methoxy styrylpyrylium salts (Table 2-3)

styrylpyrylium pyrylium salt aldehyde AcOH Yield
BF4- 0.20g(0.70mmoles) 0.25ml(2.0mmoles) 5ml 0.25g 85%
C104- 0.26g(0.85mmoles) 0.31ml(2.5mmoles) 5ml 0.35g 95%
ReO4- 2.10g(4.6mmoles) 1.67ml(14.8mmoles) 12ml 1.90g 70%

AuC12- 1.63g(3.0mmoles) 1.09ml(9.0mmoles) 10ml 1.30g -75%
SnC15- 2.0g(5.4mmoles) 2.0ml(16.5mmoles) 12ml 2.70g -80%
SbF6- 0.60g(l.lmmoles) 0.50ml(4.0mmoles) 7ml 0.50g 85%g

Hasegawa-type diolefins (Table 2-4) The salts in Table 2-4 were prepared in a

manner similar to those in Table 2-3 except in these cases, an excess of the pyrylium salt

was used. Acetic acid solutions of the pyrylium were refluxed with terepthaldehyde for 16

hours. However, each of the products begin to precipitate out of the reaction mixture after

several hours of refluxing. The hot reaction mixtures were vacuum filtered to isolate in

each case a yellow powder which was washed with hot glacial acetic acid repeatedly until

1H NMR analysis indicated that all the mono-functionalized terephaldehyde as well as the

excess pyrylium salt had been removed.

Table 2-9. Synthesis data for Hasegawa-type diolefins (Table 2-4)

terepthadehvde pyrylium salt AcOH yield

BF4- 0.268g (2.0mmoles) 1.8g(6mmoles) 12ml 0.9g 70%
C104- 0.268g(2.0mmoles) 1.8g(6mmoles) 12ml 1.0g 70%
ReO4- 0.134g(1.0mmoles) 1.4g(3mmoles) 10ml 0.7g 70%
CF3SO3- 0.268g(2.0mmoles) 2.0g(6mmoles) 12ml 1.lg 70%

Methylene-bridged diolefins (Table 2-5) Styrylpyrylium salts linked by an alkyl

chain were prepared by refluxing the corresponding dialdehyde with the appropriate

pyrylium salt. Reaction data is given below. Each pyrylium salt was refluxed with each
dialdehyde in 15ml of acetic acid for 16 hours to give the yields indicated. After 8 hours

yellow or orange precipitates were observed in each of the reaction mixtures except those of

the n=3 materials. The hot reaction mixtures were vacuum filtered to isolate the precipitates

which were then dissolved in trifluoroacetic acid and reprecipitated in acetic acid.

Table 2-10. Synthesis data for methylene-bridged diolefins (Table 2-5)

BF4- C104- ReO4-
1.92g 2.00g 2.98g
(6.53mmoles) (6.53mmoles) (6.53mmoles)

n=2 0.594g (2.2mmoles) 1.3g 0.9g 1.4g
n=3 0.625g oil 0.7g oil
n=4 0.660g 1.5g 0.9g 1.8g
n=6 0.717g 0.7g 1.2g 1.2g
n=10 0.840g 0.8g 1.lg 1.3g

Ammonium salts of cinnamic acids (Table 2-6)

The ammonium salts of the cinnamic acids in Table 2-6 were formed by first

dissolving the acid in 20ml of methanol, then equimolar amounts of the t-butylamine,

benzylamine, and piperidine were added to the solution. The methanol was then removed

by evaporation to give the acid ammonium salt. Equimolar portions of piperazine were

dissolved in methanol and added to a methanolic solution of each acid. The piperazinium

salts precipitated out immediately.

Table 2-11. Synthesis data for cinnamic acid ammonium salts

cinnamic acid amine

ct-trans- piperazine
cinnamic acid 0.43g (5mM)
1.48g (10mM)
p-methyl- 0.85g (10mM)
cinnamic acid
1.62g (10mM) t-butyl amine
0.73g (10mM)
cinnamic acid benzylamine
1.78g (10mM) 1.07g(10mM)


Styrylpyrylium triflate monomer (1-1c) and dimer (1-4c)

Micrographs. The styrylpyrylium triflate, 1-1c, crystals for the photographs in

Figures 3-10, 11 and 16 were formed by evaporation of ethanolic solutions of the

monomer on a glass slide. A clean glass slide was boiled in HMDS for 30 minutes to make

it hydrophobic, then rinsed in ethanol. 10mg of styrylpyrylium salt, was warmed in lml of

ethanol until dissolved. The solution was cooled to room temperature then several drops

applied to the prepared glass slide. Slow evaporation of the ethanol left styrylpyrylium

crystals of various sizes and optical quality adhered to the glass slide. The crystals were

loosened from the slide surface with ether. This preparation was repeated until at least 3

crystals at least 0.2-0.3mm in size and of sufficient optical quality were found in the

proximity of one another for photographic purposes.

Time vs. conversion vs. cell parameters. The styrylpyrylium triflate crystals for

these experiments were prepared from room temperature ethanolic solutions of the

styrylpyrylium triflate. 50Mg of styrylpyrylium triflate were dissolved in 50ml of ethanol

and allowed to cool to room temperature overnight. The crystals that formed were filtered

and 50 ml of ether was added to the mother liquor. 0.2mm-0.5mm single crystals of

excellent optical quality emerged from this solution overnight

Photodimer. 1-4c. The cyclobutane dimer formed from the irradiation of

styrylpyrylium triflate was recrystallized by solvent diffusion. 10mg of the dimer is

dissolved in 0.5 ml of trifluoroacetic acid. A layer of ether is carefully applied over this

acid solution. After several hours, orange cyclobutane dimer crystals are formed at the acid

/ ether interface. They must be quickly rinsed with ether when removed from the solution.

Repeated attempt sto recrystallize the cyclobutane dimer by cooling in methanol failed.

p-Methoxv-substituted styrylpyrylium salts (Table 2-3)

The recrystallizations of these compounds are summarized in Table 2-12.

Table 2-12. Recrystallization data for p-methoxy-substituted styrylpyrylium salts
(Table 2-3)
solvent color / crystal habit

2-12 BF4- acetic acid gold platyy
2-13 ReO4- acetic acid gold / platy
2-14 C104- acetic acid gold / platy
2-15 SnCl5- acetic acid orange / bladed
2-16 AuCl2- formic acid orange / bladed
2-17 SbF6- acetic acid gold / platy and red / prismatic

Slow cooling in acetic acid of 2-12 2-14 gives crystals of sufficient quality for x-

ray structure analysis. Michael Steiert at the Max-Planck-Institut fir Polymerforschung

performed many of these recrystallizations. Normal cooling in acetic acid of 2-15 through

2-17 gives crystals of sufficient quality for x-ray structure analysis. Addition of ether to

room temperature acetic acid solutions of 2-12 and 2-13 sometimes produces excellent gold

tabular crystals of those products. Cooling of hot acetic acid/ether, ethanolic and

ethanol/ether solutions, and solvent evaporation from ethanolic solutions are further

possible methods for recrystallization for compounds 2-12 2-17. The apparent

chromoisomerism of SbF6- styrylpyrylium salt may be actually due to the formation of a

mixed salt of Sb and Sn-containing counterions. This phenomenum was observed for the
ReO4- salt, 2-12, where in addition to the usual gold platelets, red bladed crystals were

obtained from the crude reaction mixture. Crystallographic analysis revealed that a mixed

salt had been formed with Re and Sn-containing counterions.

Hasegawa-type diolefins (Table 2-4)

The recrystallization of these compounds is summarized in Table 2-13.

Table 2-13. Recrystallization data for Hasegawa-type diolefins

solvent crystal / habit

2-18 BF4- *
2-19 ReO4-
2-20 C104" acetic acid orange / bladed
2-21 Triflate formic acid orange / bladed

Acetonitrile, trifluoroacetic acid / acetic acid, trifluoroacetic acid / ether, chloroform

/ carbon tetrachloride, ethanol / ether and solvent evaportation from ethanolic solutions

yielded only finely divided precipitates for 2-18 and 2-19.

Methylene-bridged diolefins (Table 2-5)
Of the distyrylpyrylium salts with the alkyl chain, crystals from only 2-33 (BF4",

n=6) and 2-34 (tyfriflate, n=4) could be obtained. Crystals of 2-33 were reached by slow

evaporation of an acetonitrile solution. A 20ml vial of the acetonitrile solution was capped

with a perforated lid. After 2 weeks, red blade-like crystals were observed to be growing

in the direction of the receeding acetonitrile solution. Crystallographic analysis revealed

trifluoroacetic acid solvent molecules incorporated into the crystal lattice. Crystals of 2-34

were obtained from cooling a hot, saturated acetic acid solution of the salt.

Ammonium salts of cinnamic acid (Table 2-6)

All of the ammonium salts of cinnamic acid were recrystallized by cooling hot

isopropanol solutions to give white needles, blades and platelets as described below. Only

the piperidinium salts failed to recrystallize.

Table 2-14. Recrystallization data for cinnamic acid ammonium salts (Table 2-6)

amine portion

acid t-butyl- benzyl piperidinium piperazinium
potion ammonium ammonium

c-trans-cinnamic acid bladed acicular -- platy
cinnamic acid platy bladed -- acicular
cinnamic acid acicular -- platy

a-trans-cinnamic acid itself was recrystallized by slow evaporation of an ether

solution to form colorless platy crystals.


