MSU

LIBRARIES
m

 

 

RETURNING MATERIALS:
P1ace in book drop to
remove this checkout from
your record. FINES w111
be charged if book is
returned after the date
stamped be1ow.

 

 

 

 

 

 

PART 1: SYMPHE‘SIS 0F STABLE FREE RIDICAIS AS POTENTIAL
ORGANIC METALS. PART 2: smmmls 0F OXYGENATED
INDACENE DERIVATIVES. PART 3: REACTIONS OF AROMATIC
P0LY(II,II-DI:R:"'RTL Rim-I5) WITH ELICTROPRILRS.

By

Mark Wayne Armstrong

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1982

ABSTRACT
PART 1: SYNTHESIS or STABLE FREE RADIOALS AS POTENTIAL
ORGANIC METALS. PART 2: SYNTHESIS OF OXYGENATED INDACENE
DERIVATIVES. PART 3: REACTIONS OF AROMATIC
POLY(N,N-DIMETHYLAM1DES) WITH ELECTROPHILES.
By

Mark wayne Armstrong

Part 1: "Organic metals" are organic compounds which
conduct electricity. Those which have attracted the most
attention are charge transfer salts, like TONQ-TTF, in which
both the donor and acceptor molecules are highly delocalized,
planar molecules. Theoretical studies suggest that neutrality
and divalency are prOperties which could be desirable to the
design of more effective conductors, perhaps even allowing

the development of superconductors.

To this end, several syntheses of diradicals deemed
likely to be stable were attempted. These included the
diradicals of 2,12-dihydro-6,10edihydrorywhfl,8H~dibenzo
(cd,mn)pyren-2-one (fig), the h,8,12-trioxa derivative of 5a_
(52), b,7-dihydroxy-1H-phenalen~1«one (g), and several
polychloro derivatives of these. The dihydroxy compounds
which were successfully prepared (g, 29) did not give stable
diradicals.

Part 2: 1,2,3,5,6,7-Hexahydro-swindacenem1,2,3,S,6,7-
hexane (§§), its b.8-dihydrory derivative (51), and the
quincne of fig were of interest as electron acceptors in

potentially conducting charge transfer salts, as carbon

(p
K.)
:3
CL
O
X
(7
(3
{-13
H
Cf
(‘
O ‘f
'. J
U.‘
Q.)
:5
CL

oxo-acid' as precursors to the interesting
polycycl'c aromatic dicyclOpentadieno(a,h}—s-indacene.
Though the synthesis of none of these compounds was achievel,
several interesting pathways were orplcred.

Part 3: Partly as a result of several reactions
explored in part 2, it was of interest to study the reactions
of aromatic N,N-dimethylamides with electrOphiles. Of
particular interest were compounds having two N,N-dimethylamino
«carbonyl groups situated ortho to one another on a benzene
ring. These were studied to determine whether the groups
react independently of each other and, if not, to what extent
neighboring group participation occurs. Though other
electrophiles were investigated, the major portion of the
study was limited to reactions of methyl trifluoromethanesul-
fonate (MeOTf) and thionyl chloride (SOClZ) because these
give straightforward chemical and spectral results. As models,
the reactions of these electrOphiles with N,N»dimethylbenzamide
(1E1), N,N,N',N'-tetramethyl iSOphthalamide (lgg), R,N,N',N'-
tetramethyl terephthalamide (121) were also studied. Both
lfié and 151 reacted with two equival'nts of MeOTf and SOC12
to give bis(methoxy dimethyliminium) salts and ois(chloro
dimethyliminium) salts respectively. N,N,N',N'-dimethyl
phthalamide (15g) reacts with only one equivalent. NMR Spec-
trOSCOpy shows that the product of lih with.MeOTf displays a
small participation by the second amide group, while with 80012

this participation is complete, or nearly so.

I hereby dedicate this work to my parents, Norman and
Florence, and to my wife, Ignn, and to my future child, who

is as yet unnamed.

ii

ACKNOWLEDGMEFTS

I am delighted to acknowledge the contributions made by
the following peeple to the completion of this work. First,
I thank Professor Eugene LeGoff and the members of his
research group for the many discussions and suggestions
offered during my years at Michigan State. I am especially
grateful to my wife, Lynn, without whom this work, literally,
would never have been finished. She served as typist and
editor of the many drafts of this dissertation, no easy task.
More importantly, she was my motivator, prodding, cajoling

and cheering to the very end.

iii

KL.” 0111.) UL‘ [U URI infiiixiiu

Synthesis of Stable free Radicals
as Potential Organic Metals

H
0?

Part

Introduction
Results and Discussion
Experimental

Part I; Synthesis of Ox,gennted Indacene
Derivatives

Introduction
Results and Discussion
Experimental

PART III: Reactions of Aromatic Poly(N,fi-

dimethyl amides) with ElectroPhiles

Introduction
Results and Discussion
Experimental

List of References

iv

12? £39

{‘3

16

128
179
192

Table

Table

Table

Table

Table

Table

Table

Table

Table

1.
2.
3.
h.

L131 OF TABLES

Conductivity of Selected Materials
PNR Chemical Shifts
Halcgcnation of 21 and derivatives

Reactions of 155 with electrOphiles
as monitored by pmr spectroscopy.

Methyl pmr resonances in N,N-dimethyl
amides and O-metnylated N,N-dimethyl
amides.

Chemical shifts predicted for possible
products of methylation of 19g,

130 chemical shifts of the carbonyl
carbons cf N,N-dimethyl amides and
O-methylated N,N-dimethyl amides.

Methyl 13C-chemical shifts for amides
and O-methylated amides(O-methyl
chemical shifts).

1 . . .
3C chemical shifts predicted for
possible products of methylation of 1SQ.

page 3

—8
*1
O
I
g
cm

9.,»
[‘0

130-133
116

1146

1h?

1&8

1h9

Figure

Figure

F i gure

Figure

Figure
Fi gure
Figure

Figure

Figure

Figure

Figure

Figure

F i gure

Figure
Figure

Figure

1.

I":

(a

3.

8.

9.
10.

11.

15.

1s.
1].

18.

]T ‘vf

1‘ .LJUEiL‘JU

iilb “1' OF
Formation of TCNQ-TT‘

TMTSF and the crystal
of TCNg—TTF.

TCNQ,TTF,
structure

of Hlec ron tianspo t
types of or roanic

Eviecl’zan ismo
for “ever
metals.

Increasing the intermOJecular
transfer integral.

J“.
p.

Preposed synthesis of diradical

Determination of is mer 11.
Preposed =ynth esis of diradical “b.
Jun-—

The synthesis of 2 as per hurtin
. (_ ' ‘-
and Smith. '

‘

Attempts c of 21.

Grignard

synthesis
Mechanism of decarboxylation of 2~.
Preparation of ti iaryl me. thane 23.

n uq‘l \ ‘ X". aj‘ljl‘rd‘l'one “'0. i0 1 find

(if "filid. J). "9.).

Syn“:usis of L,.’—dibvuroxy p: icnalone
(in) and derivatives.

Preposed synthesis of 2,),3- trichlorc-
I ,"f-di hydroxy-1H- ~phcnale n-1 one \hh,

via lg.

Chlorination of 12 with chlorine in
benzene.

Chlorination of }§_with hypocilorite
Reaction of £2 with aluminum trichloride.
Inciacene and c c ri va ti ves

vi

1O

.-. .. .. . . \
List 01 Figures (ccnt'dq

77

78

Figure 19. Gain or loss of an electron from
radical anions :2 and 60.
Figure 20. Possible sbltnesis of aromatics
61 and as from hexaketone V6.
——- nun—I— é..-
Figure 21. Oxccarbons and carbon cxc-acids.
Figure 22. gynthesis of hexakefgne 2é_by
dlciter and Schanz."
Figure 23. First approach to hexaketcne 29.
Figure 2b. Second approach to nexakctcnc 36.
Figure 25. Formal Inlechaniam of hmurerer
rearrangement.
Fig‘re 26. Proposed synthesis of 36 from [2
Figure 27. Attempted synthesis of §§_as a
modcl for 6?.
Figure 28. Generalized scheme for preparation
of hydrindacene derivatives by
cyciization of benzene bis—proyionic
acid derivatives.
Figure 29. Preparation of p-phenylene diproPionic
acid derivatives.
Figure 30. Preparation of m-phenylene propionic
acid derivatives.
Figure 31. Attempted cyclizaticns of benzene bis-
pr0pionic acid derivatives.
Figure 32. Failure of dimethoxy-p-phenylenedipro- r

picnic acid to undergo double cyclization.)

Figure 13. PrOposed cyclization of 29_via the
C-methylated salt.

Figure 3b. Reaction of 29 wit methyl triflate.
Figure 35. Scheme for the synthesis of §§_by

introduction of a carbonyl equivalent
into a tetrasubstituted durene.

vii

Figure

Figure

Figure

Figure

F 1 {3U 1"“

Figure

Figure

V

Figure

Figure

Figure

Figure

Figure
Figure
Figure

Fibzre

“igure

Figure

37.

at.

hY.

“8 0
1,9.
50.
51.
52.
S3.

Oi.
118 to dikctone 115 oya

List of Figures (cont'd)

Preparatiox- of 1,254,; L+ tra‘nis
tbrcmomethyl; ccnzene.

a ‘rBtraszMuide
malonic

Proposed convczsion
ester synthesis.

Approach to oxygenated system C”,

P013(cxymeihgl)deiivafiivos of
hydrccuinone .

.rep Tat ion or totrakistchloromsthyl)
1511:0quinoze diacetrfi e fill.

DOIgle Dials Alder epproa ch to the
oxygenated hydrindacsne sttcm

Proposed cycllzaticn of bistpropicn~
amide) to the hydrindacenc sncletcn
in the dimethoxy system.

Preparation of leUthAJhi Llide 1§8.
Attempted

y<31iza tiozx. of diamid

s _}Q.

Cyclization of imiuium chlorides
nitrilium ions.

timough

.34 7': "N‘ “ 1-‘1. .— . ‘ : . .‘01 ‘ if:
:icpostu cyciizltlon of dinltilifl 13‘
through she nitrilium salt.

Electr ophilic subs titutio on of a
nitriliun salt on an c.l Ht
deficient “roastic.

Y‘O}

'_

Preparation of dinitrilo JAE.
Benzene dipropionic ac: d and dexitati Ves.
pciy(.il, N— dimethyl)umidcs.

Aromatic

Possible products of methylaticn of
N,N,N',N'—tetramethylphthalamide 1fih.

Possible products of reaction of 18h
with chlorine electrOphiles.

Possible products oi reaction of salts
of 35g with aromatics.

viii

93

95'
96

97

Figure

Figure

Figure

Figure T

Figure

Figure

Figure
Figure
Figure
Figure

Figure

Figure

Figure
Figure
Figure

Figure

Figure

Figure

Figure

58.
S9-

60.

65.

700

71.
72.

List of Fig res (cont'd.)

Amide 135 as an entry into tn: polykcto~
hydrindaoeue system.

Preparation of 1,2,h,5~tctrakis(fi,fi-
dimethylaminocarbonyl)uenzene, 1%5.

Preparation of ester: from amides by
alkylation/hydrolysis.

Reaction of N,N~dimethylbenzamide
with methyl triflate.

Reaction of lfié with methyl triflate.

Reaction of tetrakis amide 155 with
methyl triflate followed by hydrolysis.

Dimethylation of iii: Mechanism 1.
Dimethylation of 155; Mechanism 2.
Dimethylaticn of 1&2; Mechanism h.
Reaction of 1§g_with methyl triflate.

Proposed "hybrid" structure of

methylated 15“.

Reaction of acetonitrile-d3 with
methyl triflate.

Reaction of 131 with thionyl chloride
Formation of biswamidine 16“.
Reaction of 15h with thionyl chloride.

Reactions of thionyl chloride product,

Possible structures of the product
of 16S and p-chloroaniline.

11H~dibenz(b,e)azepin-11-one.

Reactions of yellow 167.

ix

135

136

137
138

1&1
1112
1&3
1th
150

151

153
15h
157
1S9

160

162
16h

Figure

\

.L‘ :9. (3.11130

Figure

Figure

81.

List of Figures (cont'd.)

Reaction of 166 with 2,h-dinitrophcnyl»
hydrasone.

Reaction of 1&1 with 2,h~dinitrophonyl~
hydrazonc.

Reaction of 16f with methyl triflate.

Reaction of phthaloyl chloride with
p-chloroaniiine.

Possible structures of 165.
Conformation of ring closure of 16;.

Comparison of the extent of amide
group participation in the chloro
dimetnyliminium ion and in the
methoxy dimetnylimin um ion.

1
N,N,N‘,N'~tctrameth l pnthalamide or

Formation of N-meth .rnthalimide from
N,N-dimethyl phthalamoyl chloride.

Y
V.
J

Suggested mechanism of cleavage of per-H—alkyl
phthalamide or phthaiamoy]
chloride.

Formation of 2,h,6-triphenyl triazine
from N,N—dimethy1henzanide.

Preposed mechanism for the formation
of 2,h,6~triphenyl triazine from N,N-
dimethylbenzamide.

165

170
171

172

175

Part I

SYNTHESIS OF STABLE FREE RADICAIS
AS POTENTIAL ORGANIC METALS

hf.

The vast majority of org'nic compounds do not conduct
. .~ -. .. , . _- ”.12 -10 3.1 -
electIiCity to any appreciable extent (10 -10 ohm cm ).
The very few with considerably higher conductivities have been
given the Optimistic if not entirely accurate description of
"organic metals".

The word "metal" is generally reserved for elements lying

to the left of boron, silicon, germanium, antimony and polonium
1

in the periodic table. In this sense the term "organic metal"

is a misnomer as no metal atoms are present. ln contrast, the

Class of compounds known as "organometallic" does involve metal

atoms, specifically those bonded to a carbon atom in an organic

framework.

