A thesis sub~itted in part fulfilment

of the requirements fer the degree of

by

RICH.A.: D KEITH t>OLLY

Chemistry De artment

School of General Studies

Australian National University

Canberra , A. C. T . July, 1966

All the v or:( clescr 'beC.. in tl,'3 ~x1 erij mental

and discussed jn th~s thesis vas carried out by

myself, except where srecif~c reference is made to

the contributions of others .

~-X-~ -••••••••••••••••••••••

July, 1966

i

I am deeply indebted to my supervisors,

Dr . R. :b' . C. Brown and Professor f . D. 8row for their

expert guidance and advice .

I run also grateful for havin~ had the

opportunity to wor~ in the Department of Chemistry

and f'or the many helpful d"iscussions ·with its members .

My than~-cs are also r,ue to Dr . t..J . Sternhell

( Universi.ty of Sydney) for the mcasuremer1t of p . m. r.

spectra before the spectrometer v 2.s ava i J able '. ithj_n

the Department, to Dr . J . S . ~hannon (D"ivision of ~1 0·~1 J (.. -

2esearch, c . s . T. R. O.), Dr . a.s . Barnes (O . S . R.

Research Laboratories) , Dr . J . L. Occolovitz (Defence

Standards Laboratory), and to Dr . J . H. Beynon (I.J.I.

Dyestuffs Divisi.on, England; for the measurement of

mass spectra , and to Dr . G. E. S-res.m (University of

Adelaide) , Mr . J . B. Bremner ( ~ustralian Nctional

University) and Dr . J . F1 . 1 • ~rcOmie (University of

Bristol) for making availeble some of the compounds

studied in this thesis .

•

ii

SUr!J/~ARY

The react ions of or gar,.ic molecules urd er ccncJ. · t -i ori ~,

of high excitatjon ener['Y Lave been ex·1mined; iri

particular a compar1con has been mnde bet ·reen the

effects of electron j_mpact and thermal excitat-i on. It

n be argued that 1·r there is a flux of e:rier )'Y bet een ca ,

the degrees of treed om of the e., c i tea. molecule, the

reaction undergone should be s-imil&r, irrespective of

the energy source. Examples are chosen from the

literature to illustrate this concept .

Provided that the pressure of the system be

sufficiently low that molecular collisions do not

contribute substantially to the d:ssipat1 on of ener£;:r,

the decompos · tion observec on electron i .1pact should be

duplicated by heat alone . The reactions of selected

a r yl phenylhydrazones, cyclic anhydr 1 cles &nc1 cycljc

polyco..rbonyl c ompouna.s vrere compared on pyrolyr3 is and :.n

the mass spectrometer .

co1npounds and cycl j_c an..hyc1r .; des t .:1_ere ;s.s /)OC ae-·ree nent

bet reen results from the t 1'0 methods of excitat 1 on.

Hovever , in the cRse of the more flex.;-Jle phenylhy(razor es,

agreement was less satisfactory.

Anplication of the above principles h~s lead to

a new synthesis of biphenylene ~)Y pyrolys 1 G of

iii

indanetrj_one or phtr ... alic anhydride . This method

offers not only ti~e ore sten synthesis of bj phenylene

from readily available start; ng ma-:.ierials, but s.lso

the synthes:.s of substituted biphenylenes ,. ith similar

ease . As an example, the previously un1{:nov1n

octachlorobiuhen~lene has been synthesised .

iv

TABL· OF ~Gl~~NTS

. ., ............................ . 'UT\J1n RY ••••••••• • • • • • • • • • • • • • • • • • 0 • • • • • • • • • • • •

~Ji.BLE OF COITTENT . . . . . . . . . . . . .. . . . . . ........... . LIS~.1 OF 'r.l\BLES • ,. ........... 0 ••••••••••••••••••••

LIS S:.1 OF FIG U:::IBS ••••••••••••••••••••••••••••••••

I3TRODUCTION

DISCUSSION

•••••••••••••••••••••••••••••••••• 0

Aryl Phenylhydrazones • • • • • • • • • • • • • • • • • •

1, 1,4 ,4- .1.1etramethyl tetr~.l _in:-2 '::>-:J ~-0!18 • • !

Inc1anetr ione • • • • • • • • • • • • • • • • • • • • • • • • • • • •

Phthalic Anhyd~ide • • • • • • • • • • • • • • • • • • • • • •

i

ii

iv

vi

vii

1

22

L~5

51

67

1,2-DicarboYylic Anhydrides •••••.•.• • ••• 75

Sulphobenzoic Anhydride • • • • . • • • • • • • • • • • • 8~.

C oun1ar andione • • • • • • • • • • • • • • • • • • • ••• • •

Quantitative Comparison • • • • • • • • ••••••••

E- PJ:RilV~ T.\L

R 'FERElCES

• • • • • • • • • • • • • • • ••••••••••••••••••

•••••••••••••••••••••••••••••••••••••

A P ,:rnrx of T'/Iass Spectra •••••••••••••••••••••••

i

ii

ii i

iv

3enzaldehyde Phen~rlhyc.ro.zone

Acet ophenone Phen~rlhydrazone

Propiophenone Phenylhydrazone

Benzophenone Phenylhydrazone

88

96

100

135

145

V

v i

vii

viii

ix

.... ,. .I> ..

V

Acetophenonc 12ara- Tolyhydrazone

Acetophenone meta-~olyhydrazone

_,.cetophenone N- Deu tero-meta- 1 ll:.orophenylhyr=-_rG.zor e

Acetophenone, N,N-Diphcnylhydrazone

1 , 1 , 4 , 4- \.1etramethyl t etr'.11 in-2, 3-D-; one

Indanetrione

xi Sulphobenzoic nhydride

xii

xi i i

xiv

Octachlorobiphenylene

3 , S- Cyclohexadiene - cis- 1, 2-D ico.rbo::; ylic J nhyc1r· ide

1 , 2 , L. , 5 - rretramethyl - 3, S-Gyclohexadiene-cis-

1 , 2-D ic arb oxyl ic . nJ1yclr ide

xv Pyridine- 3 , 4 - Dicarboyxlic nhydride

xvi Pyrazine - 2 , 3 - Dicarboxylic Anhydride

xvii Coumarandione

xvii i 3 , 11- Benzcoumarin

xix N - }.:ethylph thal imi de

~1P...BLE 1

r1\t--0LE 2

r.rABLE 3

rrABI~ L~

TABLE 5

TABLE 6

vi

LIST OP 11ABLES

yrolysis of Phenylhydrazones and

... ela tea Compounds •••• o ••••••••••••• ~

Pyroly::~i s of Indc.netrione Solution •••

YJ.)olysis of Incl.anet1...,tone •••••••••• ~

Pyrolysis of Phthalic Anhydride •••• • •

Pyrolysis of Coumarandione and

Disalicylide ···············0••••••••!9 Quant i tative Comparison ••••••••••••••

28

56

58

70

... 1

97

vii

LIS~ (J - -11- 1u, ,s ..., _._ T -~--'

FIG. 1 . . " . • • • • • •••••••••••••••• 0 •••• 4 ~I T ri 2 ~ J. '--T• •••••••••••••••••••••••••••••• 0 • 23

IG. 3 •••••••••••••••••••••••••••••••• 2L~

FI: . L~ • •••••••••••••••••••••• 0 ••••••• ! 25

FI 1• 5 •••••••••••••••••••••••••••••••• 31

FIG. 6 ••••••••••••••••••••••••••••••• 33

FIG. 7 0 ••••• ••• 0 ••••••••••••••••••••

'IG. 8 . .............................. ~ 41 FIG 9 • ••••••••••••••••••••••••• • ;s • • • • 47

FIG. 10 • • • • • • • • • • • • • • • • • • • • • • • • • • • •••• Li.8

PIG. 11 ••••••••••••••••••••••• •••••• • 52a

Ii'IG. 12 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

FIG. 13 • • • • • • • •••••••••••••••••••••••• 62

FIG. 14 • •••••••••••••••••••••••••••••••

FIG. 15 • • • • • • • • • • • • • • •••••••••••••••• 0 • 68

FIG. 16 • ••••••••••••••• •••••••••••••••• 76

FIG. 17 ·······b•••.,•••••••••• ····~··· -·~ 82

1;7 I r' .... u . 18 • • • • • • • • • • • • • • • • • • • • • • • ••••• ! ! ~ .. 85

PIG. 19 ................................. 89

INTRODUCTION

A system of molecules may be defined as

being stable for practical pu1--ipos es if under st0.ndard

conditions they do not underr,;o any chemic81 react ion

at a detectable rate . For a system to undergo a

chemical reaction, the internal energy distribution of

the molecules must be such that sufficient proportion

have an energy equal to, or greater than, the activation

energy for the reaction. The internal ener~y of some

molecules, in particular those known as radioactive, is

sufficient to cause them to decompose to another, more

stable structure . However the internal energy of the

majority of molecules under standard conditions is belo1N

the activation energy for the lo vest energy process.

Before such molecules can undergo a reaction they must

acquire additional energy. This additional energy is

most conmonly su}Jplied in the form of heat, which

increases the population of the rotational, vibrational

and possibly electronic excited states of the molecule.

The translational energy of the molecule is also increased,

but this energy is ineffective in contri_buting to the

reactivity of the molecule until it is converted by

molecular collision into one of the above degrees of

freedom. Other means of transferring energy to the

2.

molecules include bombardment by photons, as is the

case in photolysis, bombard .. ment by high energy electrons,

neutrons or electromagnetic radiation in radjolysis,

and collision with other molecules. In thjs last cate­

gory is included the type of reaction in which energy,

as v1ell as being transferred, is also produced during

the collision by chemical interaction.

In discussing ch---mjcal react1ons it is

common priact ice to regard the structure of reacting

molecules in terms of ground state structure . However ,

the mere fact that a normal molecule is stable in its

ground state is evidence for a var iation in structure

on formation of the reactive species . It has been sho vn

for a nwnber of simple molecules, for which mathematical

calculations may be made from the absorption spectra,

that the configuration of the molecule is chan~ed by

excitation from the ground state to the first electron­

ically excited state1- 5• This direct interpretation

of structure from the absorption spectra is limited to

the low lying excited states of simple molecules. The

activation enerriy for most reactions is suffic2ently

high to enable upper rotational and vibrational excited

states of the reacting molecules to be at ta -: ned . 'J.1he

nun1ber of possible spectral transfor1na t ions in these

3.

energy ranges is so great, that a complete analysis

of the absorption spectra, even vvhen this data is

available, is not feasible . However from a knowledge

of the reactivity of a molecule and of the low lying

excited states, one can postulate possible structures

for the upper excited states . This has been done in

6 the case of acetylene by Duchesne .

Photochemical a c tivation of organic mole­

cu les is knovrn to promote either a bonding 11 electron

or a non- bonding p electron from atoms such as oxygen

and sulphur, to an antibonding excited state7. It is

then customary to discuss the consequent reactions jn

terms of these electronically excjted states . However

there are many examples in t~e literature in vhich

experimental evidence suggests thl,t vibrationally

excited states, rather than electronically excited

states, are predominant in determining the reaction

path . This is particularly so for photolysis in the gas

phase at low pressures . Thus brinivasan has sho~n that

the photolys is of 1,3,5-cycloheptatriene (I) in the gas

phase8

yields exclusively toluene (I!) if energy is not

dissipated by molecular collision. This is the case as

the pressure approaches zero. As the pressure is

increased (3,2,0)-bicyclo- 2 , 6 - heptadiene (III) is formed

a)~- a) 111 11

• Electronically Excited.

1" Vibrationally Excited.

ha,

IV

FIG. I

11

• I

II 11

5.

in an increasing proportion compared to the yield of

toluene, until in a diethyl ether solutjon with a hi r;h

frequency of molecular collisions, the rJ"i_cyclic conrpounu

is the sole product 9• Srinivasan :postulates that the

(3,2,0)-bicyclo-2,6-heptadiene is formed from an elec­

tronically excited state, while the toluene is formed

from a vibrationally excited state. Exc1.tation of the

1,3,5-cycloheptatriene by heat alone converts it into

toluene in almost quantitative yield10• The :photochemical

reaction of 1,3-cyclohexadiene to yield benzene and

hydrogen by one path and acetylene and ethylene by

another11

,12

, and of 7,12-dihydropleiadene-7,12-sulphone

(IV) to yield the acepleiadene dimer (v) 13 are further

examples of conversion of the initially formed electron­

ically excited state into a vibrationally excited

electronic ground state before decomposition to the

observed products •

Photon absorption is not the only me ans of

producing an electronically exc-; ted state. 3urnelle 7 I

has calculated that the appro::: ch of a pr1 oton to c.cet ._rlene-L/·

and ethylene15 in electrophilic addition results in a

deformation of the geometry of the molecule to a config­

uration similar to that in the first electronically

excited state. In the case of acetylene he shows that

6.

radiationless transition to the excited state is

theoretically possible if the approach of the proton

is at right angles to the plane of the molecule.

Electromagnetic radiation, or high energy

particles such as electrons or neutrons, transfer large

quantities of energy to molecules with which they collide.

This energy is usually sufficient to increase the popul­

ation of all possible low lyin~ excited states and in

many cases to completely eject an electron from the

molecule . Collision vvi th very hit:h energy "1art; cl c s

usually results in complete fra@nentation of the organic

molecule to carbon and stable gases such as hydrogen,

methane, carbon monoxide and carbon dioxide. . s the

energy of the exciting particles is reduced only the

weaker bonds are broken , and if the excited molecule

has a sufficient lifetime, molecular rearrangements may

occur . If the excitation process is carried out in a

mass spectrometer, the spectrum may be used to obtain

information of the reactions of the excited molecules.

Alternatively a knowledee of the reactions of molecules

exc~tcd in this fashion may be used to identify the

original molecule . Organic molecules in the mass

s-pectrometer are most commonly excited w·th 70 electron­

volt electrons . This energy is ample to eject an

11

7.

electron from the parent molecule, vith excess energy

of the parent ion causing further reactions to charged

ions, free radicals and/or neutral molecules . As the

possibility of the bombarding electron becoming

incorporated into the molecule to form a negative ion

is very small as compared to the possibiljty of an

electron being ejected from the molecule to form a

positive ion, mass spectrometers are normally focused

for positive ions only.

The reaction of the excited molecules to

give product ions may be folloved quantitatively, with

a precision of mass to charge ratio of better than one

unit . As this data is also obtained quic.1<:ly and from

small amounts of sample, the reactions of a great

number of compounds in the mass spectrometer have been

recorded in the literature. If the reactions of the

excited molecules produced in the mass spectrometer may

be related to the reactions of s~nilar molecules excited

b;y other means, the mass spectral information may enable

predictions of possible reaction pathways .

Prior to 1950 the use of mass spectra was

limited mainly to the investigation of the essential

features of the ionisation and dissociation of diatomic

molecules . The process of ionisation produced a vertical

8.

or Franck- Condon transition to one or another of the

potential curves of the ion. Pragmentation occurred

in that fraction of the ions initially formed above

the dissociation limits of the potential curve.

the increasing use of mass spectra for analytical

investigation of polyatomic molecules, an attempt was

made to extend the theoretical principles which yielded

successful results with diatomic molecules . Ho ~·ever,

for the larger molecules the knowledge of the potential

curves is scanty , and almost non- existent for the trans-

ition probabilities to the various curves. In 1952

Rosenstock and fellovv i,,orkers published a rate theory

which could be applied to the mass s~ectra of polyatomic

16 molecules • Since its publication this theory h2s come

under extensive rieviev as attempts uere made to compare

the theoretical predictjons with experiimental results.

The most extensive works in this field are by Chupka17 ,

who concludes that the theory is essentially corriect,

and by Steiner et al . 18

, who find that the discrepancy

between experiimental and calculated results is such to

render almost all of the assumptions of the theory invalid.

It is not intended to discuss the merits of these two

points of vievr, or to introduce other work of a similar

nature . Rosenstock and Krauss have recently published

9.

t 11 t . 19,20 th· . 'l'b . v10 exce en reviews on 1s guasi - equi i rium

theory of mass spectra . ~hese authors show that the

g_uali tat i ve features of mass s1Jectra are in accord vri th

the predictions of the equilibrium theory . Good, but

not complete , quantitative success may also be claimed

between theory and experiment .

The ori r·inal assumptions of the g_uas i ­

equil ibrium theory may be swnmarised as follovvs:

I .

II .

III .

IV.

Following vertical ionisation the molecule

has a certain amount of excitation energy

in its electronic and vibrational degrees

of freedom.

The excited molecular ion does not

decompose immediately but undergoes

several vibrations.

During these vibrations the excitation

energy is distributed in a completely

random fashion by radiationless trans­

itions amon ) the many potentjal surfaces

of the molecular ion.

The molecular ion decomposes only "Then

the nuclei are in proper conf i r~urati on

and a sufficient amount of energy has

concentrated in the necessary degree

10 .

of freedom.

On considering these assumptions it may

be seen that II and IV are also the assumptions of the

. 21 22 23 Rice - Ramsperger - Kassel ' and Slater theories of

unimolecular reactions, with the added restriction in

the latter theory of the energy being rec~uired in a

c e rtain uform" or "phase" in the critical derree of

freedom. An essential difference betv11een the two theories

is that the first theory postulates enerpy flow between

the modes of freedom for the critical mode to acquire

sufficient energy to lead to decomposition, while the

Slater theory does not consider any flow of energy after

the initial statistical distribution between the degree

of freedom.

The flow of energy between degrees of free­

dom was a contentious assumption of the original quasi ­

eg_uilibrium theory . This flow vvas postulated to occur

by radiationless transitions due to crossings among the

many potential surfaces of the molecular ion. In order

for these crossings to occur a large density of excited

electronic states near the ground state ,nras required.

2L~ It has been sho,,·n for a few molecules such as acetylene ,

propane and n - butane 25 that this density near the crounc1

state is not large . In spite of this reduced probability

11.

for radiationless transitions, only a very limited

number of molecules have been shovrn to yield products

arising from excited electronic states; benzene,

26 '?7 propylene and ethanol- are three . Hov.rever the

anharmonicity of polyatomic molecules enables rapid

energy exchange on a potentjal surface. Although these

theoretical considerations do not provide a complete

answer to the question of energy transfer, the bulk of

the experimental evi~ence indicates that rapid energy

exchange occurs among the energy levels of the excited

molecular ion. Due to the much larger number of vib­

rationally excited states accessible to the excited

n·olecular ion in its r·round electronic state, as com­

pared to the number accessible to an exc-ited electronic

state, the volume of phase space accessjble to the ground

electronic state is correspondj_ngly much vreater19 •

It follons that an excited molecule ion, for which

there is randomisation of the energy bet ·,een the access­

ible excited states, has the highest probability of

being in a high vibrational level of the ground electronic

state .

From the assumptions of the original theory

f t 16 th · · 1 °b · d · t 'b t. f o mass spec ra e quasi -e C2_Ul i rium JS ri u·ion o

activated complexes was treated as a series of competin,s,

12.

consecutive, unimolecular decomposition reactions for

which the rate constants could be calculated by means

of an appropriate form of absolute reaction rate theory.

'l.1he calculations derived from the Rice-Ramsperger-Kassel

theory for unimolecular reactions, which are eq_uivalent

to the calculations of the deEenerate oscillator model,

yield predictions which are at least jn qualitative

agreement v,1 i th experimental mass spectra results. These

calculations were originally derived as rate expressions

for react ions occurring in the gas phase which shovred a

unimolecular dependence on pressure, and are still ~idely

used for this type of reaction at the present time.

The preceding discussion would suggest that

the closest approximation to a molecule excited in the

mass spectrometer is a thermally excited molecule . Also

the decomposition of this molecule may be treated by the

same kinetics as applied to unimolecular gas reactions.

It is therefore not surprising, but rather to be expected,

that there vvould be a parallel between mass spectral

reactions and pyrolytic reactions. As the rate express­

ions are for unimolecular reactions, this re ouir e s the

:pyrolysis to be carried out under conditions for which

unimolecular decomposition may occur. Ho· 'eve care must

be exercised in the interpretation of mass s pectra reactions.

13.

A mass suectrum is a record of positive ions only; the ...

neutral and free radical fragments must be assumed to

complement the recorded fragments. The varjation in

total excitation enerfy will be reflected in the products.

The degree of bond fragmentation Yvill depend on the amount

of energy transferred to the reacting molecule.

excitation energies the excited molecule presumably vrill

have a short lifetime and break rapidly into simple

fragments . As the excitation energy is reduced the

lifetime probably ~ill be increased and molecular rearr­

angements may occur. As the excitation energy is reduced

still further, it is possible to produce in the mass

spectrometer an excited molecular ion uhi ch has a suffic­

ient long lifetime (approximately 10-5 second), for it

to pass through the mass spectrometer witV)out fragmentat-

ion.

There are many examples in the literature

of the similarity bet~een pyrolytic and mass spectral

reactions . In the majority of cases the results have

been obtained independently by workers with other aims

in view. Thus the transition state for the formation of

alkenes from esters has been investigated for pyrolyt~c

reactions by Macco1128 and P.or'<:oNski and Ausloos29 , and

in the mass spectrometer by t.TcLafferty30 • Recent

14.

isotopic work by Djerassi and Fenselau31 has shown

that the transition state is not specifically six

membered as previously postulated, but three, four and

five membered transition states also occur. A similar

pyrolytic investigation may reveal that this is also

the case for the pyrolytic transition state. The loss

of carbon monoxide observed in the mass spectrum of

maleic anhydride32 is also obtained on pyrolysis of

both unsubstituted33 and substituted34 maleic anhyc1rjc1e.

Cyclobutane has been shown to yield ethylene both on

35 36 pyrolysis and in the mass spectrometer .

The nurriber of authors ~.rho have commented

on this similarity between mass spectra and pyrolysis

is very limited . Gibson and Murray37 sho-ved horr the

9yrolysis of trimethylaminebenz imide to yield

trimethylarnine and phenylisocyanate may be related to

peaks of the corresponding mass to charge ration in the

mass spectrum. Fields and Meyerson have extended the

work published from this thesis38 on the generation of

benzyne from indanetrione, to phthalic anhydride39•

These vorkers have also extended their investigation of

the relation between pyrolysis and mass spectra to a

study of deuterated benzene40 , and also include

radiolysis in this study of excited benzene. Their

15.

conclusion was that all products could be largely

accounted for without involving electronically excited

species as immediate precursors . s those reaction

paths with the lowest activation energy, which vould oe

expec ted to be favoured in thermal processes, would also

predominate when excess energy was available, they summ­

arise that the same low energy reaction path may be

favoured in both pyrolysis and electron bombar&nent .

The obvious extension of the mass spectro­

metri c exc itation process to normal laboratory conditions

leads to molecules excited by what has been termed

radiolysis . As the energy of the exciting particles or

radiation in this field varies from millions of electron

volts down to the mass spectral range of tens of electron

volts , the degree of fragmentation of the molecL1le varies

considerably. Most comparative wor l{ has been done with

oxygenated compounds and similar1 ities have been observed

in both fragmentation and molecular rearrangements

u.1 betv1een the radiolysis and mass spectra of esters· ,

ketones42 , 43 , acetic acid and ethane44

Photon bombardment may also be used to

excite molecules in the mass s~ectrometer . It has the

advantage that a monochromatic energy beSJ11 may be

obtained , which is an important consideration in

16.

theoretical investigations, but the total enerGy is

limited by the transmission of the optical vd.ndows

used in the construction of the anparatus. It follo1s

from the quasi - equilibrium theory, in Y.ihi ch there j_s

relaxation of the internal energy from fixed energy

levels , that apart from total energy distribution, the

means of exc itation should be i :relevant . If this

relaxation process may be extended to conditions other

than in the mass spectrometer , photolytic excitation of

molecules should yield reactions comparable 1:rith mass

spec trometric and pyrolyt ic excitation. The e~uilibration

of energy is favoured by higher excitation, vrhich

inc reases the anharmonicity of the polyatomic molecule.

If the internal energy may be dissipated by molecular

col lision , the possibility of reaction from an isolated

electronically excited state are increased. This Tias

the case for 1 , 3 , 5 - cycloheptatriene as discussed earlier

in this introduction. It is therefore not surprising to

find that most successful carrel tions between photolys·s

and mass spectra are for photolysis in the vapour state.

Examples of this include the photolysis of

secondary butyl acetate to isomeric butenes29 , in whi ch

the authors investigate the effect of molecular collis­

ions on the isomer ratio by vc1.rying the vapour pressure

17.

to include a liquid phase, and by varying the excit­

ation energy. Meyerson has also commented on the

striking similarity between the mass spectra and

photolysis of ketones43 , while Turro et al. have com­

pared the mass spectrum and photolysis of small ring

ketones45 • Manning also observed the sj_m_· 1ari ty between

the photolysis and mass spectra of satu1--iated aldehydes

and ketones46 • Turro postulates that the similarity

may be due to the similar polarisation of the n;rr~

excited photochemical state and the mass spectrometric

molecular radical ion. The molecular radical ion is

considered in terms of ground eJectronic state struc­

tures . Honever Manning, in his work published in 1957,

assumes that both mass spectral and photolytic excitat­

ion give rise to a similar electronically excited state,

and he discusses the nature of this excited state and

how the observed products may be obtained from it.