Photodimer Single-Crystal Preparation

Preparative irradiations for single crystals of 1-4c, 2-12b, 2-13b, 2-14b and a-

trans-cinnamic acid were performed with a 450 Watt Xenon lamp from Photon Technology

International Inc., Mod. A 5000 (spectral output shown in Figure 2-1). A quartz diffusor

was placed in the beam focus, 17cm from the lamp housing. The filter used for each

irradiation was air-cooled and placed at 21cm from the lamp housing.

In the case of the styrylpyrylium salts, a silver mirror was used to direct the light

emerging from the filter onto a covered petri dish containing the crystals. For l-lc, a band-

pass filter at 570.4 nm (HW= 54.0 nm) was used. A significant number of the crystals

were converted to photodimer in 4 hours. The reaction was complete after 24 h of

irradiation under these conditions. For 2-12 2-14, a band-pass filter at 548.5 nm (HW=

54.0 nm) was used. Styrylpyrylium salts 2-12 and 2-14 were completely photodimerized

within 36 h, while 2-13 required 48 h of irradiation.

For the cinnamic acid, the crystals were irradiated in an NMR tube as follows: The

28% dimer crystals were obtained after irradiation in an NMR tube of 10mm diameter, with

Schott long-pass filter WG345 (transmission spectra for Schott long-pass filters shown in

Figure 2-2) after 100 h. The degree of conversion was determined by integration of the

cyclobutane and vinyl proton signals in the 1H-NMR and was consistent with the percent

conversion derived from the population parameters in the crystallographic analysis. The

40%, 67% mixed monomer-dimer single crystals and photodimer crystals were obtained

from irradiation of a batch of a-trans-cinnamic acid monomer crystals in an NMR tube of

5mm diameter with long-pass filter WG360 for 100 h. 1H NMR sampling of this batch

indicated an average conversion of over 90%. The percent conversions of the individual

crystals were derived from the population parameters of the crystallographic analysis. The

UV-VIS absorption spectrum of a polycrystalline film of cinnamic acid is shown in Figure


450 W



Figure 2-1. Spectral output for the 450W Xenon lamp used in the
photodimerizations of the styrylpyrylium salts and cinnamic acid.


The photodimer crystals pictured in Figures 3-10 and 3-11 were formed with

radiation from a Coherent Innova 90-K Krypton Laser with the 568.19 nm line as

described below under "Kinetic Studies". The crystals were exposed to 33.1 mWatts/ cm2

and were completely photodimerized in 30 minutes.

The photodimer crystals pictures in Figure 3-16 were dimerized with radiation from

a Xenon lamp passing through the 570.4 nm band-pass filter. The monomer crystals were

formed on a glass slide as described above under "Recrystallizations" and the slide

mounted on the microscope stage of a Zeiss Photomicroscope in Dr. rer. nat. Giinter

Lieser's laboratory at the Max-Planck-Institut fir Polymerforschung. The lamp was

brought as close as possible to the stage and the beam from the lamp reflected off a mirror

next to the microscope objective and onto the slide. A photo--not included in this

dissertation--of the original monomer crystals for Figure 3-16 was taken. (The microscope

was then focused on another group of crystals also on the slide. Photographs of the

photodimerization of these crystals were then taken over a period of 24 hours. These

photographs have also not been included in the dissertation.) After the photodimerization

was complete, a photo of the dimer crystals with cross-polarized light was made. This is

the top photo in Figure 3-16. The subsequent thermal treatment of the crystals was done

by removing the glass slide to a hot plate set at 100' C until a color change was noted then

reinserting it on the stage for photographing.

Kinetic Studies

Styrylpyrylium salts 2-12 2-14

Compounds 2-12 2-14 were irradiated in their absorption maxima (thus

heterogeneous conversions were effected) under identical conditions and their conversion

to photodimer measured with time. Table 2-7 lists the times and conversions of each

substance for each trial. The irradiations were done with the 450W Xenon lamp described
above with a pyrex filter (X>270 nm). 50ml suspensions of each of these compounds were

prepared in dry hexane and degassed for 5 minutes with bubbling Ar. The same 100ml

flask was used for each preparation and placed in the center of the radiating beam, 29.5 cm

from the lamp housing. Table 2-15 lists the specific amounts of the substances used for
each trial as well as the sampling volume throughout the irradiation. The samples were
rapidly stirred during irradiation and sampling. After the sample was withdrawn, the
hexane was removed under vacuum and a 1H NMR run in 3:1 CDCl3:CF3CO2H. The
signals from the protons on the p-methoxy as well as t-butyl groups were integrated to
determine the percent conversion. Figure 2-4 shows a typical 1H NMR for a partially
dimerized styrylpyrylium salt



//7/ I1 i
IA Il11
-Ifil (UJ I I- -

WG 345
------WG 360
WG 360

J- A

200 300 400
wavelength (nm)
Figure 2-2. Transmission spectra of Schott long-pass filters WG345 and WG360
used in the cinnamic acid irradiation.
(transmitted photon flux)k
(incident photon flux)k



, ,

- t-




250.0 300.0 350.0
wavelength (nm)

Figure 2-3. UV-VIS absorption spectrum of a polycrystalline film of a cinnamic

A I L,. J' L ... I -J L ... --"- .....
8 7 6 5 4 3 2
Figure 2-4. IH NMR of a partially dimerized styrylpyrylium triflate, 1-1c and 1-

Table 2-15. Time (left-hand column) vs. conversion data for the
photoirradiation of 2-12 (left-hand column under percent conversion), 2-13 (right-hand
column under percent conversion) and 2-14 (middle column under percent conversion).
The sampling volume as well as the mass of each salt used is indicated for each trial.

Trial sampling volume=2ml
0.217g 2-12 0.223g 2-14 0.303g 2-13
seconds percent conversion
10 13 11 9
20 19 20 13
30 25 25 20
40 34 29 26
60 47 40 34
80 53 55 36
100 62 60 38
140 78 -- 48
180 83 55
220 86 88 60
320 95 98 75
420 97 99 86

Trial 2 sampling volume=3ml
0.179g 2-12 0.183g 2-14 0.250g 2-13
seconds percent conversion
10 9 5 14
40 13 10 16
100 24 21 33
140 37 28 29
180 46 34 38
220 56 42 38
320 76 62 52
420 91 70 65
620 94 86 78
820 100 96 83

Trial 3 sampling volume=lml
0.360g 2-12 0.360g 2-14 0.500g 2-13
seconds percent conversion
20 -- 15 12
40 12 11 11
60 13 16 14
80 12 30 13
100 14 17 16
140 35 25 19
180 25 25
220 26 27 32
320 34 39
420 46 56 41
620 63 63
1020 84 88 62
1620 100 100 85
2420 100 -- 99

Trial 4 sampling volume=not recorded
0.100g2-12 0.100g2-14 0.100g2-13
seconds percent conversion
0 3 5 2
100 17 18 16
200 28 28 19
400 43 37 30
600 57 52 49
800 65 66 60
1000 76 77 59
1500 84 91 68
2000 90 95 80
3000 100 100 87
4000 -- 100 92

Styrylpvrvlium triflate. 1-1c

The goal was to monitor the time-dependence of the single-crystal-to-single crystal

homogeneous conversion of styrylpyrylium triflate, 1-lc, and also obtain substitutionally

mixed single crystals of known conversion for crystallographic analysis.

Initial experiments. Initially, a dye laser was chosen for the irradiation. Forty to

sixty monomer crystals 0.3-0.5 mm in length were irradiated with light at approximately

570 nm over a 1 cm2 area of uniform intensity created with a pin-hole and beam expander.

It is possible that the monomer crystals have different absorptivities, depending on the

orientation of their crystallographic axes with respect to the plane polarized light from the

laser. Therefore, a Soleil-Babonet plate (a quarter lambda plate tunable for various

wavelengths) was used to circularly polarize the beam. The power of the dye laser was

such that the crystals received 0.16 mWatts / cm2. At this power, the time required for

complete conversion was 24 h. However, the optics involving the pin-hole were not stable

over this time period. Figure 2-5 shows the beam profile at the conclusion of the

irradiation. Thus the crystals were unevenly irradiated and no useful data could be


Cross-section of irradiated area

Figure 2-5. Beam profile after 24 h. Crystals in an area corresponding to the left
side of the cross-section would recieve much less intensity than those on the right.

In the second type of experiment designed to acquire the desired kinetic data, single

crystals of the monomer were individually rotated in the laser beam to reduce the area

required for uniform intensity. This was done by mounting each crystal on a glass fiber

which was in turn placed in a goniometer head rotated by an adapted stirring apparatus.

During these experiments it was discovered that the power of the dye laser steadily

decreased with time. In addition, the Soleil-Babonet plate was no longer available and two

of the crystal faces were indistinguishable from each other, one of which had to be chosen

for mounting the crystal: if the crystal has different absorptivities depending on its

orientation, 2 conversion curves are possible. Most problematic was that a crystal small

enough for crystallographic analysis was not large enough for percent conversion analysis

by IR methods; that is, an individual crystal was too small for making KBr pellets of

sufficient concentration for reliable data, but was too thick for single-crystal IR analysis.

Final experiment. The following method finally gave reproducible data: For each

conversion, 8-9 crystals, 0.3-0.5 mm in length, were aligned along one optical axis using

the crystal habit as a guide. The crystals, lying on their one easily identified face ( this was

the [100] face which corresponds to the plane of the be cell axes as shown in Figure 2-6),

were then rotated in a beam from a Krypton laser in an area of uniform intensity created

with a beam expander. After each dosage, 2-3 of the crystals were submitted for

crystallographic study while the balance of the crystals were taken for IR percent

conversion analysis. The time vs. conversion data collected is listed below in Table 2-16.

The power from the Krypton laser was such that the optics were steady over the time

required for the conversions. Furthermore the power from the Krypton laser was steady

with time.