Alternatively, a metal may be considered to be any substance
possessing metallic characteristics, i.e., high electrical con-
ductivity, thermal conductivity, luster and ductility or malleab-
. . 1 . . . .
ility. In this sense the term "organic metal" is at least
partially accurate.

The first organic materials found to have considerable
conductivities were radical ion salts of tetracyanoquinodimethane,

T - - - , 2.3 m - 1
TChQ, with various organic electron donors. lhe most studiee

example of this class of compound is the charge transfer salt of

TCNQ with the electron donor tetrathlafulvalene, TTF.

f"
--l ‘l.-..~._. _.._ .M
-c

+
> :K

i

t
q7““"‘“‘
l, "—4/tL. _-/g
i? ‘ -- —_‘Y2
,/ cl
l m>

 

t‘

(”-
1

L.

Figure 1. Formation of TCNQ-TTF.

n -
This complex has a room temperature conductivity of 10‘ohm

1

cm“ , comparable to graphite. For a comparison of the con-

ductivity of TCNQuTTF with that of other materials see Table 1.

The preperties of the salt have been detailed by Engler.

(J

Table 1. Conductivity of Selected Materials)

 

Material (ohm;lcm“1)
metals 101*.106
TCNQrTTF 102

carbon 102

molten organic salts

silicon 10—h
alkali TCNQ salts 10~h

most molecular crystals 10-1h-10’10
ferrocene 10-13
sulfur 10-1h
quartz 10-17

anthracene 10"17

A

Since the disc very of TTF-TCNd, several variations

have been made in both the donor and acceptor molecules.
. 6 . p} _ . ,

Changes in the donor molecule have involved replaClng the
hydrogcns in TTF with methyls, polyucthylene bridges or benzo
rings, and/or replacing the sulfur atoms with selenium. The
room temperature conductivities of the compounds with TCNQ

-5 -1 -1 .3 -1 -1 +
range from 10 ohm cm to 2110 ohm cm for she tetra-

. M6 . .
seleno ana10g of dlcyCIOpentenchlr. Changes in the acceptor
has involved adding substituents to the ring of TCNQ,7
8

extending the ring system of TCNQ and/or substituting

. . i,i9 4p 11 . -
heteroatoms into the ring system of Tohu 7-49 . Uniortun-
ately, those of the above which form charge transfer salts
with TTF have room temperature conductivities less than that
c” TCNQ-TT“.

There are two further preperties of TCNQ—TTF and similar
compounds which bear mention. First, single crystals of
these complexes carry an electric current much more readily
along one axis of the crystal than along any other axis.

They are thus anisotrOpic or more specifically, one-dimensional
10 . . . . .
metals. Secondly, the conductiVlty of TCNQmTTF increases
. . . . r 0 a
with decreaSing temperature, reaching a maXimum at )8 K oi

over 1Ol‘ohmfl1

-1
cm , then falling off sharply at still lower
12
temperatures. This sudden drop-off in conductivity is
1
known as the Peierls transition. 3 It is thought that this

transition from conductor to semi—conductor or insulator is

U1

1h

due to distortions in the crystal to lower symmetry.

I
‘ , ‘ .a.‘ ‘1 “1
The marimum value 01 10 ohm cm , approaches the con-

ductivities of certain metals (tin, lead) with conductivities

15

of Sx1Ou—1x105. This sharp maximum in conductivity at low
temperature is reminiscent of the superconductors, metals
which in fact lose all resistance at very low temperatures,

15

usually within a few degrees of absolute zero.

Several recently prepared derivatives have proven fiva
more successful and have shed new light on the design of
organic metals. These were a series16 of organic cation
radical salts (ansr)qx, where TMTSF is tetramethyltetra-

C.
selenafulvalene (see Figure 2), and X is a symmetrical octa-
163 16b,c 16d
or

hedral anion16 PFg , Ang

tetrahedral anion17 BF; and £10

16d
, Sng and TaFg
’4.

superconductivity at moderate hydrostatic pressures of 1.2

The former exhibit

GPa in the O.h-1.S°K region.16 In these materials high
pressure suppresses a metal-toninsulator transition occurring

‘6 In the tetrahedral

between 10 and 200K at ambient pressure.
anion series, (TMTSF)2BFh and (TMTSF)2ReOh are insulating
below h1 and 1820K respectively, while (TMTSF)ZC1Oh remains
metallic down to 1.3-1.50K, wnere a transition into a
superconducting state occurs.1

These substances consist of two donor molecules per

anion. Thus, formally, only half of the TMTSF molecules are

charged. The structure of the crystal is as might be expected.
The planar TMTSF molecules are stacked like pancakes. These
stacks are interspersed with lines of anions.1

X-ray crystallography of TCNQ-TTF shows separate stacks
of TCNQ radical anions 'ld TTF radical cations, with the
planar ions in each stack arranged parallel to one another,
though not perpendicular to the stack.18 The distance between
molecules in the TCNQ stacks is 3.17s. These very short
interplanar distances allow overlap of the pi-clouds and
delocalization of the unpaired electrons along the axis of
the stack (see Figure 2). Combining conductivity measure-
ments along the various axes of a single crystal with X-ray
data shows that the direction of conductivity is parallel to

the stacks of TTF and TCNQ ions.

For both practical and theoretical reasons, it is
. _ , 20
clearly of interest to deveiOp new organic metals.
Practical examples include the development of materials which
may be superconductors at higher temperatures than are now
available. Theoretically, the investigation of such pheno-

mena as the Peierls transition could be aided by the avail—

ability of organic metals with higher conductivities.

There have been several papers which detail factors

important in the design of an organic metal as determined by

17,19,20,21

experimental and theoretical studies. Among these

factors are: (1) An unpaired electron.19 Clearly this is

7

necessary to allow the motion of individual electrons between
molecules. Paired electrons are held in melecular orbitals. This

is, of course, wh; most organic compounds are insulators.

(2) Uniform crystal structure. The conductign band arises

 

L) g) Q 3 Qe

L/t a
N L) Ul‘:

TONez WTF TNT?

a
.f‘x . 1
' wF' \ flu...
. / \‘ ,1‘ Int——
(J1() "R- /// l; 'l ‘
- ~ \ 2’ \\
\‘7‘ “ in //\\,_..

:/‘

la
3“

9,
fl
& .
. y)
is
_/

',
f2
//75

. f“
I

   

\.
M
0:73 .
1' ,5 " .c .'
j {J x 'I

x
/

Figure 2. TCNQ, TTF, TMTSF and the crystal structure of TCNQPTTF.

8

from a linear combination of the highest OCCincd molecuTar

10
orbitals (or conduction orbitals) on each molecule. ’ Crystal

defects (as with Peierls distortions} lower symmetry and
separate the valence hand from the conduction band. The
presence of counterjons may contribute to Peierls distortions.

or at least reduce the symmetry of the crystal. Garito,

- . 12 21. n
Heeger and co-workers nave suggested, ’ in lact, that only

one of the two kinds of chains in TTF—TCHQ actually carries

a current while the other simply supplies a framework. This

is the case in the superconductors (THTSF),W, and the stack
at

d

which carries the current is the TTF-like donor, TMTSF,

rather than a TCHQ—like acceptor as was earlier thought

6 .i . . . .
likely. Clearly ll the counterion could be eliminated,
symmetry could be increased and, perhaps, defects and dis—
. _ . >\ - fl V f - l - _ ‘ . ,_ A . , .
tortions decreaSed. \3) weak ULQLLFCHHCiQCtIOH repulSJcn or

low Ionic fluctuation enrrgy. This is the en ray requirel

‘ rl ‘ 7" I , I r ' r'- V, I. e V} « 7 4 .-~ ‘ J~“ _
to transler an electser lzom one molecule to the next, 3

\

"' . ‘ : " ‘ ' " ' ‘ ‘ "] "‘ ‘p i'r“:r'u ‘ rt
LilC‘Wll .1.“ .C‘ijtu. C: '3 L 0! 1.0 V13 [‘(1 L. tail/)1: g) (a J vfilfkgcuiy.” fi'gtax D .

J)

The .econd type shoan in Figure 3, representing incomplete

electron transfer from donor to acceptor, would appear to
have the lowest ionic fluctuation enerfi . It has been
suggested22 that complete electron transfer is not only
unnecessary, but may actually be detrimental to the stability

of the metallic state in organic :harge transfer salts. In

fact, those salts with highest conductivity contain the

. .
T’J.‘i"—VL‘CIJ type; xi" + .a' n + A" ‘
' 1 -——--9 o+Z
* 5 i’l '1' l;
incomplwto ,1 _ . ,5
charse ” + A A + ”

‘ u" . “.1 ——-9 .4. .
transier Q + h h + B
TTF—TCNo type:
neutral .
mono radical: C' + C‘ ——4 Cl + C?
type
neutral
diradic l: '0' + 'h’ ——4 D; +°D?
type

Figure 3. Mechanism of electron transport for several types
of organic metals.

3

2-

N

species in mixed valence states. This is the case for the
(TMTSF)2D superconductors in which the valence of TMTSF is
formally +i. Thus careful matching of donor and acceptor
based on the difference in their oxidation and reduction
potentials, respectively, could serve to maximise conduct-
ivity. This may eXplain why some donors only form conducting
. . 23 . . _
salts With particular acceptors. (h) Maximum intermolecular
transfer integra. The greater the overlap between pi-clouds
of adjacent molecules in an organic conductor, the greater
the conductivity. This can be achieved in two ways. First.

simply decrease the intermolecular distance (Figure ha). As

10

the molecules in a given stack of a radical ion salt such as

TTF—TCNQ have the same charge, removing the charge would re—

duce coulombic repulsion and thus decrease intermolecular

distance. Secondl as mentioned earlier the lanes of the

Y: r P
TTT and TCNQ radical ions are not perpendicular to the axes
of the stacks. Rather they are offset 'r not directly "above"
2b

one another. It is thought that this offset may be due to

coulombic attraction to the counterions in adjacent stacks as
25

well as coulombic repulsion between molecules within stacks.

Clearly, eliminating the charges could reduce the intermole—
cular offset and thus increase overlap (Figure hb).

Based on the above, Haddon19 prOposed that stable neutral
radicals, specifically odd-alternant hydrocarbons, could serve
a.

->«-—‘..— ----—. u —-.o--

-.--v-—_-..- ‘ no . an.“

 

n...—.—..——-~.4_—-. i. .. ,.

 

 

 

—-..._..._... .... ,_ -.
.a --.. .-__M-—*M

Figure h. Increasing the intermolecular transfer integral

as models for new organic conductors. Based on calculations
(ammo/3 see so) of geometry, bond order and charge densities,
he found the phenalenyl (ply) system to have many of the
characteristics (outlined above) considered important in the

design of organic metals.

ply
Specifically, it has a delocalized odd electron, no charge,
and a planar symmetrical structure. Furthermore, the redox
orbital in which the odd electron resides is nonvbonding.
Calculations show that addition or removal of an electron
changes the geometry of the system very little and the energy
requirement is equally small. Unfortunately, the phenalenyl
radical, though stable in solution undergoes dimerization on

26
attempts to isolate the solid radical.

1
Haddon 9 also mentions the triphenyl methyl system in
passing as an odd-alternant hydrocarbon, but dismissed it out

of hand due to its non-planar nature. He apparently did not

consider a bridged triphenylmethyl such as sesquixanthydryl,

1, The corresponding carbonivm ion, 2, was prepared in 1963

/L‘w’i;\ L‘ / ' E‘ ,/5
O , ‘v l / , / / ' ‘\r .
1; ‘4 I) ~ “I , Y ’K)//

,«\ ‘ ,k . K ‘ Y }/
:/' \\\/ "J" \‘\1 \\ ’ ‘..,/ \ / ‘ ‘ // .“ J ‘.
1 i H i I . t I I! \ a
\ / ,"( ~ _ 7) K\ / K a ”I E , ,' !. ‘E

b cl' , ‘lJ"
2
.1 Z. 3.

by Martin and Smith.27 Chromous ion reduction of g_gave a
dimer, probably 3, presumably through the radical 1,28

Beating the dimer in xylene to 15000 causes dissociation and
dissolution enough to allow detection of the radical by ESR.29
The ESE spectrum of l_consists of a quartet (J=3.17 gauss) of

septets (J=O.89 gauss), arising from the three identical para

protons and the six identical meta protons.

Other factors suggested by Garito and Heeger20 to stabil-
ize the metallic state in organic conductors include nominal
divalency to promote mixed valence states, and heteroatoms at
points of high spin density in order to stabilize the radical. 1
Thus it was our intention, and the object of my research to

extend the idea of planar stable free radicals to include

1-3

diradicals like g and 5,

a:X=CH2
b 3X20
P
E
l l
U r \‘V /
§\T”l‘<°l 3' I. X
' l
'L\\\l,r*£:\\r, (J' 1/” | // i; 1 ‘§?\
Id ‘a [ H 1I ;?L\
[\\ /JL\\,{3’ . O ’/ C\ .\‘ /(/fi\\¢( 0.

These diradicals include the features of the phenalemfl/
bridged triarylmethyl systems (planar, symmetrical, delocal—
ized, neutral) and add the features of divalency and hetero-
atoms at the spin-rich periphery.

That cross-conjugated diradicals of this type can be

stable is illustrated by the synthesis in 1960 ofԤ.29

 

H;

The triplet ground state for g has been studied by ESR
spectrOSCOpy,30 and justified by detailed molecular orbital
calculations.31 A similar diradical, digalvinoxyl, 1, was
studied by Chandross; and also shown to be a triplet.
Interestingly both Q and 1_are highly colored with a metallic
luster. However, neither is reported to be conducting.