This asswnption of a djstinct mass spectral electronically

excited state is not supported by the bulk of experi­

mental evidence .

In the preceding discussion examples have

been chosen to illustrate the general natur·e of excited

molecules, which cannot be adequ~ tely desc · ibed in 1/'rms

of specific energy levels . Por suitable mol0cules under

18.

suitable conditions it is postulated that the energy

of the excited molecule is not confined to a definite

energy level, but statistically distributed in a

dynamic system among all accessible energy levels. The

reactions of these excited molecules should be similar,

v.ri th the reaction being determined not by the means of

excitation, but by the energy distribution. For mass

spectral and pyrolytic excitation, the molecule has a

high probability of being in a high vibrationally

excited ground electronic state. For limited pyrolytic

excitation this probability is one, but for more vigor­

ous excitation there is a probability of the molecule

being in an excited electronic state. Homever unless

the molecule has hi ")h thermal sta.bj li t;y , and a distinct­

ive reaction which is knoy·n to occur in .'.J.n excited

electronic state, the presence of the exc~ted electronic

state will not be detected in the dynamic system.

The mass spectrometer enables the reactions

of excited molecules to be studied quickly, accurately

and from small amounts of material . As has been shown

previously, the energy distribution of a thermally

excited molecule under conditioLS in which molecular

collisions cannot contribute to the dissipation of

energy, is the closest approximation to the same molecule

19 .

excited in the nass snectrometer . Fov1ever, the PT1essure

at ·which molecular collisions become ineffective is

a matter of speculation and ,ill depend to some extent

on the structure of the molecule . If j t is assume c1,

for comparative purposes, that all the energy of the

bombarding particle is transferred to the reacting

molecule in the rnass spectrometer , then the energy is

kno\rn from the energy of the bombarding particles . The

energy of thermal]yexcited molecules may be calculated, if

the t 0 m:perature and specific heat of the nolecule are

kno 1.-m. Al though the equili"brium tem~oerature of the

mole cule may be measured, the specific heat js a function

of the absolute temperature , and this thermodynamic

data is available only for a very f e·.-r compounds over a

limited temperature range . Choosing naphthalene as s.n

0 example , and assuming the srec if"i c he~. t at 300· C

(op = 5S . 9 calories/rnole/c_egree1

+·7 ) ac the mean specjfjc

heat over the range o0 c to 6oo 0 c, the energy :ained by

the molecules over this temperature range 1 s 36

kilocalories/ mole . This is far less than the 1,610

kilocalories/mole equiv lent of the standard 70 electron

volt bombarding electrons normally used in mass spectro­

metry. Even when the mass spectral ficure is reduced

to allovr for the energy which is not transferred to the

20.

bombarded molecules, and the pyrolytic figure is

increased for compounds which are less symmetrical than

naphthalene, and so have a greater specific heat due to

the increased number of effective oscillators, the mass

spectral energy is still lilrely to be much higher. If

the number of competing reaction paths is limited, this

difference in the energy of the excited molecule vill

only have a small effect on the product distribution ,

but as the number of competing reaction paths is increc1scd,

the total energy difference for the var 7 ous means of

excitation will have a greater influence on the products

obtained.

The compounds studied in this thesis were

chosen to investigate the extent of the relation between

mass spectral and pyrolytic react ions . Kno·;rledge of a

published mass spectral reaction wa s used as the basis

of the possible similar pyrolytic reaction. The type of

reactions studied may be divided into two broad categor-

ies .

I . The reactions of phenylhydrazones of

arormtic aldehydes and ketones . These

are flexible molecules which show a

number of competing reactions in the

mass spectrometer. Although mass

spectral fragmentation predominates ,

II .

21.

rearrangement reactions also occur.

The reactions of cyclic carbonyl

compounds containing a fused aromatic

ring . These rigid molecules generally

show the initial mass spectral expuls ­

ion of a neutral molecule of carbon

monoxide or carbon dioxide. Competing

reactions are very limited with

practically no scope for rearrangements .

The second category is mainly concerned with

possible precursors of the unstable 1 , 2 - dehydrobenzene

intermediate , but the fragmentation reaction was also

extended to systems containing a fused hexadiene ring

and to possible synthesis of benzocyclobutene derivat­

ives . Although initially the pyrolysis 7as investi gated

with the known mass spectral reaction as a ruide , as

the experiments were extended to related compounds, the

pyrol ys i s data was knovrn before the mass spectral

information became available.

22.

DISCUSSION

Af YL H:BJITYLHYDRAZONES

The dominant :peal(s in the mass spectrwn of

benzaldehyde :phenylhydrazone are those with 1nass to

charge ratios of 196, 195, 93, 9~, 77 and 65 . All these

ions, with the exception of the last, may ari.se by sim­

ple fragmentation of a single bond. The ion, m/e = 65 ,

requires the fragmentation of two bonds and the elimin­

ation of a neutral molecule of hydrogen cyanide. The

presence of tv,o ni tro,,.en atoms in the molecule of

phenylhydrazones complicates the interpretation of the

fragmentation pattern, as there are tvvo pairs of non­

bonded electrons from which the initial electron may be

removed to form the parent ion. Two possible forms of

the parent ion are shown as separate identities as the

ions (VI) and (XIII ) in Fig. 2 and Fig . 3 respectively,

but they are undoubtGdly in resonance through the bonded

electrons of the nitrogen-nitrogen bond. Interpretation

of the fragmentation is further complicated by hydrogen

loss, ~iliich may occur as the initial fra~mentation, or

at a later stage. Although removal of the hydrogen atom

from the v7eaker ni trogen-hydro f'en -bond ((VI)--- ( VII)

or (XIII)---(XIV)) would seem the favoured process,

00 + • - H• y==N-~ - • • * H H m/e = 195

V 111

-Oi=N·

+ N~ I I J •

I H

m/e = 91

IX

(Q)-7=N-vo H H m/e = 196

VI

- H~ (Q)-7=N-N{)) H

m/e = 195

V II

- Oi=N· -orN•

- H• + N I H m/e = 92

X

· I -HCN

0 m/e = 65

XII

FIG. 2

- H • +. N

m/e = 91

XI

¥ Z-I ¥ z 11

+z

11

6I • I I •

-&

I I-Z

-I ¥ Z-I

lO

I m

>

+Z

Ill u

[QJ

It -X

C

l)

' E q-0

• I I

+Z Ill U-I

-I

-•Z

-I

G>

>--

...

-'

+Z

-

Ex

II

~x

-' ~

+z

¥ •Z-I

z u I

• +Z

rt)

Ill 0 -

-u

It -

[QJ~~

<..?

LL

CX) (D

X

z+

II

X

Q.)

' E (j) (D

X

• II

I-Z

+

X

Q.)

' E

~1 ••

CX) (D

I-Z

+

II X

Q

.) X

© ' E

z u I • z u I z u I •

+z

It')

I (j)

z II

>

il-I Q

.)

' E

©

~f·

•z-I

U> ·1

(j)

Z

II >

il-I ~

©

+Z

-I ~

I z II

-

~ -J:

~>

E

z u --e-z u --e-z u --e-

II

Z-J

: ~

X

+

E

X

• I Q

f'() (j)

II X

Z

-I

I-+

•

~x

E

• I 0 C\J en

>

I-Z

-I

II X

+

~

X

E

~ .

(!)

LL

26.

~-- .. ' • • _, -· , • .., J , ., -· p Y"'\ l.J1-~c.Lf::il,JlUl Jt.', c_1_·e; GO -,1~C J.O, SL.:. f.i:.

:orL1 ~c rrotcn. The mass spectrum of acetophenone

N- deutero- meta-fluorophenylhydrazone still retains a

prominent (P - 1) + peak. Evidence for the loss of an

aromatic proton from the para positjon is provided by

the ,ark of Bose48 , who sho~ed that the (P - 1)+ peak

in the mass spectrur11 of amid ines was replaced by a

(P - 2)+ peak in the mass s~·ectrum of the uara- deuter")o

analogue.

Apart from those ions vvhich arise by

fragmentation, there are other minor pea<:s jn the mass

spectrum, the ions of iNhich must be forrned ·hy molecular

rearrangement . Such procer;ses may oe rer)rer,ented by the

transitions (XIII)--- (XVII) and (:CIII)--- (XVIII)

in Fig. 3, and by the transitions shown in ~ic . 4. Bose

et al . have shown that rear")rangements of the type

(XIII) ---- (XVIII) occur on electron impact in amides

and amidines , as wel l as phenylhycb. ... azones49 • It ·was

this report which prompted the investigation of phenyl­

hydr0zones, to see if the pyrolysis could be related to

the electron impact reactions, in spite of the large

riumoer of competing pa th,JI ays from the latter means of

excitation.

27.

The pyrolysis of benzaldehyde phenylhydra­

zone , which at 220°0/720 mm yields benzylideneaniline,

trans - stilbene, ammonia, benzene and nitrosen, has

50 previously been reported by Chattaway et al. • This

vrork was repeated under similar conditions and

benzonitrile , aniline, diphenylmethane, benzalclehyde

and a trace of toluene , as well as the above products,

were obtained. The presence of the benzaldehyde indic­

ated that atmospheric oxygen was involved in the reaction.

As an approximation to the condj_t ions of the mass spectral

reaction, the pyrolysis vvas carried out by subliming the

phen;ylhydrazone at reduced pressure into a packed silica

tube maintained at a selected temperature. Under the

cond.it ions listed in Table 1, benzonitrile and an· 1jne

v1ere the major products, being contaminated v,ith only

trace amounts of other materials. The for~ation of these

products could be interpreted as a radical fission,

recombination process or alternatively as a concerted

four - centre reaction. Possible radical fission mechan­

isms are sho7n by reactions (a) to (e) in Fig. 5 (R = H)~

and possible recombination of these radicals by react­

ions (f) to (m) . This diagrammatic representation of

the fission and recombination is a simplification of

the actual process . Accordinr, to the Franck-Rabinowitsch

TABLE 1

PYROLYSIS OF PHE!\T\'LHYDRAZCHES AND R~LA'lED CON~POUr:Js

a Compound No . T°C/~(mm) 11R( sec) rroduct '1 .l clds

$JON 0FH 9"CH:N¢ ,0H -·2 -

)bCH:N . NH¢ Al 800/0 . 3 0 . 08 0 . 84 0 . 76

A2 600/0 . 1 0 . 03 0 . 82 0 . 57 0 . 08 0 . 10

A3 240/720 - 0 . 12 0 . 13 0 . 17 0 . 16

,0CH:F¢ B 800/0 . 005 0 . 01 0 . 03 0 . 004 o. 96 trace

¢CN C 800/0 . 005 0 . 0001 1 . 00

9'NH 2 D 800/0 . 01 0 . 002 -- 1 . 00 ~FHRb

CIV~ e : N. NH¢ El 800/0 . OL.J. G. 09 1 . 10 0 . 36 0 . 24

E2 600/0 . 03 0.09 0 . 94 0 . 31 0 . 43

¢NHMe F 800/0.1 0 . 02 0 . 03 0 . 28 0.39

¢CEt:N.NH¢ G 600/0.3 0 . 02 l . ?2 0 . 22 0 . 41

¢NBEt H 800/0 . 1 0.02 0 . 02 0.55 0 . 09

a Yields in Moles/Mole Pyrolysed

b From derivatives of ketones ¢-CO- R or amj_nes ¢-NHR

TABLE 1

Compound No. T°C/P(mm) TR(sec) ,0CN ¢NH 2 ,0NHRb ¢.¢ ¢.H

¢ 2C:N.NH¢ Jl 600/0.3 0.03 1.16 0.35 0.35 0.04 0.2

J2 600/0.1 0.01 1.05 0.28 0.35 0.02

J3 400/0.05 0.006 0.11 0.06 0.16

¢ 2NH K 800/0.01 0.002 - trace 1.00 trace

¢CMe:N.NH¢c L 600/0.04 0.06 1.13 0.19c 0.37c

¢CMe:N.NH¢d M 600/0.02 0.03 1.03 0.23d 0.32d

¢CMe:N.N¢2 N 600/0.05 0.07 0.61 trace ~ 0N~.re¢

trace 0.23

¢NH0 0NH¢c 0NlI0e

0.22 c.06 0 . 14

¢CH:N.NH¢ A3 240/720 - ¢Ti."e ¢CH 2¢ ,0CHOf t-¢CH:CH¢

C. 02 0.10 0 . 31 0.15 ----

C ¢ = p-Me-¢ d . ¢ = m-vre-¢ e ¢ = o-}le-¢

f Air was not specifically excluded.

3l.

F . . 7 5] r J nc ip __ e , the combinat ion of radicals j s accorLpaniccJ

by the liberation of energy ,h~ch is su f~cient to cause

s_ponte.neous f issj on of a s im:ple moJ e cule , unless some

energy is removed by a third i')O.::t: cle. r:i'he tl ircl

,.. 1 , , b ..; r} r.,- • · .,___' 1- ·' 1 .:. .L "t- C f r J.l.\,lc.r .l.--b avl.' -· C IJ, .... ~ l, .,,1 .I

f: C J , l L 1 f. C l E C 1.,: J. S

(fer res.ct icns in -'·J e J ic ... ui :phase), or the ~·,al l ( for

gas phase r ecctj_ons) . F1or Fimplicity, radi cal abstraction

reactions are also not represented in Fig . 5. In this

type of reaction , the collision of a free radical ·with

a molecule results in the abstraction of a radical from

the molecule, with the formation of a neutral molecule

from the former free radical, and a free radical from

the former molecule . Reactione (f), (k) and (h) are most

likely of this type.

If the benzonitrile and aniline arc formed

by a radical process, this Y/ould j_ndicate that the

fragmentation reactions (c) and ( a ) ( ..?j_[; . 5) are -promjnent .

The formation of only slig' t amounts of benzene shovrs

that the fragmentation reaction (e) is not significant.

As djscussed in the preceding paragraph , the - adical

coinb ina ti on reactions are not as si.mple as shown in Fig .

5. The concentration of free rad i cals in proportion to

the neutral molecules will be low. A possible

exception to this is the concentration of hydrogen

~-y==N-~-~ R H

~-C=N-N-~ I I R H

~-~=N-~-~

~-C==N. I R

~-C=N. I R

~-N I • H

~-N I • H

~-~-~-~­~-R.

R.

H

+ H.

+ R.

+ ~. + R.

+ ~-

+ H.

+ R.

+ H.

a. ~-C=N-N-~ H. ~ + I • R

b . • ~-C=N-N-~ + R. • I

H

C. ~-y==N • N-~ .. +

• I R H

d. C = N ~ + R.

e . • ~- + R-C= N

f. ~-N-H • I H

g. • ~ -~-R H

h. ~-N-~ ~

I. ~ ~-R

J. •

k . • ~-H

R-R

m .• R-H

FIG. 5

,.

32.

radicals. The coking of the walls and pacl::ing during

the pyrolysis indicates that some com~lete destruction

of the phenylhydrazone molecule occurs. The fo1--imation

of this carbon must be accompanied by the liberation of

the other a toms as free radicals, or more 1 il<:ely as

stable molecules such as n1ethane, hyc_rogen cyanide,

nitrogen and hydrogen. The concentration of the

hydrogen, either as a reactive gas or free radicals,

will be such that any other reactive specjes in the

system will have a high probability of react ing vrith it.

The most likely radical termination reactions may be

represented by (f), (k) and (m), with there being only

a slight probability of the combination reactions sho,rrn

by ( g) and (j) . Due to the relatively high concentrat-

ion of neutral molecules, hydrogen abstraction by the

free radical would be a possible process by which the

probability of conib j_nation of free radical with hydrogen

would be increased.

Al thotF"h the pyrolytic fragment at ion ( c) is

identical v1ith the mass spectral fragmentation

(VI) : (X) or (XIII) ~ (XV), a pyrolyt.ic rearran[;e -

ment is also possible which parallels the similar mass

spectral rearrangement. Possible sterically alternative

four - centre concerted mechanisms are sho ,n in FiE . 6

H H H

R N ¢ a I R----N ¢' R N ¢ ~ I I

I I I I I +

¢ I I

C N 1¢ I c----N I ¢ C N

'

R = H, Me, Et, ¢

H H H

¢ ' ¢ b ¢ N ¢

N I ¢- - - -N ¢1 I I

I I I + I

R C N IR c----N I R C N

FIG. 6

I

I

11

34.

(R = H) . The observed products are fully explained

in terms of reaction (a), with no evidence of the less

sterically favoured (b) occurrin~ to any extent.

As the pyrolytic reaction of benzaldehyde

phenylhydrazone may be satisfactorily explained in

terms of a free radical reaction, or a concerted four­

centre react ion , the pyrolysis of acetophenone

phenylhydrazone was undertaken in an attempt to diff­

erentiate between the tTio processes . The formation of

N- methylaniline in 43% yield at 6oo 0 c is clear evidence

that the four - centre mechanism (react-ion (a), Pj_F . 6)

is operat~v e . This amount of N- substituted aniline

cannot be formed by a radical mechanism due to the lo

prob ability of a methyl radical com1J ininr-; vri th an

aniline radical . The presence of these radicals vrould.

be indicated by the formation of methane and aniline.

As aniline is formed in 31% yield at 6oo0 c and 36% yield

at 8oo 0 c the radical mechanism is not absent at either

temperature , but becomes more prominent with increasing

excitation energy. As no benzene or diphenylamino was

obtaine in the products, neither the fragmentation

reaction ( e) (Fi~. 5) nor the rearrangeme~t reactjon

( b ) ( Fig. 6) can be considered.

Although the formation of N-methylaniline

35.

is inconsistent with the forrrution of free rad~cals,

a radical-cage type of mechanirm cannot be cl..iscountecJ.

In this mechanism the radical remains wjthin the

electronic influence of the parent nolecule and does

not exit as a free radical. For this case, arguments

on the statistical recombinatton of the radicals are

invalid, and this type of radical mechanism may apply.

However the difference betv,een the radical-cage and the

concerted mechanism is only in the formal representat ­

ion of the bond breaking and formation. In the later

case this is considered as a concerted process, vrhile

in the former it is not. As such, for the results under

discussion, the radical-cage mechanism may be considered

as a variant of the four-centre mechanism.

The absence of other compounds from the

statistical combination of the free radicals, as expected

from the reactions as shown in Fiq . 5, sugGests the

existence of the four-centre mechanism. hS no atte~pt

was made to detect and identify ["8.ses under the low

pressure experimental conditions hovrever, reactions ( 1)

and (m) for R = H and Me v.rould not have been detected.

This difficulty was overcome by examination of the

reaction in which R =¢,the products in this case are

biphenyl and benzene respectively, both easily identified

11

I

in small quantities by ga s liquid c hromBtography. The

pyrolysis of benzophenone phenylhyorazone confirmed the

predominance of the four-centre mechanism over the free

radical mechanism for the migration of rroup - from

carbon to nitrogen. The format-Lon of aniline and benzene

in significant quantities indicated that the radical

fission reactions (c) and (d) (Fis . 5) occurred to an

appreciable extent . The format ion of only mi nor amounts

of biphenyl , as compared to the yield of aniline and

benzene , is definite experimental evidence for the pres ­

ence of the laree excess of hydrogen in the reaction

system. The absence , or presence only in very small

amounts , of products of radical combination with mole­

cules other than hydrogen , is further· evidence that the

forrration of N- substituted anilines as a major product

is primarily due to a concerted mechanism. For the

phenylhydrazone of the symmetrical benzophenone, the

t-vro sterically alternative four - centre me chanisms (Fig .

6)are equivalent .

The experimental evidence for the exjstence

of the four - centre mechanism is stronr, . Hor,ever for a

flexible molecule a six- centre transition state would

seem to be favoured as the deformation of the bond ar.gles

would be less . Another factor 1vvhich must be considered

37.

in the formation of the cyclic transition state is the

proximity of the reacting centres. This factor forms

part of the probability or entropy effect, and would

favour the smaller transition state. In the case of

acetophenone phenylhydrazone , the existence of the six­

centre transition state vuuld be manifested in the

forrra.tion of ortho - toluidine (Fig. 7), althouch once

again this could arise by a radical process . This prod­

uct was not detected, and it was of interest to see if

the larger transition state was involved in the pyrolytic

reaction of propiophenone phenylhydrazone, in which it

is not necessary to break the conjugation of an aromatic

ring to form the transition state (Pig. 7). Prom the

results ( G, H , 'I1able 1) there is no evidence for the

six- centre process, the high yield of N-ethylaniline

being explained in terms of the four-centre transition

state ( reaction (a) , Fig . 6, R = Et), and the aniline by

a radical process (reactions (c), (d) and (f), Fig. 5).

A puzzling feature of the experimental yields

is the forrm.tion of more than one mole of benzonitrile

per mole of parent phenylhydrazone, in the case of

acetophenone , benzophenone and propiophenone. The excess

benzonitrile must presumably arise from the substituted

anilines by a radical process as sho\m in reaction (c),

+ xrcN

+ 0CN

b +

H

FIG. 7

39.

F1·a 7 Evidence for this lies in the formati on of C> . •

benzonitrile from the pyrolysis of the substituted

anilines themselves, and the first stage of the

proposed pyrolytic reaction is akin to a published

mass spectral cleavage of amines51a. The thermal

isomerisation of phenyl isocyanide to benzonitrile -2

is also known:J • Hovrever, attempts to detect the

partial conversion of the aromatic nucleus from the

original phenylhydrazine into benzonitrile, by the

py:-eolysis of tolyhydrazones vrere not successful.

Although benzonitrile was still obta1ned in greater

than mol8.r quantities from the pyrolysis of a

proportionate mole of the meta and uara - tolyhydrazone

of acetophenone, a careful examination by gas liqujd

chromatography and proton magnetic resnnance

spectroscopy failed to detect any meta or para-tolunitrile

in the products . However it j_s possible that the choice

of a methyl substituted phenylhydrazone vvas a bad one,

as the methyl group r.rould be lost more readily as a

radical than hydrogen.

The factors controlling the direction of the

reaction must remain a matter for sreculation. The

non- bonding electrons of the nitrogen may initiate the

migration of the group from car½on to nitrogen. The high

thermal stability of the benzonitrile is undoubtedly the

40 .

reason for its formation in large yields . Hov1ever

product stability may not be the only factor, as sho·.-·n

by the case of' the :propiophe none phenylhydriazone, the

more stable ethylene and an.:_line are not forr11ed in

preference to N- ethylaniline. Althou~h the le Es strained

six membered transitjon state is not formed in preference

t o the four membered transition state, stertc factors do

play a part , as shovrn by the absence of the less sterically

attractive of the t\YO four - centre mechanisms. To see if

bul ky groups on the nitrogen would hinder the migration

t o th i s atom , the pyrolysis of acetophenone N, N- diphenyl­

hydrazo ne ·was investigated. The results (N, Table 1)

indi cated that radical fission- recombination, as evident

in the yield of cliphenylamine , was no more pronounced

than for the smaller hydrazone . The extent of the

methyl migration appeared not to have decreased, although

the yield of the N- methyl amine was less . This

discrepancy is due to the operation of a secondary methyl

migration of the Hofmann- ~Jartius typeS~,S4,~S. T11is

rearrangement (reaction (a) , Fi~ . 8) has been studied

mechanistically in more recent years and all evjdence

indicates that it j s intermolecular in nature, al ~.,hough

there is a difference of opinion on the nature of the

attacking species . Hughes and Ingold in a reviei,-, of the

R/

R-~-x-- R+

C-N +

/

R-N-X

Reaction a

Lewis acid

l

N H

Reaction b

FIG. 8

R+ or RX

+ H3C

/ R-N-H

I I R + x-

0 and Q

H3C-N

+

C=N

N H

II

1,

42.

work published up to 195256 favour a carbonium ion as

the reactive species in an electrophilic type substitution.

A similar species has been proposec1 in the rearrangement

of methylphenylskatylamine57 ancl of N- cL -phenethylaniline 58 •

Ogata et al . by an intensive analysis of the products

from the pyrolysis of N-methylaniline with halogen and

Levvis acids. concluded that the formation of polymethylanil­

ines and free aniline was indicative of an intermolecular

rea.ctj_ on59 • HoHever , as they obtained a similar product

distribution from the pyrolysis of aniline hydrochloride

in the presence of methanol, and rom the py1,olys is of

N- methylaniline in the preqence of hydrochloric acid, the

alkyl halide was proposed as the attacking specjes .