Table 2-16. Time vs. conversion data for the single-crystal to single-crystal
conversion of the styrylpyrylium triflate, 1-1c. The conversion in italics are those of
uncracked crystals; those underlined are conversions of cracked crystals.

time (s) trial 1 trial 2 trial 3

840 96 87 83
780 84 86 89
720 89 82 78
630 77 61 83
540 62 51 65
450 49 42 57
360 47 5 35
270 32 30 31
180 23 23 27
90 8 12 10
30 0 1 -2

The Krypton laser was a Coherent Innova 90-K, with a range of 350.7-799.3 nm,

800mW. The 568.19 nm line was used for this work. A beam splitter directed a portion of

the beam to the head of a Coherent 210 Power Meter hooked up to a 5000 Digital

Multimeter by Prema. This in turn was connected to a personal computer installed with

software that transferred the power data from the multimeter to a file. For each conversion,

a file containing the power vs. time data was created and integrated with Lotus Symphony

software to give the total relative energy dosage for each time period. A plot of time vs.

dosage was linear. The beam not collected for power analysis was passed through a #8

Oriel beam expander and then onto a mirror which directed the beam onto the rotating

crystals. The power of the beam was 21.7 mW/cm2. The laser setup was constructed by

Bernhard Zimmer at the Max-Planck-Institut ftir Polymerforschung.

Figure 2-6. This projection of the crystal structure of the 67% monomer-dimer
substitutional mixed crystal shows the orientation of the styrylpyrylium salt, 1-1c, in its
unit cell. The [100] face, or be plane of the unit cell was perpendicular to the radiation


Calibration curve. The percent conversion for each batch of crystals was
determined by integrating the C=C-H out-of-plane bend absorption band between 920 and
967 cm-1 and referencing it to an internal standard, the -S02- absorption band of the

counterion at 1030 cm-1. The calibration curve shown in Figure 2-7 was created by
irradiating suspensions of the monomer with the Xenon lamp, no filter, and determining
percent conversion of a given sample by 1H NMR. IRs of samples of known conversions
could then be measured and the ratio of the peaks described above determined for each
conversion. Initially, the large C=C stretching peak at 1583 cm-1 was chosen as the peak
to monitor for conversion analysis but identically handled monomer crystals repeatedly
failed to give the same ratios for this peak area to that of the internal standard. Figure 2-8
shows the IR spectra for the 0, 20, 40, 60, 80 and 100% conversion points on the curve.

0.7. 0.7

0.6 0.6

0.5 0.5

I 0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 ,,0I, I ,I ,,,l I,.,E 0
0 20 40 60 80 100
Percent Conversion

Figure 2-7. Calibration curve for IR analysis of single-crystal-to-single-crystal
conversion of 1-lc. y=0.668-3.84x10-3-4.60x10-5+2.58x10-7, r-0.999. The peak ratio
is the ratio of the C=C-H absorption at 954 cm-1 to the -S02- absorption at 1030 cm-1.


Figure 2-8. Series of IR spectra for the conversion of 1-lc to 1-4c. The rearmost
spectrum is 1-lc unirradiated while the foremost spectrum is photodimer 1-4c. The
intermediate percent conversions are 20, 40, 60 and 80%.


Selected elemental, infrared, mass spectral and nmr analyses for the compounds in

Tables 2-1 2-6 are presented in the balance of this chapter as follows. Tables 2-17

through 2-22 list elemental analyses for all the compounds reported in this dissertation.

Talbes 2-23 through 2-26 present infrared spectral analyses for all the compounds except

the methylene-bridged diolefins. Tables 2-27 through 2-28 contain data from the mass

spectral analyses of the pzrzlium and ammonium salts. Tables 2-29 through 2-33 report the

NMR analyses.

Elemental Analyses

The leftmost column is the carbon analysis and the adjacent column is hydrogen.

Table 2-17. Elemental analyses for Table 2-1 compounds

SnCl62- theo.

AuCl4- theo.



2-6 exp. 37.86 5.11
theo. 37.95 5.23

6.30 (
6.21 (

4.21 (
4.88 (

F 25.46
F 25.73

1l 26.07
C1 35.22
:1 28.51

1 26.22
1 14.92
l1 26.00

Sn 17.25
Sn 23.58
Sn 15.91

Au 36.35
Au 41.45
Au 36.07

Sb 27.70
Sb 27.48

Table 2-18. Elemental analyses for Table 2-2 compounds

2-7 exp. 70.83 5.29
theo. 71.10 5.22

2-8 exp. 55.25 4.41
theo. 71.82 5.67


exp. 73.42 6.96
theo. 73.60 6.79

2-11 exp. 75.20 7.92
theo. 75.36 7.91

2-9 exp. 72.25 6.16
theo. 72.47 6.08

Table 2-19. Elemental analyses for Table 2-3 compounds

2-12 exp. 64.02 6.99
theo. 64.09 7.09

2-13 exp. 45.79 5.06
theo. 45.90 5.08

2-15 exp.
SnC162- theo.

2-16 exp.
AuCl4- theo.




F 18.22
F 18.43

Re 32.45

4.93 C
4.70 C
5.95 C]

4.67 C
4.93 C
4.37 C]

exp. 47.23 5.11
theo. 47.08 5.21


B 2.80
B 2.62

2-14 exp. 62.28 6.87
theo. 62.19 6.88

1 23.25
1 38.53
1 21.65

1 11.95
1 21.39


Sn 19.60
Sn 19.10
Sn 12.08

Au 28.30
Au 33.20
Au 29.67

Sb 21.75
Sb 21.69

Cl 8.34
Cl 8.34

Table 2-20. Elemental analyses for Table 2-4 compounds

2-18 BF4-1

2-19 ReO4-1

exp. 59.25 6.37
theo. 63.00 7.05
exp. 41.60 4.69
theo. 42.68 4.78

2-20 C104-1

2-21 Triflate

exp. 55.37 5.96
theo. 60.76 6.80

exp. 56.07 5.87
theo. 56.29 5.97

Cl 8.71
Cl 9.96

F=13.83 S=7.75
F=14.06 S=7.91

Table 2-21. Elemental analyses for Table 2-5 compounds


exp. 63.40 6.57
theo. 64.25 6.86

2-33 exp. 58.41 6.14
theo. 59.12 6.20


2-23 exp. 43.95 4.57
theo. 45.98 4.91

2-24 exp. 61.17 6.30
theo. 62.33 6.66

2-28 exp. 61.74 6.43
theo. 62.71 6.78

2-29 exp. 51.86 5.08
theo. 58.74 6.08


exp. 62.96 6.67
theo. 64.96 6.99

2-31 exp. 46.65 5.07
theo. 46.93 5.14

exp. 64.44 7.03
theo. 65.62 7.34

2-35 exp. 47.38 5.35
theo. 47.83 5.35



exp. 63.33 6.85
theo. 63.78 7.14

exp. 66.04 7.45
theo. 66.82 7.76

2-39 exp. 48.66 5.47
theo. 49.51 5.75

2-40 exp. 64.46 7.39
theo. 65.06 7.56

Table 2-22. Elemental analyses for Table 2-6 compounds

2-42 exp. 68.91 6.79
theo. 69.09 6.85



exp. 68.24 8.31
theo. 72.07 8.21

exp. 75.36 6.74
theo. 75.27 6.71

2-45 exp. 70.58 8.55
theo. 70.56 8.65

2-46 exp. 70.06 7.32
theo. 70.22 7.37

2-47 exp. 69.34 8.46
theo. 72.84 8.56







2-48 exp. 75.76 7.15
theo. 75.81 7.11

2-49 exp. 71.61 8.86
theo. 71.46 8.99

2-50 exp. 65.02 6.83
theo. 65.14 6.83

2-51 exp. 65.60 7.76
theo. 68.42 8.04

2-52 exp. 71.46 6.76
theo. 71.56 6.71

2-53 exp. 66.78 8.34
theo. 66.91 8.42







Infrared Analyses

Many of the following IR analyses were performed by Frau Elke Muth at the Max-

Planck-Insitut-fiir Polymerforschung.

Ammonium salts of cinnamic acids (Table 2-6)

On making the ammonium salt from the free acid, the shifts listed in Table 2-23 are


Table 2-23. Prominent shifts in the IR absorption spectra of cinnamic acid
ammonium salts compared to the free acid.

cinnamic acid
C=O shifts from to C=C shifts from to
2-45 1680 1529 1381 1628 1643
2-44 1530 1355 1639
2-43 1548 1376 1639
2-42 1534 1379 1641

p-methoxy-cinnamic acid
C=O shifts from to C=C shifts from to
2-53 1686 1532/1513 1381 1624/1598 1641
2-52 1512 1383/1373 1643
2-51 1562 1401 1637
2-50 1511 1374 1640

p-methyl-cinnamic acid
C=O shifts from to C=C shifts from to
2-49 1681 1530 1361 1622/1605 1643
2-48 -1530/1500 1360 1641
2-47 1555 1374 1641
2-46 1512/1490 1372 1642

All the ammonium salts of the p-methoxy cinnamic acid also shows an absorption at

1605 cm-1. The ammonium salts of the primary amines, benzyl amine and t-butyl amine,

each show absorption at approximately 2200 cm-1.

Table 2-24 lists the strongest IR absorption bands for the cinnamic acids and their

ammonium salts (Table 2-6).