It may well be that much of the stability of é_and 1.13
due to the steric efforts of the bulky t-butyl groups. The
same steric effect would be expected to keep the molecules
far enough apart in the crystalline state to prevent overlap
of the pi-clouds and thus preclude conductivity. In addition
it was thought that the rings in §_and 1 were not c0planar,
but rather prepeller-like, thus preventing complete delocal-
ization. It was our hope that the planar molecules g_and‘i
would stabilize the diradical enough that the bulky t-butyl

groups would be unnecessary.

 

 

H fig c We
/ -M.el---_, L/ i __1_3ff 2 }
\ OH ”3“ \. Me “314 \/()Me
§ 2 Br _1__(_)_
1’)
AgOUCUFY
OMG
I
/
\xy/[L\/Dfik%
UMH or 12
4J\\
I u _‘ (3M8
Met) \\ \\/lhue 1)Mg, ether L\
j W %2)oc(ost) / i
/ l' ” \/ ' 2 \ 0M»
1
/\\\/"’ ‘\ . .
”PU "’ UN 3
K ‘L I 11
1 3 cm *-

H0

   

Figure 5. Proposed synthesis of diradical 5a.

16

RESULTS AND DISCUSSION

The proposed synthesis of diradical ;a_is given in

The methylation of 3~hydroxybenzyl alcohol is straight-
forward and occurs in quantitative yield with sodium hydride
and methyl iodide.

The halogenation step is interesting in that the product
could be one of two isomers, _1_or lg, ignoring the isomer in
which the halogen atom is substituted between the other two
substituents. Bromination (BrZ/bClh) gives a mixture of
isomers, including h-bromo-B—methoxymethylanisole (l_).
Iodination (using 12 and AgOOCCFz) appears to give a single
isomer as determined by nnr and TLC. Normal instrumental
methods of structure elucidation all tail to distinguish
unequivocally between the possible isomers. The pmr spectra,
though clearly resolved cannot differentiate between isomers
(see Table 2). The structure cf 11 was finally determined by
the following sequence of reactions.

OMe ONO OI'Ie

Figure 6. Determination of isomer 11,

17

The structure of lactone lé was proven by mass spec.,

nmr and IR.

Thus the monoiodo compound must be 11, as 1;

would not give a balactone in the above reaction sequence.

As the nmr spectrum of the isolated monobromo isomer is almost

identical to that of.ll, it is assumed that it is 19.

 

 

Table 2. PNR Chemical Shifts (ppm downfield from TMS)
cs-ocn- cameos,
d 5 c J
. ' 4
110 fifi; ‘Vb X ‘// I”La
~~ \\
11 ULII, ,, //”‘\ .‘
b ' “b j OcH3
i. H
C
X=H X=Br X=I X=COOH
“CH2“ 14025 14030 1‘02? “079
BzOCH3 3.23 3-3h 3.33 3.uh
191100113 3.62 3.63 3.63 3.79
Ha 6090 6.80 7.10
(J=-3Hz) (J=3Hz) (J=3Hz)
Eb 7016 7036 7'93
(J=8Hz) (J=8Hz) (J=BHz)
Hc 6.15 6.30 6.70
(J=3Hz,8Hz) (J=3Hz,8Hz) (J=3Hz,8Hz)

 

18

Table 2 (cont'd.)

Ht Ha

 

 

 

-—_——
.‘o—W
'- -.—.....——— —_

 

1 r \
ch(”o) ”10(“nfl
3.11:3,4 values
JO: 6-10 Hz
Jm: h-B Hz
J : 0-1 Hz
P

Treatment of iodide 11 with magnesium followed by di-

ethyl carbonate did not give the desired triaryl methanol, 13.

Instead only ethyl h-methoxy~3~methoxymethylbenzoate, 11.and

the diether, 2 were isolated, along with some starting

material.

[\llb'z

 

./
7/ H 1)Eg ‘
Lar/K\/oc15 2)(Et)cCo
(.
I 3)H23;

+

no

An approach to molecule Eb is shown in Figure 7.

 

Li

 

lg

ClCOOEt
or

2}“ W... 3'
(itL)Z(o

Naifii

 

 

Figure 7. Preposed synthesis of diradical 53,

20
Martin and Smith27 have prepared the bridged triaryl
carbonium ion g,by the reaction of the lithium salt of
resorcinol dimethyl ether with diethyl carbonate to give the

triaryl methanol, followed by heating with pyridine hydro-

chloride. It appears, therefore, that the first step of the

\ ’/€ \ , - ether .% // -
%- OCH; CH, O I OCH5
CI'LI) / Ll
1)(ri,;)jfit
3) Tzr L

 

 

27

Figure 8. The synthesis of §_as per Martin and Smith.

sequence shown in Figure 8 should be straightforward. Upon
addition of an ether solution of butyllithium to an ether
solution of 1,3,S-trimethoxy benzene, a light yellow color
results and darkens slowly over a period of time. No precip-
itate is observed. Upon addition of diethyl carbonate in
benzene, the color changes very rapidly to a dark yellow and
eventually turns muddy green. The tarry substance obtained

is not soluble in water and only slightly soluble in ether.

It is, however, soluble in dilute acid, turning a brilliant
deep blue color. The blue material Was shown to be the tri—
aryl carbonium ion gl_with the counterion X- depending on the
acid used in the work-up. When dilute hydrochloric acid is
used the counterion X- is chloride. The blue triaryl methyl
carbonium ion g; is analogous and quite Similar in its
pr0perties to the hexamethoxy triaryl methyl carbonium ion

g5 of Martin and Smith27 (Figure 8).

The blue material can be extracted into chloroform or
methylene chloride but is insoluble in benzene. The yield
of blue material is only about two percent in this reaction.
However, if ethyl chloroformate is used instead of diethyl

carbonate yields up to 80% can be obtained.

Reaction of the aryllithium 12_with methyl 2,h,6-tri-
methoxy benzoate gives the blue material g1 but in poor yield.
This ester was prepared from the corresponding acid using
sodium carbonate and methyl iodide rather than via the acid
chloride, because 2,h,6-trimethoxybenzoic acid readily de-

carboxylates under acid conditions, as discussed below.

Upon treatment of the blue carbonium ion g; with hydroxide,
it becomes muddy yellow and can be extracted into cepious
amounts of ether. It is assumed that this material is the

triaryl methanol 29,

N
P\. 1‘

Using a Grignard reagent instead of the organolithium as
outlined above is unsatisfactory for several reasons. First,
preparation of the organometallic reagent requires two steps
in the case of the Grignard and only one in that of the
organolithium. Second, monobromination of trimethoxybenzene
is no simple task as di- and tri-bromination tend to occur
quite readily. Third, formation of the Crignard is sluggish.
Finally, the only products isolable from the reaction of the
Grignard reagent with ethyl chioroformate are ethyl 2,h,6-
trimethoxy benzoate, g1, and starting material 19,

The same products are obtained on heating 2,h,6-trimeth-
oxybenzoic acid gt, with thionyl chloride followed by evapo-
ration and refluxing in ethanol. The production of 1,3,5-
trimethoxybenzene in this reaction is apparently due to acid
catalyzed decarboxylation in the thionyl chloride step. The
same process occurs in phOSphorus oxychloride. The mechan-
ism of this decarboxylation may involve catalysis by traces
of proton present in the reaction mixture, aided by the

unusual stability of the trimcthoxybenzenium ion.

A second approach to the triaryl methyl system (21)
involves the phosphorus oxychloride induced condensation of

2,h,6-trimethoxy benzene_l§ as shown in Figure 7.

When 2,h,6-trimethoxybenzoic acid is heated with two

equivalents of 1,3,S-trimethoxybenzene in phosphorus oxy-

"0
\JJ

chloride as solvent, the reaction mixture became deeply
colored. werk-up involves carefully diluting the reaction
mixture with ice, washing with benzene, and extracting the
deep blue product into chloroform. Yields range from 9% to
20% and seem to depend strongly on the initial ratio of
starting materials used. The only other organic material
isolated from the reaction mixture is 1,3,S-trimethoxybenzene,
and its quantity accounts for the portion of bgth_starting
materials not going to product. Hence, 2,h,6-trimethoxy-
benzoic acid apparently decarboxylates under the reaction
conditions as discussed earlier for reaction with thionyl
chloride. The best yield (20%) was obtained when a ratio of

h:1 trimethoxybenzene:trimethoxybensoic acid was used.

   

(.2 )\:1CO(’}‘L\ +
”' 1 '.
3) L, f-
(f.
’
n
K“ . .// 15

Figure 9. Attempted Grigmird synthesis of 21.

 

(m. (2) V (H‘i'

 

11 j +
K 3/ \on \ d 1
i ---* * *
rm / . 002
CILO / mm my; OCH,
:3 J 1' 2

Figure 10. Mechanism of decarboxylation of 3.5.4.-

M
\"l

The product is a dee~ blue-black solid which decomposes
at 110°C. The NMR is as expected and the mass spectrum shows
a large parent ion at m/e:513, corresponding to the tris(tri-

methoxyphenyl)methyl carbonium ion.

Reduction of the blue triaryl carbonium ion ngwith
lithium aluminum hydride gives the expected triarylmethane gg,
This material can also be prepared by adding formic acid to a
hot solution of 1,3,S-trimethoxybenzene in phosphorus
oxychloride. Large amounts of formic acid are needed as most
of it decomposes in phosphorus oxychloride to hydrochloric
acid and carbon monoxide. Treatment of 1,3,S-trimethoxybenzene
with DMF in POCl gives only the Vilsmeier product, 2,h,6-

3

trimethcxybenzaldehyde.

The diphenol 2g.was prepared from the carbonium ion 21_
by the method of Martin and Smith,27 that is by fusion in
pyridine hydrochloride (prepared by passing hydrogen chloride
through a dry solution of pyridine in ether).35 (See Figure 7).
Similar treatment of the triarylmethane gg with pyridine
hydrochloride gives the same product gg. Apparently incidental

oxygen in the reaction mixture serves as the oxidizing agent.

The diphenol gg_is a highly insoluble brick red material
whose characterization is difficult. Thus one or more soluble

derivatives were necessary.

l'\)
O\

Several attempts to prepare such soluble derivatives of
gg_led to the eventual preparation of the diether 39, charac-
terized by nmr (SO.90,t,6H;Xl.lO-1.67,m,2hH;53.90,t,bH;JS.97,s,
23;é6.2§,s,hfi) and sass spec. (m/e 556. hhhv 332); Further-

more, the diacetate }1_was prepared (acetic anhydride, pyridine)

OCH}

    

 

f‘

Figure 11. Preparation of triaryl methane a

CD

27

 

a R = H a»
}_O_ R = n-C8H17

Figure 12. Sesquixanhydrone-diol and derivatives

and an nmr spectrum of the crude, chloroform-soluble material

showed the pcesence of acetate, but it was not purified.

Oxidation of gg_in an attempt to prepare 5p was carried
out in aqueous KOH under N2 atmosphere using Ker (CN)6. A
white solid was obtained in 80% yield which was totally in-
soluble in all solvents, though the solid turned blue in
acidic solvents. The material gave no detectable signal in
the mass spectrometer. It was assumed that the radical had

polymerized.

One attempt to chlorinate gg was made using NaOCl. This

reaction gave a red solid, but it could not be characterized.

Figure 13 shows the route used to prepare h,7-dihydroxy-

phenalenone gg, as well as the diether 31,and the diacetate 1Q.

 

 

alffl,‘
benacnc

 

 

UH

lv:
0\

 

Figure 13. Synthesis of 14,7udihydr0mhenalone (3g) and
derivatives.

29

36
All but the first step were taken from M. Jarcho.

Methylation of commercial 1,6-dihydroxynaphthalene gave
the corresponding diether in 80% yield. Friedel-Crafts acyl—
ation with trgggycinnamoyl chloride (generated igggitg from
Eggggrcinnamic acid and phosphorus pentachloride) in benzene
gave a 71%»yield of the single isomer h-cinnamoyl-1,6-dimeth~
oxynaphthalene. Cyclization occurs in 85%»yield in polyphos-
phoric acid. Treatment of 35,with anhydrous A1013 in
refluxing benzene results in elimination of the elements of
benzene, as well as demethylation, to give a quantitative
yield of }_6_. Methylation of 33' gave only a 30% yield of
diether 31, The method given by Jarcho for preparing the
diacetate }§_(acetic anhydride, catalytic amount cf con. HZSOh)
failed in my hands, but treatment with acetic anhydride and pyr-

idine gave the diacetate in 70% yield.

At this point an oxidation of 3p was attempted. An
oxygen-free solution of K3Fe(CN)6 in water was added to an
oxygen-free solution of’}é in aqueous KOH. The whole system
was kept under nitrogen atmosphere. The solution gradually
turned from light orange to dark red, which may indicate the
fermation of a radical or diradical. The material appears to
be stable in basic solution, but defies all attempts to
isolate it. On introduction of even traces of oxygen, an

intractable red solid appears.

3O

’

Much time was spent in attempts to halOgenate 3g, 11 and
jL, This problem was approached from several angles. The
dihydroxy compound 3§,was treated with halogen or hypohalite
in basic media. The dimethoxy compound 31 was treated with
halogen in various organic solvents, and the diacetate with
chlorine in acetic acid, with and without added acetate.
Because }§_and its halogenated derivatives are insoluble in
organic solvents other than pyridine and DMSO, the products
of halogenation of 3Q were converted to the corresponding
diethers or diacetates for purification and characterization.
Table 3 lists the reactions carried out and the results of

each.

Since we had little success with the direct introduction
of three halogen atoms onto the delocalized ring system 3g,
31, jfi, the feasibility of halogenating intermediate }5_in
two or more steps was studied. Figure 1h illustrates the
possible pathways. Treatment of 35 with chlorine in benzene
at room temperature gave a mixture of g§_or Qé_(mp 159-165 dec)
and Q1_(mp 1hh-1h7) in a 60:h0 ratio. Note that it is very
difficult to distinguish between isomers g§_and pg, The mass
spectrum of Q1 also showed a small peak at m/e 350 correspond-
ing to g§_or an isomer. The same reaction in the presence of
FeCl3 gave mostly tar but a small amount of Qfi_or y§_was

isolated (less than 10%).