The isolation of K- phenyl-ortho and nara ­

toluidine from the pyrolysis of acetophenone

N,N-diphenylhydrazone (reaction (b), Fig. 8) is unique,

in that no previous examplesof this type of alkyl

n1igration have been published, in v•hich a nucleophilic

species has not been present in the reaction mixture.

s~eculation on the mechanism of this pyrolysis may be

premature at this stage, as the investigation is being

extended to other N,N-disubstituted phenylhydrazones.

The absence of a nucleophilic species eliminates an alkyl

43 .

halide as the attac1cinP;; a3;ent . Also there is no ev~dence

for an intermolecular reaction as no polym, thyl­

diphenylamines v1eL·e detected. An j ntramolecular reaction

would seem the best explanation of the available evidence,

r1ith the development of quaternary bonding to the

nitrogen of the intermediate in reaction (b), Fi·. 8

being followed by methyl migrat·on into the phenyl ring.

This later migration may involve the TT electrons of the

phenyl ring , or a radical cage mechanism of the type

proposed for the acid- catalysed rearrangement of the

nitro group in P- nitro - N- methyl-aniline60 • In connection

with the ortho/para isomer djstribution of the

N- phenyltoluidines , resonance structures of the type

(VII I) ( Fig . 4) , as proposed in the mass spectral

rearrangement, may be postulated.

An examination of the mass spectra (appendix

of the ketone phenylhydrazones, vvh i ch ~_,,e 'e the subject

of this pyrolysis study, sho\'S that fragmentation of the

N- N bond is the major reaction. Therr; j s ev 1 dence for

the molecular rearrangement ions corres1-,onding to the

reactions as shorrn in Fig . 4, but these are represented

by minor pea\':s . Alkyl and phenyl migrations of the type

postulated in the pyrolytic reaction, are a rarit~.- in

mass spectra reactions, and almost all of the published

44.

examples have involved an oxyeen atom in the reaction

61 62 63 scheme ' ' • The energy of the 70 electron volt

bonfuarding electron~ used in obtaining the analytical

spe c tra may cause f ragmentation of the excited phenyl­

hydrazones before molecular rearrangement can occur .

If this is the case , the peaks corresponding to the

rearrangancnt ions should increase in intensity as the

exc itation energy is decreased.

45 .

1 , 1 , 4 , 4 - TETRAI ~-h""Y'LT~TRALIN- 2, 3- DIONE

dhereas the mass spectra of the aryl

phenylhydrazones show a nurriber of renctions which may

be explained by simple bond fragmentation and molecular1

rearrangement , the mass spectrum of l,l, 1,,Li-­

tetramethyltetralin- 2 , 3-dione is much less complicated.

The forrnation of the ions may be explaired by multiple

bond fragmentation to eliminate a neutral molecule of

carbon monoxide , follo 1red by competing paths with

further elimination of carbon monoxide, simple fragmen­

tation and elimination of a methyl radical , and frag ­

mentation rith molecular rearrangement and ring closure.

Interpretation of this mass spectrum is more facile than

for the phenylhyc1razones; the number of competing paths

is less , and fragmentation reac-:ions result in the

expulsion of small neutral or radical molec~les, leaving

the bllk of the molecule as a detectable positively

charged fragment . From the possible reaction scl:.emes

Cr.7ig . 9) , it may be seen that all rea.ction pathv1ays le[.._d

through only t 10 products, the benzcyclopcntadiene ion

( ]{XII) and the na:_hthaler:e racl.; col :.on (XXXV! 1 • lhe

· rork of Bo .rie et al . 64 provides strong support for this

react ion scheme as the ions (XXXI), (AXXII), (X r-1) and

(XXXVI) ere postulated to be for~ed by similar reactions

46.

in the mass spectra of 2,2,3,3-tetramethyl­

naphthoquinone. It is difficult to estimate the relat­

ive importance of each of these competing reactions

from the mass spectra, as the pea 1: due to the product

ions depends not only on the stability of the final ion,

but also on the instability of the intermediate ions .

Thus the fact that the relative intensities of the pea':s

at m/e = 115 and m/e = 128 are almost e~ual, does not

necessarily imply equal imports.nee of path,.rays .

Excitation of the l,l,lt,4-tetramethyltEJt­

ralin- 2,3- dione (XXXVII) by passing it through th

pyrolysis tube at 600°0 and 0.05 mm pressure resulted

in the formation of 1,1,3,3-tetramethylinclane-2-one

(XXXVIII) in 49% yield . Increasing the excitation energy

by passing the aiketone through the pyrolysis tube at 0

800 C and 0 . 2 mm pressure caused the yield of the mono-

ketone to decrease to zero, with naphthalene (XLII)

being formed in 39% yield. The higher excitation energy

also greatly increased the amount of complete fragmentat­

ion of the molecule, as indicated by the formation of

carbon in the pyrolysis tube. The formation of the

pyrolysis p-..coducts may be directly related to the mass

spectral reaction scherie . At a temperature of 600°0,

the energy of the parent molecule is suffic·ent for bond

CH3 CH 3 +• +• +. CH 3 0 - co - co CH 3

0

0 • • CH 3

CH 3 CH 3 CH 3 CH 3

CH 3

m/e = 216 m/e = 188 m/e = 160 XXV XXVI XXVII

CH3

+ CH 3 H - co - C2H4

0 • CH 3 • H

CH CH3 3

CH 3 H m/e= 173 m/e = 145 m/e = 117

XXVI II XXI X XXX

• l -CH3 •l - 2H H H

CH 3 +•

- CH• H 3 .. H

CH3 C H3 CH 3

H m/e = 144 m/e = 130 m /e = 115

XXXIII XXX I XXX 11

• j- CH 3

H H +•

- CHi H

- H• H ... • • H

CH 3 H CH 3 m/e = 129 m/e = 128 m/e = 143

XXXIV XXXVI XXXV

FIG. 9

-co

X XXVI I

• -2 CH3 <

'°CH 2 X LI

FIG. I 0

XXXVIII

(b)

XX XIX

XL 11

XLII I

=O

fragmentation anc.. tne elim::..n.o.tion of one molecule of

car~ ,on monoxide. :-nc1~ er s · n ~ tb e c;ner~: y b \ PY .colys is

at Soo0 c, resulted in fu~ther reaction of the mono ­

decaroonylatecl s1ecieB to naphttalene. The !'"ost likel:,

route for the formation of' thef,e profluctc is s1 1 0 nn by

( '-CY rVTI) ( ;rIJIT_) ( ~, . .ru. • .£'>. --)'ll>- .. ~':l lg • 10) 0 }he re~ctions have been

sl.o 1n in a concertec. for,~, as the st:-cuctures (X./XIX),

(XT..i; and (XLI) ar'e fa.vou~·ec1 s.s intermed-Lates. Of theBe

ttree com )ounds 1, 1, 2, 2-tetrcrnctl~ ylbenzc -clot.Jut ene

(.tCXXIXb J L.nd 1, 2-dimethyJ.-:JenzcyclobutacU er:e ( XLb) 8re

not knovrn, 1v h:.le the rec...ct j on of 1, 2-d.iffif;t:1;'> ler.e­

benzcyclobutene (.LI) to na 11':tl a1 .,r1 f', ( rL_·1) ~-~ 700°(.)

6h :1as ~Jeen reported by Cavr et c..J • . .1 • • 1•

1 ·c ,-;-irft,~CEJ.

(XXXIXa) is also formed on rr..o,_olvsjr of' t: e ncno r>"t.,one

(X -:XVI I), but undergoes hy1..· ro eL t .. 1 e.:1s::·cr t.r ro..__1,.·h a

six memberF,cl. trans it ion sto.te (X-x .. I c) to yie1.d (XLIII)

in high y: eld 66:

It i.s alro posf i '1le th, t t"te l\Y'i'olys · C' }:'I·oceec1s

iYl t.. stepmise fash:i.on, djr1 ectly pu~c.ll.el to t.1.e

ffiL r, S s uectral -l • reac 1j ions. Thus the loss of a methyl

cod.ical from the monokctone (X.1 XVIII 'Jefore the los C! of

e. second mo lee ule of cc..r~Jon monoxi a e c::.n( subse ucr~t

rec.ctions of -L:1e type (.1C-XVIII) ------­

c::i.:rnot ~_,c discountec. ~1herc , ·as no uviclerce for' t:1e

50.

second mass s-pectral decomposit 1 on throu~h the

benzcyclopentadiene cation in the pyrolytic reaction,

as no benzcyclopentadiene v:as identified in the products.

This reaction may occur to a small extent as thorough

examination of the minor py1')olys is products was not

made. It is not surprising that the formation of the

benzcyclopentadiene structure is more predrnninant in the

mass spectra than on pyrolysis . The added stability of

the benzcyclopentadiene cation, as opposed to the

benzcyclopentadiene molecule, would favour the mass

spectral reaction. Similarly, the greater stabjlity of

the naphthalene molecule, as opposed to the radical ion,

would favour the pyrolytic reaction. The effect of the

lesser number of electrons of the prominent mass spectral

int erme d ia t es ( XXVI I I ) and ( XXIX) is ref l e ct ed in the

absence of si,e:nificant derivatives from these ions . in the

pyrolysis products. An anomaly bet,. cen the mass spectraal

and pyrolytic reactions is the low intens-ity of the ion

due to the monoketone (XXVI), as compared to the parent

ion (XXV), whereas on pyrolysis at 6oo 0 c, the monoketone

(XXXVIII) is the stable product. It v_;ould be of inte1')E::st

to see the relati.ve intensity of the parent peak in the

mass spectrum of the monoketone itself, but this inform­

ation is not available .

51.

INDAN"ETRIONE

1 , 2 - Dehydrobenzene, also known as benzyne,

has attracted considerable theoretical and experimental

interest in recent years . This short lived intermedj_ate

has never been isolated, although it has been detected

physically by ultraviolet67 and mass68

s~ectroscoPy.

1,2- Dehydrobenzene has been postulated as an intermed·ate

in a great number of reactions, in v1h ich it may react

as an electrophile, a nucleophile, or a dienophile. The

interest in this species has resulted in the publication

of a great number of papers, 'vhich have been revie;, ,ed by

Bunnett69 , Heaney70 and more recently by ~ittig71•

Biphenylene, the stable dimer of 1,2-dehydrobenzene, is

also of theoretical interest as a derivative of the

highly strained cyclobutadiene system. Hovrever

experimenta1 72 ,73 and theoretica1 74 , 75 evidence suggests

partial bond fixation in the benzene rings, with only

slight cross conjugation across the four membered ring.

V'li th the intense activity in this field, many

syntheses of biphenylene have been reported. However

these are tedious 76 , require unstable reagents77 , or

make use of relatively irn.ccessible starting materials78 ,7 9.

On turning to mass snectral react ions, there ).re a larc ·e

number of accessible, stable compounds · hich on electron

52.

impact, yield an ion with mass to charge ratio equivalent

to 1,2- dehydrobenzene. These include benzonitrile80 ,

phthalic anhydride32 , indanetrione (appendix), fluorenone,

anthraquinone , naphthoquinone80 and ~ulphobenzoic

anhydride (appendix). Of these compounds, indanetrione,

with its cross-conjugated arrangement of carbonyl groups

if a five membered ring, would be expected to um.ergo

more readily the loss of the carbonyl groups on excitation.

From the mass spectrum of indanetrione

(appendix x) it may be seen that the scheme as sho~nrn in

Fig . 11 represents almost the sole mass spectral fragment ­

ation. Early attempts to reproduce these reactions by

pyrolytic excitation vrere conducted by addj ng a solution

of indanetrione in hot benzene dropwjse to the top of a

,r~rtical silica tube pac1<:ed with Pyrex .ti1 enske helices and

heated with an external electric furnace. ... stream of

nitrogen was used to exclude air from the system and to

sweep the products into the collection traps. Analysis of

products from the :pyrolysis at 500° showed the :presence

of bi:phenyl (1.3%, based on indanetrione) and naphthalene

( 5. 5%) in the unchanged benzene, vvhile acetylene was

detected in the gaseous products. At 600° the apparent

yield of biphenyl (based on indanetrione) was increased to

56% and that of naphthalene to 23%. As benzene itself is

0

0

m/e = 160

XLIV

+•

©J1 I

©~ m/e = 76 • XLV II

+•

0 - co

•

- co

•

[ H-c- c-c==c-H] +•

m/e = 5 0 XLVIII

FIG. 11

m/e = 132 XLV

0

• - co

0

l m/e = 104 XLVI

+•

+•

53 .

0 81 converted rapidly into biphenyl at 700 C , blank

p~yrolysjs showed 1010 and 81% of the biphenyl obtained,

but no naphthalene, was formed from the solvent alone

at 500° and 600° respectively . The formation of these

hydrocarbon products suggeststhe loss of three molecules

of carbon monoxide from the indanetrione resulting in

the format ion of 1 , 2 - dehydrobenzene, vrhich reacts vri th

the large excess of benzene. Partjcipation by the

solvent was confirmed by the formation of

1 - chloronaphthalene (13~) and 2-chloronaphthalene (3~0)

from the pyrolysis of indanetrione in the nresence of

a large e cess of chlorobenzene at 500°8. The

naphthalenes are most likely formed by the elimination

of an acetylene molecule from the addition compounds

(LII and LIII) of 1 , 2 - dehyclrobenzene (L) and solvent (LI)

(F'ig. 12) . The format1on of benzocyclooctatetraene

(LII b) and benzobicyclo- (2,2,2)- octatriene (LIII) by

the addition of (L) (generated from benzenediazonium-2-

carboxylate) to benzene at 45°c, has previously been

reported by Miller and Stiles82 • These authors also

reported the conversion of (LIII) to naphthalene by

heating in a sealed tube at 300°0. The direct pyrolysis

of (LII) to naphthalene h~e ~ot been reported, but

naphthalene is formed as a b;yprocluct ~n the s;:/ntheeis of

=O - 3 co ©J1

/_'-......../ '-........ .,,.O Ac

,, 0

XLI X L LVIII

~ ~R

insert ion

R LIV R -. /R ~R

~

©J1 1,2 addition /" /I R + . )

I ~ -a C LI L R LIi + Ill

R C 1,4 addition _ rhlfi 'I I rnTnT I

H 1,.....,_~.A.'\ ~ ............ ------ ,,,,,,,,,,... ............ --- .,,,,,,,,,,...

R = H or CI LIii

FIG. 12

( ) J t L ( T , ... - -1 °3 LII by the pyrolysis of che ace ·2: c 1.1 : -~ • 1 •

0 Cyclooctatetra.ene its elf, on ~;hoiolysis above 20

t t I 1 8L eliminated acetylene v;i l;. he forrc.9.-cion of )en:.7cre ·•

The ronaom ~ddition of 1,2-~ehy~robcLzere (J 1 to 811

possible bonds of the chlorobenzene, follo'l~/cd by the

random elimination of acetylene or chloroacetylene from

(LII) and (I.III), v1ould ~-ec.d to the formatjon of

nc.i:,hthalene., l - chJ.or:oP.c.phthr.10ne 2n{ ~-cL~ o::. onc:~1~th,.. lcr•e

in ey_uaJ. nr: unts . However these compounds were formed

in the ratio 3 : 4 : 12 respect ·i.ve:;l~r . Obviously the

electronegative c~lorine atom is infl~cncing the

direction of addition and elimination. It is difficult

to interpret these c:;~pe:eimental ratios in terms of

electronic mechsnisms due to the fifteen possible

addition- elimination modes, but the use of n.ore s;/1..rnetr ~ cal

benzenes with substituted inc1a.netr1 j_ones, may pr?vj de

more mechanistic irrformation on these 1.,eactions.

Although the pyrolytic experiments provide

evidence for the formation of a s~ectes equivalent to

the mass spectrometric ion (XLVII), there is no ev 1 dence -of products derived f'rom the ecuivalent of the partially

decarbonylated ions (XLV) and (.lLVI). In order to

decarbonylate the indanetrione in the absence of potential

djenes, it was introduced into the heated pyrolysis tube,

Experi nent -

T°C

Ind~netrione (mg)

Benzene (ml)

Chlorobenzene (ml

-------· :; a:phthalene

Chloronaphthalenes

Bi:phenyl

A

5co

250

60

11

3

1 A..,.....I r·, 2 ~- ... .-:$ .J~

PY.a.OL "''!S ~s OF Il'iDAl'ETH 101-~E SCL"J'=1I or·

B

500

60

C

600

270

60

D

600

oO

Yields or Products (mg)a

50

0 03 +L~ 7 120

.i1.

500

150

60

15

80b

tr-ace

F

500

60

trace

a Yjelds esti:Y:ated by analytical gs.s c1n·omatorra:phy.

b 1 - chloro : 2-chloro = 1 : 3, estimated b~r :p.rn.r. and i.r.

spectr•osco:py.

----

57.

packed with Pyrex helices, by slov· sublimation under a

moderate vacuurn. Pyrolysis of the indanetrione vapour

at 600° and o. 7 mm pressure caused extensive decomposition

of the molecule , as indicated by coking of the tube and

packing, and gave a product from which a large hydrocarbon

fraction was separated by chromatography on silica.

This vas further fractionated by chromatography on alumina

and preparative gas chromatography , and gave as the najor

product the dimer of 1 , 2- dehydrobenzene, biphenylene (23~)

together vi th tr iphenylene and a number of other aromat j_c

hydrocarbons . In order to reduce the degree of coking

vithin the pyrolysj_s tube , the Pyrex heljces w-ere

replaced by a pac 1ring of short lengths of silica tubing,

vhich provided less restriction to the gas flow. ~he

affect of variation in temperature and pressure uas also

studied, the latter factor being only slightly subject

to experj_mental control, and depended mainly on the rate

of sublimation. The results of these exper~ments are

shovvn in Table 3 and provide (efinite evidence for the

stepvise extrusion of three molecules of carbon monoxide

from indanetrione on thermal excitation. This evidence

is seen in both the carbonyl and hydrocarbon products,

as so1.1e of the hydrocarbons are .,ostulated as be:.ng

derived from the carbonyl analogue by reduction.

L~ 3

PY . LYSIS. uF _ IlrD.:...::....br:_· __ :I0: . .c.

----- --.J..!J:;:f er iment A j rt D v

~10v/P( rrm) 600/0.2 8CO/O . h 600/C . 7

,... C/0.1 J

Pac (ing silica si lie a pyrex silica tubing tubing hel-ices tu·oing

.,.inhydr in ( g) 1 . 0 5.0 5 .0 1.0

Pyrol . Time (hr) 1 1.5 2.5 1 --

Yields of Hydrocarbons ( rr~) a --

Naphthalene 12 26

Biphenyl trace 85 Lj.8

Biphenylene 170 610 490 ( L.010) ( 28;0) ( 23;1c)

Pluorene 9 21 1 mg

nthracene 16b 158 of Phenanthrene total 0

14 fract j_on C 2 3 hydro-

j fraction C 16 5 carbon '--'15 0 16 fraction C traces 8 3 c17 fraction C 18 5 cl8 (total)

fractionc 2li9 31 Triphenylene 18 185 20

3,LL-Benzphen- 10 anthrene

1 , 2- Benz- 18 anthr&cene

Chrysene 36

--Yields of Carbonyl Jompounds (mg)

--Exr,er irP,nt B C D

----Benzocyclo- 125 not 4 but enedione ( 1 71~)

Fluorenone 9 65 jnvest -

Anthrag_uinone 14 43 igated

Phenanthrene- 16 g_uinone

a Yields of c18 hydrocar~ons are crude fraction

weights , but yields of other hyc1...r·ocarbons were

estimated by analytical r·as chromatography.

d

b The ratio of anthracene to phenanthrene was a}Jpro r_

imately 1 : 3 in both experiments .

c Carbon numbers assigned f.eom gas cllromatographic

retention times .

d Yields are weights of crude fractions having correct

infrared spectre , except in the case of fluorenone.

60.

Pyrolysis of indanetrione vapour at 600°/u.2

mm yielded substantial amounts of both hydroca:cbon and

carbon~r1 product . The isolation of benzocyclobutenedione

(LXIII) ( 17%), the stable cyclic form of the intermediate

(LVIII), indicates this monodec2rbonylated specjes must

have a finite lifetime under the experimental conditions.

~·urther evidence for (LVIII ) is the formation of

anthraquinone (LX), by the addition of the intermediate

(L), as shown in Fig . 13. However anthraauinone may

also be formed by the dimerisation of the bj sdecarbony­

lated species (LIX). A careful investigation revealed

no evidence for the presence of benzocyclopropenone,

the cyclj_c form of the monocarbonyl species (LIX) , in

the products , but the formation of fluo renone, and most

likely phenanthrenequinone, r•equires at least the transj_ent

existence of (LIX) . Although phonanthrenequinone (LXI)

could also be formed from benzocyclobutenedjone through

the diradical (LXIV), this form of fragmentation of the

cyclobutene ring is not electronically favoured . Irocesses

of this nature, which do not seem likely on theoretical

grounds , are shown by bro~en arrows in Fig . 13 and Fig .

14. The failure to isolate benzocyclopropenone may

merely indicate that this unknown compound is unstable

at 600° . It may be stable at a lo½er temperature, as the

61.

parent hydrocarbon benzocyclopropene l'1as recr,ntly l)een

isolated85 . Attempts to dee, 1-1~)onylate the a.·one (LYJTJ J

6, 0_7 photochemically in the liq_uj_d phase "'ere unsuccessf'v_l ,

but provided further evj_dence for intermediate vii th

structure (LVIII) .

The hydrocarbon fraction from pyrolysis at

600°/o. 2 mm was shown by gas liouid chromato~·raphy to

consist almost entirely of biphenylene (LXV) and

triphenylene (LXVI) , ·both derivable from polymerisation

of 1 , 2 - dehydrobenzene . Pyrolysis at 500°/0.1 mm produced

very little decomposition of the indanetri one, the only

products formed being a slight amount of benzo-

cyclobutenedione, and a trace of hydrocarbons. Increas-

ing the pyrolysis temperature to 800°/o . L~ mm, reversed

the relative amounts of car:Jonyl products to hydrocarboYls,

vvi th the proportion of hydrocarbons beinr by far the

larger . _t 600°/o . 2 mm this proportion v,ras approxjmately

equal . These ratios sug~·est that the fragmentation

rep1'lesented by (XLIX)-- (LVIII ) ---+- (LIX) -----+- (L)

(Pig . 13) is the prirrary process, and cyclisation or

addition of these intermediates lead to what may be

regarded as prj_rnary products. 11 the products from the

pyrolysis at 500° a1)d 600° , ould be in this category.

iith the greater excitation of the indanetrione and

-co XLIX

~-:::::-0 -CO ©J>=o a ~C:-... ~o t t

~C-;::::.O ~~=O :

c~o L VI 11

!+L

11 0

LX ... I

'+L I

1/0 OC' -----+ o· ·-~ C-0

LXIII LXIV ~ FIG. 13

-co ©1

11 0

LXII

L

LIV

0 11

LXII

LX

11 0

+L

reduction

reduction

reduction

©1 L ' ' ' ' ' ' '

C15 and C17 Fraction

+

LXVII

LXVIII

LXIX

FIG. 14

+2 L

' ' ' ' ,,+C4

' ' ' ' ' ' ' ~

" / LV / /

/ /

,, ,, + C4 /

/

~

c,a Fraction

I, 2-benz anth racene

+C4 3,4-benzanthracene

triphenylene

chrysene

I

ll II ll

6LL .. 0 subsequent product molecules at 800 C, secondary

fragmentation and reduction of the primary products

may occur .

No benzocyclobutenedjone or phenanthreneq_uin­

one was obtained from pyrolysis at 800°/c.4 mm. The

benzocyclobutenedione vould be expected to readily

decarbonylate at this temperature, but scarcity of the

dione precluded experimental corrf irmation. Hov!ever

phenanthrenequinone was shown to decarbonylate to

fluorenone in 76% yield at 8co0 /c . li. mm. li'luorere ,

anthracene and phenanthrene were 1 de rrt·if·i ed in the

hydroco.rbon fracti.on, and t:bese are poc~tulatecl as being

secondary product6, obtained by reduction of the

corresponding carbonyl compounds . The nature of the

reducing agent is uncertain, al thou .,~h t vro por sibili ties

are the carbon deposited liberally on the packing and

·,alls of the pyrolysjs tube, and the hydrogen Vihich is

lil)erated in the forrn.ation of this carbon. The formation

of biphenyl also requj_res abstraction of hydrogen by the

species (L) or (LXV) from other molecules or from the

wall deposit .