Table 2-24. IR absorption bands for cinnamic acid ammonium salts (Table 2-6)
The numbers listed are wavenumbers and have the units cm-1. The absorptions are
arranged in columns with the use of italics for easier comparison.

cinnamicacid] 1577 1494 1449 1313 1284 1222 1098/1072
979 943 875 846 768 709 682 588 541

994 968

991 964

1497 1450

1285 1250 1228 1073
775 717 689 584 534

1577 1496 1450 1324 1286 1249 1102/1071
882 847 778 718 690 586 535

1495 1447
1034 979 953 879 841

1247 1087/1072
774 724 685 585 532





p-methoxy cinnamic acid 1429 1314 1255 1171 1028
975 944 828 773 566 528
2-53] 1607 1422 1306 1257/1242 1174 1030
983/974 832 775 705 555
2-52 1606 1418 1302 -1255 1170 1025
993/970 934 835 775 728 705 552
2-51 1607 1424 1305 1258/1239 1174 1032
977 825 774 710 553
2-50 1605 1419 1305 1251/1241 1173 1038 1009
983/964 829 776 706 554

p-methyl cinnamic acid
1511 1421 1311 1284 1222 1176 1115 1040
941 877 812 771 735 685 547
2-48 1453 1284 1251 1220 1175 1116 1069
966 897 884 814 752 704 542






2-47 1512

1284 1249 1225 1180 1117 990
883 854 817 778 742 705 544
1313 1292 1254 1182 1096 1033 1010 987
885 854 818 774 741 706 540
1283 1250 1178 1083 1033 987
883 817 775 740 705 551

1494 1446 1313 1295 1248 1095/1073
1032 980 963 880 844 776 721 687 584 537

Compound 2-50 has a prominent absorption at 620 cm-1. a-trans-Cinnamic acid

has a prominent absorption at 1419 cm-1, not seen in its ammonium salts. 2-42 has a

prominent absorption at 622 cm-1.p-Methyl cinnamic acid has a prominent absorption at

1421 cm-1, not seen in its ammonium salts. 2-46 has a prominent absorption at 621 cm-1.

Pvrvlium and styrylprvlium salts

Table 2-25 is grouped by counterion. Pyrylium salts, p-methoxy-styrylpyrylium

salts and selected Hasegawa-type diolefins were subjected to analysis. The counterion is

indicated and the compound number is underlined. The strongest absorption bands are

listed as wavenumbers in units of cm-1 and have been arranged roughly in columns for

easier comparison. Bands due to counterions as well as double bonds have been indicated.

In each of the p-methoxy-substituted compounds, a C-O-C aromatic ether absorption

around 1022 cm-1 (in the case of 2-16, the gold-containing counterion it is at 1036 cm-1)
can be seen except for the compounds with a triflate counterion. In these cases the -S02-

absorption falls in this region.

Table 2-25. IR absorption bands for pyrylium and styrylpyrylium salts. The
numbers listed are wavenumbers and have the units cm-1. The absorptions are arranged in
columns with the use of italics for easier comparison.

3050 2974 2873 1633
1538 1498 1463 1445 1369 1241 1200 1036
910 (ReO4-)

3064 2976 2841 1638 1613
1525 1514 1459 1428 1366 1311 1256 1174 1113 1022
981 949 831 528
900 (ReO4-) 1588 1567 (C=C)

photodimer 2-13b
3051 2974 2837 1623
1527 1514 1462 1431 1368 1306 1255 1176 1117 1032
909 (ReO4-)

Table 2-25--continued.

3065 2982 2882
1538 1499 1469
1052 (BF4-)
3071 2975 2846
1530 1515 1462
978 949
1058 (BF4-) 1589
photodimer 2-12b
2973 2842
1057 (BF4-)

3061 2980 2880
1536 1500 1467
622 1092 (C104-)
3070 2975 2844
1529 1515 1460
977 949
622 1081 (C104-)

1452 1373 1244 1201

1429 1370 1312 1258 1176 1118
830 521
1568 (C=C)





1097 1034


1251 1180


1367 1312
1568 (C=C)


1244 1200

1257 1176 1118 1097 1022

2976 2875

623 1595 (C=C)

1428 1313 1369
miscellaneous bands: 1781,1741,1556, 690

3059 2975 2875

1608 (C=C)

1528 1515 1460

1579 1568 (C=C)



1313 1262
1351 1278 1220 1156

1350 1309


1259 1174


1029(-S02-) 636


1155 1030(-S02-) 638


Table 2-25--continued.

3099 2983 2880 1633
1536 1497 1465 1374
658 (SbF6-)
2978 2875 1641 (2845)
1522 1516 1462 1431 1306
979 946 836
658 (SbF6-)
1580 1556(C=C)

1238 1198


1264 1175 1109

3048 2971 2870 1629
1533 1491 1461 1448 1372 1240 1200
3064 2973 2871 1638 1615 (2837 1693)
1529 1514 1463 1429 1307 1261 1172 1113
983 947 836 523
1583 1567 (C=C)


3048 2971 2870 1629
1533 1491 1461 1448 1372 1240 1200
2973 2871 1641 (2837)
1537 1486 1460 1442 1368 1314 1258 1173 1116
949 852 528
1588 1563 (C=C)







Methylene-bridged dialdehydes (Table 2-2)

Table 2-26. IR absorption bands for methylene-bridged dialdehydes 2-7 2-
11.Note that the data in Table 2-26 are not arranged in a columnar fashion, rather they are
simply listed as wavenumbers, in units of cm-1.

2-7: 3046 2954 2943 2922 2852 2810 2758 1697 1681 1601 1579
1508 1247 1222 1212 1156 832 655 625

2-8: 3074 2952 2882 2842 2810 2756 1696 1601 1580 1510 1314
1265 1247 1215 1159 1049 828 616

2-9: 3040 2956 2933 2888 2861 2843 2757 1683 1605 1577 1509
1248 1216 1155 1045 972 833 651

2-10: 3071 2945 2916 2855 2842 2805 2755 1687 1598 1309 1255
1214 1156 1010 833 614

2-11 3052 2942 2922 2870 2852 2737 1687 1596 1316 1274 1265
1155 836 615

Mass Spectral Analyses

Mass spectra were obtained using a Varian MAT 7A with El ionization (70 eV) for

compounds in Tables 2-1 and 2-6. The compound number is underlined and the mass to

charge ratio listed with the intensity of the peak relative to the base peak: m/z (rel. int., %).

Table 2-27. Mass spectral analyses for pyrylium salts (Table 2-1)
206.20 (M*-HReO4, 100), 191.17 (40.29, M*-CH3), 91.11 (15.29, M*-2xtBu), 57.18
(34.47, -tBu)
206.00 (M*-HC104, 100), 191.17 (39.88), 91.00 (14.14), 57.00 (41.67)
206.20 (M*-HBF4, 100) 191.00 (41.67), 91.00 (12.21), 57.18 (37.07)
206.19 (M*-HSbF6, 100), 191.14 (44.87, M*-CH3), 169.12 (26.28), 105.06 (9.78),
91.09 (10.82, M*-2xtBu), 69.12 (10.98), 57.16 (29.17, -tBu)
206.17 (M*-H2SnCl6, 100), 191.13 (42.21, M*-CH3), 91.08 (10.76 M*-2xtBu), 69.09
(11.68), 57.16 (29.51, -tBu)
276.12 (24.22), 274.11 (38.02), 240.13 (30.21), 206.14 (M*-HAuC12 41.41), 191.17
(19.53, M*-CH3), 91.03 (19.01 M*-2xtBu), 69.08 (30.47), 57.17 (100, -tBu)

Table 2-28. Mass spectral anayses for ammonium salts of cinnamic acids (Table 2-6)
148.00 (M*- NH2tBu, 71.55), 147.00 (M*- NH3+tBu, 100), 103.00 (148 CO2H),
91.00 (20.47), 77.00 (34.27, Ph), 51.00 (32.33)
58.00 (NH2tBu CH3, 76.72)
148.00 (M* NH2CH2Ph, 50.82), 147.00 (M*- NH3+CH2Ph, 71.43), 103.00 (35.99),
91.00 (30.22), 77.00 (42.31), 51.00 (33.24)
107.00 (59.62, NH2CH2Ph), 106.00 (100)
148.00 (M* NH2C5H10, 70.35), 147.00 (M*- NH3+C5H10, 100), 103.00 (48.84),
91.00 (23.11), 77.00 (35.47), 51.00 (32.41)
84.00 (NH2C5H10, 90.12), 57.00 (21.80), 56.00 (84.00 C2H4, 29.65)
148.00 (M* NH2C4H8NH2, 70.43), 147.00 (M* NH3+C4H8NH3+, 100), 103.00
(47.17), 91.00 (24.13), 77.00 (39.57), 51.00 (35.22)
44.00 (NH2C2H4, 37.61)
162 (M*- NH2tBu, 52.31), 161 (M*- NH3+tBu, 31.54), 147 (162-CH3, 22.69), 115
(33.27), 91.00 (19.62)
58.00 (NH2tBu CH3, 100)
162 (M* NH2CH2Ph, 33.23), 161.00 (M*- NH3+CH2Ph, 19.94), 147 (15.43), 115
(20.57), 91.00 (24.37)
107 (56.96, NH2CH2Ph), 106.00 (100),
162.00 (M* NH2C5H10, 64.47), 161.00 (M*- NH3+C5H10, 40.10), 147.00(30.08),
115 (45.69), 91.00 (26.65)
84 (NH2C5H10, 100), 57.00 (27.41), 56.00 (84 C2H4, 32.99)
162.00 (M* NH2C4H8NH2, 100), 161.00 (M* NH3+C4H8NH3+, 63.36), 147.00
(47.52), 115.00 (74.81), 91.00 (24.13)
44.00 (NH2C2H4, 51.91)
178.00 (M*- NH2tBu, 38.07), 177.0 (M*- NH3+tBu, 12.27), 161.00 (14.11), 89.00
(12.56), 77.00 (12.84), 63.00 (11.35)
58.00 (NH2tBu CH3, 100)
178.00 (M*- NH2tBu, 49.65), 177.0 (M*- NH3+tBu, 14.96), 161.00 (17.25), 89.00
(17.08), 77.00 (36.97), 63.00 (14.26)
107 (63.03, NH2CH2Ph), 106.00 (100),
178.00 (M*- NH2tBu, 100), 177.0 (M*- NH3+tBu, 31.08), 161.00 (44.14), 89.00
(21.28), 77.00 (15.77), 63.00 (11.71)
84.00 (NH2C5H10, 19.26)
178.00 (M*- NH2tBu, 100), 177.0 (M*- NH3+tBu, 29.69), 161.00 (33.33), 89.00
(21.61), 77.00 (21.09), 63.00 (16.67)
44.00 (NH2C2H4,30.21)