31
Treatment of 25_with 5% sq. NaOCl/methanol in a hetero-
geneous reaction gave a 60% yield (the remainder being unreacted
starting material) of pg, Due to contamination by starting
material it was impossible to tell if either or both diaster-

iomers were present.

Reaction of gg_with aluminum trichloride in refluxing
benzene gave an insoluble red material which on treatment with

acetic anhydride and pyridine gave a 50% yield (from pg) of 59.

Table 3. Halogenation of 3g
Starting Conditions
material

36 3) Br 2, {syridine
2) (CH wh, NaOH
36 1)Br2, aq. NaOH
2)A020, pyridine
36 :) NaOCl
2) (CH3)2SOh, NaOH
36 1) NaCCl
2) 08H171, Na)H
36 1) NaOCl
2) Ac20, pyridine
37 Br2, CH3002H
37 Br2, CH3COZH,
CHBCOZNa
38 012, CH3002H, 0°
38 012, CHBCOZH, 70°

b

amp 2a0.5-2u2.5, m/b

mp 195-197. m/6 330

398

and derivatives.

Results

10% yield of mix--

ture of mono-, dim,
and tri-bromo diethers
by mass spec

tar

‘ l l l
powdégso ub e b ack

insoluble black
powder

uncharacterizable
oil

insoluble black
oil

30% yield of a
dibromoethera }2_

70% yield of a
monochlorodiacetate

A2.

insoluble black

powder

 

1:. 1’
C l
w 44

 

C)
/
/
”Q
/\./

  

cs- ‘ 0:331

IQ) ‘ j r . 0 3
u' I

/' Cl C}

OCH, 48 OCH.j (trace)

Figure 15. Chlorination of fij_with chlorine in benzene.

 
 

 

 

C1
0
5% sq. NaOCl. , H
methanol // I C“3
OCH.j OCH?

Figure 16. Chlorination of }§_with hypochlorite.

1)A1:;13 c

>
\
‘S.

 

V
, K/

2 )AC2O, ‘ .i' ’, fi 3?

pyr '

 

no t

A
‘)

 

Figure 17. Reaction of hi with aluminum trichloride.

It is interesting to note that though one might expect to
obtain a mixture of isomers 59, §l_and 5g, nmr and melting
point data indicate the presence of only one isomer, apparently
59. In any case, the elimination of HCl from Qg_though not

unexpected, is not the desired result.

36

EXPERIMENTAL

Measurements. Boiling points and melting points were
uncorrected. Melting points were determined on a Thomas-
Hoover melting point apparatus. 1H NMR spectra were recorded
on a varian T—60 or Bruker 180 NMR spectrometer. 13C NMR
spectra were measured with a varian OFT-20 spectrometer.

Both types of NMR spectra were obtained with tetramethylsilane as
the internal standard unless otherwise noted. Chemical shifts
(.8) in parts per million from the internal standard are given
as positive values for downfield shifts in all cases. Infrared
spectra were recorded on a Perkin-Elmer Infracord 137 spectro-
photometer. Ultraviolet-visible spectra were recorded on a
Unicam 800 or Lambda 3. Mass spectra were obtained on a Hitachi
EMU-6 mass spectrometer by Mr. Mark Weidner or Mr. Ernest
Taylor, to whom I hereby acknowledge my gratitude. Elemental
analyses were performed by Galbraith Laboratories, Knoxville,
TN.

Solvents. Benzene was purified by distillation from
sodium metal. Tetrahydrofuran (THF) was dried by distillation
from potassium benZOphenone ketyl. Anhydrous diethyl ether was
prepared via distillation from lithium aluminum hydride. Each

of the above distillations was performed under nitrogen.

37

Anhydrous dimethylsulfoxide (DMSO) was dried over and distilled
o
from calcium hydride at reduced pressure and stored over u.A

molecular sieves under nitrogen.

3:Methoxymethylanisole(9)37. A dried 250 ml, 3-necked
round bottom flask equipped with magnetic stirring bar,addition
funnel, condenser and drying tube was charged with 2.5 g
(20 mmol) of 3-hydroxybenzyl alcohol (Aldrich) dissolved in
100 ml of dry tetrahydrofuran. To the mixture was added 10 ml
(5.0 g, 35 mmol) of methyl iodide. The mixture was cooled in
an ice Oath and, while stirring, 2.h g (100 mmol) of sodium
hydride (previously washed with benzene) was added. The
mixture foamed as hydrogen was liberated. After stirring at
room temperature for 20 hours, an additional 10 ml of methyl

iodide was added.

The excess sodium hydride was destroyed by the careful
addition of saturated sodium sulfate. The THF was evaporated
and the residue extracted thoroughly with ether. The combined
organic phases were dried, filtered and evaporated. Short-
path distillation gave 2.63 g (86%) of 3-methoxymethylanisole.
Pmr:.$3.23,s,3H;.33.62,s, 3H;,§h~23,s,2fl;,S6.S-7.2,m,hfl.

h-Bromo-3-methoxymethylanisole(10). To a solution of
2.00 g (13.2 mmol) of 3—methoxymethylansiole (2) in no ml of
carbon tetrachloride was added very slowly and with stirring

and cooling, a solution of 2.10 g (13.2 mmol) of bromine in

38

200 ml of CClh. The rate of addition was determined by the
decolorization of the bromine. Hydrogen bromide was evolved
throughout the reaction. After the addition was complete,

stirring was continued for one hour.

The reaction mixture was extracted twice with 30 ml
portions of 5% sodium bicarbonate and twice with water, dried

filtered and rotary evaporated. The resulting dark

(Meson).

oil contained at least three components, one of which was

later shown to be unbromo-3-methoxymethylanisole(lgfl. A

second spot with a similar rf value may be the other isomer.

Chromatography (CHC13/alumina) gave as the first band (0.9 g,

30%) of a single monobromo isomer. Pmr: 3'3..3h,s,3H; 53.63,s,3H;
Jh.30,s,EH; 3 S.h5,dd(J=3Hz,J::5Hz),1H; (16.9O,d(J=3Hz),1H;\§ 7.16,

d(J=8Hz),1H. From the dark oil crystallized a few milligrams

of white needles (mp 73.2-73.50C), assumed from mass spec to

be a demethylated anaIOg. Mass spec: m/e 216 (96%), 21h (100%),
185 (12%). 183 (13%).. 63 (81%)-

h-Iodo-3-methoxymethylanisole(ll). Under a nitrogen

atmospherea200 ml, 3-neck, round bottom flask was equipped

with stirring bar, addition funnel and condenser. It was flame

dried, charged with 2.80 g (13.0 mmol) of silver(I) trifloro-
acetate and redried. To the salt was added 2.00 g (13.0

mmol) of 3-methoxymethylanisole. To the stirred slurry was

added through the addition funnel a solution of 3.30 g (0-13

39
mol) of iodine in 125 ml of chloroform. As the addition prOCoeded
the slurry gradually dissolved and a yellowish precipitate of
silver iodide formed. After the addition was complete the mixture

was stirred for an additional hour.

The mixture was filtered to remove the Agl, washing with
chloroform. The combined chloroform solution was evaporated
to give a yellow oil which was distilled under aspirator
pressure (bp 1580-16000). Yield 3.3 3 (90,13) of colorless oil
which on cooling solidified, mp 30-3SOC. Pmr:é}3.33,s,3H;§§3.63,
5,3H; i‘u.21,s,2n; A-Vé.30,dd(Jz3HZ,8Hz),1H; 3' 6.80,d(J==3Hz),1H;
37.36,d(J=8Hz),1E.

AfMethoxyw2—methoxymethylbenzoic acid‘lfi). A 100 ml,
3-necked round bottom flask was fitted with stirring bar, two
rubber septums and condenser. The apparatus was fitted with
a nitrOgen atmosphere and flame dried. It was charged with
0.30 g (0.125 mol) of magnesium turnings and redried. Enough
dry ether was injected to just cover the magnesium. A portion
of a solution of 3.0 g (10.8 mmol) of h-iodo-3-methoxymethyl-
anisole(11) in 20 ml of dry ether was added. When no reaction
occurred a few drops of methyl iodide were added. When the
mixture warmed slightly and bubbles began to form more of the
reaction proceeded. After the addition was complete and the
reaction appeared to slow down, the mixture was heated at

reflux for several minutes.

L0

Carbon dioxide dried by passage through concentrated
sulfuric acid was introduced under the surface of the solution
causing masses of white solid to precipitate. When no more
white solid formed, the addition was stopped and 30 ml of 3%
aqueous sulfuric acid was added slowly to the etherial mixture

causing the white solid to dissolve.

The ether layer was removed and the aqueous layer extract-
ed twice more with 15 ml portions of ether. The combined
organic layers were extracted three times with 15 ml portions
of 10% aqueous sodium hydroxide. The alkaline extracts were
combined, made acidic with 9% aqueous hydrochloric acid and
extracted again with ether. After drying'(MgSOh), filtering
and rotary evaporation there was obtained a yellow powder.
Recrystallization from water gave 1.1 g (55%) of white needles,
mp 1523—15145. Pmr. 33.hh,s,3H; S3.79,s,3H;Sh.79,s,2H; $6.70,
dd(J=3Hs,8Hz),1H; g 7.10d(J=3Hz),1H; s 7.93,d(J=BHz),1H. Mass
Spam 13/9 196 (111%). 151 (17%). 153 (100%). 1149 (3%). 135 (17%)-

i—Methoxyphthalidew. A solution of 100 mg (0.510 mmol)
of h—methoxy-Z-methoxymethylbenzoic acid in 10 ml of concen-
trated sulfuric acid was allowed to stand at room temperature
for three days. It was then added drOpwise to 100 ml of
crushed ice. After melting the aqueous solution was extracted
thoroughly with ether. The combined organic extracts were

dried (MgSOh), filtered and evaporated to give a dark powder.

bl

Recrystallization from a minimum of water gave white needles,
mp 113.0n11h.500. Yield h? mg (56%). Mass spec: m/e 16h (72%),
135 (100%), 107 (9%), 92 (1173), 77 (19%); meta stables: 111
(16u-135). 85 (135-107). IR: 3025 cm'1(w). 1775(3), 1625. 1505.
1370, 135, 1275, 1160, 1110, 1060, 1025. Pmr:s3.80,s,3s;¢\‘5.78,
3,213 $6.77, br 8,111; 57.03.,br d(J:8Hz),1H; $7.67,br d(J=8Hz),‘H.
Ethyl h-methoxyyZ-methoxymethylbenzoate(11). The Grignard
reagent was prepared from h—iodo-B-methoxymethylanisole(11) as
reported above for the preparation of 15, The quantities of
starting materials used were 5.5 g (19 mmol) of 11 and 0.50 g
(21 mmol) of magnesium in 50 ml ether. To the Grignard solu-
tion was added, through an addition funnel, a solution of 0.93 g
(7.0 mmol) of diethyl carbonate in 10 m1 of dry ether. After
the addition was complete the mixture was stirred at reflux

for two hours.

To the cooled mixture was slowly added 30 ml of cold 10%
aqueous sulfuric acid. The aqueous layer was separated and
extracted with three 50 ml portions of ether. The combined
organic extracts were dried (MgSOb), filtered and evaporated
to.give an oil which on chromatOgraphy (CHCl3 on silica) gave
in order, 3-methoxymethy1anisole(2), h-iodo-3-methoxymethyl~
anisole(11) and ethyl h-methoxy-Z-methoxymethylbensoate(ll).
Pmr: S1.33,t(J=7Hz),3H; £336,833; S3.80,s,3H; Sh.2h,q(J-=’?Hz).

es; g‘u.80,e,2s; $6.70,dd(J=3Hz,asz),1H; S 7.83,d(J=8Hz),1H.

h2

Tris(2,h,6-trimethoxyphenyl) carbonium chloride5121)_and
Tris(2,h,6-trimethoxyphenyl) methanol(20). A 250 ml, 3-necked
round bottom flask equipped with septum,condenser, magnetic
stirring bar and condenser was charged with a solution of 6.0
ml of gig butyllithium (13 mmol) in 20 m1 of dry ether. To
this mixture under an atmosphere of nitrogen was added slowly
a solution of 2.0 g (12 mmol) of phloroglucinol trimethyl ether
(1Q) in 20 ml of dry ether. The solution was stirred overnight
and a solution of O.h7 g (h.0 mmol) of diethyl carbonate in 20
ml dry ether was added slowly. The solution was stirred at

room temperature for 12 hours.

To the mixture was added 20 ml of 10% aqueous hydrochloric
acid, causing the mixture to turn deep blue. The aqueous
layer was separated and the ether layer washed once with 10 ml
of 10% 301 to remove the last of the blue color from the ether.
The combined aqueous layers were washed five times with 10 ml
portions of ether to remove starting material. The blue mater-
ial stayed in the aqueous layer while the ether extracts
remained colorless. The blue material was extracted from the
aqueous solution with chloroform. Drying (MgSOu) and filter-
ing the organic solution followed by rotary evaporation gave
0.050 g (2%) of blue-black solid, mp 135-1h0 dec. When this
material is treated with 20 ml of 10% aqueous sodium hydroxide
the color is discharged. Extraction with ether followed by

drying (MgSOh), filtering and evaporation gave a trace of

h3
material assumed to be g9, Pmr of gl;.§3.57,s,1dh;.33.97,s,9H;
.36.00,s,6H. Mass spec of 21; m/e 513 (62%), h99 (1mg), LEB
(113%). 3116 (100,16). 31S (58%). UV/vis of g1: 581. nm, 150 nm,
366 nm, 323.