All the products discussed in the above

section may be derived from the a-opropriate decarbonylated

species by simple addition, follo1Hcd by reduction or

hydrogen abstraction in particular c ases . ~he hyrlrocarbon

l

I

65 .

products from the pyrolysis at Soo0 contain several

molecules which require the addition of one o4 or two

0 fragments to the primary carbon skeleton. raphthalene 2

~s the smallest of these molecules, ard is rro8ably

formed by the adcli t ion of four carbon atoms to the six of

the 1 , 2- dehydrobenzene unit . The hydrocarbons larger

than 014 , with the exception of 018 , were not fully

investigated. Hov~ver it is significant that both the

015 and c17 fractions gave a positive Vansheidt colour

test in pyridine vvi th metha.nolic potassium hydroxide88 ,

v1hich suggests they contain the fluorene type structure .

This indicates that they may be formed from the fluorene

primary skeleton by the addition of one or tv o c2

( or

one o4) fragment respectively. The presence of the odd

c c.rbon hydrocarbons is further evidence for the

participation of the odd carbon srecies (LIX) in the

react ion scheme . Badger anc. co - , or7(ers her.; sho"'n that

these odd carbon hydrocarbons cannot be obtained from

the pyrolysis , even under more energetic conditjons

81 89 90 than used in this study, of an even carbon skeleton ' ' •

The major component of the 0 18 fraction was

al~"1ays triphenylene, and the yield of this and other

C1 3 hydrocarbons increased with increasing exc ·i tat ion

of the indanetri.one . The fo1,,mat ion of triphenylene at

I

66.

600° is probably by a prime.ry process equivalent to the

trimerisation of 1,2-dehydrobenzene, but the increase

0 in yield of this and other c18 ~ydrocarbons at 800,

is most likely due to the introduction of secondary

fragmentation and addition of the c4

fragments tc:, the

phenanthrene and anthracene si:eletons . Speculation on

the mode of formation of these higher h;srdrocarbons vvill

not be attempted, as they are most likely due to

secondary reactj_ons of the type fully investjgated in

00 the pyrolysjs of anthracene and phenanthrene 7

•

...

II

l

,,

67.

PHTHP..L I~ A::-HIYDR IDE

The mass s-oectrum of phthalic anhydrj_de is

very similar to that of indanetrione. The initial

fragmentation of the phthalic anhydride molecular ion

(LXX) results in the expulsion of a neutral molecule

of carbon dioxide, and the formation of the same ion

(XLVI) 9 as is formed by the expulsion of two molecules

of carbon monoxide from indanetrione . This ion undergoes

further decarbonylation in the formation of the

1/2-dehydrobenzene ion (XLVII) as has been previously

shovm in Fig. 11 .

Due to the similarity of the mass spectra ,

the pyrolys;s of phthalic anhydride vvould be expected to

yield similar products to those obtained from indanetrione ,

with products derived from the dicarbonyl spec 1 es (LVlII)

naturally being absent. Another· factor which inf'luences

the product distribution is the energy reauj_red for

initial fragmentation of the molecule . '.l:his aspect will

be discussed in more detail in a later section , but in

the pyrolysis it is ref1ected in the temperature necessary

before decomposition of the molecule occurs. 1"!ith the

packing of short pieces of silica tubing in the pyrolysis

tube and a pressure of 0 . 2 mm , a temperature of 600°

resulted in the decomposition of almost all the

Cl

Cl

m/e = 148

LXX

LXXI

Cl

Cl LXXII

0

+•

Cl

Cl

=O

C=O +

m/e = 104

XLVI

Cl

Cl

=O

C=O •

LIX

Cl

Cl LXXIII

FIG. 15

+•

l

Cl +

Cl

Cl

XLVII

L

Cl

Cl

l

Cl

LXXIV

J

ll

indanetriore sublimed into the heated tube, but a

0 temperature of 830 'tras necessary for a similar

decomposi.tion of the phthalic anhydride . This greater

stability of the phthalic anhydride is to be ex~;:>ected

from the stability of the anhydride F:roup as con1p'?-ried

tn the tricarbonyl system of incanetrione . The higher

temperature reguired to open the anhydrjde ring results

in further decomposition of the less stable primary

friagment s. 11his increases the EJ11ount of coi<:ing of the

pyrolysis tube and packing and decreases the total

amount of condensate recovered. The irc~eased amount

of reducing agent in the form of carbon and hydrogen

is most likely the main reason for t1e absence of

identifiable amounts of cs.rbonyl products fro.:n the

pyrolysis of phthalic anhydride. 1-1~hough phenan-

threnequinone vas decomposed at 830°, fluorenone and

anthrag_uinone were recovered in almost quantitative

amounts um.er similar conditions.

Evidence for biphenylene and triphenylene

as prinary products from polymer:_sation of 1,2-dehyclro­

benzene, is obtained from the pyrolysis of phthalic

anhydride through a narro if ( 8 mm I . D. J unpac',;:ed tube at

830°/0.2 mm (:1able 4) . Under these conditions

b iphenylene ( 18%) and triphenylene ( l"?) farmed almost

I I

1,

~L1ABLE J+

PYROLYSIS OF PHTHALIC ANHYDRIDE

---·------------------- ·-------.!..:Jxperiment

Pa.eking

Phthalic Anhyd. (g)

Pyrol . Time (hr)

760/0 . 3 a none

5

1

B

830/0 . 2 b none

1

0. 5

C

830/0.2

s i 1 ica tubing

1

1

Yields of H~/drocarbons ( ng) c

D

700/10

s j_ 1 ic a tubing

1

1

----·---------------------------Naphthalene

Biphenyl

Biphenylene

Fluorene

C fractiond lL1.

015

fractiond

c17

fractiond

c18 (total) d fraction

'l1 r iphenylene

6

19

240

( o//o)

7

traces

45

40

trace

trace

90 ( 18%)

5

5

8

26

69 ( lL1.10)

17

6

2

39

35

2

6

2

5

15

Unchanged Phthalic ~ nhydride (g)e ·------·-----··-------

3. 6 0 . 53 0.39

I

I

I

I

Ii

TABLE bl

a 20 mm I . D. pyrolysis tube contaj_ning 8 mm O D.

silica sheath.

b .l!Jmpty 8 mm I . D. pyrolys j.s tube.

c Yields estimated by analytical ·;:.s cl''romato;~raul"'y.

d Carbon nwnbers as signed. fro n gas cr.roma tcgra nhi c

retention times .

e 1: eight of liquid chromatographic fraction.

--

II

I

72.

the sole crystalline products . 1 . i th slightly more

restricted ras flow and a temperature of 760°,

biphenylene (9%) and triphenylene (1.5%) were still

the major crystalline products, but naphthalene,

biphenyl and fluorene were also detected. Increasing

the temperature to 830°c and restr 1 ct 1 ng the gas flov,

still further by packing the pyrolysis tube with short

pieces of silica tubing resulted in complete decom­

position of all the phthalic anhydride introduced into

the tube. Although no material passed through the

pyrolysis tube unchanged, the increased decomposition

of phthB.l ic anhydride only resulted in a slight

j ncrease in the amount of crystalline product ( Table L~) .

The yield of biphenylene was only slightly increased

to 14%, but increased amounts of naphthalene, b-iphenyl,

fluorene and triphenylene were formed, and c14

, c15

and c17 hydrocarbons were detected by gas liquid

chromatography. Reverting to the narrow unpacl{:ed tube

at a temperatu~e of 700°, but increasing the pressure

to 10 mn, substantially reduced the yield of biphenylene

( to 3%), but only slightly r·educed the formation of the

other hydrocarbons.

The results indicate that the energy necessary

for fragmentation of the anhydride group is a]most

73 .

suffic.ient for further fragmenta.tjo11 of the l,2-dehycc110-

benzene species . This further fragr1entat ion to Cl~ s.rc1

c2 fragments and eventually carbon, does occur unle2s the

cyclic c6 fragment js protecie~ from further e~citation

by rapid removal from the hot pyrolysjs tube as a stable

polymer . Thus the largAst and cleanest y -· eld of the

dimer was obtained \ hen the py1,olys j_s ' fa8 CE rr j eC:. 01-.:.t i:r.

c r1c.. .. r1 O\ unre[tr-ic-tec. tube. Hu\ ever ur:.c.er tl:cse

conC.i t ions the problem is that less than 5G~.) o:::' the

phthalic anhydride undergoes decomposition. The c.tten_.1t

to j_ncr·ec.ce t},e ever[ 11 clecompo8 i tion by increasing the

pressure and hence the residence t · me in the hot tube,

but decreasing the total excitation energy b31 cJ.ecrer-i sing

the teurerature , v1a.s unsuccessful in increasing the yield

of biphenylene. The likely rarticipetion of derivatives

of the monocarbonyl species (XLVI) is again sho1,.n by the

:t orma tion of c13 , c15 and c17

h;rdrocarbons.

This synthesis of biphenylene from indanetrione

and phthalic anhydride is at lJresent the nost convenient

synthesis of this materic.l. 'rhe cJ:;_r·o.dvrntage of the sms ller

yield from phthalic anhydrj de is counteracted b;s, the

fact that this mD.teriG.l ts more recd: ly available. Due

to the number of substituted phthalic ari.llyd1·ides which

are readily available, or easil;? synthesised, the

1,

74 .

possibilities for preparing previously unknovrn

bj_phenylenes are extensive . As an example, the pyrolysis

of tet rachlor ophthalic anhydride (LXXII) at 700° and

0 . 2 mm pressure gave the symmetrical octachlorobipheny­

lene (LXXIII) in 28% yield. This reaction was not

investigated in detail and the yield may be increased

by a var.iat ion in temperature and pressure. ~r.he

identity of ( LXIII) was confirmed by the mass spectra

( Appendix xii) , which showed the correct molecular

weight and fragment ions corresponding to the stepwise

loss of chlorine atoms . ~I.1his frE e rad· ca.l loss of

chlorine atoms is reflected in the formation of

hexachlorobenzene in 21% yield on py~olysis . This

must arise by the reaction of the chlorine ·rith the

l , 2 - dehyc1rotetrachlorobenzene intermediate . As only

chlorine and carbon atoms were present in the system,

the loss of a chlorine radical must subsequently result

in recapture , or the complJte decomposition of the mole ­

cule to chlorine and carbon. This is evident in the

c oking of the pyrolysis tube and paclcing, and the

condensation of chlorine gas in the liquid nitrogen

cooled type . The preparation of substituted biphenylenes

is being continued oy Dr . J . F . -~- . r c0mie of Bristol

University, and 2 , 6-lichlorobiphenylene and

2 , 3 , 6 , 7 - octachlorobiphenylene have been synthesised9l •

75 .

1...,_2 - DICARBOXVL! C AlTE:!D.,_~~:.DES

The expulsion of t .he anhydride ·r our, and

the introduction of an additional pa~r of bondjng

electrons in the junction carbon- carbon oond is shovrn

by a large number of 1,2-dicarboxylic anhydri.des on

electron tmpact. Phthalic anhydride has been discussed

in the previous section , in view of the synthetic value

of the parallel pyrolytic reaction. The mass soectral

reaction of 3 , 5- cyclohexadiene- cis-1,2- dicarboxylic

anhydride ((LXXV), R = H) to benzene and of

1,2, 4 , 5- tetramethyl - ~ , 5 - cyclohexadiene- cis - l,2-dicarboxylic

anhydride ( (LXXV) , R = ~.'Ie) to durene has no synthetic

value , but it is of interept to see if a correlation

can be found in the pyrolytic reaction, as aromatisation

of the ring should enhance the resction. The additional

pair of bonding electrons do not need to be introduced into

a carbocyclic ring, as the mass srectral reaction of

maleic anhydride to acetylene 32 , has been 1;.,revj ously sho, 'n

to have an equivalent in a pyrolytic reactjon33 .

Returning to reactions of possible synthetj c vs.lue, the

elimination of the anhydride group in the mass spectra of

pyridine (LXXIX) and pyrazine (LXXXI) dic,..rboyylic

anhydrides, if follo1·,ed in the pyrolytic reaction, may

offer a route to the corres~onding azabiphenylenes.

-

R

LXXV

0 11

Co 11 0

+•

m/e = 98 L XX VI I

m/e = 149 LXXIX

m/e = 150 LXXXI

+•

FIG. 16

R R

R

LXXVI

+•

~ C Ill C

~ m/e = 26 LXXVIII

+•

m/e = 77 LXXX

~I N

m/e = 78 LXXXII

+•

11

j 11

77.

These anhydrides (.J.?ig. 16) ,·ePe selected for thej_r

mechanistic or synthetic velue and are not meant to

be reprerentative of the type of dicarboxylic c.nhydride

Fhich uno.ergoes this elimination. S::i.mile~rly no attempt

has been made to investigate dicarboxylic nnhydrides.,

such as 4 - cyclohexene-1,2- dicarboxylic anhydride,

t d th · t 1 1 · · t . 9 2 ~1hich do no un ergo 1s mass spec ra e 1m1na ion •

The liquid phase photolysis of (LXXV), R = H

and Me, has been investigated by van Tamelen and Pappas93 ,

and 1 !arrener and Bremner9L~ res:pectivel~y. Wherec .. s the

unsubstituted compound unler~·oes disrotatory ring

closure of the dj_ene system., excitation of tLe rnethyl

derivati.ve resu1·ted in a novel J,,hotochemicc:il .:.ragmentatj_on

to durene ( (LXXVI) , R = 1'.:e) in hO'f~ y:i.elcl. T:·s.rrener

and Bremner consider this reaction as bein.g due to a

multicentre vaJence- electron redistribution of the

excited- state diene 95 , rather than a sjmple f'ragment at ion

of the anhydride group . Hov ever no ex1;lana t ion is

advanced for the absence of the similar reaction in

the photolysis of the unsubstituted an.hydride. In vi.e,.n,

of the similar pyrolytic fragmentation of phthalic and

maleic anhydride , the pyrolysj_s of these tvro cyclohexadiene

an.hydrides v,as investj_gated in collaboration with these

authors .

-

I

I

I

I

I

I

I

11,

I

II

78.

Sublimation of the anhydride through the

pyrolysis tube pacl ed vvi th pieces of silica tubing

at 800° and 0 . 1 mm resulted in the expected elimination

of the anhydride group and the aromatisation of the

cyclohexadiene ring . The yields in each case were

relatively high, 58% benzene being obtaine c1 from the

unsubstituted anhydride, and L~O;o durene from the

tetramethyl derivative . 'rhe lovrer yield of clurene is

most likely due to the ease at v~ich methyl rad~cals

may be lost from the initial anhydride, or from any of

the intermediate fragmentation products . These

demethylated products condensed as an oil in the liquid

nitrogen cooled trap , and compounds ~ith retention time

corresponding to mesitylene , xylene, toluene and benzene

were detected by gas liquid chromatography.

It is most likely that the nyrolytic reaction

is initiated by fragmentation of the anhydride group

in the vibrationally excited molecule, while ring closure

of the unsubstituted anhydride occurs through the

electronically exc~ted state95 • It may be postulated

that the anomalous photochemical fragmentation of the

tetramethyl anhydride is also initiated by a vibrationally

excited state . This could be a vj_brationally eycited,

electronic excited state, or possjbly a hi r·hly vibrationo. l

ground electronic state, formed by racliationless transition

-

I

I I

I

I

I I

11

79 .

from the excited electronic state . However as discussed

in the introduction (Page 3), this vibrational excjted

state control of the course of photochemical reactions

is less likely in the liquid phase than in the gas phase,

due to the dissipation of the vibrational energy by

molecular collisions . In this respect, the photolysis

of the anhydrides i.n the gas phase may yield fu11 ther

information on the course of the reaction, but the

experimental difficulty in obtaining the anhydride as a

stable species in the gas phase , could complicatP. the

interpretation of the results . An alternate explanation

for the difference in reaction path on photolysis of

the unsubstituted and substituted anhydride, 1ould be

that the activation energy of the fragmentation reaction

is decreased below that of the cyclisation reaction by

methyl sub st i tut ion. A kinet j_c s tuav would clarify

this aspect of the reaction, but the instability of the

anhydrides V'ould again be a complj_cating factor .

Just as 1 , 2 - dehydrobenzene is produced by

the elimination of ortho substituents from the benzene

ring , so 1 , 2- dehydroheteroaromatics may be formed from

the hetero c yclic ring . These species, which he grouped

under the title , ,,Hetarynes" , rere the subject of a

recent review by Kauffmann96 • Although this aryne

-

I

!

80 .

nomenclature has the advantage of sirnplicity for naminr;

species , it has been avoided in this thesis as it

implies the formation of a triple b ond. Thjs is a

customary diagrammatic representation, but the acldj_tional

bonding electrons are localised in an aro111at 1 c system due r,

to the sidewise overlap of their sp~ orbitals . The

bond forrred by this overlap is very weak, ;vhi ch accounts

for the enhanced reactivity of the 1,2-dehJdro species

as compared to true arynes . In heteroaromatics the

inductive effect of the heteroatom distorts the 11 dehydro"

bond and reduces the triple bond character as compared to

carbocyclic arornatics . Although stron.;,; evidence exists

for 1 , 2- dehyrl.roheteroaromatic intermediates in

cine - substitution and Diels- lder addition reactions,

no stable heterocyclic analogue of biphenylene has been

isolated .

The mass s:oectrum of pyrazj_ne- ? ,3 - dics.rboxylic

( ) ( + \ anhydride LX:,.xr shows strong peaks at m/e 15G r: . ; ,

106 (( 11 - co 2 ) ; ) and 78 ((I1·• - 002 and co) -t 1 corres~Jond;n2;

to those in the mass spectrum of phthal1c anhydride .

Due to the symmetry of the molecule any dimeriso.tion of

a ? , 3- deh:rclropyrazine intermediate ,1ill result in the

forna t ion of only one isomeric tetraazab iphenylene. In

an attempt to synthesise this unkno· n compound, the

anhydride (LXXXIII) was sublimed into the pyrolysis tube

-

81 .

packed with short pieces of silica tubing at 830° and

o. 05 rm pressure . The pyrolysate was separated by

liquid and gas liquid chromatography and the ultraviolet

spectrum of all fractions recorded. No evidence for any

biphenylene type structure could be seen in these spectra ,

but as t he modifications which the nitrogen atoms will

introduce into the u1-~raviolet spectra are unknown ,

this is not conclusive evjdence for the absence of

azabiphenylenes . The product, obtained in 80~ yield,

was an approximately 1 : 1 mj xture of maleo:1. i trile

(LXXXVI) and fumaroni trile ( ·J: ~x_,r --=r) .

cor.f i1-·me. t ioL of tLe r.1ec an ism of tr1 i.,::. Pet.ct iu..: .;culc

re\iuire icotore or sub st i tuent lubelled pyr·azinc-

.!( ! o--,·'· l -·l',-f., 7 "':r -/-lr.r-t ..._,L, ..,_ __ l ____ v i.l_c:, the

c1E:r1~ c r·o Lr tcr·rr.ec' j '- tt.. (LXXXIV) is formed, but undergoes

fission to form t.1e b~· o- isorjtrile (IX }CV), ,h 'ch

re--·.r·rang-f:s t _e~mc1ly to rnaleonitrjle (o.nc1 subc.-e½.uently

f'um2.ronitr·ile by further isomerisation) . Although

precedent-=; have been r,1 tbJ i sr'"ec fer -L?-.e firs ion l,f i,l P

:'1d~F}:1ec~U[Lte (LXXXIV) to (Lxxxv) 97 , 98 , 99 and for the

isomerisation of j_;=.,onitriles to nitrileslCC, mo:-ce

concerted processes are possible. Thus concerted ri0si n

cf c:. ccr1 or.- cc- r·1.1 on bcncl. r ... n..:i_ c CE rte n-ni trcf!E.T. c::.a t:-i

the tricycli c resonance form of the intermediate (I XJ.J~·v1111

fC=:N +

N=C

LXXXVII

0

,-- -r(c=N ....._: _ AJ l(C=N ~

-LXXXVI LXXXVIII

t ...- -

())jo N 11

- CO2

-co . .

N

0)1 N

0)! 0

LXXXIII

fC=:N +

N=:C LXXXVI I

N --

r(C:::N ....___~ _

l(c:::N LXXXVI

FIG. 17

N -LXXXIV

l

- -LXXXV

I --

:

11 11

II

v j_ll yield maleoni trile vi:.. t 1out the inter·n:ec"l_j E cy of

t 1- P: .: r 1· ~ r: -' t ·n -L' l e - I-, J ;:, - _ l., - .L - . •

'"''h .L • ., -- eor r., 'J. C' ..... Ll'c1 - - c~·.!l erta l ev·1· dence cu.r·'')' -,·j·c· _. ..!._ - - •- . · .L.:. u c., b ~ 0 -' D

3,4-dehydropyridine is the most stable 0£' the

- } - • - • • J d. t 96 ue :iyQl'OI\Yr·1o_- ne inuerme 1.a es • The "dehyd:.c 0 11 bend of

this ir_t er·n ed iate is lees distorted by the inductive

effect of the nitrogen a·torn.. 1?;yr·-i_d-:ne-.::,LJ.-cL~.c2.rboxylic

8JJhyc1ride :.J(XJX 1 E,J.fio Eflo-. s E1. ctror_g r,eak corresponding

to (M - CO2 and CO ) t in the mass s1,ectrum. Exploratory

attempts to synthesise the t,No isomer ic aza.biph~Y(/le::~es

~:::-·011 this anhydride were unsucces2ful, as no identifiable

material was obta ined in -Lhe arno1·phous condensat e from

a pyrolysis at 800° . However improvements in techni aue

and a1)1x.:i.ratu.s may enable sufficient of these compounds

to be isolatec an~ ch2rEcterjsed.

I

,,

I

8L~ .

SULPHOBENZOIC ANHYDRIDE

Due to the simplicity of the lJ iphe:nylene

synthesis from indanetrij_one and phthal ic anhydr:.de, the

mass spectral and pyrolytic fragmentation of structurally

related ortho substituted benzenes v1as investigated. nhc

elimination of a small stable neutr·al ri1olecu ..... c j s 1:::t r:rocesf'

which occurs readil;r in the mass Ci")cc·'.:1,.,omc.i~or, and. on

:9yrolys is. This is seen j_n tLe r :recerl i n,r cl-; scuss ion in ......

the elimination of hydrogen cyanide, carbon monoxide and

carbon dioxide . Nitrogen and sulphur dioYide are further

small stable molecules which are known to be readily

eliminated on pyrolysis67 , 13 . If a CO group of phthalic

anhydride is replaced by an N2 group, the resulting

molecule , benzenediazonium-2-carboxylate is known to yield

1 ,2-dehydrobenzene with very mild pyrolysis67 • Replace­

ment of a CO group with so2 however, yields the much

more stable sulphobenzoic anhydride molecule .

The sulphur atom of the anhydride may be

eliminated as part of ti,ro possible stable molecules,

sulphur dioYide or sulphur trioxide, vii th corresponding

loss of the anhydride carbon as carbon dioxide or

carbon monoxide . The initial elimination of both

sulphur trioxide and sulphur dioxjde is seen as competing

pathways in the mass spectrwn of' sulphobenzoic anh~rcl.ric_e

0 •

m/e = 64

LXXXIX

* - co

rg:o· m/e = 92 LXXXVIII

m le = 76 XLVII

-co •

- co

*

0 m/e = 148

XC

- co

~o

~c ~ 0

0

m/e = 104 XLVI

FIG. 18

0

+•

+•

m/e = 120 LXXXVII

- so 3

0

0

rnr+ o' ~c/

11

0

0

m/e = 184 LXXXVI

+•

l

1 86.

,1

II

I

(Fig. 18). The loss of sulphur trioxide leads to the

format ion of the base neak in the spectrum, due to the

monocarbonyl species (XLVI), ancJ. to the 1,2-dehydrobenzene

ion (XLVII) by the further loss of car½on monoxide . ~1he

alternate loss of sulphur dioyicle and ti-1.e fort1U tion of

the ion (LX)CXVII), leads to the forrn tj_on of the

cyclopentadiene ion (LX .. XIX). The frs.gmer .. t,ition of

(LXXXVII) to (LXXXIX) is also seen in the r.iass s-oectra

of sa.l icylates, in which the ion (LXXXVII) is forr:~ed by

the loss of an alcohol fragment from the pa.1- ent ion101 •

The products from the pyrolysis of the anhydride

at 730°c and o. 05 mm pressurie confirmed. the e.x:r:,ected

parallel with the mass spectral fragmentation. The

fo11 ma tion of 1, 2-dehyclrobenzene was sho~:1n ·by a similar

distribution of hydrocarbons in the product to that

obtained from indanetrione and phthalic anhydride0

Bjphenylene was obtained in 5. 5% yield vrith smaller o.mounts

of biphenyl (0 . 5~o), naphthalene (1%; a11d fluorene (0.7f'o 1,

r.;ith tr•aces of ell+' c15 , c17

and c18

hydrocarbons .