NMR Analyses

Table 2-29. NMR analyses of pyrylium salts (Table 2-1)
1.492 (s; 18H, C4H9), 2.759 (s; 3H, -CH3), 7.683 (s; 2H, pyrylium ring-H)
1.492 (s; 18H, C4H9), 2.773 (s; 3H, -CH3), 7.716 (s; 2H, pyrylium ring-H)
1.487 (s; 18H, C4H9), 2.760 (s; 3H, -CH3), 7.715 (s; 2H, pyrylium ring-H)
1.495 (s; 18H, C4H9), 2.889 (s; 3H, -CH3), 7.781 (s; 2H, pyrylium ring-H)
1.543 (s; 18H, C4H9), 2.760 (s; 3H, -CH3), 7.715 (s; 2H, pyrylium ring-H)
1.509 (s; 18H, C4H9), 2.800 (s; 3H, -CH3), 7.747 (s; 2H, pyrylium ring-H)

Carbon CDC13
25.08 (-CH3), 28.15 (-C(CH3)3), 39.23 (-C(CH3)3), 120.53 (pyrylium ring C-H),
176.55 (pyrylium ring C-CH3), 186.90 (pyrylium ring C-tBu)
24.62 (-CH3), 27.84 (-C(CH3)3), 39.05 (-C(CH3)3), 120.09 (pyrylium ring C-H),
176.78 (pyrylium ring C-CH3), 186.85 (pyrylium ring C-tBu)
24.00 (-CH3), 27.68 (-C(CH3)3), 39.04 (-C(CH3)3), 119.91 (pyrylium ring C-H),
176.77 (pyrylium ring C-CH3), 187.10 (pyrylium ring C-tBu)
24.10 (-CH3), 27.73 (-C(CH3)3), 39.07 (-C(CH3)3), 119.97 (pyrylium ring C-H),
176.69 (pyrylium ring C-CH3), 187.04 (pyrylium ring C-tBu)
24.05 (-CH3), 27.73 (-C(CH3)3), 39.05 (-C(CH3)3), 119.92 (pyrylium ring C-H),
176.81 (pyrylium ring C-CH3), 186.97 (pyrylium ring C-tBu)
24.12 (-CH3), 27.77 (-C(CH3)3), 39.18 (-C(CH3)3), 119.83 (pyrylium ring C-H),
176.65 (pyrylium ring C-CH3), 187.42 (pyrylium ring C-tBu)

Table 2-30. NMR analyses of p-methoxy styrylpyrylium salts (Table 2-3)
1.527 (s; 18H, C4H9), 3.942(s; 3H, -OCH3), AA' BB' -signal (BA=7.047 6B=7.811
4H, C6H4), 7.636 (s, 2H, pyrylium -H), AB-signal (8A=7.21 8B=8.18, JAB=16Hz;
2H, ethylene-H)
1.493 (s; 18H, C4H9), 3.921(s; 3H, -OCH3), AA' BB' -signal (8A=7.028 8B=7.753
4H, C6H4), 7.597 (s, 2H, pyrylium -H), AB-signal (8A=7.153 5B=8.118, JAB=16Hz;
2H, ethylene-H)
photodimer, 2-12b

Table 2-30--continued.
1.348 (s; 18H, C4H9), 3.794(s; 3H, -OCH3), AA' BB' -signal (8A=7.271 8B=6.872
4H, C6H4), 7.644 (s, 2H, pyrylium -H), 5.06 (m; 4H, cyclobutane ring-H)
1.492 (s; 18H, C4H9), 3.914(s; 3H, -OCH3), AA' BB' -signal (8A=7.018 5B=7.764
4H, C6H4), 7.609 (s, 2H, pyrylium -H), AB-signal (8A=7.166 8B=8.125, JAB=16Hz;
2H, ethylene-H)
photodimer, 2-14b
1.352 (s; 18H, C4H9), 3.795(s; 3H, -OCH3), AA' BB' -signal (5A=7.304 8B=6.880
4H, C6H4), 7.632 (s, 2H, pyrylium -H), 5.110 (m; 4H, cyclobutane ring-H)

1.490 (s; 18H, C4H9), 3.919(s; 3H, -OCH3), AA' BB' -signal (8A=7.019 8B=7.764
4H, C6H4), 7.598 (s, 2H, pyrylium -H), AB-signal (8A=7.156 8B=8.097, JAB=16Hz;
2H, ethylene-H)
1.502 (s; 18H, C4H9), 3.926(s; 3H, -OCH3), AA' BB' -signal (8A=7.039 8B=7.748
4H, C6H4), 7.589 (s, 2H, pyrylium -H), AB-signal (8A=7.138 8B=8.128, JAB=16Hz;
2H, ethylene-H)
photodimer, 2-13b
1.348 (s; 18H, C4H9), 3.794(s; 3H, -OCH3), AA' BB' -signal (8A=7.242 8B=6.908
4H, C6H4), 7.589 (s, 2H, pyrylium -H), 5.06 (m; 411, cyclobutane ring-H)

1.508 (s; 18H, C4H9), 3.922(s; 3H, -OCH3), AA' BB' -signal (8A=7.027 8B=7.806
4H, C6H4), 7.611 (s, 2H, pyrylium -H), AB-signal (8A=7.194 6B=8.131, JAB=16Hz;
2H, ethylene-H)
photodimer, 2-15b
1.362 (s; 18H, C4H9), 3.824(s; 3H, -OCH3), AA' BB' -signal (8A=7.242 8B=6.598
4H, C6H4), 7.544 (s, 2H, pyrylium -H), 5.160 (m; 4H, cyclobutane ring-H)

27.80 (-C(CH3)3), 38.70 (-C(CH3)3), 55.84 (-OCH3), 112.36 (pyrylium ring C-H),
115.70 (phenyl ring C-H), 119.73 (ethylene CH), 127.19 (phenyl C-OMe), 133.10
(phenyl ring C-H), 151.99 (ethylene CH), 164.64 (phenyl ring C-CH=), 164.85
(pyrylium ring C-CH=), 184.36 (pyrylium ring C-tBu)
photodimer of 2-13
27.57 (-C(CH3)3), 39.25 (-C(CH3)3), 46.68, 49.23 (cyclobutane ring C), 55.82 (-
OCH3), 115.29 (pyrylium ring C-H), 118.46 (phenyl ring C-H), 128.64 (phenyl C-
OMe), 129.45 (phenyl ring C-H), 159.12 (phenyl ring C-CH=), 176.31 (pyrylium ring
C-CH=), 187.23 (pyrylium ring C-tBu)
27.95 (-C(CH3)3), 38.71 (-C(CH3)3), 55.89 (-OCH3), 112.72 (pyrylium ring C-H),
115.59 (phenyl ring C-H), 120.20 (ethylene CH), 127.39 (phenyl C-OMe), 133.31

Table 2-30--continued.
(phenyl ring C-H), 151.90 (ethylene CH), 164.64 (phenyl ring C-CH=), 164.85
(pyrylium ring C-CH=), 184.18 (pyrylium ring C-tBu)
photodimer, 2-15b
27.79 (-C(CH3)3), 39.27 (-C(CH3)3), 46.68, 49.31 (cyclobutane ring C), 55.89 (-
OCH3), 115.37 (pyrylium ring C-H), 118.76 (phenyl ring C-H), 129.0 (phenyl C-OMe),
130.24 (phenyl ring C-H), 159.1 (phenyl ring C-CH=), 176.5 (pyrylium ring C-CH=),
186.92 (pyrylium ring C-tBu)
27.79 (-C(CH3)3), 38.67 (-C(CH3)3), 55.80 (-OCH3), 112.64 (pyrylium ring C-H),
115.54 (phenyl ring C-H), 120.07 (ethylene CH), 127.40 (phenyl C-OMe), 133.06
(phenyl ring C-H), 151.76 (ethylene CH), 164.57 (phenyl ring C-CH=), 164.96
(pyrylium ring C-CH=), 184.22 (pyrylium ring C-tBu)
photodimer, 2-14b
27.56 (-C(CH3)3), 39.19 (-C(CH3)3), 48.95, 46.07 (cyclobutane ring C), 55.80 (-
OCH3), 115.09 (pyrylium ring C-H), 118.66 (phenyl ring C-H), 129.11 (phenyl C-
OMe), 129.67 (phenyl ring C-H), 158.72 (phenyl ring C-CH=), 176.68 (pyrylium ring
C-CH=), 186.91 (pyrylium ring C-tBu)
27.76 (-C(CH3)3), 38.68 (-C(CH3)3), 55.82 (-OCH3), 112.51 (pyrylium ring C-H),
115.61 (phenyl ring C-H), 120.05 (ethylene CH), 127.33 (phenyl C-OMe), 133.06
(phenyl ring C-H), 151.87 (ethylene CH), 164.98 (phenyl ring C-CH=), 164.73
(pyrylium ring C-CH=), 184.33 (pyrylium ring C-tBu)
photodimer, 2-12b
27.50 (-C(CH3)3), 39.18 (-C(CH3)3), 45.99, 48.44 (cyclobutane ring C), 55.82 (-
OCH3), 115.10 (pyrylium ring C-H), 118.61 (phenyl ring C-H), 127.33 (phenyl C-
OMe), 129.60 (phenyl ring C-H), 158.75 (phenyl ring C-CH=), 176.9 (pyrylium ring C-
CH=), 186.97 (pyrylium ring C-tBu)
28.06 (-C(CH3)3), 38.79 (-C(CH3)3), 55.94 (-OCH3), 112.94 (pyrylium ring C-H),
115.68 (phenyl ring C-H), 120.02 (ethylene CH), 127.5 (phenyl C-OMe), 133.37 (phenyl
ring C-H), 152.21 (ethylene CH), 164.8 (phenyl ring C-CH=), 164.8 (pyrylium ring C-
CH=), 184.13 (pyrylium ring C-tBu)