Method B. Various ratios of 2,h,6-trimethoxybenzoic
acid were mixed with 1.0 g of 1,3,5-trimethoxybenzene and
dissolved in phosphorus oxychloride. The solution was
stirred for 8 hours over a steam bath under a condenser
protected by a drying tube. After cooling the solution was
poured over crushed ice. The dark aqueous acidic solution
was extracted thoroughly with benzene, which removed starting
material, as well as a reddish purple material. The blue
product stayed in the aqueous layer until extracted into
chloroform. Drying (MgSOh), filtering and evaporating gave
yields ranging from 6% to 20% of dark blue solid g1, From the
benzene layer is obtained mostly 1,3,5-trimethoxybenzene.
Anal. Calcd. for 028H33C109: C, 61.27; H, 6.06%. Found: C,
59-93; H, 6-1S%-

Tris(2,h,6utrimeth0xyphenyl)methane(g§). A small amount
(20 mg) of the blue triarylmethyl material gl_was dissolved
in 5 ml of dry THF and added to a slurry of 20 mg (0.53 mmol)
of lithium aluminum hydride in 5 ml of dry THF. The blue color
was discharged immediately. The excess LiAIH)4 was destroyed by
adding saturated aqueous sodium sulfate drOpwise with stirring

in an ice bath. The mixture was filtered and the white solid

uh

washed thoroughly with THF. The organic solution was dried
(MgSOh), filtered and evaporated giving a brown oily residue
which was crystallized from 95% ethanol. The brownish needles
(mp 178-1820C) gave an nmr which suggested the desired

triaryl methane, though clearly impure. Pmr: S3.h0,s,18H; $3.70,
s,9H; 5' 6.00,s,6H; 556.17,s,1H. IR. 2920 om’1, 2810, 1575, 1150,

1a00, 1315(w), 1125(st,br), 1050, 9a5.

Method 3. Into a refluxing solution of 2.0 g (12 mmol)
of 1,3,5-trimeth0xybenzene (1§) in 20 m1 of phosphorus
oxychloride was dripped 20 ml of 80% aqueous formic acid. The
resulting reaction is quite vigorous, evolving quantities of
gas due to the reaction of P0013 and water as well as POCl3
and formic acid (giving hCl and 00). After the addition was
complete the red solution was diluted carefully with water.
The resulting aqueous acid solution contained some solid
precipitate. It was extracted thoroughly with benzene. The
combined organic extracts were washed with brine, dried
(Nazsoh), filtered and evaporated to give 2.3 g of dark solid
residue. washing with 95% ethanol/chloroform gave 1.2 g (60%)

of triaryl methane, mp 205-207°C. Pmr: See above.

2-Bromo-1,3.5-trimethoxybenzene (23). A solution of 3.0
ml (9.0 g, 56 mmol) of bromine in 50 m1 of carbon tetrachloride
was added dropwise to an ice—bath chilled stirred solution of
10.0 g (60 mmol) of 1,3,5-trimethoxybenzene in 50 m1 of CC1b°

After the addition was complete, the solvent was evaporated

QS

and the residue recrystallized from EtOH/HZO to give 11.6 g
(787:3) of white needles, mp 92-9u.5, iii.38 99°C. Fmr: :1, 3.'{S,s,
3s; 553.81,s,6H; 2?13.07,s,2H.

Ethyl 2,h,6~'rimethoxybenzoate(gl . in oven-dried
apparatus consisting of a 100 ml, three-necked round bottom
flask, condenser, addition funnel, septum and magnetic
stirring bar was charged with 0.50 g (21 mmol) of magnesium turn-
ings and redried under nitrogen atmosphere. A portion of a
solution of 3.6 g (1h.6 mmol) of 2-bromo-1,3,S~trimethoxybenzehe
in 30 m1 of dried tetrahydrofuran was transferred into the
reaction flask by syringe. As the reaction did not commence
immediately, several drOps of 1,2-dichloroethane were added
Soon the mixture began to form small bubbles, to warm slightly
and to turn a golden yellow shade. The remainder of the aryl
bromide solution was added gradually and the mixture was
stirred at room temperature for five hours. After most of
the magnesium had been consumed and the solution had become
dark brown, a solution of 0.60 ml (0.5h g, 5.0 mmol) of ethyl
chloroformate in 15 ml of dry THF was added. The mixture was

heated at reflux then stirred at room temperature overnight.

To the reaction mixture was added 50 ml of 20% aqueous
hydrochloric acid. The mixture turned a very deep purple.
It was extracted several times with 30 ml portions of chloro-
form until the color was removed from the aqueous layer. The

combined organic extracts were washed with brine, dried (MgSOb),

L6
and rotary evaporated to give 2.9 g of a dark oil. Chromato-
graphy gave 1.LS g (h1%) of ester as indicated by pmr. The
rest of the material stayed on the column. Pmr:§§1.31,t(J=7Hz),

3H; 530699593H;$307149396H3'31:027:Q(J27H2)y2H3 86°00’892H°

Method B. A solution of 2.0 g of 2,h,6-trimethoxybenzoic
acid and 15 ml of freshly distilled thionyl chloride was
heated at reflux for four hours. The thionyl chloride was
removed under reduced pressure. The resulting yellow solid
was dissolved in 25 m1 of absolute ethanol and the solution
was heated at reflux for two hours. Rotary evaporation gave
a colorless oil whose pmr spectrum showed it to be a mixture
of the desired ester and 1,3,S-trimethoxybenzene.

2,h,6—Trimethoxybenzaldehyde(32). A solution of 3.0 g
(18 mmol) of 1,3,S-trimethoxybenzene in 20 ml of phosphorus
oxychloride was chilled in an ice bath. With stirring, a
solution of 5.0 g (69 mmol) of dimethylformamide in phosphor-
us oxychloride was added dropwise. After the addition was
complete, the solution was allowed to come to room temperature

and stirred for three hours.

The solution was poured carefully over 100 m1 of crushed
ice. The aqueous solution was washed several times with
chloroform, neutralized with 20% aqueous sodium hydroxide and
allowed to stand overnight. From the solution crystallized
2.9 g (90%) of beautiful white needles, mp 117-119.5°c, iit.38

118-120°c. Pmr:§3.9o,a,9s; S‘6.o7,s,2H; 2.10.26,a,1s.

b7

Methyl 21h16—trimethoxybenzoate(26). A mixture of 1.0 g

 

(b.75 mmol) of 2,h,6-trimethoxybenz010 acid(gg), 2.0 g (11.5
mmol) of potassium carbonate, 2.0 g (1h.1 mmol) of methyl
iodide and 20 ml of dry acetone was stirred at room temperature
for 2h hours. The mixture was then filtered and the acetone
evaporated to give 0.97 g (91%) of a white solid. Pmr: $3.66,

39.33; 83°72’8a6H3 $307998’383 85°98’592H0

Eyridine hydrochloride(}j). Hydrogen chloride gas was
passed through a solution of 79 g (1.0 mol) of dry pyridine
dissolved in 300 m1 of dry ether. The white product precipi-
tated. After no more solid formed, the precipitate was filtered
and washed thoroughly with ether. Yield 11; g (100%), mp 1th-1L6OC,
lit.35 1L3.h. The salt is very hygrosc0pic, but can be stored
indefinitely in a dessicator. Pmr:§‘6.17-7.33,m.

S,Q-Dihydroxy-1H—3,1,11—trioxatriangulenn1uone(22). A
mixture of 8.60 g (15.7 mmol) of tris(2,h,6-trimethoxyphenyl)
carbonium chloride(gl) and 30.0 g (261 mmol) of pyridine
hydrochloride were sealed in a tube and heated in a 180°C oil
bath for six hours. After cooling, the dark fused mass was
taken up in water. The brown insoluble solid was washed
thoroughly with water, alcohol and chloroform giving 5.1 g
(98%) of brown powder, mp3oo°c. Pmr (CF3C02H):S 6.7,s. Pmr

(D20,K0D):.§5.1,s. uv/vis (Kori/320): 1.16 nm.

Method B. The same procedure as above starting with

tris(2,h,6—trimethoxyphenyl)methane gave a good yield of the
same material.

5,9«Dioctyloxy-1H-3,7,11-trioxatrianguleg:1-one(39). A
mixture of the brown powder(gg) (1.0 g, 3.0 mmol), potassium
carbonate (1.2 g, 9.3 mmol) and octyl iodide (2.0 ml, 2.5 g,
10.h mmol) in 25 m1 of dimethylformamide was stirred and heated
at reflux for five hours. After cooling, the dark solution
was diluted with 100 ml of 5% aqueous sodium hydroxide and
extracted with three 30 ml portions of chloroform. The com-
bined organic layers were dried (MgSOu), filtered and evapor-
ated to give a brown solid. Chromatography (CHBCl on silica)
gave as an intermediate yellowugreen fraction 0.5 g (30%) of
orange solid, crystallized from acetone/chloroform, mp 2h?-
2so°c. Mass spec: m/e 556 (100%). mm (100%). 332 (119%).
30L. (10%). UV/vis: has, 150, h20(w). Pmr: $0.90,br t(J=6Bz),
6H; 8'1.1o-2.00,m,2l.s; 83.90,br t(J=7Hz),hH; $5.97,32H; 86.25,br
s,hH.

Attempted oxidation to diradical 3. A solution of 0.50 g
(1.5 mmol) of brown material 12 and 1.0 ml of 5% aqueous potas-
sium hydroxide was purged of oxygen by refluxing under nitro—
gen while passing nitrOgen through the refluxing solution.
To the solution was added an oxygen-purged solution of 1.00 g
(3.0 mmol) of’potassium ferricyanide in 10 ml of water. The
mixture was heated at reflux for five hours, during which the

color of the solution lightened considerably, but the visible

M9
spectrum s;ayed the same (i.e., identical to a Spectrum of the
starting material dissolved in aqueous potassium hydroxide).
Gradually an off-white precipitate formed which was isolated
by centrifugation and vaccuum dried. lt amounted to 0.h0 g
(80%), did not melt below 30000, was insoluble in all solvents
including aqueous acid and base, and gave no signal in the

mass spectrometer.

1L6-Dimethoxynaphthalene(33). In a 3003ml, three-necked
round bottom flask equipped with addition funnel, overhead
stirrer and condenser, 100 g (0.625 mmol) of 1,6-dihydroxy-
naphthalene was dissolved in 500 ml of 23 aqueous sodium
hydroxide. To this was added, with ice bath cooling, 126 g
(1.00 mol) of dimethyl sulfate all at once. The ice bath was
removed and the exothermic reaction warmed the mixture above
room temperature. After the temperature had fallen back to
room temperature, an additional 280 ml of 2Q sodium hydroxide
was added, followed by 60 g (0.h7 mol) of dimethyl sulfate.
The mixture was stirred for one hour with heating on a steam
bath. After cooling to room temperature, the mixture was
extracted five times with 75 ml portions of chloroform. The
combined chloroform extracts were washed twice with 50 ml
portions of 23 sodium hydroxide and once with brine.
Drying(MgSOh) followed by filtration and evaporation gave 111 g
(gap) of a dark oil which eventually solidified. Filtration

through a column of silica with chloroform, followed by two

50

recrystallizations from petroleum ether (bp 60-900C) gave 9h.0 g
(80%) of colorless needles. Mp 57.5—59.00, lit.39 50-6100.
Pmr: $3.78, 3,311; 53.85.3311; 56.3-6.7,m,1a; 56.9-7,m,mi;
7.9-9.2,m,1H.

1-Cinnamoylyh,7-dimethoxynaphthalene36(3g). A 1-liter,
3-necked round bottom flask was fitted with overhead stirrer
and condenser with drying tube. It was charged with 55 g
(0.37 mol) of trans-cinnamic acid, 70 g (0.37 mol) of 1,6-
dimethoxynaphthalene and 500 ml of dry benzene. The solid
did not completely dissolve. The mixture was cooled in an
ice bath, and 80 g (0.38 mol) of solid phosphorus pentachloride
was added by spatula and the mixture was stirred vigorously.
After about five minutes the solution had turned light green
and there remained clumps of solid. The ice bath was removed
and the mixture was stirred at reflux over a steam bath for
five minutes until all the solid was dissolved and the solution
was dark green. It was again cooled in an ice bath and, while
stirring vigorously, 52 g (0.39 mol) of anhydrous aluminum
chloride was added in portions. Upon addition of aluminum
chloride the mixture immediately turned deep red. After all
of the aluminum chloride had been added, the mixture had
become semi-solid and difficult to stir. After refluxing for
an additional ten minutes, the mixture was cooled to room
temperature and a mixture (50:50) of ice and concentrated

hydrochloric acid was added slowly. After the hydrolysis

was complete (about 100 ml of aqueous solution had been added)
the benzene layer was removed and the aqueous layer extracted
twice with 75 ml portions of benzene. The combined organic
extracts were washed successively with two 50 ml portions of
10% aqueous hydrochloric acid, 50 ml of 10% aqueous sodium
hydroxide and water until neutral. The organic solution was
dried (MgSOh), filtered and evaporated to give a dark red oil
which slowly solidified. Crystallization from benzene/pet.
other with the use of decolorizing charcoal gave 85 g (71%)

36 96’9800 o

of light yellow-green needles, mp 95.5—97.50C, lit.
Pmr: S3.87,s,31i;§3.97,s,3H; 86.56.d(J-‘-8HZ).1H;56-9-7-5m98H;

57.78,d(J=BHz),1H; 58.01,d(J.—.3Hz), 1H; 88.11 ,d(J=3Hz),1H.

2,3-Dihydro~h,Zrdimethoxy~3-phenyl-1H-phenalenm1-one36(3§).
in a one~liter beaker, 66 g (0.21 mol) of 1-cinnamoyl-h,7-
dimethoxynaphthalene, 33, was mixed with 300 ml of polyphos-
phoric acid, and the mixture was heated on a steam bath while
stirring the thick, syrupy mixture by hand with a stout glass
rod. After stirring and heating for 20 minutes, the brick red
mixture had become much easier to stir and essentially homo—
geneous. As much of the material as possible was poured over
500 ml of crushed ice in another beaker, and that remaining in
the reaction beaker was treated with 500 ml of water. Event-
ually the syrupy insoluble material became hard and lumpy.