If the pyrolysis exactly follo,. ·s the mass snectral

decomposition (LXXXVI) -----'~- (:\.LVII), sulphur trioxide

a.nd c2.rbon .1onoxide would be formed. . I:o··rcver the sliE:;ht

amount of white solid v·hich concle:iced in the l.icuicl

nitrogen cooled trap, melted ar..cl ooilecl on ~ ·o.rminr,' to ~

86a •

room temperature, and he.d a pungent odour of sulphur

dioxide Qualitative tests on a solution of the gas

in dilute sodium l"'ydroxide confirmed the presence of

sulphite ions , but no sulphate ions, •rhich 1 1ould be

formed from sulphur trioxide, 1ere detected.

The failure to detect sulphur trioYide would

indicate that t~1 is gas , if it is for>:1ed, is not being

condensed in the liquid nitrogen cooled trap under the

lon pressure experimento.l conditions, or ~ho.t it is being

removed chemic ally in the reaction system. T~ is latter

explanation would be in agreement v1ith the greater

c hemical reactivity of sulphur trioxide as compared to

carbon monoxide and carbon dj oxid.e . It ·~ rould be expected

to add rapidly to the 1 , 2 - oehydi•obenzene intel'med iate ·. i th

the eventual format ion of benzenesulphonic acid. by

hydrogen abstract ion. Hovlever no benzenesulphonic a.c id

was detec ted , although 55% sulph6benzoic acid v.ras obtained.

after chromatography on silica. . The most simple

explanation for the formation of the latter acid would

seem to be hydration of the anhydride during the

chromatography.

The presence of sulphur Cio.xio.e .. ould suggest

that the pyrolytic equivalent of the mass spectral

elimination (LXXXVI) ~ (LXXXVII) is occurring.

Although there is evidence that such a :rragmentation

I

87 .

may lead to 1 , 2- dehydrobenzene, the combination produc~s

normally obtained are disalicylide, xanthone and

3 ,4 - benzcownarin. The peak due to the ion (LJ(:XXVJ~)

forms the base peak in the mass spectrum of coumarandiore,

and evidence for the formation of this species on py1~01ysis

is presented in the next section. The lac:c of s imilc.r j ty

bet ireen the products would indicate that an intermediate

equivalent to (XLVI) is formed on pyrolysis of ""ulphobenzoj_c

anhydride , and an intermediate equivalent to (LXXXVII)

on pyrolysis of co1.urw.randione . An alternate explanation

for the formation of sulphur dioxide, which may also

account for the lac1c of sulphur trioxide, is that the

sulphur trioxide is reduced to sulphur dioxide by the

active reducing agents , which have been previously

postulated in the system.

i

88.

C OUiv. .. L\.RA.ND IO ITE

As an extension of the study of compounds

which appeared to offer a s;~rnthesis of bj_phenylene,

the pyrolytic and mass spectral fragmentation of

coumaranclione was investigated. 'l'his molecule (XCI)

is isomeric with phthalic anhydride, and asain the

elimination of the small stable cc.rbon dioxide and

carbon monoxide molecules ,·.,ould seem favourable .

The major t11ass snectral fragr.1entat ion ( ~l· r;• .... b .

is the loss of carbon monoxide and the formation of the

ion (LXXXVII). This ion is stabilised by several

resonance structuries and forms the base peak in the

spectrum. It undergoes fu1~ther lo 0 s of successive

molecules of carbon monoxide, as was previously observed

18)

in the mass spectrum of sulphobenzoic anhydrjde. Although

the 1,2-dehydrobenzene ion (XLVII) could be formed by

the loss of carbon dioxide from (LX.1::XVII), the absence

of any significant peak at m/e = 76 indicates that this

process does not occur to any apprP.c iable extent. .s

carbon dioxide cannot be lost directly from the resonance

structure (b), it would seem that this form makes the

largest contr•ibut ion to the stabilised ion.

11

he pyrolysis products (rriable 5) indicate

that loss of a molecule of carbon monoxide and the

.......

OCoM 11 0

XCVIII

XCI

0 11 0

-MX XCIX

l+L

-co

©le-0-11 0

XCIII

1 0 '/0

~ ~o XCII

l ©tl

0 XCIV

FIG. 19

-ArI

-CO

+L

+

rAYI'Ar

~0-11 0

C

+L

XCV

R = H, Me

R R

11 Cl 0

90.

formation of the stabilised spec~es (XCII 1 , (XCilI)

and (XCIV) is again the most facile process . Hovrever

the mode of recombination of these species with the

formation of the major products, xanthone (CI) and

3 , 4 - benzcoumarin (XCIX) is not apparent . Two routes are

possible, one involving dimerisation to disalicylide

(XCVII) and subsequent loss of carbon dioxide, the other

invoking the addition of 1,2- dehydrobenzene .

The formation of xanthone from c1isc.licylide

102 hds been reported by Anschutz , and it is also formed

10~ in the preparation of disalicylide from phenylsalicylate J.

A detailed study has not been made of this mechanism,

but it appears that the loss of carbor: dioxide is follo'Jed

by inversion of one of the phenyl rinss . Thus

4 , 8- dimethyldisalicylide ( (XCVII), R = e), formed by

the distillation of ortho- cresyl - ortho-cresotinate ,

decarboxylates with ring inversion and 4,5-dimethylxanthone

((CI),~ = Me) is formed104• ~o other isomeric xanthones

.-1ere obtained, indicating that the process is intrarnolecular.

Ho 3 , 4-benzcoumarin has been isolated as the product of

the pyrolysis of disalicylide prior to this wor~ , but it

has been postulated as an intermediate in the formation

f nth . t l . t . 10 5 o· xa one in 1is rer:ic 1 on • As 3 ,1-/.-ben?:coumar in

itself is only very slightly decomposed under similar

--

....

n

II 1,

1,

II

TA3L~ 5

Cownarandione Djsctlicylide

Experiment A 3 0 D --T°C/P(mm) 600/0 . 2 820/0 . 4 600/0.1 800/0.2

Amount (g) 0 . 5 1 . Li. 0.9 1 . 0

Pyrol . Time (hr) 4 2.5 1.75 2

Yields of ryc1rocarbo11s (mg)a ---Naphthalene 6 36

Biphenyl 1 6 traces traces

Fluorene 0 . 5 3

c14 fraction 2 8

---Yields of Carbonyl Compounds (mg)

Biphenylene Oxid~ 1 10 2 -:z;o

..,/ Xanthone 14 40 152 132

3 , 4 - benzcoumarinb 35 128 291 191

Polysalicylidec 30 53

a Yields estimated by analytical gas chromatography.

b Yields obtained by chemical separation and verified

by analytical gas chromatography.

c Weight of liquid chromatographic fruct ion.

92.

experimental conditions, and a trace of hydrocarbon

was the only material isolated f'riom the unchL.nged lac tone ,

this contention seems unlikely. Ho\.ever it is possjble

that both xanthone and 3,4- benzcou1narin are formed from

the same j_ntermediate, which resembles the open for·m of

the lactone . This does not imply that xanthone and

3 , 4 - benzcoumarin should be in eoujlibrium under the

experimental condi t j_ons, as the energy necessary to open

the conjugated carbonyl rings j_n each case, is most

likely .IDre than sufficient to cause further fragmentation

of the excited intermediate. This is a similar case to

that previously discussed for the energy required to

open the cyclic rings of indanetrione and phthalic

anhydride . VVhereas for phthalic anhydride the more

energetic pyrolysis was sufficient to cause almost complete

fragmentation of the anhydride ring, but also resulted

in substantial fragmentation of the intermediate

( 1 , 2- dehydrobenzene), similar energy was not su1·f ic ient

to open the more stable heterocycl~c rings of xanthone

and 3, ~-- benzcoumarin.

The formation of all products is also

explained by the addition of the 1,2- dehydrobenzere

molecule to the appropriate inter:nediate as shov.n in • ig.

19. This mechanism assumes the formation of

93.

1 , 2 - dehydrobenzene (L) by the loss of carbon dioxide

from a resonance form of the species (XCIV). However

as the yield of biphenylene vvas below the detection

limit of 0 . 01% , the concentration of free (L) in the

system must be small. This fact is supported by the very

lovr abundance of this ion (XLVII) in the mass sDectrum.

As the total yield of addition compounds is as high as

501u in some experiments, the overall format ion of (L 1

cannot be low. These experimental details are consistent

vrith an er_uilibrium betvreen the resonating species (XCII),

(XCIII) and (XCIV) , the phenolic species (XCV), and the

dehydrobenzene intermediate (L) . The equilibrium must

be strongly displaced towards the resonating snecjes,

with only a small amount of (XCV) and a very low

concentration of (L) . Hor~rever this lovv concentration

of (L) is being constantly replenished as the s ystem

moves towards restoration of the equilibrium, which is

displaced by the removal of (L) in the formation of the

stable 3 , 4 - benzcoumarin (XCIX), biphenylene oxide (XCVI)

or xanthone (CI) molecules .

The species (XCIII) has also been generated

by the pyrolysis in vacuo of the metal salts of

ortho- halobenzoates (XCVIII) . Xanthone is the main

product from the pyrolysis of the majority of benzoates ,

94.

but the phenyl benzoates and 3,4-benzcoumcrin are

also formed in varying amounts depending on the actual

halide and metal atom. Althoup;h a nucleophi.lic

substitution mechanism b7 the formation of a ring

opened106

, or ring closed107 disalicylide, followed by

the loss of carbon dioxide with the inversion of one

ring to yield xanthone, has been proposed for this

reaction, the use of substituted benzoates suggest that

a 1,2-dehydrobenzene intermediate (L) is involvedlO?,lOB.

The trapping of this intermediate with dienophiles

provided further evidence for its for.ma. ti on107 • The

pyrolysis of solutions of 2-phenyliodoniobenzoate (C)

1 · ld th d -z 4 ., · l09 a so yie s xan - one an _) , -oenzcoumar1n • 1.lthough

these products could be formed by intramolecular

nucleophilic substitution , an alternative postulate is

the addition of the resonance structures (XCII) and

(XCIII) to 1,2-dehydrobenzene (L). Dienophile trapping

of (L) again furnished evidence for its formation.

In most experiments g_uoted from the liter­

ature xanthone vvas the main product ,. with only minor

amounts of 3 , 4 - benzcoumarin being formed . In all

exneriments of this present investigation, the reaction

was in the vapour phase and 3,4-benzcoumarin was the

major product . The pyrolysis of appropriately substituted

95.

disalicylides and an investigation of the substitution

pattern in both the 3 , 4 -benzcoumarin and the xanthone

should enable a distinction to be made between the

intramolecular mechanism, and the j nt ermolecular process

involving the formation of 1,2-dehydrobenzene. Carrying

out the pyrolysis in the presence of a larCTe excess of

diene (such as benzene or chlorobenzene) should provide

further information on the concentration of the

1,2-dehydrobenzene species during the short reaction

time of less than 0.1 second .

96 .

~arly in the discussion of the reactions of

indanetr ione , a nwnber of covnpounds vere mentioned which

yielded 1 , 2 - dehydrobenezene ions in the mass spectrometer.

IIost of these were pyrolysed under conditions si1ilar to

those used for phthalic anhydride, indanetrione and

sulphobenzoic anhydride , but gave no (or only"\ery slight)

fragmentation to 1 , 2 - dehydrobenzene . Buu- Hoi et al . llO

found that benzo- 2 , 1 , 3- selenadiazole also yielded

1 , 2- dehydrobenzene ions on electron impact, and then

stated that it ,muld be exnected to generate benzync and

related species under drastic nyrolytic process . In view

of the negative results obtained for some compounds in

this work , a ta1Julation v'as made of the yield of biphenylene ,

c ompared to the ratio of the peak height due to the

1 , 2 - dehydrobenzene ion and the peak height of the

molecular ion ( Table 6) . A far more meaningful comparison

vuuld be the yield of the 1 , 2- dehydrobenzene species

produced on pyrolysis , compared \lith the proportion of

the total ion current due to the 1,2- -ehydrobenzene ion.

However this pyrolytic information is difficult to obtQin

due o the ambiguity of the mode of formation of some

compounds . The mass spectral information is most

accurately obtained electronically when the spectrum is

Indanetrione

Phthalic Anhydride

Sulphobenzo ic Anhydride

N - I\1ethyl -Ph thal imide

Fluorenone

1,4- Naphtho­quinone

3enzonitrile

TABLE 6

m/e - 76 m/e - P ;

26 .1

1 . 87

1 . 84

0.45

0.33

Pyrol . a

(%)

40

28

0

14

5 . 5

0

trace

5

0

0

Benzo- 2 ,1, 3 - 0.25 Selenadiazole

'I' 0 cjP( mm)

600/0.2

800/0.4

600/0.05

830/0.2

730/0.05

800/0 .05

830/0 .05

1100/0.1 C

800/0 .05

800/0 . C05

a 20 mm I . D. pyrolysis tube packed v,rith pieces o:f

silica tubing .

b Rn t i o of' ( ( m/ e = 7 6 ) + ( m/ e - 154) ) / ( m/ e - Pt ) c Porcelain bead packing91 •

98.

being recorded, but this information is not available

for the compounds studied in thi.s thesis.

From the variables 2.s tabulated in Table 6,

the correlation of the yield of biphen;rlene, yj th the

ratio of the intensity due to this ion in the mass

spectrum is quite good. All three com-pounds ,.Yhich yielded

more than trace amounts of biphenylene under· the

experimental condi tions described in this thesis, had a

ratio greater than one. It1rom this table it Hould seem

most likely that the yield of biphenylene from sulphobenzoic

anhydride could be increased by more drastic pyrolytic

conditions, such as a hi£her temperature. The mass

spec-+,ral ratio is also an indication of the stability of

the excited molecular ion. As vould be expected, this

mass spectral stability is related to the pyrolytic

stability , and all compounds with a ratio less than unity ,

were stable, or only very slightly decomposed under the

most drastic conditions (850°c) with the present apuaratus.

Obviously these compounds will oe <iecomposed by higher

temperatures, but this is likely to lead to further

fragmentation of the 1 , 2 - dehydrobenzene and yields of

biphenylene would be lo 1. Fluorenone is one compou11d

which was subjected to 1100°c and found to yield a small

ctmount of biphenylene91 . s the dccomposjtion of

fluorenone in the mass spec~urometer :proceeds through

99 .

the biphenylene ion , the height of this peak has been

included wl th t11.e height of the peak due to the

1, 2-CLehyo.rouenzene ion , in calculating the .n&E r s ,ectr::.l

ratio . In connection with the production of

1,2-dehydrobenzene by pyrolysis of benzo- 2,1 ,3-selenadiazole ,

~twas of interest to note that this molecule has the

lo.· est mass s:i;:ect1,,al ratio of those tabulated. rrhus

although it may be possible to produce 1,2- dehydrobenzene

by very drastic pyrolysis of this material, it would be

expected to yield very little, if any biphenylene.

Throughout this thesis there have been many

correlations between the 0ualitative as··ects 0f mass

spectra and pyrolysis . In Table 6 reasonable quantitative

agreement has been shown between these t -.,o modes of

excitation , in spite of some unsatisfacto11 y aspects of

the variables chosen. The scope of the compounds in

this comparison is very limited, even foi,, compounds

vhich are knovvn to produce 1 , 2-dehydrobenzene, either

in the mass spectrometer, or by other means . However ,

it is sufficient to show trat mass spectral information,

particularly of compounds ·rrhich do not have a large

number of co~peting reactions , can prove valuable in

predicting the likely reaction initiated by other means

of excitation.

Joo .

Melting po-:nts are 'Jncorrcc ,ed .

petroleum' refer·s -Lo the frs.cL::.cn of b . p • . iC-60°0.

~ ilica gel (100 - 200 mesh) v'as obtained from L . Lirht

and Co ., and c~lum-i na for chromatography from r·ay and

Baker . nalyses are by the ..... ustralian i\~icroanalytical

Service , Melbourne . Infrared s ectra of gases \r.ere

recorded with a Grubb Parsons I .odel DBl/60L1. grating

,.., cctrometer , and spectra of Nujol mulls c.nd solutions

with a Unicam SP200 instrument . Quantitative ultra­

violet rfectra were measured with a Beckman DK- 2A

spectrophotometer and qualite.tive spectra vrj_th a Unicam

S1-800 spectrophotometer (for 901, ethanol solutj ons unless

other vj se stated) . 1-'roton ma[~net j c reson2.nce spectra

rere recorded in cerbon tetrachlorjde or deuterochloro­

form solutions with tetramethylsilane as an internal

standard on a Perkin ~lmer 10 spectrometer . An AtlRs

CH - 4 spectrometer 1as used to record the masQ spectra

unless otherwise stated. !..l1£.lytic2] VCcpour phase

chromatography results wer·e obto.. inec1 \ ·i th the ~ and r

Iv:odel :)00 gas chromatograph with a } · odel 16 9 f larne

ionisation detector, which 1:.ras cali"b1'lated 1.·1itt. standard

solutions . Preparatjve gas chromatot:;r•aphy; as carried

out with the F and J: D'"odel 500 go.s cl1romatoc1')am for sm&.11

lGl.

samples, or with an Aero~raph 1 o~el 705 for lar~er

samples .

n£olytic App~ratus

Pyrolysis vrere carried out :ising a s i 1 ica

tube ( l+O cm x 2 cm I . D. ) of 1 hi.ch the central 30 cm

sect j_on was usually pac <:ed 1:1;i th 5 mm lens·ths of

5 mm I . D. /7 mm O. D. silica turJing, c.lthough P~rrex

2 mm Fenske helices vere used in a few experir.1ents .

silica tube ,·rj th the centr-..:i.l section of F3 mm T . D.

v·c.s used un:pac 1-ed in one cxperimr.,nt . 1.I.1his central

sect ion was heated .. i th an exterEaJ. elcctrj c furn£..ce

(type 70 - r~ , Hevi - Duty Heat~n:)· Eruipment Jo., .is., U.S .. ; ,

the tecperature lJeing measured beforehc.nd with a

s ~-1 ica- shea thed ch::1omel - alumel thermocouJ•le rios i ti onec

j_ n the centre of the s i 1 i cc:, tu~> e .

Por pyrol;?s=:.c in VL.cuo ~1 1E, -J_v.rol -rr~,:: c-; tu·)e , as

mounted t.orizontally and the sol:ic1 , · s c-·J.o, ,J_y SUJ)l 11:1ec

int o the t u·0 e fr om e. s ria 11 ::-1 , ~ L 1 a: by he,:, t :_ n · · i v h ri

oil batt. or a f:;_exiole heat int.; mc...11tle . L.:ierc·ent va ;o-.trs

· ere collected in succer~si ve tra s cooled by "ice c nD.

liquid nitro:en .

containinf, degassec molecular sieve (Ljnde 1~'ype 5_, 1-"

rellets , 150 g) coolet in liquid ai~, as used to trap

102 .

gaseous products . The rrer,sure -rt..s )'Vle' 2 1 -reel rith a

T.ay differ from pressurie -,j thin the p~yrol~rf-: r- tube

itself .

0olutions ·ere l.dded drop\!::.se from a prersure ­

compensa ted funnel controllecl wi t:1 e. need1e vnl ve to

the top of the vertjcally- mounted pyrolysis tube ,

pac1 -ed ith Fens~<:e helices . A stream of ni.trogen

( 150 ml/min) sv1ept the solution down the tube c nd the

__Jroducts ~·ere collected in tv·o i.ce-cooled tra~s . T:1e

gs.s stream · ras then -oas sed t1~.ro 1J.gh a cuprous chlor·ide

1 t . 111 so u · 10n •

Pre a.r.gtion of the henylhyarazones

Stoichiometric amounts of the carbon;7l

compound in methanol and phenylhydrczire in lOo ace~~c

acid were mixed, and the product ·cecri: s t':tll:sccl ... ·rom

aqueous :::netha·.1.ol until tle m. p . sl10 ec ·ood a~ree .ent

·ri th previously recorded values 112 • ::'"n t 1 ~ cc se of

propiophenone, the derivat2.ve (recorded in t.e 1~ter·avu:::--e113

·:J_s an oil) ··,as obtair1ed as cclourless crystals m. p .

~8- ..J0° ( decorn:p; . The :p . m. r . srectrurr: \ &s accord in~~ly

measured, and shor"ed only the S7<..;n8.ls to be ex ected

from the phenyl 1ydrazone of prop:..ophenone.

Ii

l

10~ -- .

Pyro1x..§is of the Phenylhyd1'"'azones

The phenylhydrazones (usu~lly about 5 g)

rere su~)l imed from a 5 cc flask heated by a \ rap- round

mantle into the heated pyrolysis tube pac~zed with pieces

of silica tubing , and the emergent vapours trapped and

q_uant i tat i vely analysed b:r gas c~1romatography ( 6 ' x ! "

Ca.l'bowax 20E (C . 2510) on glass beads (0 . 177 mm) using

the F and ~ 1609 (flame ionisation detector, N2

florr

rate 25 ml/min , temper8.ture pr·ogrammed at 11 ° /m·i n from

100 to 200°). The anilines, vhich were poorly resolved

on Curbowax , rere se-carated on a 12 ' x 4 n 11 ~'Y..~Ja_rh:1.se11

silicone oi l column ( 0 . 2510 on O. 177 mm gl2ss ·oeac1s,

N2 25 ml/ min , temperature 125°) .

Table 1 , vvere calculated from peak areas, as the cletector

gave a linear mass response curve for these com~ouncs .

Qualitative separation of the condensate into c hydro­

carbon fraction (benzene, toluene etc.), benzonitrile,

anilines , and biphenyl, diph(nylamine &nd toluidines

( .her·e present) , v1as achieved by pre-parative v . p.c .

( Aero graph 705 , 10 ' x ~" varb o,, ax 20]\ ( 25;)) /YOH ( 10/o)

on Embacel ).

In the part :icular cs.se of the condersed-phaoe

pyrolysiQ of benzaldehyde phen;ylhydrazone at atmospheric

pressure ( 720 mm at Canberra ) , the co.npound (l .• 7 6 ) ,·as

heated in a metal bath (200- 260°) anc the dist ' lling

104 .

vapours condensed and collectea . No precautions were

tG~en to exclude oxygen, as this -~s a dunlication of

1-..,1 ~ ..... ,1ed ·ork . The cloudy, colourless distill&te

( 1 . 0 g) sas separated on the C2.rbowax 2or /KOH column

(N2 ~00 ml/111in, temperature 80 to 240°- nonlineur) into

benzene (0 . 3 g), toluene (C . 05 g), ber.zonitrjle (0.2 r;),

aniline (0 . 2 :,) and c1iphenylmethane (0 . 1 g;. lhe

residue in the flas 7{ ( 3. 1 s) ;;:as dissoJ ved in be!lzP.ne

and cbromatographec1 on silica (.?. .: 21 cm). ~lut · on , r~ th

benzene gave a brovrn oil ( 2 . 8 c), follo 'Cd )Y 8 lj ·hter

oil ( O. l 1:..·) , whj ch was shown to consj st m2.inly of

aniline by v . p . c . No other tractable proc'uct s could lJe

recovered from the column. 7he 7.Jrown oil first eluted

was freed from less volatile materials by short - path

distillation at 200°/0 . 005 mm, and the distillate

separated j_nto benzaldehyde ( o . 8 g), dj_phenylmethane

(0 . 3 ~), trans - st1lbene (0 . 6 g), benzylideneaniline

( O. 75 g) and several unidentified products on the

Carbo ,ax 20} /hOH column (n 2 3 0 ml/min, temperature

100 to 200°-nonlinear) .

:vrolysis of 1, 1 , 4 , 4 - Tetrameth:vl tetralin-2, -:S - D · or.e11l.J-

T:te djone (100 rr~ 0·) ·was sublimed in 35 m:.ntttes

from a 5 ml flask into the p:rrolys~ s ~~ube pac1--ed "\v ith

105 .

pieces of silica tubing at 6oc 0 /G . 05 mm. The condensate

in the exit tubes a1x1 ice and licujd nitrogen cooled

traps 1as V'ashed out Tith methylene chlor 1 de. ~vaporation

of the solvent yielced 74 rr,g of oily bro1·rn solid which

vras dissolved in light petroleum ana added to a colwnn

of silica (1 x 16 cm) . Elut~on rith light petroleum

gave a yello·· bro· n oil (8 mg) . Gas chromatography

(2 ' piezon L, He 100 ml/min, temperature 100 to 275°

at 11°/min) of this fraction showed the major component

(.Q.§ . 3 mg) to have a similar retention time to that

of naphthalene . Further elution of tt.e silica. column

7ith benzene yielded fr act ions of pale bro\ ·n f'ol-; d,

follo red by oil and amorphous gum. Continued elution

·ri th chloroform gave no furthe1., c1:1:rsts.lline mr tc,.,~ 01.