27.79 (-C(CH3)3), 38.68 (-C(CH3)3), 55.83 (-OCH3), 112.50 (pyrylium ring C-H),
115.54 (phenyl ring C-H), 120.19 (ethylene CH), 127.50 (phenyl C-OMe), 133.06
(phenyl ring C-H), 151.60 (ethylene CH), 164.45 (phenyl ring C-CH=), 165.07
(pyrylium ring C-CH=), 184.29 (pyrylium ring C-tBu)
Table 2-31. NMR analyses of styrylpyrylium salts (Table 2-5)
1.493 (s; 36H, C4H9), 4.485 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.060 5B=7.758H,
C6H4, 9Hz), 7.607 (s, 4H, pyrylium -H), AB-signal (6A=7.161 6B=8.116, JAB=16Hz;
4H, ethylene-H)

Table 2-31--continued.
1.496 (s; 36H, C4H9), 4.484 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.054 8B=7.768H,
C6H4, 9Hz), 7.618 (s, 4H, pyrylium -H), AB-signal (8A=7.179 5B=8.049, JAB=16Hz;
4H, ethylene-H)
1.505 (s; 36H, C4H9), 4.497 (s; 4H, -OCH2-), AA' BB' -signal (8A=7.078
8B=7.756H, C6H4, 9Hz), 7.601 (s, 4H, pyrylium -H), AB-signal (8A=7.153
8B=8.127, JAB=16Hz; 4H, ethylene-H)

1.489 (s; 36H, C4H9), 2.036 (t; 4H, methylene-H, alkyl chain), 4.184 (t; 4H, -OCH2-),
AA' BB' -signal (8A=7.003 8B=7.745 8H, C6H4, 8Hz), 7.585 (s, 4H, pyrylium -H),
AB-signal (8A=7.147 5B=8.118, JAB=15Hz; 4H, ethylene-H)

1.492 (s; 36H, C4H9), 2.037 (t; 4H, methylene-H, alkyl chain), 4.189(t; 4H, -OCH2-),
AA' BB' -signal (8A=6.991 5B=7.751 8H, C6H4, 9Hz), 7.594 (s, 4H, pyrylium -H),
AB-signal (8A=7.160 BB=8.121 JAB=16Hz; 4H, ethylene-H)

1.499 (s; 36H, C4H9), 2.044 (t; 4H, methylene-H, alkyl chain), 4.190 (t; 4H, -OCH2-),
AA' BB' -signal (8A=7.020 8B=7.740 8H, C6H4, 8Hz), 7.579 (s, 4H, pyrylium -H),
AB-signal (8A=7.133 8B=8.127, JAB=16Hz; 4H, ethylene-H)

1.491 (s; 36H, C4H9), 2.042 (t; 4H, methylene-H, alkyl chain), 4.186 (t; 4H, -OCH2-),
AA' BB' -signal (BA=7.018 6B=7.737 8H, C6H4, 9Hz), 7.565 (s, 4H, pyrylium -H),
AB-signal (8A=7.137 8B=8.103, JAB=16Hz; 4H, ethylene-H)

1.496 (s; 36H, C4H9), 2.355 (m; 2H, methylene-H, alkyl chain, 6Hz), 4.305 (t; 4H,
-OCH2-, 6Hz), AA' BB' -signal (8A=7.036 6B=7.730 8H, C6H4, 9Hz), 7.561 (s, 4H,
pyrylium -H), AB-signal (8A=7.127 6B=8.092, JAB=16Hz; 4H, ethylene-H)

1.486 (s; 36H, C4H9), 2.337 (m; 2H, methylene-H, alkyl chain, 6Hz), 4.294 (t; 4H,
-OCH2-, 6Hz), AA' BB' -signal (8A=7.019 8B=7.750 8H, C6H4, 9Hz), 7.595 (s, 4H,
pyrylium -H), AB-signal (8A=7.156 8B=8.115 JAB=16Hz; 4H, ethylene-H)

1.492 (s; 36H, C4H9), 1.577 (m; 4H, methylene-H, alkyl chain) 1.870 (m; 4H,
methylene-H, alkyl chain), 4.128 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.015

Table 2-31--continued.
5B=7.745 8H, C6H4), 7.581 (s, 4H, pyrylium -H), AB-signal (SA=7.143 8B=8.118,
JAB=16Hz; 4H, ethylene-H)
1.491 (s; 36H, C4H9), 1.564 (m; 4H, methylene-H, alkyl chain) 1.867 (m; 4H,
methylene-H, alkyl chain), 4.121 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.005
8B=7.757 8H, C6H4), 7.592 (s, 4H, pyrylium -H), AB-signal (6A=7.157 8B=8.123
JAB=16Hz; 4H, ethylene-H)
1.500 (s; 36H, C4H9), 1.514 (m; 4H, methylene-H, alkyl chain) 1.875 (m; 4H,
methylene-H, alkyl chain), 4.126 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.027
8B=7.738 8H, C6H4), 7.571 (s, 4H, pyrylium -H), AB-signal (5A=7.126 8B=8.124,
JAB=16Hz;4H, ethylene-H)
1.345 (m; 12H, methylene-H, alkyl chain), 1.492 (s; 36H, C4H9), 1.827 (m; 4H,
methylene-H, alkyl chain), 4.105 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.019
8B=7.739 8H, C6H4), 7.573 (s, 4H, pyrylium -H), AB-signal (8A=7.132 8B=8.112,
JAB=16Hz; 4H, ethylene-H)
1.350 (m; 12H, methylene-H, alkyl chain), 1.490 (s; 36H, C4H9), 1.823 (m; 4H,
methylene-H, alkyl chain), 4.101 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.013
8B=7.751 8H, C6H4), 7.585 (s, 4H, pyrylium -H), AB-signal (8A=7.150 8B=8.119,
JAB=16Hz; 4H, ethylene-H)
1.371 (m; 12H, methylene-H, alkyl chain), 1.499 (s; 36H, C4H9), 1.830 (m; 4H,
methylene-H, alkyl chain), 4.105 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.027
8B=7.736 8H, C6H4), 7.569 (s, 4H, pyrylium -H), AB-signal (8A=7.123 8B=8.124,
JAB=16Hz; 4H, ethylene-H)
1.493 (s; 36H, C4H9), 4.485 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.060 8B=7.758H,
C6H4, 9Hz), 7.607 (s, 4H, pyrylium -H), AB-signal (8A=7.161 8B=8.116, JAB=16Hz;
4H, ethylene-H)
1.496 (s; 36H, C4H9), 4.484 (t; 4H, -OCH2-), AA' BB' -signal (5A=7.054 8B=7.768H,
C6H4, 9Hz), 7.618 (s, 4H, pyrylium -H), AB-signal (8A=7.179 8B=8.049, JAB=16Hz;
4H, ethylene-H)

Table 2-31--continued.
1.505 (s; 36H, C4H9), 4.497 (s; 4H, -OCH2-), AA' BB' -signal (SA=7.078
6B=7.756H, C6H4, 9Hz), 7.601 (s, 4H, pyrylium -H), AB-signal (8A=7.153
81=8.127, JAB=16Hz; 4H, ethylene-H)

1.489 (s; 36H, C4H9), 2.036 (t; 4H, methylene-H, alkyl chain), 4.184 (t; 4H, -OCH2-),
AA' BB' -signal (SA=7.003 8B=7.745 8H, C6H4, 8Hz), 7.585 (s, 4H, pyrylium -H),
AB-signal (8A=7.147 8B=8.118, JAB=15Hz; 4H, ethylene-H)

1.492 (s; 36H, C4H9), 2.037 (t; 4H, methylene-H, alkyl chain), 4.189(t; 4H, -OCH2-),
AA' BB' -signal (8A=6.991 8B=7.751 8H, C6H4, 9Hz), 7.594 (s, 4H, pyrylium -H),
AB-signal (8A=7.160 8B=8.121 JAB=16Hz; 4H, ethylene-H)

1.499 (s; 36H, C4H9), 2.044 (t; 4H, methylene-H, alkyl chain), 4.190 (t; 4H, -OCH2-),
AA' BB' -signal (8A=7.020 8B=7.740 8H, C6H4, 8Hz), 7.579 (s, 4H, pyrylium -H),
AB-signal (8A=7.133 8B=8.127, JAB=16Hz; 4H, ethylene-H)

1.491 (s; 36H, C4H9), 2.042 (t; 4H, methylene-H, alkyl chain), 4.186 (t; 4H, -OCH2-),
AA' BB' -signal (8A=7.018 8B=7.737 8H, C6H4, 9Hz), 7.565 (s, 4H, pyrylium -H),
AB-signal (SA=7.137 6B=8.103, JAB=16Hz; 4H, ethylene-H)