The water was decanted and, while still hot, was extracted

52

I

with hOO ml of benzene. The solid remaining in the two
beakers was combined and boiled with hOO ml of benzene. This
was decanted and added to the benzene extracts. This process
(heat in water, extract with benzene, heat in benzene) was
repeated three times, or until all the solid had dissolved.
The combined benzene extracts were evaporated to two liters,
washed with 500 ml each of water, 5%iaqueous sodium hydroxide
and water, then dried (MgSOh), filtered and evaporated to
give 71 g of orange solid. Two crystallizations from pet.
ether/benzene gave h5.2 g (70%) of light yellow needles, mp
169.5-171.5°C, lit.3b173-17h. Pmr: 8‘3.0-3.3,m,28; 53.80,s,3H;
3'3.99,s,3H;.Sh.9-5.1,m,1H;§‘6.67,d(J=8Hz),1H;S'6.93,br 8,53;

5 6.97, d(J=9Hz), 1H; J 8.10,d(J=8Hz), 1H; 8 8.12,d(J=9Hz),1H.

h,7-Dihydroxy-1H-phenalen-1-one36(}§). A two-liter,
three-necked round bottom flask was equipped with overhead
stirrer, condenser and drying tube and was charged with hS.O g
(0.1h1 mol) of 2,3-dihydro-h,7—dimethoxy-3-phenyl~1H-phenalen—
1-one(}5), and 500 m1 of dry benzene. To this solution was
added, in 10 g portions, 91.0 g (0.681 mol) of anhydrous
aluminum chloride. The mixture was stirred at reflux for
two hours. After cooling it was poured into iced concentrated
hydrochloric acid. The pieces of solid material were rinsed
out with water. The solid was broken up as much as possible
and separated by filtration. Thorough washing with benzene

and water gave an orange silt-like solid which was dissolved

53

in 350 ml of 5% aqueous sodium hydroxide, washed with benzene
and precipitated with acetic acid. Washing with water and
vaccuum oven drying gave 29.9 g (100%) of orange solid, mp
300°. Pmr(DMSO): 56.87,d(J~.=8Hz),3H; $8.20,d(J=BHz),3H. Mass
spec: m/e 212 (100%), 18L, (593-3), 155 (2273), 128 (22%), 92
(20%); Uv/vie(DMSO); 509 nm. h79. ass. u29. tea. 363.
9‘132imethoxy-1H-phenalen-1-one36(}1). In a 50 ml, 3-
necked round bottom flask fitted with stirrer and condenser,
2.0 g (9.1., mmol) of h,'l-dihydroxy-1li—phenalen-1—one(}_6_) was
dissolved in 17 ml of 13 sodium hydroxide and cooled in an ice
bath. To the mixture was added 2.0 ml (21 mmol) of dimethyl
sulfate. The mixture was allowed to come to room temperature
and stirred until it solidified. An additional 9 ml of 1‘3
NaOH was added, dissolving the solid, followed by 1.0 ml (10
mmol) of dimethyl sulfate. After stirring for one hour the
mixture was heated on a steam bath for 10 minutes. The con-
tents of the flask were poured into methylene chloride and
the solid material washed out with 1!,NaOH and methylene
chloride. The aqueous solution was extracted three times
with 20 ml of methylene chloride. The organic phase was dried
(MgSOu), filtered and evaporated to give 0.7 g (30%) of yellow
solid. Recrystallization from benzene/pet. ether gave yellow

36

needles, mp 173.5-175.5, lit. 176-1780C. Mass spec: m/e

21,0. Pmr: S 3.97.8.3H; 5 h.OO,s3H; ‘3 6.50,d(J=10Hz),1H; S 6.78,

Sb
d(J=8HZ).1H;5S7.O1,d(J=9Hz),1H;:,7.98,d(J=10Hz),1s;§f8.18,d(J=9
Hz).1H;518.u3.d(J=8Hz)- UV/vis: tau. u30. L10. 388(sh). 365.

8
3u7(sh). 273; lit. uss. u32. L10. 366. 270.

6
hJ-Diacetoxy-1nghenalen-1-one3 (38). A mixture of 1.0

 

g (u.7 mmol) of b,7-dihydroxy-1H-phenalen—1-one(19), 20 ml of
acetic anhydride and 5 ml of pyridine was stirred at room
temperature for 18 hours. The solvent was removed at room
temperature under reduced pressure. Crystallization of the
residue from benzene gave 1.09 g (75;)of golden crystals, mp

177-17900, lit.36

180-18200. Pmr:'32.u3,s,3a;*§2.u§,s,3H;536.62,
d(J:10Hz),1H;~s7.28,d(1=9nz),is;§ 7.u5,d(J:8Hz),1sg S7.7F,d(J-

10Hz),1s;537.97,d(J=9Hz,1H;558.51,d(J:8Hz),1a.

Attempted oxidation_tp the diradical(h); In the addition
funnel of a reaction apparatus which also included a 100 ml,
3-necked round bottom flask and condenser with nitrogen atmo-
sphere, a solution of 1.55 g (h.71 mmol) of potassium ferri-
cyanide in 20 ml of water was purged of oxygen by bubbling
nitrogen through the solution. This solution was added to a
similarly purged solution of 0.5 g (2.36 mmol) of h,7-dihydroxy-
-1H—phenalen-1-one(}§) in 5 ml of 5% aqueous sodium hydroxide.
The solution turned bright red within a few minutes and the
color deepened as time went on. This red material could not
be dissolved in water or any organic solvent. 0n introduction

of oxygen, a brown solid precipitated.

55

5,8-Dibromo-h,7-dimethoxy-1H-phenalen-1-one(32). In a
100 ml round bottom flask equipped with condenser and magnetic
stirring bar, a mixture of 0.50 g (2.1 mmol) of h,7-dimethoxy-1H-
phenalen-1-one, 0.82 g (10 mmol) of sodium acetate, 0.32 ml
(1.0 g, 6.3 mmol) of bromine and 10 ml of glacial acetic acid
were stirred while heating at reflux for one hour. After cooling,
the mixture was diluted with 30 ml of water. The resulting solid
was filtered out, washed with water, and filtered out, washed with
water, and dissolved in methylene chloride. The organic solution
was washed with 5%»aqueous sodium hydroxide, 1H hydrochloric
acid, water and brine. It was dried (MgSOh), filtered and
evaporated to give 0.55 g of a dull yellow solid. This was
boiled in carbon tetrachloride and the resulting orange solid
was filtered out, chromatographed and crystallized from benzene
to give 0.13 g of golden needles. From the carbon tetrachloride
solution, 0.h2 g of needles crystallized. Total yield 0.55 g (60%).
MP ZhO-S-ZhhoS- Mass specz m/9 1400 (10%). 398 (100%). 396 (19%).
357 (13%). 355 (2h%0. 353 (12%). UV/Vis= h76 nm(8h). hhS. h25(8h).
3&2. 327(sh). 27s.

2-Chlggo-h17-digcetoxy:1H—phenalen—1-on§(yh). A solution
of 0.30 g (1.01 mmol) of h,7-diacetoxy-1H-phenalen—1-one(3Q)
in 25 ml of glacial acetic acid was added drOpwise to a solu-
tion of 0.2 g (2.8 mmol) of chlorine in 25 m1 of glacial acetic

acid (prepared by bubbling chlorine gas into a weighed amount

56

of acetic acid, reweighing, and taking an aliquot to give the
desired amount of chlorine). The addition was done at the
freezing pOlnt of acetic acid (16°C) by freezing the chlorine
solution in an ice bath, then adding the diacetate solution
with stirring as the acetic acid in the flask melted. The
reaction mixture was stirred at room temperature overnight

and the solvent was evaporated to give a golden oily liquid
which eventually solidified. Recrystallization from ethanol
gave 0.21 g (6%) of golden needles, mp 195;.-197°0. Pmr: Saul,
5,63; S7.23,d(J=8Hz),1H; S 7.h2,d(J=8Hz),1H; ,8 7.87,s,1H; S 7.93,d
(J=8Hz),1H; 38.51,d(J=on),1H. Mass Spec; m/e 332 (3%), 330

(5%). 290 (533). 288 (121%). 21:8 (31.73). 21.6 (100%).

2-Chloro-2,3-dihydro-h,7-dimethoxyr3-phenylw1H—phenélen-
lzgng(gg). A mixture of 0.5 g (126 mmol) of 2,3-dihydro-h,7-
dimethoxy-3—phenyl-1H-phenaien-1-one(35), 2.h ml of 5% sodium
hypochlorite (0.12 g, 1.65 mmol) was diluted to five ml with
methanol. The heterogeneous mixture eventually turned to a
gel. After three hours at room temperature, the mixture was
dissolved in benzene/water, the organic layer separated, dried
(Nasth) and evaporated to give 0.57 g of yellow solid.
Recrystallization from benzene/pet. ether gave 0.5h g of a
60:h0 mixture of starting material and product in two creps,
the first being mostly product. Analysis was carried out by
nmr. Pmr: S3.78,s,3H; Sh.57,d(J=2Hz),1H; Sh.00,s,3H; 55.17.d

(J=2Hz),1H;€56.6-7.2,m,7H;238.17,d(J=8Hz),2H. Mass spec: m/e 35h

57

(10073). 352 (37%).

yij-Diacetoxy-3-phenyl-1H-phenalen-1»one(50). A 5 ml
flask fitted with stirrer, condenser and drying tube was
charged with a solution of 200 mg of the first crop from above
(0.60 mmol) in h ml of dry benzene. To the mixture was added
all at once hOO mg (3.0 mmol) of anhydrous aluminum chloride.
The mixture was heated at reflux for 1% hours by which time it
had become deep red, and allowed to stir at room temperature
overnight. A red solid was filtered out and washed with water
and benzene. After drying this solid was dissolved in 20 ml
of acetic anhydride and 5 ml of pyridine. The solution was
stirred for four house at room temperature, then the solvent
was rotary evaporated. Recrystallization of the residue from
benzene gave 100 mg (1.731.) of yellow needles, mp 1995-20200.
Pmr: S2.b0,e,3H; 3'2.L.3,e,3a, .S 7.0—8.0,m,9H; 5'8.50,d(J.—.Hz), 111.

Mass spec: m/e 372 (2773). 330 (3076). 288 (10075). 287 (91%)-

Part II

SYNTliES IS OF
OXYGENATED INDACENS DERIVATIVES

5:3

59

INTRODUCTION

Indacene, 5}, an anti-aromatic, unstable, red liquid was

h2

prepared in 1963 by Hefner. It is a “benze” derivative of
pentalene, the anti-aromatic hydrocarbon, 53. Though penta~
lene itself has resisted synthesis due to its great propensity

L3

to dimerize, its 1,3,5-tri—t-butyl derivative and its hexa—
phenyl derivativehh have been prepared. Hydrindacene, 55, is
the hexahydro derivative of indacene. It was my goal to pre-
pare several oxygenated hydrindacene derivatives, Specifically

the hexaketone, 59 and its dihydroxy derivative 51.and the

octaketone 5Q.

 

 

Figure 18. Indacene and derivatives.

60

These molecules are interesting in themselves due to their
highly electron deficient nature. For example, they could
serve as electron acceptors in charge transfer salts. Their
small size, ceplanarity and highly delocalized nature suggest
that the radical anions of 5§_and 2Q (see Figure 19) might
serve as replacements for TCNe‘ in the organic conductor
TCNQ-TTF and similar charge transfer salts. The molecules
Eé_and 51_are also envisioned as precursors to novel poly-
cyclic aromatic compounds. Examples of aromatics which could
possibly be synthesized from 56 are 65 and 66 (see Figure 20).
The first step is simply a double aldol condensation and will
be explored further in the results and discussion section.
The second step, though not further explored, was envisioned
as a Clemmcdsen or Holf-Kishncr tyne reduction to the hydro-
carbons §3_and 65, This would be followed by aromatization
with a dehydrogenating agent such as chloranil. Though
aromatic, the molecules 65_and 69 incorporate two pentalene
systems. They thus could be considered benZOpentalenes, and

could add to the growing body of knowledge about this elusive

anti-aromatic system.

Molecules 51_and 5§ are of additional interest as each
is a member of a class of compounds which has received consid-
erable attention recently, specifically the carbon oxo-acids

15.1.16

and the oxo-carbons respectively.

61

Carbon oxo-acids are species containing only carbon,
oxygen and oxygen—bound hydrogen. Simple examples are car-
bonic acid and oxalic acid. The cyclic carbon oxo~acids,

s uaric acid, b , croconic acid, "0 and rhodizonic acid 1,
J...’ v _

5 others ‘

 

7 others

 

Figure 19. Gain or loss of an electron from radical anions
§2_and éQ.

// ‘Ij
c I -' o
\ , "‘\.\
o o
0
CH //n\\ CH

 

 

(0) (0)

Figure 20. Possible synthesis of aromatics él and 9_2_ from
hexaketone 5g.

 

 

 

 

 

531
HO /c
HO 0

£9.

Figure 21. Oxocarbons and ca bon oxo-acids.

are strong acids as their mono— and dianions are stabilized by
delocalization. In fact, the dianions are delocalized to such
an extend that they are considered to be non-benzenoid aromat-
ics.)“s’1J6 The oXo-acids 6 , 19, and ll_can be thought of as

reduction products of the corresponding oxocarbons cyclobuta-
tetrone, cy010pentapentaone and cyclohexahexaone. Oxacarbons
are, of course, compounds containing only oxygen and carbon.

The hexaketone-hydroquinone, 51, is one of a series of carbon

oxo-acids which includes §1 and octahydroxyindacene, §§, all

ON
\J-L

reduction products of the oxocarbon fig. For a review of carbon

oxo-acids and oxocarbons see References hS and us.