Co11parison of the infrarec.:. s ectra of t·le cryc--,tt..lline

benzene fractions shoved they cont2ined starting rrBter-ial

(8 mg), and a material ,,,ith an infrared absor~pt~on at

17~-8 cm- l (nujol)(G.3 mg)(l,l,3,3-tetramethylinc~ane66

- 1) 1750 cm •

Sublimation of the dione (5 C mg in 40

minutes) into the pyrolysis tube at 800°/0.2 mm yielded

a dar~ ~rorn condensate, vfu~ch ~as washed from the

collection tubes rith benzene . The residue from

evaporation of the benzene was extracted ith light

106 .

petroleurn ( 3 x 15 ml), and the extract 8f ced to a column

of cilica (1 x 17 c0) . Elution\ ith lifht petroleum

yielded 198 mg of light bro·,n oily solid. rurther elution

\'ith light petrolewn, follo· ed by benzene and c:1lo:eofc2m,

gave no fur0 ther crystalline rno.ter1 io.l . ~)>::a.:.'ination of the

ini'1"ared sl;)ectra of all fr'act ions sho\~red that the

crystalline mate1'lial eontained no car~bonyl cc~r:pouncts,

while no absorption cor·respondiDt; to the tet11 r.lin-2, - -dione

or the indane - 2 - one could be detected in the snectr& of

the remaining fractions ori t e lie·ht }JetroleJJ;" irsolu,1lc

material . -the

me. terial on aluriina '!rj_th ljght petroleun1 prod1.,-:.ccd y~o

further seiarntion of the constituents. u0-nt i tat i ve

anc-:. preparative gas cl:~_,.,omc.tography (2' .t*p·ezon L, He

100 ml/min , temperature progro.nuned at 11 ° /min from J 00

to 275°) shoYed the major component was napht~alene

(118 m~) , with the remainder consistint.: of sever·al

compounds, most of 111 hich had rctent ion times less than

that of naphthalene .

PyJ."olysis of Indanetrione Solutions

Solutions of indanetrjone in benzene and

chlorobenzene v.ere prep:....red by azeotro··)ic dehydration

of ninhydrin ri th exce' s o ' the sol vent. The dror:ring

107 .

funnel used for addition of the solutions to the

pyrolysis tube was rapped ri th a heatj_ng tape to

prevent crystallisation of the sparin°;ly soluble

indanetrione .

Experiment A (Table 2) Indanetrione (250 m~) in hot

benzene (60 ml) ~~s added dropwjse to the top of the

vertical pyr olysis tube pa c ked vri th Pyrex Penslce

helices at 500 + 10° . The benzene solution , whi ch

collected in t he ice- cooled traps, was evaporated and

the dark tar ·hich remained was chroma to graphed on a

column of silica (15 x 1 cm) . Elution with light

petroleum yielded a hydrocarbon fraction (21 mg) , while

further elution wi th benzene and chloroform gave a more

polar fraction (114 mg), consisting largely of unchanged

indanetrione (infra red spectrum) . The hydrocarbon

fraction was separated by gas chromat ography (6' Carbo ·ax

201: , He flo ii/ rate 100 ml/min, temperature programme 100

to 250° at 11°/min) into naphthalene (11 mg) , bjphenyl

( 3 mg) and t vo other minor uni.dent ::..fied components .

Experiment C Pyrolysis of indanetrione (?70 mp) in

benzene ( 60 ml) at 600 ± 10° gave a proouct wh, ch vvas

se1,arated as for Experiment A into a hydrocarbon fraction

(210 mg) containing naphthalene (50 mg) and biphenyl

( 147 mg) followed by amorphous tar whic h was not examined

108.

further .

~xreriment E Pyrolysis of indanetrione (150 mg) in

hot chlorobenzene (60 ml) at 500° and evaporation of

the resulting chlorobenzene solution at .Q.§.. 60°/20 mm

(rotary evaporator) gave an oily residue (185 mg) . Some

loss of the more volatile components (naphthalene,

biphenyl) possibly occurred during this evaporation.

The residue was chromatographed on silica (1 x 16 cm)

and a non- polar fraction (130 mg) was eluted with light

petroleum. Elution with benzene gave a polar fraction

( 43 mg) consisting largely of unchane;ed indanetr1 one.

The non- polar fraction vas separated by gas chromatography

on Carbo ax 20M into biphenyl, naphthalene, and a mixture

of 1-chloro - and 2- chloronaphthalene . A number of

unidentified minor products vrere al so forimed, but these

could be shov n by gas chromatography to 'J e present also

in the pyrolysate from chlorobenzene alone ( experiment .f.l1 ) .

~YPeriments B , D and F The products from blank

ex~eriments on the pyrolysis of the solvents alone were

examined by direct gas chromatography of the condensate

in the ice-cooled traps .

~he nitrogen gas stream from the ice-cooled

traps ras passed through a cuprous chloride solution111

before venting to the atmosphere . The red precipitate

of cuprous acetylide which formed in Ex_periments _ , C and

109 .

E Has collected and decomposed by warming with aqueous

1~otass ium cyanide solution. The libers.tec1 acetylene

vras rassed through a calcium chloride drying tube into

an infrared gas cell .

Pyrolysis of ~inhydrin in vacuo

The products and fractio:-c1s isolated from

pyrolysis of njnhydrin under four· (1-if/er•ert se~s of

conditions are listed in ~ able 3. y-r·ol:ys-is of

ninhydrin (1 g) by sublimatjon throu"·h t:--e )yrolysis

tube pac1~ed vrith pieces of silica tubing at

Li.OO ± 20° /c . 1 mm gave only- starting materj aJ., obtained

as a mixture of indanetri one and j ts hydrate, ninhydrin.

The approximate lower temperature lim~t for the pyrolysis

is set by the experiment at 500° /0 . 1 mm, ~rhich gave

only a trace of hydrocarbon , a little benzocyclobutene ­

dione , and starting materj_al . r:I1he methods of se-oarat ion

are outlined for the t~o most corrwrehensive exreriments

... and B Crable 3) . In experiments A,.B and C the

packing 8.nd v:all of the tube bec2-rne coated "j_th a hard

film of carbon.

Ex1Jer iment A Hinhydrin (1 . g) ,. ras sublimed from a

5 ml flo.sk into the pyrolys j s tube rac1 ".:ed , :. th nieces

of silica tubing at 600°/0 . 2 nYn durin.f~ one' our. 'r~-:e

I

110 .

condensate in the exj_t tube and ice- cooled trap was

dissolved in benzen8 (15 ml) and chromato rraphed on

a column of siljca (2 x 24 cm) . Elution v:rith benzene,

gave fractions containing success~.vely :bydrocarbons,

fluorenone , anthraquinone, benzocyclobutenedione,

phenanthrenequinone and indanetrione. Final elution

with chloroform yiel~ed a small fraction containing

ninhydrin. Individual carbonyl compounds v.rere purified

by further chromatography on silica. The hydrocarbon

fraction was separated by gas chromatography

(2 ' Apiezon L , He 100 ml/ min , temperature 100 to 275°

at 11°/min) into biphenylene and triphenylene, with

only trace amounts of other hydrocarbons .

Exneriment B Ninhydrin (5 . 0 g) was pyrolysed over

pieces of silica tubing at 800 ± 20° /0 . L~ ± o. 2 mm during

one and a half hours . For the first 40 minutes of the

experiment a trap cooled in li~uid nitrogen and filled

1ith molecular sieve was connected to the end of the

train of traps . At the end of the eX!)eriment this tra_p

was connected to a 10 cm j_nfrared gas cell and the gases

which were evolved on heating the trap to 1co0 vere

collected in the cell . In thjs way the pres nee of

methane and carbon monoxide was detected by their

characteristic infrared absorption. Passage of a stream

.......

-

-

111.

of nitrogen over the molecular sieve at 250° and then

1 . 1 t· 111 d through a cuprous ch oride sou ion gave a re

precipitate which was probably cuprous acetylide, but

the amount obtained was too small to :permj_t the

regeneration and identification of acetylene.

The condensate from the first trap ~as

subjected to preliminary chromatography on silica

(18 x 2 cm) from benzene solut:..on, which serlar>ated the

hydrocarbons from the mor>e polar fractions containing

fl uorenone, anthraq_uinone, and phenantLreneq_uinone. 11he

combined hydrocarbon fractions were extracLed repeatedly

with warm light petroleum, and a small q_u, nti ty of

insoluble dark tar ras discarded. 'I1he light petroleum

extract was chrornatogro.phed on alumina ( 2 x 18 cm).

Elution ·with 1 ight petroleum gave (a) b iphenylene

containing small amounts of naphthalene and biphenyl,

followed by (b) a series of small fr>actions containing

fluorene , anthracene and phenanthrene. Light netrolewn­

benzene then eluted a fraction (c) contc...ining hi~her

hydrocnrbons in the range c14

- c18

• Fractions (a) and

(b) Here separated by gas chromatography on a G' column

of Cs.rbo · ax 201: ( e 100 ml/min, 100 to 250° at 11 °/min) .

l1

he higher hyd. ocarbons (fraction c) vvere chromatogrnphed

on a column of partially acetylated cellulose (LO x 3. J

with methanol- toluene- water (10: 1 : 1) as eluant 115.

'\ Cu J

........

112.

No serarat ion of hydrocarbons below .:::18 "s.s achieved,

but the c18 comi;,ounds, 3,l1.-benzphena17 threne,

triphenylene, 1,2-benzanthracene, arcl chrysene rere

satisfactorily separated. The first fractions,

a:9proximately c14 - c18 hydrocarbons were then sera rated

by gas chromatography on the 2' colurrm of Apiezon L.

The fractj ons cor .. taj nir..a carbonyl compounds

from the original column v.rere re-chromatographed on

silica in benzene, and incli vidual components were then

recrystallised, or 1 in the case of fluorenone, converted

to the 2 ,L~- dini trophenylhyd..razone .

Pyrolysis of Phthalic Anh:ydr1de

1I1he phthalic anhydrjde was PYI'olysecl under

similar conditions as used for the pyrolysis of ninhydrin

in vacuo . The results under four sets of ex,erimental

conditions are shov n in j_1 able Ji, while the methods of

sepnration for experiment A are reoresentative of the

group.

Ex,~ er iment hthalic an..hycl.J::j_de ( 5 g) vro.r su1J1i1ned

from a 5 ml flask into the pyrolys j s tu1Je, unpac 7{ed

except for the silica sheath of the thermocouple placed

concentricc.lly in the central heated section of the tu1Je .

After an h our the \ hole had sublimed. through the tube 2.t

760°/0. 3 mm and t:ne condensate in t11e exit tube and

-

113 .

ice cooled trap was extracted with light petroleum

(150 ml) . The extract 17as chromatographed on a silica

column ( 22 x 2 cm) from which elut j_on ri th light

petroleum follovrec1 by benzene yielded firstly a pale

yello7 solid (no carbonyl absorption in the i . r . ) ,

fol lovred by a yellov.r solid ( no carbonyl absorption), a

light brown solid ( i . r . absorption corresponf .. ing to

phthal ic anhydride) and f ine.lly a br•own r;um ( no c s.rbonyl

absorption in the i . r . srectrwn) . The f'rst fr3.ction

¥as senaratecl by gas chrom~tography (2' Ap'ezon L, He

100 ml/r,1in, te1 _perature 100 to 275° at 11 ° /r'.lin) into

b iphenylene containing small a.nount s of naphthalene,

b iphen;?l , fl uorene , phenanthrene, arid anthracene. By

similar gas chrorr..a tography the second fraction was shovrn

to consist largely of triphenylene contaminated with

only small amounts of other hydrocarbons .

The petroleum insoluble fraction was

diL~olved in benzene from -hich chromatography on silica

(19 x 2 cm) separated it into phthalic anhydride as the

ffi<..,jor corstituent, follov1ed by a small amount of phthf lie

acid . The liquid nitrogen cool eel trap vras alloi,,ed to

··arm to room temperature so :ihat the expanding gas

bubbled through barium hydroxide solution. A v1hite

precipitate, soluble in dilute ac 'd, indicateC:. the li'ce1,/

--

presence of carbon dioxide .

Py~olysis of Fluorenone

114 .

Fluorenone (5 g) was subljmed from a 25 ml

flask into the pyrolysis tube pac 1,,.ed ;rith pieces of

s"i.lica tubing at 830°/c . 05 mm durj_ng 50 mj_nutes . The

apparently unchanged condensate v'as extracted ,·r"i th

licht petroleum (3 x 20 ml) and after reducjng the

volume of sol vent to 15 ml, the extract v1as added to a

silica column ( 11 x 1 cm) . ~lut ion vrith petroleum

gave a yellov: oil ( 3 mg) vvhile further elution wj_th

benzene gave only unchan~ecl fluorenone. r:_,:1e oj_l ;_·ras

shoi,·n to contaj_n several covnponents by gas chromatoP~ra·ohy

on a 2' column of Apiezon L (:-Ie 100 ml/n in, te[-1pero..ture

100 to 275° at 11 °/min). The t ~,o major components

were separated prepg.ratjvely and the ultravj_olet

s:0ectra showed absorption characteristic of oiuhenylene

and fluorene . I\iir..or pea,cs had retention times the

same as naphthalene, bjphenyl and c14

hydrocarbons.

Pyrolysjs of Anthrg_guinone

Anthraqujnone ( 5 g) vvas pyrolysed in the

same a.-rparatus as used for the fiuorenone at 830°/o. 05 mm

over 80 minutes . The condensate was extracted '7ith

115.

l 1 µ,·ht petroleum ( 3 x 20 ml) and the extract

chromatographed on a column of silica ( 2..J x 2 cm).

Elution with petroleum r:ave first] ·::,r a srrEll amount of

yellow oil (..£§ . 1 mg) followed hy nnthracu 1 none.

Further anthraq_uinone vvas the only material obtained

by change of elu tant to benzere . Gas ch.roma:tography

of tl'1e o j_l on the 2 ' Api ezon L column sho r1ed it cort airied

components with the same retention tj_me as naphtha.lene,

biphenyl, biphenylene, fluorene, anthracene and

phenanthrene . Tli.e petroleum insoluble material was

extracted with benzene (2 x 25 ml) and the extract added

to a silica column ( 15 x 2 cm) . Elution ri th benzene

gave only unchanged antbraquinone. The infrared spectra

of all fract-ions was recorded but no absor· tion which

could be attributed to fluorenone was detected.

Pyrolysis of Phenanthrenequinone

The qui none ( 1 g-) \ as sublimed through the

pyrolysis tube pac'red wjth pieces of 2ilica tubing

at 800°/o . 4 mm during l,5 minutes. 'l1he product, vh-i ch

i.vas formed cleanly without ceposit"ion of cRrbor. on the

packing, vas dissolved in benzene rrrl cl-iromato ·raphed

on a colwnn of silica (18 x 2 cm). ~lution ,rith benzene

gave fractions containing a trace of bro\·n oil ( ca. 1 mg),

116 •

fluorenone (660 mg) and unchanged phenanthrenequinone

( 130 m,c;) . Gas chromatography ( 2' Apiezon I) of the

oil showed that 1t contained only very slight traces of

hydrocarbons larger than the petroleum solvent.

P:yrolysjs of Tetrachl oro-ohthalic Anhydride

Tetrachlorophthalic anhydride (5 ) ~,as

sublimed from a 25 ml flask into the pyrolysis tube

packed ··i th pieces of silica tubing at 700°/o. 2 mm over

ti o and a half hours . The condensate in the exit tube

and ice cooled trap was recovered into ether·, y 1 elc j nn·

2 . 68 g of dark, yellow- green meter-ial on ev~norat~on of

the solvent . Cru.1 omatography of the l 1 °;ht petroleum

extract on silica (20 x 2 c~) gave a pale green fraction

(520 mg) on elution vith petroleum, but further elutjon

with benzene and chloroform yielded only a ffilall amount

of &morphous material . The crystalline fract~on ~as

shown to be principally lriexachlorobenzene ( 375 mg) by

gas chromatography (2' piezon L, ~ e flo~ rate 150 ml/min ,

temperature 100 to 275° at 11°/min) . The benzene extract

of the light petroleum insolubles -ras chromatoFraphed

on a silico. colwnn ( 19 x 2 cm). P_)nzene eluted a green

bro- rn solid ( 630 mz), but further eluL :.on with chloroform

follo red by ethyl acetate gave no further material .

Hexachlorobenzer.e (J50 ~ 0 ) and octachlorobinhenrlene

117 .

(15 mg) vere obtained from the solid fraction by gas

chromatography ( 6 11 Apiezon L, He 2or ml/r:1in, tern- ,erature

program.rne 100 to 275° at 11 ° /.nin). The cxtract 1 on­

chromatography process wa.s re.eatecl , ,·ith cl1loroform

and yielded,. after gas chroua to12:ro:ohy, hexc:1.cr1lorobenzene

(98 mg) and octachlorobi-ohenylene (?3 mg) . The remaining

inso:uble residue (1 . 33 g) was recrystallised from ortho­

dichlorobenzene to yield octachlorob-i-ohenylene (992 mg)

as yellow- green crystals .

Many of the separatjon steps are redundant ,

and a single extraction and sj 1 ica chroma tor;raphy v ith

chloroform 7ould pi obabl3 se:rarate the hexachlorobenzene

from the bulk of the octachlorobiphenylene. Overall,

l o02 g of hexachlorobenzene and 1 . 03 g of octachloro­

b1nhenylene were obtair:ed from the 5 g of tetrachlorophths.lic

anhydride .

Pyrolysis of 3, 5- C;rclohexad-i ene - c i s-1, 2-DicQ.£boxylic

Anhyd.r ide 116

The anhydride ( 500 mp;) as subl-; xed from a

5 ml flask into the pyrolysis tube ps.c· ed 11i th ;:: 1 eces

of silica tubing at 800°/0 . l mm during one hour . 120 ffiO' 7

of colourless liquid, ~1hich was shoirvn to be substantially

benzene by gas chromatography (6' Carbowax 20P (0 . 251°b)

118.

on glass beads ( 0 . 177 mm), N2 flov rate 25 ml/min,

temperature 75 to 150° at 11°/min), was poured from

the lig_uid ni t rogen cooled trap. A further 30 mg of

lig_uid was washed from the tr~ap vri th ether . A pale

brown solid ( 40 mg), which sho\red absorption in the

infrared spectrum correspondin[" to a mixture of the

original anhydride and phthalic anJ--ydride, , as r·ashed

from the exit tube and ice cooled trap with eth " .

Pyrolys j s of 1 , 2 ,4 , 5- tJ.1etramethvl- 3, 2- )ycJ.ohcxad j_.§Jl§­

c is - 1 , 2- D icarboxyl ic Anhydriae117

The anhydride (500 mg) ·,as sublimed through

the same apparatus as used for the unsubstituted

anhydride at 800°/0 . l mm in one hour. The white

condensate (136 mg) in the exit tube and ice cooled trap

was shown by gas chromatograrhy ( 2 ' Apiezon L, He 75 ml/ nin,

temperature 75 to 150° at 11°/min) to consist of durene

(129 mg), contaminated with o~ly small amounts of other

materials . The lig_uid nitrogen cooled trap vas allowed

to warm to room temperature so that the expanding gas

bubbled through barium hydroxide solution. A white

precipitate , soluble in dilute acid, 'Nas indicative of

the likely presence of carbon dioxide . The yelloi. .. oil

( 9L~ filL), which remained in the trc.p at room temperatur"le

... 10 ..L _.I .

rras shovrr.. by similar gas chro1natography to t}1at &,Jove,

to cont a in compounds with the sc1. ne retention ti 1 e as

benzene , toluene , xylene , ucs i tylene ancl durene , vvi th

the last compound constituting less than~~ of the

mixture .

;vrolysj_s of yrazine- 2, 7 - Dicarboxylic hnh~l-P ic1e118

'~I.1he anhydride ( 620 m6 ) · 'o.S su'Jlirr ec'.. fro!':.

a 5 ml flas 1

:( into the pyrol~/s :.s t L1 )e cJ.c 1 ed , it h short

pieces of silica tuoi.n'~· at c3c O / • s m:-: ovr;[' 75 r 1iriut -:s .

'11he condensates vere ch:-comato~~rnr,:1~c1 on sil:c. '16 x 1 c .. ) ,

from v.hj_ch eluvion with benzerie yielr eel fi.rstly •r,·hite

crystals ( 116 n-iu ;, f ollored by a pale ycllovv oj 1 ( 152 ms) .

Further elution o:· the culo.n1n \ i t.1 chlorof or·m ar:rJ ethyl

acet&te gave o:::ily a sl ieht &. ount of ~--.1or:ohoL1s um. Gas

c .1.romat0Lr1 L,.phy ( 6' LAC 72 _, Le 75 11/min , te.!'.perature

175°) of the t;ro crystalline fractions, s 11.ov.ed e&ch vr:as

a mixture of two cot 1poun( s, al thouch each f ract:on

contained principally one cox.pour_c:, v:ith a 8111&11 anount

o-r the other . The com~ouncJs were separated pre u.r& ti vely

on tne sa1 'e colu.nn into iumaronitrile ( 130 ff r) L.ncl

ma 1 e or.. i t r : 1 e ( 13 0 1 1 1 • lhe ultraviolet srectra of all

cl:romato~r ~Jhic f'ract-;ons &s : 1 .corded, but croved ro

che.ract(! j st; c 'bsorption ans.rt :·ro.;1 tr_e 1:itr.:.lc::: . Ln

removino the second trap ~·rom t ie l i nic-;_ r: i trc ·er:, t ri '";

120.

expanding gases vere passed through barium hydroxide

solution and again produced a white precipitate,

soluble in dilute acid .

Pyrolx.§is of Pyr7-dine - 3J±.::Dicarboxylic Anhydride

The anhydrlde (1 g) \"as sublimed through

the same apparatus as used in the previous exr:er 1 ments

at 800° /0 . 3 mm over li-5 minutes . The darl greenish

condensate in the exit tube and ice cooled trup v·as

extracted 1JTit h ether and ca . 50% dissolved . ':he

rern2.inder was coluble in 2 molar hy~rocrloric ac.;d;

and the resulting solution bas'fied , .. .;t soc7..j_u..1:. hydrox:..de

solution. Extraction of the basic solution with

chloroform ( 3 x 15 ml I gave a dar1<:: greer: ar:iorphous gw11

on evaporation of the solvent . 1111,.. • ~1...ls vas chrornatcgranhed

on alum:..na (30 x 1 cm) , but elut::i.on 1v·ithbenzene,

followed by chloroform an~ ethyl acetate, gave no

cr;ystalline ma ter:.al . The infrc.red and ultra.violet

spectra of all fractions rere recorded, but no

distinctjve absorption could be recognised

P:vrolysis of SulLLobenzoic nhydr:-:_de

Sublimation of the e.nhydride (1 g from a

25 ml flas 1<:: :.nto the pyr·olyt:is tube uac 1(eo. · ith p.; eces

-

121.

of sjlica tubjng at 730°/0 . 05 mm during one hour

produced a brown condensate in the exit tube and ice

cooled trap. The liouid nitrogec ccoLe~ trap ,as

allovved to warm to room terperature so that the

expanding gases passed into a sodium hydroxide solution

(2~o) . The addition of barium chloride solution

produced a white precipitate which vas soluble in dilute

hydrochloric acid. Boiling the acid solution resulted

in the evolution of a gas vvith the pungent odour of

sulphur dioxide . The boiled solution was basified with

dilute ammonia but no precipitate was formed, indicating

the absence of sulphate ions .

The bro~n condensate (779 mg) vas extracted

with benzene and chromatographed on a silica column

(18 x 1 cm). Elution of the column with benzene

produced a hydro carbon fraction ( 36 mp) which V;,as

separated by gas chromato~raphy on the 2' colu~n of

Apiezon L into naphthalene ( l~ mg), biphenyl ( 2 mg),

biphenylene (23 mg) , fluorene (3 mg), c14 hydrocarbons

( ca. 1 mg) . c15 hycrocarbons (ca . 0. 5 m~), c17 hydro ­

carbons ( ca . 0 . 5 mg) and c18 hydrocarbons (trace).

Further elution of the colwn_n with chloroform followed

by ethyl acetate gave a dark gum (591 mg) . This

material, as with the benzene insoluble residue (111 mg),

--

122 .

·ras soluble in vvater . The addition of ethaDolic sodium

hydroxide to the aqueous solution caused t~n3 precipit ­

ation of the sulphobenzoic acid e.s the sodium salt .

P 1 . f n :i. 119 yro ysis o · .Joumar·ano.1one

The products from the pyrolysis of

coumarand.ione at 600° and 820° are shown in Ts.ble 5.

For all experiments in this table the pyrolysis tube

was packed ·vvith short pieces of silica tub tng .

Elxper iment A Coumarand 1 one ( 500 mg) Vias sublimed in

four hours through the pyrolysis tube at 600° /G . 2 mn1.