1.496 (s; 36H, C4H9), 2.355 (m; 2H, methylene-H, alkyl chain, 6Hz), 4.305 (t; 4H,
-OCH2-, 6Hz), AA' BB' -signal (8A=7.036 6B=7.730 8H, C6H4, 9Hz), 7.561 (s, 4H,
pyrylium -H), AB-signal (8A=7.127 8B=8.092, JAB=16Hz; 4H, ethylene-H)

1.486 (s; 36H, C4H9), 2.337 (m; 2H, methylene-H, alkyl chain, 6Hz), 4.294 (t; 4H,
-OCH2-, 6Hz), AA' BB' -signal (8A=7.019 8B=7.750 8H, C6H4, 9Hz), 7.595 (s, 4H,
pyrylium -H), AB-signal (8A=7.156 8B=8.115 JAB=16Hz; 4H, ethylene-H)

1.492 (s; 36H, C4H9), 1.577 (m; 4H, methylene-H, alkyl chain) 1.870 (m; 4H,
methylene-H, alkyl chain), 4.128 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.015
8B=7.745 8H, C6H4), 7.581 (s, 4H, pyrylium -H), AB-signal (8A=7.143 8B=8.118,
JAB=16Hz; 4H, ethylene-H)
1.491 (s; 36H, C4H9), 1.564 (m; 4H, methylene-H, alkyl chain) 1.867 (m; 4H,
methylene-H, alkyl chain), 4.121 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.005

Table 2-31-continued.
8B=7.757 8H, C6H4), 7.592 (s, 4H, pyrylium -H), AB-signal (5A=7.157 6B=8.123
JAB=16Hz; 4H, ethylene-H)
1.500 (s; 36H, C4H9), 1.514 (m; 4H, methylene-H, alkyl chain) 1.875 (m; 4H,
methylene-H, alkyl chain), 4.126 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.027
8B=7.738 8H, C6H4), 7.571 (s, 4H, pyrylium -H), AB-signal (8A=7.126 8B=8.124,
JAB=16Hz;4H, ethylene-H)
1.345 (m; 12H, methylene-H, alkyl chain), 1.492 (s; 36H, C4H9), 1.827 (m; 4H,
methylene-H, alkyl chain), 4.105 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.019
8B=7.739 8H, C6H4), 7.573 (s, 4H, pyrylium -H), AB-signal (8A=7.132 8B=8.112,
JAB=16Hz; 4H, ethylene-H)
1.350 (m; 12H, methylene-H, alkyl chain), 1.490 (s; 36H, C4H9), 1.823 (m; 4H,
methylene-H, alkyl chain), 4.101 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.013
8B=7.751 8H, C6H4), 7.585 (s, 4H, pyrylium -H), AB-signal (8A=7.150 8B=8.119,
JAB=16Hz; 4H, ethylene-H)
1.371 (m; 12H, methylene-H, alkyl chain), 1.499 (s; 36H, C4H9), 1.830 (m; 4H,
methylene-H, alkyl chain), 4.105 (t; 4H, -OCH2-), AA' BB' -signal (8A=7.027
8B=7.736 8H, C6H4), 7.569 (s, 4H, pyrylium -H), AB-signal (5A=7.123 8B=8.124,
JAB=16Hz; 4H, ethylene-H)
Table 2-32. NMR analyses of styrylpyrylium salts (Table 2-4)
1.559 (s; 36H, C4H9), 7.880 and 7.842 (s: 4H, C6H4; 4H, pyrylium -H), AB-signal
(5A=7.48 5B=8.15, JAB=16Hz; 4H, ethylene-H)
1.533 (s; 36H, C4H9), 7.812 and 7.867 (s: 4H, C6H4; 4H, pyrylium -H), AB-signal
(8A=7.442 8B=8.147, JAB=16Hz; 4H, ethylene-H)
1.519 (s; 36H, C4H9), 7.810 and 7.852 (s: 4H, C6H4; 4H, pyrylium -H), AB-signal
(8A=7.437 8B=8.129, JAB=16Hz; 4H, ethylene-H)
1.523 (s; 36H, C4H9), 7.821 and 7.839 (s: 4H, C6H4; 4H, pyrylium -H), AB-signal
(8A=7.442 8B=8.133, JAB=16Hz; 4H, ethylene-H)

Table 2-32--continued.
28.10 (-C(CH3)3), 39.05 (-C(CH3)3), 114.32 (pyrylium ring C-H), 124.91 (ethylene
CH), 130.75 (phenyl ring C-H), 137.96 (phenyl ring C-CH=) 148.9 (ethylene CH),
165.08 (pyrylium ring C-CH=), 186.19 (pyrylium ring C-tBu)
27.74 (-C(CH3)3), 39.06 (-C(CH3)3), 114.17 (pyrylium ring C-H), 124.80 (ethylene
CH), 130.75 (phenyl ring C-H), 137.97 (phenyl ring C-CH=) 148.95 (ethylene CH),
165.14 (pyrylium ring C-CH=), 186.35 (pyrylium ring C-tBu)
27.80 (-C(CH3)3), 39.11 (-C(CH3)3), 114.10 (pyrylium ring C-H), 124.78 (ethylene
CH), 130.80 (phenyl ring C-H), 137.97 (phenyl ring C-CH=) 148.94 (ethylene CH),
165.01 (pyrylium ring C-CH=), 186.48 (pyrylium ring C-tBu)
28.03 (-C(CH3)3), 39.19 (-C(CH3)3), 120.03 (pyrylium ring C-H), 125.06 (ethylene
CH), 131.08 (phenyl ring C-H), 138.0 (phenyl ring C-CH=) 149.23 (ethylene CH), 164.9
(pyrylium ring C-CH=), 186.3 (pyrylium ring C-tBu)
Table 2-33. NMR analyses of methylene-bridged dialdehydes
4.412 (s; 4H, -OCH2CH20-), 7.023 (m; 4H, aromatic), 7.816 (m; 4H, aromatic), 9.859
(s; 2H, aldehyde)
2.313 (p; 2H,-OCH2CH2-), 4.234 (t; 4H, -OCH2-), 6.987 (m; 4H, aromatic), 7.618 (m;
4H, aromatic), 9.848 (s; 2H, aldehyde)
1.502 (m; 4H, -OCH2CH2CH2-), 1.795 (m; 4H,-OCH2CH2-), 3.992(t ;4H, -OCH2-),
6.919 (m; 4H, aromatic), 7.753 (m; 4H, aromatic), 9.804 (s; 2H, aldehyde)
1.335 and 1.465 (m; 12H, alkyl chain), 1.808 (p; 4H,-OCH2CH2-), 4.027(t;4H, -OCH2-
), 6.979 (m; 4H, aromatic), 7.814 (m; 4H, aromatic), 9.867 (s; 2H, aldehyde)
25.87 (methylene C, alkyl chain), 28.97, 29.22, 29.36 (methylene C, alkyl chain), 68.33
(-OCH2-), 114.68 (phenyl ring C-H), 129.72 (phenyl C-OCH2-), 131.89 (phenyl ring C-
H), 164.19 (phenyl ring C-CH-O), 190.66 (CH=O)
25.67 (methylene C, alkyl chain), 28.88 (methylene C, alkyl chain), 68.10 (-OCH2-),
114.66 (phenyl ring C-H), 129.75 (phenyl C-OCH2-), 131.87 (phenyl ring C-H), 164.06
(phenyl ring C-CH=O), 190.64 (CH=O)

Table 2-33--continued.
28.92 (methylene C, alkyl chain), 64.51 (-OCH2-), 114.72 (phenyl ring C-H), 130.08
(phenyl C-OCH2-), 131.95 (phenyl ring C-H), 163.75 (phenyl ring C-CH=O), 190.67

66.47 (-OCH2-), 114.84 (phenyl ring C-H), 130.35 (phenyl C-OCH2-), 131.93 (phenyl
ring C-H), 163.35 (phenyl ring C-CH=O), 190.64 (CH=O)

Figure 2-9. C-13 Solid state NMR of styrylpyrylium triflate, 1-1c. Note that the
spectrum has been expanded so the top of the t-butyl signal near 30 ppm is cut off.

~M I~ r-- __

180 150 140 120 100 80 60 40
Figure 2-10. C-13 Solid state NMR of styrylpyrylium perrhenate, 2-13a.


S I I I I I I .I
180 160 140 120 100 60 60 40 20
Figure 2-11. C-13 Solid state NMR of irradiated styrylpyrylium perrhenate
monomer, 2-13a. The two small signals between 40 and 50 ppm are from the cyclobutane



This chapter is concerned with the topochemical photodimerization of a

styrylpyrylium salt. The introduction to this chapter describes how a topochemical reaction

may be viewed as a special type of phase transition. If this phase transition occurs

simultaneously over the full dimensions of a single crystal, the transformation is said to be

single-phase or homogeneous. It is then explained how homogeneous product formation

within a topochemically reacting single crystal is believed to lead in turn to single crystals

of the product. The significance of this is discussed as well as a method to induce such

transformations in UV-VIS-photoreactive crystals.