66

RESULTS AND DISCUSSION

After much of the following work was completed, a syn-
thesis of hexaketone §§_was reported.“7 It involves an
oxidation of tetraketone 1g (see Figure 22). It was also
reported that fié has a half-wave potential measured on
polarographic reduction of 0.15 V (in acetonitrile vs. 30132)};7
This approaches the value of 0.19 V found for tetracyanoquino-

h8
dimethane. In addition, 5Q was found to form a 1:1 complex
with pyrene,h7 lending some support to the idea of §§_and

similar compounds serving as TCNQ replacements in organic

conductors.

 

 

 
 

O o
\
. To“. B’F/ t5L‘—> N2
EtOH
0 .
.l
t-BuOCl
HCOOH
U 0 (J:
A \ 1
0 {III 0‘366” “L C
-HCl nooo>2> cone
0 0 Q 0

EE .6222.

Figure 22. Synthesis of hexaketone 5§_by Gleiter and Schanz.)47

 

Con. HCl

 

 

 

Figure 23. First approach to hexaketone fig

68
Our first approach to hexaketone 2g is outlined in
Figure 23. This synthesis is analogous to a preparation of
ninhydrin, 12_(Figure 27), from dimethyl phthalate, as out-

b9

lined in Fieser and Fieser.

Meoec

lfleO,(3

 

O Ql. 0

Figure 2h. Second approach to hexaketone 5g.

Condensation of dimeth sulfoxide (DMSO) with tetra-
methyl pyromellitate, 1}, gave a dark material, assumed to be
the disodium salt 1g, Without further purification this
material was treated with concentrated hydrochloric acid to

give a yellow solid shown to be the tetraketodichlorobis-

CA
a. 3

(thioether), ZS. This interesting reaction, known as the
SO. . .
Pummerer rearrangement is an intramclecular redox reaction

or, formally, a shift of hydroxide from sulfur to carbon.

0 (-1
/ 1 /
s ______L___. >.
Ch;-5 / \ M
’ (Al-,3
;2
011 Cl

Figure 25. Formal mechanism of Pummerer rearrangement.

The yellow material 15-is apparently readily hydrolized.
When it was left exposed to air for any period of time, it
darkened and emitted a foul thiol-like odor. Treatment of
15_ with water gave a red solid. This material was decolorized
with bromine, but always quickly regained its color. All of
this suggests that the red material may be a reduced form of

the desired hexaketone 5g, such as 11, In an attempt to pre-

70

vent such a reducing process the dichlorobis(methylthio)
compound 15 was treated with ethylene glycol and mercury(II)
sulfate. It was hOped that the ketal 1§.would result.

Unfortunately, no product could be isolated from this reaction.

A second approach involves tetraketone 1g as in the
scheme of Gleiter and Schanzh7(see Figure 22). It was prepared
by them and by us by the method of Neilands and Vavere.51
(see Figure 2b). Sodium hydride-induced Claisen condensation
of ethyl acetate with tetramethyl pyromellitate gave a red
solid assumed to be the salt g9, Treatment of this salt
with concentrated HCl in glacial acetic acid gave a 50% yield
of silvery flakes. This reaction involves protonation of the
salt, hydrolysis of the ester and decarboxyaltion of the
resulting acid. When the reaction was carried out in the
presence of benzaldehyde, aldol condensation gave the bis

benzylidene derivative (cis or trans isomers) §l, Similarly,

treatment of lg_with benzaldehyde and acid gave @1,

At this point let us pause for a moment for a discussion
of solubility aspects of this project. The desired product,
5g, or its dihydrate, 1Q, would be expected to be quite
insoluble in organic solvents, but fairly soluble in aqueous
solvents. Thus the reaction in which one of these is formed
must not involve inorganic by-products which would be
difficult to separate from the product. For example, selenium

dioxide is a convenient method for converaicn of the methylene

0‘

71

group alpha to a carbonyl group.52 However, it is unsuitable
for oxidation of fig to 6, as separation of products would be
difficult. Thus to carry out the conversion of 1g_to fifi,
Gleiter and Schanz pyrolyzed ޤ_(Figure 22) so that the only
by-products were the gases hydrogen chloride and carbon
monoxide.

Similarly oxidative cleavage of the bis-benzylidene fil_
would give the ketone fifi. However, cleavage reactions
involving periodate, osmium tetroxide or other inorganic
oxides cannot be used for the same reasons as above. Ozon-
olysis of fil, on the other hand, would not give inseparable
by-products and thus would seem to be a reasonable alterna-
tive. However, the bis-benzylidene fil is so insoluble that

ozonolysis is not possible.

It was our plan to carry out a multi-step oxidation of
1§_to 5fi_in which the last step involved pyrolytic cleavage
of volatile molecules, thus providing the necessity for
separation of the product from by-products. This reasoning
is, of course, the same as that of Gleiter and Schanzf‘7
Our scheme (see Figure 26) was based on a ninhydrin synthesis
given in Fieser's undergraduate lab manual.S3 Nitration of
the active methylene groups in 1g gave the dinitro compound
fi}, Bromination gave the dinitrodibromotetraketone fig,
Pyrolysis of this compound should give the hexaketone fifi,

However, heating the material in o-dichlorobenzene (bp 1800C)

gave only a black intractable solid.

One of the key steps in the preparation of the polycyclic
aromatics fi5_and fifi is condensation of hexaketone 59 with
acetone. As a model, ninhydrin 12 was dehydrated (reflux in
benzene under a Dean-Stark trap) then treated with acetone
and a catalytic amount of toluenesulfonic acid. The only
product isolated was gg, as characterized by pmr, IR and mass
spec. This is an intermediate along the desired pathway.

Dehydration followed by an intramolecular aldol condensation

would give fifi.

 

Figure 26. PrOposed synthesis of §fi_from 12,

 

C Q
\ i" //x A]
J ' -__. . ) \ r‘
U;\ 7,; benzene ; 0
¢ 1" -H2O \\/
0 U
[9 ; ‘r'?LC}l
‘5 raent
v
\\ J?
B; ._.____
/ \ \L
//\v
86 '

 

Figure 27. Attempted synthesis of fifi as a model fer fig,

7h

A different approach to the oxygenated hydrindacene
system §fi_involves cyclization of benzene bis-propionic acid
derivatives and gives rise not only to the indacene skeleton
SQ but also to the as-indacene skeleton of 9} (Figure 28).
Cyclizations of the para and meta derivatives were explored,
but the ortho derivatives 9fi.were not, as they can only
produce the less desirable as-indacene skeleton. In any case,
cyclization of the ortho derivatives should be similar in

nature to that of the para derivatives.

Preparation of p—phenylene diprOpionic acid derivatives
filf 99 and m-phenylene diprOpionic acid derivatives 2g and 95.
are outlined in Figures 29 and 30, respectively. Treatment
of terephthaladehyde, 199, with malonic acid in refluxing
pyridine,Sh followed by esterification of the resulting bis
(acrylic acid), 9‘, gave 1,h-bis-(2-carboethoxyvinyl) benzene,
19g.in 75% yield. The same diester was prepared directly from
terephthaladehyde by reaction with the Wittig reagent, 19},
prepared from ethyl bromoacetate, albeit in lowerxyield (29%)
since separation of the product from triphenylphosphine oxide
was difficult. Hydrogenation to the bis(ethyl propionate),
91 and hydrolysis to the bis(propionic acid), §§_, both
proceeded quantitatively. Similarly, reaction of isOphthal-
dehydeSS with malonic acidS6 gave the diacrylic acid 199 in

76%iyield. Palladium catalyzed hydrogenation in aqueous

ammonia gave a nearly quantitative yield of diacid 24.
Figure 31 shows the results of various attempts to
cyclize these derivatives. Warming of diester Q1 in poly-

phosphoric acid followed by hydrOiysis gives keto-acid 101

Q /\ O\ /r ’9
COX , :\,’~ ~\ ’ r/ \"v” )\\/I \\/\\
OE l——-—> u /-—-*:57/|I : s
, ,/‘\ ,/” ‘\ \\ /¢V“‘w
\ / , \V , N ,, \, u
+
87:X=OEt Q /Q
.__ O '_',
§§=X=OH 0 \‘t‘"\ 0 \
g2 ; x=c1 ‘K /; \ ,1" /‘~\\/: 0
ggzx=NMe2 ( ' \\i —> O b
\/’I\ )1 2 ' /
‘7 0‘ 22
9
q C
cox .' cox ‘ //o\
' \\. v A“ \\.//
g H 1 g ———-> / H % I) —" 56
K\V//§\5§Q\ v// \//J\‘:¢/‘ / -
91
94:A=OH U
— \\ /
2:}X=01 x/“7(’\‘N ~—> 22
\\,/3'.\ K)“ \
5~——Jl
C :21

\/ (1L4 40 .9;
\\—————

2a.

Figure 28. Generalized scheme for preparation of hydrinda-
cene derivatives by cyclization of benzene bis-
prOpionic acid derivatives.

   

 

/

\
Lee
W

-. I‘v-“, , ._
—KJ£1\J'\J‘\4) shat

CM, (Goo :
C.

W v

103

   

//
/
I tin/i t
\ C. :
<,ULM1 Utm.;t 57
101 in, i
i. .1“
1)rxui,

n ‘ fi ' "'
, ,o. LU'l C‘fii
(Kiln-18>. V l/\ DJ

_.

}:£.'1'~'l€:-; ,,
C-
COLfihgf
90 89 E33

Preparation of p—phenylene diprOpionic acid

Figure 29.
derivatives.

78

 

 

on, \g
j Cr't:3 \\. th
—.-(=——-> I
H2004 //
'l on ———105
CH3 ‘ C ‘ two
104 ‘
CH2(COOH)?,
‘ -
C5h5N
colt»:
El 1/l‘d/(J
ao. NHUH
. “(9'0“ coon
2.9. \
SOC-l

Figure 30. Preparation of m-phenylene prOpionic acid deriva-
tives.

I

t)

79 g Q
/
. A) + / /\ITJ
HOCO - EtOCO \

 

 

33 Reaction

‘1?

 

 

   

i2
/ i/
I V / +
HOCO ~\ \ t l r
\ a “
\ Heto \\
31 109 EEC

 

c001 \ 0001 ”“2 +
o ’ ' .
3 ”25 110

Figure 31. Attempted cyclizations of benzene bis-prOpionic
acid derivatives.

80

in 9t% yield (along with a small amount of keto-ester log).
The product 191_was recovered unchanged after heating in PPA
for 2h hours. Treatment of the bis(acid chloride), Q2, with
aluminum chloride in carbon disulfide gives only the ketc-
acid‘1;1. (Apparently after the first electrophilic substitu—
tion, the newly formed ketone carbonyl deactivates the ring to
further substitution, especially since the only positions
available for the second substitution are ortho and para to
the newly formed carbonyl. Thus it was thought that deriva-
tives of 1,3-bis(pr0pionic acids) might lend themselves more
readily to double cyclization. However, such was not the
case. Warming of diacid 2g with PPA followed by hydrolysis
gives a 92% yield of a mixture of keto~acids 192 and 119.

This mixture was converted to a mixture of the corresponding
acid chlorides 111_and llg_by treatment with thionyl chloride.
Treatment of this mixture of acid chlorides with aluminum
chloride in carbon disulfide followed by hydrolysis gave only
keto acids 192 and 119, Fusion of the same mixture of acid
chlorides in a melt of aluminum chloride, sodium chloride and
)57

potassium chloride (mp 100°C gave only unidentified

aliphatics.

Later deve10pments in our laboratory showed that failure
of our system to undergo a second cyclization is not unusual.
In an unrelated project, Dr. L. L. Klein58 had reason to
desire to perform a double cyclization of the dimethoxy-p-

phenylenedipropionic acid 113. He found that only one cycliza-

81
tion would occur even in this system which is considerably

more electron-rich than ours, and thus would be expected to

undergo electrOphilic substitution more readily.

31::l33
COOli \\\ ”\\1
I l/

//' Ctdfii

 

Figure 32. Failure of dimethoxy—pfphenyleggdiprOpionic acid
to undergo double cyclization.

One final attempt to cyclize a p—phenylene diprOpionic
acid derivative was based on the known59 methylation of
N,N-dialkylamides to give N,N-dialkyl methoxyiminium ions
using strong'methylating agents such as trimethyl oxonium
salts of methyl trifluoromethone sulfonate. we considered
the possibility that the methylated amide 11g_might be a
strong enough electrophile to substitute on the aromatic ring.

If so, the product (115) would not have a deactivating group

82
and a second substitution should occur almost as readily as
the first. Hydrolysis would then give the desired diketone
21.
Warming of 29 with two equivalents of methyl triflate

in chloroform, followed by hydrolysis gave an essentially

83

 

 

 

TfO
MenNCO ,» §;.//\ . Men; OMe ,/~ ,/\
‘ [\J l l 4 er ‘u’
--—--—-¢ 5? 4
/ OOH}. V I \
‘. JVider ”x A ‘
‘ a *C‘/ Moo NMe2
+
29. 114 -
TfC
V
( Tro‘
. ) ze
\ he i DWe
\\ /'\ 2 ~ /‘ 0 \\
/ ‘ “\t’ \ 2 ' 7’ X
\/’ 2 "I v ‘~ ‘\
v’ /\\ V /A\
. = MN: WV:
“to “‘2 Med A“t2
41"
‘, Y.‘ A
\ “,‘\un
\K'i’
Figure 33.

PrOposed cyclization of 29 via the O-methylated
salt.

8h

quantitative yield of 1,h-bis(2-carbomethoxyethyl)benzene,
11g, As esters are the expected product of hydrolysis of
O-methylated-N,N-dimethylamide59, formation of 11g in this
reaction indicates that no cyclization occurs, though the
salt 11g is formed. Apparently O-methylated-W,N-dimethyl-
amides do not deve10p sufficiently reactive electrOphiles to
effect electrophilic aromatic substitution, at least not on

unactivated rings.