'l1he condensate in the exit tube and ice cooled trap

was chromat ographec1 on s j_l i ca ( 25 x 2 c11) and elut-; on

with benzene gave fractions containing o. brovrn oil

( 3 m<..;) , a pale brown so 1 id ( 5 5 .t11,i; 1 , ar1d 2 1 · rt.t bro ~rn

sol id ( 30 mg ) .

cum ( 1 ,4 mg ) •

Elutjon wit: ethyl acetate y 1 elded a

Gas clirom1:-i.togru.pL:, ( 2 ' 1)~ e?on I., re 1cc

ml/min , temperature 100 to 275° at 11°/min1

of the o;l

shored it -,as mostly biphenylene oxide cor..tuininf· traces

of other hydrocarbons, but none 1 ith a retention time

similar to that o:· b j_pheny1ene .

shoved infrared absorption corres rondinc to o. m~xtm'"'e of

xantr.one, and 3 ,L1- benzcoumarin, o.nd these were separated

by dissolving the solid in a hot solution of sodium

123 .

hydroxide ( _7 g) , methanol ( 5 ml) and , ,ater ( 10 ml/.

The methanol Nas removed on a rota:::1 y eV' por'"'1.tor and

the aqueous solution extracted ·,rj th methylene chlo:-ide

( 3 x 10 ml) . Evaporation of the methylene ct ... lol· ide

gave xanthone (14 mg) . Acidification of the aqueous

solution with hydrochloric acid (10 IT) and extraction

vri th methylene chloride yielded 3, 4 - benzcoumarin on

evaporation of the solvent . The third fraction showed

absorption in the infrared spectrum corresponding to

a mixture of polysalicylides . Thj s ,~,as confirmed by

hydrolysis of an aqueous sus·l ension of the solid in

2 H sodium hydroxide for 16 hours at 100°, v:h i ch gave

salicylic ac 2.d .

The gum eluted with ethyl acetate from the

original chromatography ,7as rechrom2.tographed on silica

(11 x 1 cm) . Salicylic acjd (14 mg) vas obtained by

elution ~ith chloroform.

Exner iment B Pyrolys j_s of the cnumf.::_'"'c_nd 1 one ( 1. L~ g

at 820°/0 . 4 m.m gave a dar 1-: coloured condensate vh~ch

\ ras ror ked up as for the prev. ous experiment . Gas

chromatography of the first s::_lica fraction sho··ed j_t

was largely biphenylene oxiCe, but ia.entifiable amounts

of naphthalene, binhenyl and fluorene o/ere precent . The

xanthone and 3 . 4 - benzcoumarin ·ere again separated by

1:

their sel e c tive solubility in alkaline solution •

. emaining fractions , vhich contained no crystalline

material , and sho,,·ed no carbonyl absorption jn the

infrared spectra, vrere not investigated fw.-.ther . 'l1he

lig_uid air cooled trap 112.s allo ,reel to PP_rm to room

temperature so that any libeJ...,ated gas bubbled through

barium hydroxi.de solution. No noticeable chanve ,·as

produced in the solution , indicat in[, the o.b£ rn...,e of

carbon dioxide .

1:xrolysis of JJ1salic_ylide120

The products are she 'Yl ·\rjth those of ~re pyrolysis of coumarand~one in ~able 5. lhe p~/rolys j s

tube was again packed \yjth 1;ieces of silica tu1Jj_nc-,; .

The separation of the products as j_dent ical 1; i tr. tl:.& t

descr jbed for the ccumaranciione, -, i th the except ion

that the silica chromatography fraction co~aining the

mi -ture of xanthone and 3 , L:.- benzcoumarin vvas separated

into its const ~ tuents by gas chromatography on a 2'

column of Apiezon L (He _flovv rate 150 rnl/min, terr:periaturc

200 o) .

Pvr 1 . f ~ I. B . 121 . o ys is o 'J , 4 - enzc.Qll.me.r in

Pyrolysis over one hotL' o-:! 3,L1.-ccnzcoumc.r~n

( 2 1 through the pyrolysis tube nc. c\:ec jth pieces of

silica tubing at 630°/0 . 2 mm produced no roticeable

change in the com.pound . The cond 0 rs2.te 1·s.s cx.,cracted

-

...

125.

r1i th benzene ( 20 ml I and the extrt1 ct e.ddecl ~o a c ""'ltFrLn

of silica ( 22 x 2 cr;1) . 'lution ·ith benzcre gave f-irctlJ

n yello-r oil (5 lY.~) f'ollorrcd oy unchan°·cc. 3,J~-ber.7,cownc . .i."in.

G&s chromatography (2 ' piezon L, He l0C ml/ ln,

temperature 100 to 275° at 11 ° /min) of t}1e o 1 l f./1c·1 ed

it contained several components, ·\ j th ~l'e ri~~ in one

coLst i tut ing less than 1 mp o: L 1e .;1 1 xt:i::·e .

Pyrolysis of Xanth.Q.D.§

Xanthone ( 1 g) v:as sublimed through -Lhe

pyrolysis tube packed with pieces of sili.ca tub:inf; at

830°/0 . 1 mm dur ~ n; 75 minutes . The ar::;~e.r1 e::1tly unchant.~ecJ

condensate was extracted with benzene (25 ml and the

extract added to a silic2 column (19 x 2 cm) . ~lution

V'i th benzene yielded. firstly a brown oil ( 2 m~) f ollov ed

b;r unchanged xanthone . Gas c~romatogranhy of the oil

( 2 ' Apiezon L , He 100 ml/min , temperature 100 to 275°

at 11°/nin) showed i.t contained sever8.l components,

none of v1hich v.rer·e more than 20% of the m ·xtur e .

Deto.ils of Identification of Products -·-=---------------"-~----'--..::..--~-=-;-=--~-:::=-...;;_::...;::.-=.,;:;: __

r~I.1he means of identifying the exLe:r·irncnts,l

products is corrnnon to all sec tionr· ,. ·hej·e [,_ no.r~icv.l .r

product :: s formed unles2 other\ .; se sto.tec.

Benzonitrile

retention tin1e, infra .... ed, ultraviolet d.nd ;rotor.. ffi.'.JP,Y"letic

--

l

126 .

resonance s1-,ectrc. vi th those of c.n authentic sc.mple .

Ar:iline and r - Substituted Anilines Obtained as a mixed

fraction (from pyrolysjs of 1retone phenylhydrazones) by

gas chromatography on Car··Jo' ax 20flf-K0H . Relative

proportions were calculated from integrator traces of

the proton m2~r:net=i.c resonance s1Jectra &rd these were

in good agreement with valaes from o.nal~/t-; c·,tl gas

chromatography on silicone oil. 'Jcmpc.rir:..,on of the

pro on magnetic resonance ana jnfrarea spectra of the

experimental mixtures 8.f ·ain showed good ~ ~reement ,_ j th

the spectra of the appropriate synthet1c miyture .

D iphenylamine, ortho and meta-Tolu.id i ne,

N- ,.ethyldiphenylamine , N-Phenyl-o-r>tho and Ds.ra- Toluidj_ne ,

Benzaldehyde, Diphenylmethane, trans - Stilbene and

Be~zylideneaniline

benzoni tr ile.

Benzene and Toluene

Isolated and ident i.f ied as for

Isolated by chroma to::·raphy on

Carbov ax 2Gr'- r0F and identified by comparison of the

retention time and ultraviolet spectra ~,ith those of the

authentic compound.

3 i}2henvl (from pher.ylh~ ,- _,s.zone s)

as for benzonitrile .

Biphenyl (all other sectio~c;

~solated and iden~ified

:- s o 1 at e d by ·as c 11.r 0111,.., t o -

graphy on Carboi.vax 20~.I; m. ~ • and n--i xecl rn . lJ . , Ss-70°,

....

l?.7 .

)_ :ax 24 7 m r . :N'aphthalene Isolated by sas er romsto[·1,r- ,11_:y on .. r iezon

Lor Carbowax 20K , and identified by the m. p. and mixed

~ p 70 80° and by the ultraviolet Ll e e ' _1 - ' 1 s}:;e ctrum, .A.. .nax 220 , 248 , 257 , 266, 27:J, 285, and 312 m ~ •

1 - Chloro an~ 2- C:loronapht~alene 0bt&ined &s a mixed

fro.ct ion by gas chroma t ogrc.phy on Corbowa.; 2or • The

ratio of the tv·o isomers v•as calculated from the

integration of the p . m. r . spectrum of the mizture, as

the si6nalc centred at Y 1 . 74, presQ,ably d~e to H 8 in

the spectrum of l - chlo1·onaphthalene, 1 ere ·rell c1ov:nfield

of all other sign .ls . The inf'rared spectrwn or' a

synthe tic mixture of this calculated ratio of chloro ­

naphthalenes, showed good agreement \Tith the spectrmn

of the experimental ~ixture .

Acetylene Trapped as a red prcc :.:pi tL te of cupr·ous

acetyl ide . Ic.ent j_ f i.ed by the err ractcri ist-; c -· nfrared

122a ( ~6

-1 abs or pt ion of the gas bet·. een 730 c..nd -) 0 cm ;

·1hich ,as liberated from the acetylide by vrar;1 inr v?j th

aqueous potassiurn cyanide solution.

Carbon Monoxide I dent if ied after· desorption frcm

molecular sieve by its infrared absorption pattern12 2b

- 1 betv:een 221·0 and 2C00 cm •

Methane Identifjed after desorption from molecular

---

l

12cc .

t . tt 122c sieve by its infrared absoru 10n pa ern

- 1 3200 and 2900 cm and between 1370 and 1200

bet,reen

- 1 cm •

Carbon D ·oxide Condensed as a solid in the liq_u~d

nitrogen cooled tra.p . Identif::.ed by the forn1:'3.tion of a.

whj_te c arbonate , soluble in dilute 8.cid, with barium

hydroxide solution.

Bi nhenylene Purified by gas c:hromatography on Carbo -rax

201.: follorved by sublimation at 50°/0 . 03 mm, and identified

by the m. p . and mixed m. p ., 112- 113°, 8.ncJ. by compar~.son

of the infrared and ultraviolet s ,.ectr>a 76 ·i th those of

an authenti c sample ( ). for .,,resent, SL.tt1nle ct 23(), max

2L~8 , 338 , 343 and 358 m ~ ) .

Fluorene Isolated by gas chromatography on Carbov~x

20 ~ ri th m. p . and 1::ixed m. p . , 114- 115° arcl wj_th A max

262 , 287 , and 301 m LA identical with A . for an / max

authentic sample .

Phenanthrene and Anthracene A mixed fraction 1as

isolated by gas chromatography on Carbo, 8.X 20k and

sho 7ed -Amax ( cyclohexane) 221 , 24Li, 253, 274, 281, 293,

323 , 333 , 356 , 370, and 374 m ~ • The proportion of

anthracene present vras calculated from the value of

at 374 m r' , and this v-as checked by preparation of a

synthetic mixtu~e of anthracene and phenanthrene which

showed ultraviolet and infrared snectra identjcal vith

those of the original mixed fraction . The rctio

-

I

-

129.

anthracene : phenanthrene rvas approxjmately 1 : 3 j_n

each case .

c1 ) Fraction The carbon numbey, , as c1.ss ~ rned from the

gas c}7..romatographic re tent ion ti me on A n:_ezon L.

Vansheidt test88

for the presence of hydr·ocarbons of

fluorene or indene type ·~s positive for the fraction

from Experiment B of 'l1able 3 . ~~his fraction sho 1ed

.A max 234, 252, 261, 271,, 281 , 285, 291'-, 321,340, 357,

and 386 mr' . c17 :i?ract ion The carbon number v;as ass ::..gned from the

gas chromato graphi c retention time on Apiezon L. The

fraction from Exneriment B, rrable 3 showed A max at 230,

253, 262, 300, 314, 327, 333, 3'+7, and 366 myt • The

maYima at 253, 262, 314, 327, and 31+7 m }-A surgest the

possible presence of benzo(a)fluorene123 , and in agreement

88 with this the fraction gave a positive Vanshetdt test •

Benzclc)phenanthrene Isolated by chromato~raphy on

partially acetylated cellulose, and ide~tlfied by the

ultra violet spec i-rum vh i ch sho. red A 23c, ?7C, ?80, '18.X

302, 3llt, 326, 352, and 4 71 m /A in r;ood ap·reemert ·· i th

12 7

the published spectrum -J .

Triuhenylene Isolated by chromatography or. alumina.

and se:r::arated from other c18 hydrocarbons(· here nrcsent)

by chromato graphy on partia1ly acet;ylated cellulose .

I

130 .

Identified by the m. p . and mixed m. p . , 196-198°,

1.'hich ,as not depressed by admixture v ith an authentic

sa.nple, and by comparison of infrared and ultraviolet

spectra ( .A 2L1.8 , 257, 273, 28' 1 , 321, 3?6, and 333 m Ll • • ~ax {

Chr:ysene Isolated by cbr oma togra:ph;y on ·Jart iEiJ ly

acetylated cellulose and identif~ed by the ultrav~olet

spectrum vvhich s·1owed ~ 221, 21~2, ?SP, 257, 2Q3, max

294 , 306, 320, 3L~4, 353, a1 1d 362 m~ in ~ood ae·reement

123 vith the published spectrum •

Benz ( a)anthracene Isolated by chromatography on

partially acetylat ed cellulose, and j dent if i ed by

comparison of its ultraviolet spectrum , .A max 222, 228,

256 , 268 , 27 8, 287, 300, 326, 341, 358, 37Li., and 38L~ m ~ ,

\ri th the publ j shed spectrum123.

F1luorenone Isolated by chro~atography on silic2 . It

v.ra.s obtained contaminated with an oj ly impurity, al though

the infrared spectrum was very similar to that of an

authentic sample . The srunple ~as converted to the

2 , 4 - dinitrophenylhydrazone which had m. p. 282- 284°, not

denressed by ad1ixture with an authentic derivative.

Anthraquinone Isolated br chromatography on silicr,

and recr~rstall ised from ethanol to give yello,, -bro "'Yl

crystals, m. p . and mixed m. p . 285-286°, ard ~urther

identified by comparison of the infrared s=_ectrum.

131.

Benzocyclobutenedione I cola ted by chrom..a tography on

silic&, and purified by recrystallisation from methylene

chloride/light petroleum and sublirm. tion at 50° /0. 005 rnrn

to give pale yellow prisms, m.p. 130-132° (lit. 124

132-135°) (Found: C, 72. 7%; H, 3. 2%. c8

tt4o

2 requires

C, 72. 71; H, 3 . 1%) . Infrared spectrum ( rjs2

solution):

6 - 1 lJ 1811, 1782, and 175 cm • Ultraviolet spectrum: max

.A max (90% ethanol) 225, 286, 292, 301, and 424 mf (log E 4 . 67 , 3.62, 3 . 79, 3 . 82, and 1.57). The intensity

of the long wavelen'- th band is lower than that found

by Cava124

( log E 2. Li-4) . Oxidation of c. smal 1 sample

with hydrogen peroxide in acetic acid124 gave phthalic

acid, identified by conversion to phthalic anhydride, 0 m. p . 131- 132 .

Phenanthreneouinone Isolated by chromatography on

silica and recrystallised from ethanol to gjve orange

crystals of m. p. and mixed m. p . 205- 206°, further

identified by comparison of infrared s·--:ectra.

Hexachlorobenzene Isolated by chromatography on silica

and purified by gas chromatography on piezon L.

Identified by the m. p. and mixed m. p . , 225-227°, which

was not depressed by admixture with an authentic sample,

and by comparison of inf' rared and ultraviolet spectra .

Octachlorobiphenylene Isolated by removal of the more

132.

soluble impurities by extraction with light petroleum,

benzene and. chloroform and purified by recrystallisation

f rom ortr o-dichlorobenzene to give yellow-green crystals

m. p . 296-298° (decomp.). (Found: C, 33.9%. c12

c18 ( new compound) requires C, 33 . 7%). Ultraviolet spectrum:

.A (chloroform) 269 (inflex.), 279, 290, 333, 352 , max

369 ( jnflex.), 374, 393, 424, and 453 mf ( log E 4. 34,

4.72, 5.13, 3. 22, 3.31, 3. 44, 3.53, 3. 56, 3. 03, and

2.95 respectively). The mass spectrum showed a group

of lines corresponding to the molecular ion, with the

ion of lowest M/ e at 424 (12c12 35c1 8) and the most

intense line at M/e = 428 corresponding to the chemical

molecular weight .

Durene Recrystallised from ethanol as white crystals,

m. p . and mixed m. p., 79-80°. Infrared and ultraviolet

spectra were identical with an authentic sample.

_aleoni tr ile Isolated by chro.:a tography on silica

and purified by gas chromatography on LAC 728. Had

m. p . 30- 32° (lit.125 30- 32°), lJ 3100, 2240 cm-l

max (nujol), , 3. 78 in chloroform.

Fumaronitrile -- Isolated by chromatography on silica

and :purified by gas chromato craphy on LAC 728 and

crystallisation from ethanol . Had m. p. 95-96° (lit. 125

96°) , V max 3080, 2220 cm-l ( nu jol), T 3. 72 in chloroform.

l

133.

orthQ- Sulphobenzoic Acid Isolated 1Jy c:1roma tocraphy

on silica and purified as the sod-ium salt "'Jy orecjpit ­

ation from ethanolic sodium hycroxide solu~ion.

Identified by comparison of the infrared spectrum of the

sodium salt with an authentic sample.

Biphenylene Oxide Isolated by chromatography on silica

and purified by gas chromatography on Apiezon L .

Identified by comparison of the retention time and the

infrared and ultraviolet spectra with those of an

authentic specimen.

Xanthone Isolated in admixture with 3,4- benzcoumarin

by chromatography on silica. Purified by removal of the

aqueous sodium hydroxide soluble impurities and

recrystallisation from ethanol, or by gas chromatography

on Apiezon L. I dentified by m. p . and mixed m. p.,

173-174° and by comparison of the tnfrared and ultraviolet

spectra with those of the authentjc compound.

3,4-Benzcoumarin Isolated in a&nixture with xanthone

by chrom9.tography on silica. Pur::_fied by gas chromato­

graphy on Apiezon L , or by removal of the a0ueous sodium

hydroxide insoluble material and recrystallisation from

light petroleum. Identified by m. p . and mixed m. p., 0

93- 91+ and by the good agree111ent sho1i1n by the ultraviolet

and infrared spectra with those of an authentic sample .

134.

Polysalicylides Isolated by chrorratography on silica.

Structure suggested by the infrared absorption at

1724, 1730, 1750 and 1760 cm- l (lit . 126 gives strong

-=l absorption for tetrasalicylide at 1730 cm ,

trisalicylide at

-1 ) and 1764 cm •

- 1 1730 cm , and disalicylide at 1751 ,

Hydrolysis ,-ri th 2 N sodj um hydroxide

and acidification of the resul.tant solution gave

salicylic acid.

l

135 .

R.:£FERJNCi!iS.

1 . C. K. Ingold and G.~r. :~inr;, J. Chem. Soc .,

2725 (1953).

2. K. K. Innes, J . Chem. Phys . , 22, 863 (1954).

3. P . G. ':lilkinson_ and 1:~ . S . Hull iken, J. Chem.

Phys ., gj, 1895 (19j5).

4. P . G. 'ilkinson , Can. J. Phys., Jlj., 6L~3 (1956) .

5 . G. \ . --:ilkinson and V. E . Di Giorgio, Can. J .

Chem. , ..l§" 31 ( 195°) .

6 . J . ~uchesne , J . Chem. Phys., 18, 1120 (1950).

7. P . .. . Leermakers and G. P . Vesley, J . Chem. Educ.,

hl, 5 3 5 ( 19 64) .

8 . ... • Srinivasan , J . Am. Chem. oc., 8Lb, 3Li-32 (1962) .

9 . \!. G. Dauben and R . L . Caugill, Tetrahedron 12 ,

186 (1961).

10. r G ,-- ::i J . • • .JOOCLS, • 0rg . Chem. , .Q .. 110 ( 1958) .

11. R. Sr in:.vasan , J . Am. Chem. Soc., 82, :)063 (19;0.

12. R. Srinivasan, J . f\.m. Chen:. Soc ., 84 , 3982 1 1sS~1

•

13 . .1.P . Cava, R . H. Schle Jsinr 1;er s.nd cT.P. Van :,:eter ,

J. Am. Chem. Soc ., 86, 3173 (1964) .

lL~. L. Burnelle, r:retrahedron, 20, 2l.i.03 ( 1964).

15. L . Burnelle, Tetrahedron , 21, 49 (1965).

16. H. K. osentstock , M. B. 1 /allenstein, A.L. "fahrhaftig

and H. Eyrin~, Proc . Nat . Acad. Sci. (u.s.), .l§, 667 (1952).

136 .

17 . .A. Chupka , J o Chem. Phys., ..3.Q, 191 (1959) .

18. B. Steiner , C. P . Giese and I\~ . G. Inglam, J . G11.em.

Phys • , J1± , 1 8 9 ( 19 61 ) •

1 9. H. M. Rose 1.G tuck and H. Krauss in I\,Iass Spectrometry

of Organic Ions , (Edl tor : P • . • r~cLafferty) ,

cader.1ic , New Yor 1( , 1963.

20. H. 11

1I . Rosenstock ancl M. Krauss in Advances in

rv:a s s Spec tr ome try , Vo 1 ume 2 , (~di tor :

R. M. Elliott) , Pergamon , London , 1963 .

21. 0. L. Ri ce and H. C. ..=?.ar~spei,.,,_~er, ;; . .. m. Chem. a oc . ,

113. , 1617 (1927) .

22. L. S . Kassel , J . Phys . Chem., J_g , 225 (1928) .

23 . N. B. Slater , Theory of Unimolecular 3eactions ,

Cornell Universit:,T , revv Yorlr , 19_)9 .

24. A. D. i: cL e an , J . Chem. r hy s • , J_g , 15 9 5 ( 196 0 ) •

25 . T ,, • A. Chupka and ~,: . Kaminsky , J . Chem. Phys . , ...12 ,

1991 ( 1961) .

26 . R. E . F1ox and \! . I. Hickam , J . Chem. Phys ., 22 ,

2059 (19511) .

27 . H. Von Koch and E. Lind:11.olm, Ar 1civ Fysi'c , 19 ,

123 ( 1961) .

28. A. IIaccoll , J . Chem. Soc , 3398. (1958) .

29. ...1. . Borkowski and P . Ausloos , J . _ m. Chem.

fil , 1053 (1961) . oc . '

137.

30 . F . 1·. McLafferty, Anal . Chem., 31, 37 (1959) .

31 . c. Djerassi and c. Fenselau, J. Am. Chem. Soc.,

fil, 57'.:.>6 (1965) .

32 . J . H. Beynon, Mass Snectrometr~ and its Applications

to Organic Chemistry, Elsevier, msterdam,

( 19 60) .

33 . F . O. Rice and r1 . rri . r"uriphy., z. ,im. C ~err.. S0c.,

.§E, 896 ( 1942) .

34. 1 • J . iddleton and 1 ·. H. Starkey, J. Am. Chem.

SOC . , 81, '30 3 ( 19 5 9) •

35 . C. T . Genaux, F . Kern and v:. D. "'alters, J . Am.

Chem. Soc . , 75, 6196 (1953 1 •

36 . R. D. Doepker and P . usloos, J . Chem. hys . , hl ,

3814 ( 1965) .

37 . M. S . Gibson and A. V/. I·:urray , J . Chem. Soc . , 880

( 196 5 ).

38. R. P . C. Brown and R. K. Solly, Chem. and Ind.

(London), 181 and 1462 (1965) .

39. E . K. Fields and s . Jt""eyerson, Cherr.. Co?:;1111 • • 474

( 1965) .

40 . E . K. Fields and s . ,!eyer son, J. Am. Chem. Soc., ·

88, 21 (1966) .

L!-1 . A. . S • N e"t on and P . St r om, J. hy c•. Ch em. , 6 2 ,

24 (1958) .

138.

42 . J . N. Pitt s and /\ • D. 9 s u or n e , J . Lm. <'J hem. S o c • ,

fil , 3011 ( 1961) .

43 . s . Tu~eyerson , J . Phys . Chem., 68, 968 (1964) .

L~L1- . J . G. Burr, J . Phys . Chem. , 61 , 1483 ( 1957) .

45 . N. J . Turro , D. C. N ecJ ers, l • • Leermakers,

D. Seldner and P . D. Angelo , J . Am. Cher .

Soc., fil , u09 7 (1965) .

46 . . P . Mannine; , J . 1m. Chem. Soc ., fl, 5151 (1957) .

47 . A. L . ilcClellan anc1 G. c. Pimentel , J . Chem. Phys .,

gJ , 2L~5 ( 19 5 5) .

L18 . A. K. Bose , I . Kugojevsky , P . T. Funke and K. G. Das ,

Tetrahedron Letters , 3065 (1965) .