The results reported in the balance of the chapter offer the first proof of a single-

crystal-to-single crystal topochemical reaction that undergoes a single crystal-to-single

crystal back reaction. The forward photodimerization reaction is demonstrated to proceed

through the formation of a solid solution of monomer and dimer molecules in the reacting

crystal over the entire range of conversion of the monomer crystal to the dimer crystal. That

is, the photodimerization is shown to occur homogeneously throughout the crystal to

produce substitutional mixed crystals of the monomer and dimer. By full crystal structure

analysis of the as-reacted dimer crystal and single crystals of intermediate conversion, it is

proven that the photodimerization under investigation proceeds under strict control of the

crystal lattice. The crystal structures are the first direct proof of the verity of Schmidt's

Topochemical Principle. The irradiation conditions are the key to inducing the

homogeneous mechanism in this photoreaction.

Topochemical Reactions: A Type of Phase Transition

A phase is a region of uniform chemical potential, that is, uniform chemical

composition and uniform physical properties and phase transitions are driven by applying

some type of gradient. For example, Figure 3-1 shows how a phase change may be driven

by a temperature gradient. In a certain temperature range, phase 1 has a lower potential and

is therefore the stable phase that is observed, at a higher temperature range, the potential of

phase 2 is lower and therefore this is the phase observed. There is a distinct transition

temperature at which this transformation takes place. Figure 3-la is an example of a first
order transition because its first derivative, (dG/dT)p, entropy, is discontinuous at the

transition temperature. An example of a first order transition is the phase change
undergone in H20 at O'C and atmospheric pressure.

Figure 3-lb shows the temperature dependence of the chemical potential for a

second order conversion. In this kind of transition, the first derivative of the potential with

respect to temperature changes continuously. While no sudden change in the entropy of the

system at the transition temperature is observed, there is a sudden change in the second

derivative, that is, the heat capacity.

Another measure of the order of a transition is the order parameter. Figure 3-2

shows how the order parameter of a system varies as the system undergoes a second order

transition. The order parameter changes continuously below the transition temperature and

then vanishes at the transition. An example of an order parameter would be a certain

crystallographic dimension describing the packing in the unit cell. In these types of

transitions, though the dimensions describing the unit cell change continuously, the

symmetry of the crystal changes discontinuously.

To illustrate this type of second order phase transition, consider the behavior of
SrTiO3, a perovskite characterized by an oxygen octahedra.106,107 Below the phase


Tt Tt

Figure 3-la.

G S p

Tt Tt T,

Figure 3-lb.

Figure 3-1. Behavior of thermodynamic parameters with temperature during classic
phase transitions.
a) A first order transition is discontinuous in the first derivatives of the Gibbs free energy,
that is, entropy, enthalpy and volume.
b)Second order transition continuous in the first derivatives but discontinuous in the second
derivatives of the Gibbs free energy, for example, heat capacity.

transition temperature, SrTiO3 is tetragonal; call the unequal cell parameters, x and y, x>y.

As the transition temperature is approached, the oxygen octahedra rotate which leads to a

corresponding change in the cell parameters. In this case, the angle of rotation of the

oxygen octahedra is the order parameter. The cell parameters change such that the rate of

expansion of y with increasing temperature is greater than the rate of expansion of x. Thus

as the temperature increases, y will become equal to x. At that point the symmetry of the

crystal suddenly changes to cubic, even though the configuration of the atoms have been

changing continuously.

A similar situation is observed for the phase transition of the AsF6- radical cation

salt of fluoranthene. At the transition temperature, the symmetry of the crystal changes

corresponding to a rotation of the fluoranthenyl cation as shown in Figure 3-2b.108 In both

cases, the continuous change in the atom configurations results in a continuous change in

the first derivatives of the free energy of the system. But because the change in symmetry

is discontinuous, they are second order transitions. These types of symmetry changes are

characterized by a group-subgroup relationship; that is, the transition is between phases of

related symmetry. For example, in the case of the fluoranthenyl cation, the high-

symmetry, high temperature phase is A2/m and the low-temperature, low-symmetry phase

is P21/c.

Among the best investigated examples of crystalline-state reactions are the [2+2]

photodimerizations of cinnamic acid derivatives.66,92-13 G.M.J. Schmidt noted in his

landmark studies on the photodimerizations of cinnamic acids that the only dimers formed

in the crystal are those whose symmetry is already present in the monomer packing.55 This

observation led to the proposal that there is a distinct type of crystalline-state reaction which

is characterized by the following: the atoms involved in bond formation approach one

another by well-defined rotations of the reacting molecules on their lattice sites and in such

a way that the smallest atomic displacements prevail. Hence the product is always

crystalline, and also, will have the same type of crystal packing (symmetry) as the parent

crystal. The reaction is controlled by the parent-lattice. Such reactions are known as

topochemical reactions, and this "least motion" rule is Schmidt's Topochemical Principle.

To contrast a topochemical reaction with other types of crystalline state reactions,

consider the following. Reactions of crystals may occur, instead, by diffusion of the

reactants to centers of reactivity or initial product formation may destroy the parent lattice.

In addition, the parent lattice may act as a template for the orientaed growth of product

nucleii leading to a crystal modification with axes having a certain orientation with respect

to the crystallographic axes of the parent crystal (topotactic). Although other types of

organic solid state reactions involve two or more phases, topochemical reactions may be


Thus a topochemical reaction may be viewed as a special type of phase transition.

According to the definition that a phase is a region of uniform chemical potential, a solution

of two substances is also a single phase. Consider substance m constituting a mother

phase M reacting to form substance d, consituting a daughter phase D. If the two

substances are miscible to form a single phase, the system may be converted from phase M

to phase D by the conversion of substance d to substance m. In this case, the gradient that

drives the phase transition is concentration. The behavior of the free energy with the

gradient, concentration, can be roughly estimated as the curve shown in Figure 3-3.109

The first derivative of this function with respect to the gradient is continuous throughout the

transformation as in a typical second order transition.

Now consider that this conversion of M to D may take place in the crystalline state

through a topochemical reaction. Thus superimposed on the chemical transformation of

phase M to D, is a type of second order structural phase transition such that we may write,
crystal CM is converted to crystal CD through a continuous series of crystals CMD which

are solid solutions of the reactant and product (in the case of a photodimerization, monomer

and dimer). In a topochemical reaction, the reaction proceeds through the rotation of

molecules on their lattice sites and thus there is no diffusion or displacement of the

molecule from the lattice site. Although the center of mass of the molecule is not shifted

from its lattice site, side groups projecting from the reacting centers, become reoriented

accounting for the continuous shift in lattice parameters throughout the conversion. The

continuous changes in cell parameters that have been observed for topochemical systems

due to side group reorientations as the system proceeds from the reactant phase to the
product phase, strongly recalls the second order transitions of SrTiO3 and the fluoranthyl

radical cation salt discussed above. However, in contrast to these types of second order

tetragonal < cubic; SrTiO3

P21/c < i A2/m; fluoranthenyl
radical cation salt

Figure 3-2a.

Figure 3-2b.

Figure 3-2. The rotation of fluoranthenyl radical cations during a phase transition.
a)The change in order parameter in a second order transition is continuous. One example
of an order parameter might be an angle of rotation of defined in relation to the unit cell.
b)The top projection shows the crystal structure of the fluoranthenyl radical cation T>Tt.
The bottom projection shows the structure at T low temperature structure.



structural phase transitions, the symmetry of the system is never changed in a topochemical

reaction. The symmetry in the reactant crystal is identical to that in the product crystal.

An interesting system highlighting the difference between a structural phase

transition which involves the displacement of molecules from their lattice sites and a phase

transition characterized only by side group reorientation, is the polymerization of the

diacetylene, 1,6-Di-(n-Carbazoyl)-2,4-hexadiyne (DCH).82 This system proceeds through
the series: CM to CMO, which is a phase transition of a monomer crystal reacting to

form a mixed crystal of monomer and oligomer, then CMO to CMO* which is a

displacive structural phase transition where the concentrations of monomer and oligomer do
not change; and finally from CMO* to Cp* which is a continuation of the topochemistry

to form a polymer. The cell parameters vs. conversion curves show that up until a

conversion of approximatley 25%, a topochemical reaction proceeds in the crystal as

evidenced by the smooth change in lattice parameters. Then a concentration driven phase

transition takes place which is manifested by an abrupt change in the lattice parameters (as

seen in Figure 3-4) and a large increase in the rate of the reaction. Thus the strictly

structural phase transition in the crystal results in a monomer packing that favors the

continued topochemical phase transition.

Homogeneous vs. Heterogeneous Product Formation

Viewed at the macroscopic level, topochemical reactions may be seen as occurring

heterogeneously or homogeneously within a single crystal.83'11.7.57 For example, with

one exception up until now,12 topochemical [2+2] photodimerizations of olefin crystals,

have been observed to proceed with disintegration of the crystal. No single crystal of the

product could be obtained. In the cases where the crystal disintegrates upon reaction, it is

believed that the product forms heterogeneously within the crystal, that is, the spatial

distribution of product within the crystal is not uniform.

------ AGofMixing
Heterogeneous Mixture
Homogeneous Mixture GD

GM...----------.-- ...


Figure 3-3. Free energy of the system as a function of dimer percentage.

The reason a crystal disintegrates during heterogeneous product formation is the

following. When a reactant molecule converts to product molecule, van-der-Waals contacts

are convened to comparatively shorter chemical bonds. Thus there will be a difference in

the dimensions between the reactant and product lattices. As the concentration of product

builds up locally in the monomer crystal, a new daughter lattice, characteristic of the

product concentration, evolves. Eventually this daughter lattice reaches its limit of

solubility in the mother lattice and nucleation of a new phase occurs. Phase boundaries

between discrepant phases causes the disintegration of the single crystal into polycrystalline


However, a homogeneous transformation--that is, a single-phase transformation--

produces single crystals of product. When the product is formed homogeneously, a solid

solution of the reactant and product is formed over the entire dimensions of the crystal.

Thus, the product and reactant form a mixed