L~9 . • K. Bose , K. G. Das and P . T . Punke , Am. Chem. Soc .

Abs tr . , 31s ( 1964) .

50. F . P . Chat taway , C. L . Cumming and rr . ..dil sden , J .

Chem. Soc ., 99 , l9j0 (1911 ,.

51 . J . F1r1 anck and E . ... alJinovritsch, ---1ranc . l·'araday

Soc., jQ , 120 (~93~) .

51a . J . Collin , Bull . Soc . {o~r . ci . Liege, 21 , ~46 (1952;.

52 . G. Kohlmaier and B. s . Rabinovitch, J . Fhys . Chem. ,

B, 1793 (1959) _

53 . A. v!. Hofmann and C . ... . .. artius, Chem. Ber' . , g ,

7u.2 ( 1871) .

54 . A . W. Hofmann , Chem. Ber., j, 70L1. , 720 (1872) .

55. A. l . TT ofmann , Chem. Ber. , 1, 526 ( 187 LL) .

1'

u

139.

56 . E . D . Hughes and C. K . Ingold , Quart . Rev. (London) ,

6, 43 (1952).

5 7. J. 'l1hes ing and H. ~~ayer , Chem. Ber . , 87, 1084

( 19 54) .

58. H. Hart and J . R. Kosak, J. Org. Chem., n_,

116 ( 1962) .

59 . Y. Ogata, H. Tabuchi and K. Yoshidc:..., ;'etruhedron ,

_gg, 2717 (1964) .

60 . v:.N. "Jhite, D. Lazdins and H. S . r~-hi te, (T. lm.

Ch em. So c . , 86 , 151 7 ( 196 4 ) •

61 . F . Komitsky , J.E . Gurst and • Djerassi, J. ~m.

Chem. Soc . , _§j, 2· oO ( 1963) .

62 . J.H . Bovvie , R. Grigg, D. H. ·.~!:..lliams, s .e. Lav 1esson

and G. Schroll, Chem. Comm., 403 (1965).

63 . D. G. B. Boocock and E . S . ·.raight, Chem. Co-0101. , 90

( 1966) .

64. J.H. Bowie , D. r. Cameron and D. H. Uillia1ns , J. _tun .

65 .

66.

67 .

68.

L. P.

G "7 • _e,i .

n.s.

R. S .

Chem. Joe ., .fil., 509l~ (1965) .

Cava , l • J . ohl and I . J . T1:i tchell, J . Am.

Chem. Soc . , Ju, 2080 (1963).

Gream, personal communication.

Berry , G. N. Spokes and .( . M. St i 1 es, J . ..t,1.

Chem. Soc . , 82, 524 ( 1960) .

Berry , J. Clardy and '!I , )Chafer, J. -''...111 . l • ....:..J .

Chem. a uoc . ' 86, 2739 (1964).

...

69 . J . F . Bunnett, J . Chem. ~due ., j_§, 278 (1961).

70 . H. 1-eaney , Chem. Rev., 62, 81 (1962) .

71 . G. 1

fj_ttic , Angew. ~hem. Intern. Ed., 1±, 731 (1965) .

7 2 1, ! B 1 J F 1 T " 0 . D R P t d • . • a{er, •• . • Ic mie, •• res on an

V. Rogers, J . Chem. Soc., 414 (1960) .

73. A. R. Katritzky and R. E. ~eavill , Rec . Trav. Chim.,

fil, 1230 ( 1964) .

74. rr (; 1 r r rr · 1 d T '11 t t · - · ,J . • .~a.1:c an • ro er, J . Chem. ooc. , 1 ( 196 2) .

75. C. A. . Coulson and M. D . Poole, Tetrahedron, 20 ,

1859 ( 1964).

76. 1! . Baker , Iv~. P . 1. Boarland and J . F .·r. rtcOmie,

J . Chem. Soc ., l/~76 (1954) .

77. G. Wit tig and L. Pohn1er, Chem. Ber., 89, 133Li. (1956J.

78. C. D. Campbell and c . .. ··ices, Chem. Corrun., 192 (1Si'~.:)) .

79. G. ittig and R . n, . ::Io:f"fma.nn, Chem. Ber· . , 95 ~

2718 ( 1962) .

80 . J . H. Beynon , G. R. Lecter and A. E . H,~illiams , J .

81 . G . • 1.

Phys. Chem. , .§J, 1861 ( 1959).

Badger and J . 1Tovotny , J . Chem.

(1961).

oc ., 3400

82. R. G. t:iller and 1:. Stiles , J . Am. Chem. Soc., fil, 179 8 ( 196 3 ) •

83 . G. Yittig , H. Eggers and P . Duffner, Ann. Chem.,

fil , 10 (1958) •

lhl .

81.J.. G. J . Fonken, Chem. and Ind. (London), 1625 ( 1963).

85 . E . Vogel , 1.r. Grimme and s . Korte, Tetrahedron

Letters , 3625 (1965) .

86 . R.F. c. Brovn and R . K. Solly, Tetrahedron Letters,

169 ( 1966) .

87 . H. A. Staab and J . Ipaktschi, Tetrahedron Letters,

583 (1966) .

88. A. A. Vansheidt , J . Gen. Chem. (u . s . S . .1.{ . ), g,

8 7 5 ( 19 34 ) • Chem. b st r . , 22 , 216 0 ( 19 3 5 ) •

89. G. M. Badger , S . D. Joland and 11.E. Spotswood ,

Aust . J . Chem., 17, 771 (1964) .

90. G. M. Badger , J . K. Donnelly and T. M. Spotswood ,

Aust . J . Chem., 17, 1138 and 1146 ( 1964) .

91. R. F . c . Brovm , J . F . N. lJcOmie and R. K. Solly ,

unpublished results .

92 . S . J . \Veininger , V. T., Mai and E . R. Thornton ,

J . m. Chem. Soc ., _§6 , 3732 (1964) .

93 . E . E . van Tamelen and S . P . Pappas, J . Am. Chem. Soc.,

.§5. , 3297 ( 1963) .

94. R. N. Viarrener and J . B. Bremner , Angev1, Chem.

Intern. Eel. , 5 , 311 ( 1966) .

95 . J. . N. ~larrener and J . B. Bremner, ~ ev. ure ~ -upl .

Chem., in the press .

96 . Th. Kauffmann , Angew , Chem. Intern. Ed. , l±,

543 ( 1965 •

' 1

142.

97. 'Y. We ith , Chem. Ber., _§ , 210 (1873) .

98. K. Nakaga18 and H. Onoue , Tetrahedron Letters ,

l ~-53 ( 1965) .

99 . Idem, Chem. Comm., 396 (1965) .

100 . C. D. Csrnpbel l and c_- ~·. Rees , ibid , 192 (1965) .

101. E . 1. Emery , Anal . Chem., ,lg , 1495 (1960 ).

102. R . Anschutz , Chem. Ber ., E , l88L1. ( 1919) .

1 03. \l. Baker , 11f. D. Ol is and T. s . Zealley , J . Chem.

Soc ., 201 (1951) .

10h . M. Schopf , Chem. Ber . , ..25, 1t136 ( 1892) .

105 . E . Strohbach , Chem. Ber . ,.Yd, 4136 (1901) ~

106 . J . H. Koch, J . Org . Chem., 26, 932 (1961) .

107. ~ . }/cNelis, J . Org . Chem., 28 , 3188 (1963) .

108. G. Koebrich and W. Drischel, Ann. Chem., 664 ,

88 ( 1963) .

10 9 . F . M. Beringer and S . J . Huang , J . Org . Chem., 29 ,

445 ( 1964 ).

110. N. P . Buu-Hoi , P . Jacqugrion and rvr . Man£ane ,

Chem. Comm., 624 ( 1966) .

111. B. J . Pur·ser , Analyst , Tu , 237 ( 19L~9) .

112. R. Criegee and G. Lohaus , Chem. Ber., .filt, 219 (1951) .

113 . O. empel and G. Schmidt , Chem. Ber., 1.2 , 2896

( 1886) .

' l

114. Supplied by Dr . G. E. Gream of the University of

Adela id.e and prepared b;r the method of

A • . • Burgsta.hler and TuJ. . o. Abdel - Rahmc_n, J . Am.

Chem. ooc., fil, 173 (1963) .

115 • T . 1( . 8 pots wood , J . Chroma t o g . , j , 101 ( 19 6 0 ) •

116. Supplied by Mr . J . B. Bremner of the Australian

Iational University and prepared from

the method of

A. Ba eyer , nn. Chem. , 1.§2 , 1L~5 ( 1892) .

117. Su pplied by Mr . J . B. Bremner and prepared by the

method of

K. Ziegler , German natent 875573 (1953) .

118. Prepared from N. Rab john, Organic Synth__es "i_s,

Collective Volurne IV, John \"'ilcy , New Yor1c ,

( 1963), P . 824.

119. Supplied by Dr . J . F . l.J. :·comie of the Universi t;

of Bristol and pre1~red by the method of 1 v. Logemann , G. Cavagna and G. Tosolini, Chem.

Ber., ..2.§ , 221+8 (1963) .

120. Prepared by the method of

N. Baker . 1". D. Olis and T. s . Zealley, J . Chem.

Soc., 201 (1951) .

121. Supplied by Dr. R . ' . C. Bro,·/n of the Australian

National University and prepared by the

method of

G • . ittig and G. Pieper , Chem. Ber., lJ., 295 (1940) .

f l

u

14)~.

122 . I . U. F •• c., Tabl§JLof_avenumbe~c for the

Cal -ibrat ion of Infra- ed __§pectrometers,

Butter--orths, Lonclon, 1961, (a) lJ . 699,

(b) p . 691+, (c) p:i;, . 690, 691, 696, 697.

123. R. A. Friedel and 1:. Orchin , UltI'aviolet Snectra

of Aromatic Compounds, John 1 'iley , New

York, 1951 .

124. 1: •• Cava, D. R. Napier and R . J . ohl,

J . Am. Chc-m. Soc . , 85, 2076 ( 1963) .

125 . A. T?.. . Badger, R . P . Buckley, ·i, . Leavitt and Tu: . Sz\-18.11 0,

J . Am. Chem. Soc ., 13., 562? (1957) .

126 . L . t . Short, J . Chem. Soc. , 206 (1952, .

145 .

of r.~ass Spect1'"}a

Peaks are listed in ctescendj_ng order of m/ e

ratio, v~·ith j_ntensitjes exnressed as a percentage of

the base pealc .

( i)

:n/ e

I

a Benzaldehyde Phenylhydrazone ( t • l • - 196)

m/e

I

m/e

I

m/e

I

197

1.6

104

6.7

78

2. 9

50

3. 3

196 195

100.:{r 2 . 2

103 98

6. o 6 . o

77 76

20 . L1. 3. 3

28

5. 6

169

1 . 1

95

3. 1

66

lL~ . 5

27

2 . ~

168

4 . 9

94

35 . 6

2A. 9

167

5. 1

92

62 . 2

7 0 , . • l)

119

3. 3

91

lC 2 •

3. 6

118

1. 3

90

2 . 7

52

( ii) Aceto12henone_RhenylhycJrazonea (r, •• _ 210)

m/e 211

I 1. 3

m/e 132

I 1 . 0

m/e 102

I 0. 8

m/e 65

I 21 .7

210 200 195 194

lOOH 1. 2 4 . 0 1. 7

119 118 117 1C7

7. 2 55 . 0 2. 2 2. 2

94 93 92 91

1 . 0 14. 5 25 . 0 10. 7

3. 7

63

3. 5

52 51

2. 3 13 . 7

193 16R

1. 2 2. 7

106 1C5

2.5 5.7

78 77

8.0 76 . 7

50

3. 0

167

~ 7 __,I • .:J

104

5 . 2

76

2 . 2

28

5. 0

105

2. 2

89

3.3

1.0

103

6. o

66

'-!-. 0

27

2. 5

(_·ii) ro :Q_j_ o nhe no Y"le Phenylhycl.J.,..,azone a ( 1 > r 22LL) • • -

m/e 225 224 223 210 209 208 207 206 1°6 ./

I 1. 5 100&- 1. 1 1. 9 5. 1 1.7 2. 6 1.5 1 . 1

m/e 195 19LL 168 167 133 132 131 130 l?

I 6. 4 1. 3 1.7 3. )_t- 6. 2 52 . 3 ) ' . 3 5. 5 ~ ? ....I .

m/e 119 118 117 116 115 112 107 106 105 I 3. 8 1.7 5. 3 0. 9 3. 4 3. 2 1. 1 7.C 6 . 8

m/e 104 103 94 93 92 91 79 78 77 I _36 . 2 6. o 3. 6 39.4 29 . 8 14. 9 1. 1 5.7 65 . 9 m/e 76 66 65 64 63 52 51 50 I 3.0 )-+ . 9 23.4 3. 6 3 . 0 2. 1 13.4 2.1

m/e 29 28 27

I 4. 3 5. 5 4 . 5

( :v) !:'.1 Benzo:Qhenone Phenylhydrazonec. ( n. '7 . - 272)

m/e 273 272 271 196 195 19~- 193 182 181 I 6. o 1001{ 6. 3 1.7 4. 3 3. 3 1.3 1.7 10 .7

m/e 180 179 178 169 168 167 166 165 164 I 73.3 2. 0 1.0 1.0 2.0 4.C 2. 3 8. 3 1 .7 m/e 163 139 1 "' 7 1 7 6 135 115 1C5 lOL~ 103 I 1. 0 1 . 3 0 .7 3.3 1 . 0 1. 0 1. 3 5. 3 2. C

r1/e 93 92 91 78 77 76 66 65 ~L~

I 6. 3 6.7 9.3 t3 . 3 7 7 • 3 2. 3 l . 3 10 . 0 1.7 m/e 63 51 50 28 27 I 1. 0 20 . 0 2. 0 65 .0 2 . 7

1

147 .

( V) Acetophenone para- l1c l,tr·1ydrazone a ( ~ . ? /""\}' • :.::

m/e 225 224 223 2~? 21C 209 2C 0 2C7 168

I 12 . 5 100-M 11 . 5 1 . ~- 1 . 0 1 t . 8 1.9 '""' 9 ~:.. . 1 . 9

m/e 167 120 119 118 117 112 111 108 7.. 07

I 5 . 8 8. 6 7. 7 ~-2 . 3 2 . 9 3. 8 2. 9 2. 9 ?.) '· . 0

r-./e ]CJ6 105 104 103 102 92 91 90 ~9

I 84 . 5 11 . 5 ~- 6 7. 7 1 . 9 2. g ~. 6 1. 4 1 . 9

:x/e 80 79 78 77 76 75 66 65 6L~

I 2. 9 2L~ . 0 l ~- - 4 63 . 4 4. 8 1 . 4 1 . 9 7. 7 1 . 9

m/e 63 53 52 51 50

I 3. 8 4 . 8 5. 8 15 . L~ 3. 8

( vj_) 8. , 224) Acetophenone meta- Tolyhydrazone t. r: ., -

rr/ e 2~5 2?L~ 223 210 209 208 207 206 182

I 11 , 8 52 . 2 8. 1 2. 2 J ~. L~ 2. 2 2. 9 1 . 5 1. 5

rn/e 181 180 168 167 166 165 122 121 120

I 0 .7 1. 5 1. 5 4 . 4 1 . 5 2. 2 1 . J 5.1 )?. • ~-

m/e 119 118 117 112 111 108 107 1( 6 105

I 7. 3 27 . 9 2. 2 7\. 7 ? . 9 ? 0 • 7 1.8 50.0 80,8

m/e 104 103 102 93 92 91 QO _,,. 89 EO

I 9. 6 6. 6 2. 2 2. 2 9. 6 17.6 2 . 9 ;:: 0 1.5 •

m/e 79 78 77 76 75 74 65 64 63

I 16 . 2 16 . 9 lCO~ 5. 9 1,

3. 7 _3 . 7 ~. 8 1 . 5 5. 9 11

Ii m/e 62 53 52 51 50

I 1. 5 3. 7 6. 6 :3 . 8 11. 8

(vi i)

1"1/e

I

m/e

I

m/e

I

m/e

I

m/e

I

m/e

I

m/e

I

m/e

I

m/e

I

,I

148.

a Acetophenone- N- ~eutero-rreta- Fluo1-1ophcnylhydrazone

( r.: • r . = 2 2 9 )

230 229 228 227 214 213 212 211 187

13 . 1 65 . 4 23 . ~- 5. 6 2. 8 1. 9 1 . 9 0. 9 1. 9

186 185 152 151 150 11i9 137 136 126

2. 8 3.7 0. 9 1. 9 2. 8 2. 8 5. 6 1. 9 l . S

125 124 123 122 121 120 119 ::i.18 117

3 .7 1. 9 2. 8 1~ 1 • r:· 6 :J • 6. 5 1c, 6 - . 57.8 ) ~- 7

114 113 112 111 110 109 ll., ~ 107 106

2. 8 8. 4 10 . 3 12. 1 Li .• 7 0 3 .:; . 11 . 2 3. 7 2. 8

105 lOu. 103 97 96 95 94 93 92

11. 2 8. 4 7. 5 1. 9 2. 8 5. 6 5. 6 2. 8 L1 .• 7

91 85 84 83 82 81 Bu 79 78

6. 5 4. 7 6. 5 21. 5 2. 8 3. 7 L~ . 7 5. 6 17. 7

77 76 75 74 73 72 71 69 ~7 100k 5. 6 5. 6 3. 7 6. 5 2 . 8 1 . 9 1 . 9 2. 8

66 65 64 63 62 60 59 58 57

4.7 4. 7 6 . 5 4. 7 7. 5 3. 7 LL . 7 2. 8 9. 3 56 55 53 52 51 50

2. 8 2. 8 2. 8 l+ . 7 24 . 3 7. 5

r,

1) ~9 .

(viii) __Q_§tophenone N,N- D1phen.Ylhyor_z.1._§ a ( ~ . 286) 1V • J -

m/e 287 286 271 209 19Li- 18C 170 169 1S8

I 7. 9 37 . 8 1. 4 1. 0 1. 0 1. 3 1 . 1 1 8 100.:\r •

m/e 167 143 141 140 139 128 119 118 116

I Li7 . 8 1 . Li- 1. 4 2 . 1 2. 1 1. 1 3 . 6 1 ~l. 4 J. . L1.

m/e 115 lOL~ 103 91 89 78 77 76 65

I 4 . 3 1.1~ 2 . 9 3. 6 1 . 1 5.7 59.2 ? . l 5. 7

m/e 6h 63

I 1. L~ 2. 9

( ix) 1 , 1 , 4 , L1-- T et r amet hyl t et r al in- ~ 3- D i one b (L . v' . - 216)

m/ e 216 173 160 146 lL1.S 128 115 91

I 100 2L~ . 5 31 . 2 10. 2 86 . 1 10.2 10. }_~ 11. 5

( x) Indanetrione C (, .·,v. 160) -m/e 160 132 105 101+ 77 76 75 74 73 I 3. 1 13. 9 3 . 1 40 . 0 6. 6 100 7. 7 13 . 5 3 . Li.

m/e 62 61 60 53 52 51 50 LL o ,7 Li.4

I 1. 2 2. 3 0. 8 1 . 5 ~. o 3. 1 71 r ,- 11 . 2 O .:)

m/e 38 37

I 16. 6 10. 1+

II (1i) Sul hob..§nzo i c Anhyc1r i deb ( r/T T r - 18)-1-) • •

m/e 184 120 l0lL 92 76 75 74 50 38 I 49 . 5 34. 9 100 13. 6 91 . 1 11. 3 19 . 9 5f . 3 13 . 4

I I

150 .

(xi.,) Octachlorobiphenylerc a ( j' 1I. Yv . = 428)

m/e 436 434 432 L~31 430 L~29 L4-28 )~ 27 1~_26

' I 5. 3 2. 0 24. 0 7. 3 60 . 6 12. 0 100:tr 10 . 7 87 . 4

m/e 425 42L~ 398 397 396 .395 394 393 ?;02 / ,I

I 3. 3 3L~. 6 1 . 3 L1 . 7 ? . O 6. 7 9.3 12. 7 l(.J . 7

m/e 391 390 389 362 361 7;60 .,/ 359 358 357

I 12 . 0 h. O Li- . 7 2. 0 1 . 3 10 . 0 3. 3 2l, . 0 L~ . 0

m/ e 356 355 354 ~2LL - '

~03 Jt _322 321 320 319

I 750 .0 2.0 14.7 1. 3 3. 3 2. 0 Li- . 7 1 . . 3 2 . 0

m/ e 288 287 286 285 284 251 250 249 216

I 7. 3 2. 0 16 . 0 1 . 3 11. 3 2. 0 1 . 3 2 . 0 7. 3

m/ e 215 214 213 212 197 . 5 196 . 5 195. 5 181 180

I 6. 7 17. 3 9. 3 2. 7 1. 3 3. 3 3. 3 2. 0 8. 0

m/ e 179 178 177 161 . 5 160 . 5 159 . 5 lLi-5 144 143

I 20 .0 23 .3 12.0 5. 3 8.7 4. 7 1.3 10 . 0 9. 3

m/e 142 125 . 5 124. 5 108 107

I 8. 7 4 . 0 4 . 0 8.7 11. 3

(xii i ) 3 , 5- Cyclohexadiene- cis - 1 , 2-Dicarbo.xylic nhydrideb

( M. ,- • = 150)

m/e 150 1L~8 122 106 105 104 80 79 78

I 40 . 5 42 . 9 13 . 6 18. 6 31 . 5 45 . 2 13. 5 100:{,t 58

m/e 77 76 74 52 51 50 L~L~ 39

I 56 44 13 lC . 8 19 . L1 34 . 5 31.9 10. 1

' I

I

151 .

(xiv) l,2,4 , 5- Tetra~ethyl - 3,5- Cyclohexa~ien.e-~-

1,2- Dicarboxylic Anhydy,j_deb (I. ~. - 206)

!::/e

T .....

206

26 . 2

135

29 . 7

120

26 . 3 10 .7

m/e 51 41 39

I 10 . 6 16. 7 19 . 0

119 1 4

looH 3- ,. :J . 4

91 77

19.8 14.?

65

lC.8

(xv)

ri/e

I

P:vridine- 3 ,4- D:carb xylic .Anh:vd.rided (. ..• - 149)

m/e

I

150

7

75

10

149

72

74

3

106

11

57

4

105 l0L•

lOOH 3

55 53

4 5

79

3

52

5

78

36

51

13

77 76

Bu. 10

50 h9

lOOM 29

'& ears off scale, 2bso1u-e intensity un.1<::nol n .

( xvi)

m/e

P;vraz ine-2-t...3.:::8 ica.rboxyl .:_c __ nhvc:-(' j de . -

I

m/e

I

m/e

I

151 150 149

2. 9 29 . 4 3.7

78 77 76

86 . 7 14. 7 4. 4

56 55 54

2. 9 4.4 ( . . C - . b :~v11; oumaranctlone

m/e

I

149

0 . 3

138

50

121

17 . 8

.. :/e 55 44 39

I 1 . 3 81 . 8 25 . 6

141

2 . 9

71

2 . 9

53

131 lC 7

2. 9 6 . 6

70 69

2 . 2 3 . 7

52 51

106 101~

1001! 4. 4

67 6c

3. 7 5 . 9

50

4. u 2 ~ . 0 l ~4 . 2 11 . 8

( 1 =. • - 148)

94

-7 r :J •

92

LL8 . 6

66

11. 2

64

20.1

l:5C)

7S

5.9

57

63

13.6

f

152.

( YViii) 3 , 4 - 3enzcoumarinb (r,t. . r. - 196)

m/e 197 196 168 139

I ll+. 7 1oo'r 28 . 4 25 . 4

( xix) N-Methylphthalimideb ( r ·~· • • • - 161)

.. n/e 162 161 160 133 132 117 105 1011 . 77

I 10 . Li. lOOtr 13 . 8 11 . 6 10.2 18.1 12.1 L~2 . 3 1-:z: ? •

m/e 76 75 66 52 50

I 45 . 3 11 . 2 2L~ . 2 lC . l 26. 7

Indicates base peak .

a Recorded by Dr . J . L . Occolovritz or an Atl:J.s CE4

mass spectrometer . All io:1.r- hs.vin'~ a r·reater

abundance than 1% of tho.t of the base peak are

recorded.

b Recorded by Dr . J.H . Beynon. All ions of freater

abundar:ce than 1o;0 of the base peak are recorded.

c Recorded by Dr . J . S . Shannon on an Atlas CH4 mass

spectrometer. Ions of ~reater abundance than 1%

of that of the base peak are recorded.

d 2ecordecl by Dr . C.S. 5arnes on a e"('~<.:in-Elmer

Hitachi RMR - 6D mass st~ctrometer . Ions of greater

abuncl&nce than 1 ~ of the buse pea1< are recorded.