 

/’\
// M82N O ' L" 24‘ // MGOCO
UHLLL§7 Ti
29 11.6.

Figure 3h. Reaction of 29 with methyl triflate.

Another approach to the polyketohydrindacene system
involves substitution of a carbonyl equivalent into a tetra-
substituted durene. Tetrakis(bromomethyl) benzene, 115, can
be prepared in almost equal yields (50%) by either of two
methods. Free-radical bromination of durene,6O 112, gave a
rather impure product. Reducing tetramethyl pyromellitate,

13, with lithium aluminum hydride gave tetrakis(hydroxymethyl)

85
benzene,61 129, which, due to its appreciable water solubility
is difficult to separate from the metal hydroxide by-products.
Treatment of this 1,2,h,S-tetrakis(hydroxymethyl)benzene
with HBr in acetic acid gave a quantitative yield of 1,2,h,5-
tetrakis(bromomethyl)benzene,‘119.

It is reported62 that treatment of 1,2-bis(bromomethyl)
benzene with cobalt octacarbonyl in the presence of lithium
chloride gives 2-indanone. However, we found that similar
conditions gave no reaction with tetrabromide 119,

Treatment of 11§_with malonic ester and sodium ethoxide
gave tetraester 121 in 85% yield. Hydrolysis to the tetra-
acid 1gg_was quantitative. The plan was to convert the
tetraacyl azide 1231_followed by Curtius rearrangement to
the isocyanate and hydrolysis to diketone 11163. Treatment
of tetraacid 1gg'with'thionyl chloride gave a yellow solid
which is insoluble in organic solvents. Heating in acetonitrile
with activated sodium azide, in a heterogeneous mixture gave,
after evaporation of the solvent, a glassy yellow solid, pre-
sumably the tetraazide 1_3, This material was highly insoluble
in organic solvents and appeared not to change on short heating
in chloroform. Prolonged heating gave only black intractable
material. Heating of 1g}_in ethanol gave a complex mixture
of unidentified products.

After the report of the preparation of 59F7, we set

aside further work on this system. However, the quinone and

86

hydroquinone 51 and 59 are significantly different and were
deemed worth pursuing. A convenient entry to this system is

shown in Figure 35.

Treatment of hydroquinone with aqueous formaldehyde and

sodium hydroxideéb gave a 70% yield of 1,2,h,S-tetrakis(hydrox-

 

——L
-J
\
J
g.)
—-3

.;;‘\\ ,
0‘ H to
/ Q>;//hn\\
0 O
:32

Figure 35. Scheme for the synthesis of’fié by introduction of
a carbonyl equivalent into a tetrasubstituted durene.

p,

MeOCO ,COOMe

   
 

 

\\ '
/d\ uiaih, ,
MeOCO COOMe f \
0H 0H
12 120
Ihhr,
ACCJU
Er Br
C11 Gil ‘
’ \ ’1 13" \
VJ J \
n t‘ I' «;o f’ I
(It /\/ (If ‘- 'T b 1 ‘\ .J—jt /’
.ir hr
1.15%

Figure 36. Preparation of 1,2,h,S-tetrakis(brcmomethyl)benzene.

b8

   

.3

.3
_s
m
N

R=COOH
R=CON
. RzNCO

.0

MI

Figure 37. PrOposed conversion of tetrabromide 11§_to

diketone 11§_by a malonic ester synthesis.
ymethyl)hydroquinone 129, Treatment of 1g§_with HBr in acetic
acid gave tetrabromide 1_§, However, this material tends to
polymerize or oxidize to intractable material especially in
the presence of base, so methylation of 199 was impossible.
Thus it was thought to methylate the phenolic hydroxy groups
of 1g§ giving 12 , then prepare the tetrakis tosylate 199.
(Substituting bromide for the hydroxyl groups in 121 with HBr
would cleave the phenolic ethers and give 129). Unfortunately,

the only product isolated from several methylation attempts

on 125 was a small amount of trimethyl compound 1 2. It

89

should be noted that the preparation of 12 was not reported

 

 

 

 

OEi Oti
/L\ HO ; on os
l “ 2’ \ ‘F’\\T HBr [/1\ J
/ 7 ‘ . ,‘I’
ha/H J\,5J\ HOAc ‘7 1 ;
I ‘f 1 [/\\fik‘
OH HO on i ‘
On Dr on Br
12% 126
V ‘3
I182u04
‘anCO
v ‘¢ 3
Che
1‘
n0 ,c‘ TSO one ris
\ /\ \/ ' l\ ’\ g
Y! TsCl 17 \\I
/ 1 '
/\ / \\ ' /’J‘\
Iior Y "L V5 5 ‘ I] \f/ 1
‘ " r7150 OMS
\ve CH9 *
12” 123

 

Figure 38. Approach to oxygenated system 51,

in reference 6h even though the preparation of 13__from 131.
was reported, suggesting that there may be a special difficul-
ty in this case. The apparent lack of formation of 121_may

in fact simply be a failure to isolate it from aqueous solu-
tion, as evidenced by the high water solubility and low
organic solubility of the ana10gous 1,2,h,S-tetrakis(hydroxy-

methyl)benzene.

90

When 125 was treated with HCl in acetic acid, followed by
acetic anhydride in acetic acid with a catalytic amount of

sulfuric acid,6h there was isolated the tetrachloro diacetate in

 

 

0 he 0 M e
H OH
OH
OH a, o1 a
“he UMe H
OH
.119 ill.
.Egi
Figure 39. Poly(oxymethyl) derivatives of hydroquinone.
_ (3H
OH on Cl OACCl
liCl
ACCMI
Ofi Oz! 1
on Cl OAC L1
12.5 1.32 m

Figure hO. Preparation of tetrakis(chloromethyl)hydroquinone
diacetate _3},

91

h2% yield. Treatment of 133 with malonic ester and sodium
ethoxide led only to an oil which was apparentlv a mixture

of esters.

Another approach to the oxygenated system involves
annelation of two three-carbon moieties onto benzo-uinone
via two Diels-Alder reactions as outlined below. This is

’5 5

based on a paper by Diels and Alder themselves.

”—0

n-

fie

 

\ i ,
\tgo‘M‘L/n “ I
oca’ ;
OAC
136
Figure h1.

92

 

 

2)Zn,hgo
L.

   

 

Double Diels-Alder approach to the oxygenated

hydrindacene system.

93

Reaction of benzoquinone with cyclopentadiene (freshly cracked
from dicyc10pehtadiene) gave monoadduct 139. On heating 119
in refluxing acetic anhydride, the diadduct 1}§_was obtained
in 33%»overall yield, along with some of the by-product
hydroquinone diacetate. Initital attempts at ozonolysis cf
1}§_failed, perhaps due to attack of ozone on the electron-
rich aromatic ring. Were the desired tetraaldehyde 1}9_to
be isolated, it could conceivably be converted to tetraketone
111_via any one of a number of multi-step oxidations.

Another approach involves cyclization of some derivatives
of the dimethoxy diacid 11}, This route is ana10gous to that
explored earlier (Figure 31) for the same system without the

aromatic methoxy groups.

  

(I‘E'i‘? + (11".
. 7 ,. ' 1V1» 3\ \
thw», //\W 1 H
i; .31 V .m ‘ '."- 6", / x
. .,.. gyfi‘j‘i 1.1 w ‘w“
Ubi\i"€r) —_—-—-9 M80 Lt Vie!)
l. + «be
_ T10
LAM:
.131 13.2
i?
,, w me> ,,
1‘: _ N v.1-

, iMe
MOO $460 2

 

1.119.

Figure h2. PrOposed cyclization of bis(propionamide) to the
hydrindacene skeleton in the dimethoxy system.

9b

58 explored the

As was mentioned earlier, Dr. L. L. Klein
cyclization of these derivatives under acidic conditions,

and found that after the first cyclization the newly formed
ketone carbonyl deactivates the ring to further substitution.
In addition, we found (Figure 33) that O-methylated-N,N-
dimethyl amides were not powerful enough electrophiles to
effect aromatic substitution in the non-functionalized
system 119, However, it was of interest to investigate the
same reaction starting with 119 in which the aromatic ring
should be considerably more active. Again, the rationale
behind such a reaction is that, once the first cyclization

has occurred, the ring is not deactivated to a second

cyclization.

To this end 139 was synthesized via the following route.
The bis(chloromethyl) compound 1gg_was prepared in nearly
quantitative yield by the reaction of p-dimethoxybenzene
with excess parafOrmaldehyde and concentrated HCl. It was
initially isolated in an attempt to prepare the tetrakis
(chloromethyl) diether 191 which, despite all attempts, re-
mained elusive. The malonic ester synthesis and hydrolysis
proceeded in h5%»and 82%.yields respectively giving tetraester
1g}_and tetraacid 199, The decarboxylation to diacid 113_was
carried out in L6% yield by heating 1gg_to melting (220°C)
over a low flame until evolution of gas subsides. The bis—

amide 138 was prepared in nearly quantitative yield by reaction

95
of the acid chloride of acid 11} with aqueous dimethylamine.

Warming of 138 with methyl triflate in chloroform, follow~

ed by addition of water gave only the diester 1h5. In addition

0PM}
l I

 

   

 

1A1
-- 138: Xad‘ri

Figure L3. Preparation of dimethoxydiamide 38.

96

 

 

    
 

DUI3
OCNM82 11MeOng OOMe
OCNMe 2)H O '
2 2
OCH
1 8 3 ‘—

H2O SOC12

+ Me

Megh OMc
Cl— C1
OMe
1 6

Figure uh. Attempted cyclizations of diamide 38.

97
the diamide 199 was heated with thionyl chloride to give the
bis(chloroiminium) ion.199. Hydrolysis of the residue from
evaporation of the thionyl chloride gave a low yield of

starting material 13 as the only isolable product.

The failure of these cations to cyclize is predicted by

66

Fodor and Nagubandi. They claim that cyclization of imid-
ium chlorides occurs only through the nitrilium ion, as shown

in Figure LS.

A \‘\ . \
//’\\D -—9 I 1 7%?) ‘ ”\\r//\\\l
Cfi/ JH /C \f/Iifl / .NH
R R

\ /

Figure hS. Cyclization of imidium chlorides through nitrilium
ions.

98

As it is not possible to form a nitrilium ion from an N,N-
disubstituted amide unless an N-alkyl group is removed, the
cyclizations of these amides is not lilely. However, the
dinitrile 1. , upon alkylation should cyclize once, though a
second cyclization is problematical. Nitriles can be alkylated
by strong methylating agents such as trimethyloxonium tetra-

fluoroborate, methyl fluorosulfonate and methyl trifluoro-

 

methanesulfonate.
OMe V9 OMe
i N
‘ m
CN MeOT ‘ C \
N ‘_"_”’ C
C / m
H -+
0M6 OMe Me
OMe
Me—N
?
C O H c
M
N-Me § +
UMe OMe Me

Figure h6. Preposed cyclization of dinitrile 195 through the
nitrilium salt.

99

Though the second cyclization may seem unlikely due to the
electron withdrawing imino group (perhaps protonated in the
reaction mixture), there is precedence for nitrilium ions
attacking electron deficient aromatics. For example, the
electron deficient nitrilium ion 199 underwent cyclization.(
Thus the already prepared diacid 11}_was converted to the
diamide 199_via the acid chloride. The diamide was converted

to the dinitrile by dehydrating with P905°b9

N02 N‘

2
l)\
/’
11+ -—-—» t
\ )3 EV‘ 4’
Q J
I ..// K

Figure h7. Electrophilic substitution of a nitrilium salt on

an electron deficient aromatic.66

()CH

COOH
' OCfli

100

1)soc12 OCNH

 

(DCH

21Nn4oh’ CONH2

1A2
P297//’

1N

OCH3

191

Figure h8. Preparation of dinitrile “1.

Treatment of dinitrile 1h] with methyl triflate in

carbon tetrachloride gave a complex mixture of products which

was partially separated by column chromatography. All of the

fractions showed pmr absorbsnces in the aromatic region.

This suggests that, at most, one cyclization had occurred,

since double cyclization would require substitution at all

six aromatic carbons.

In a related deve10pment, it was discovered that some

70

Argentine chemists

claimed to have obtained the diketone

101

151_in 59% yield as the only isolated product of the reaction
of diacid 159 with PC]5 and A1C13. Although no experimental
details were given, it was also stated that, if 80012 and
AlCl3 were used, there was obtained besides the diketone 151,
a 22% yield of the keto acid.15_, In my hands, reaction of
the diacid §§_in a melt of PC15 and AlCl3 gave only a 29%
yield of the keto acid 191 along with an intractable tar.
Similarly, treatment of the diacid 111 with a melt of PCI

5

and AlCl3 gave no soluble products, perhaps due to cleavage

of the ether linkages followed by lactonization, oxidation

and/or polymerization.

 

)1 9 CH3 R
COOH [:J W <\rl/§‘i
9 COOH ,,L \7 \f‘
o
88: R==H M 191: R—H
133: RzCH3 ’2'" 12: R=CH3

- . . . f -f w mi...- "14-: .
Figure h9. denzene diprOpionLc doid ani derivatives.

102

1111(I- Hi If 1.1.1- 4 TA L

Instrumentationl solvents, etc. See experimental section
for Part I.

2,6-Bis(methylthiol-gi6-dichloro-143,5,7-hydrindg9ene
tetraone (75). A flame dried, 250 ml, 3-necked round bottom
flask under a nitrogen atmosphere was equipped with overhead
stirrer, addition funnel and condenser. It was charged with
1.5 g (0.063 mol) of sodium hydride. Through the addition
funnel was added to ml of dry DMSO (distilled from KDH and
stored over molecular sieves). After stirring for 15 minutes,
a solution of 6.2 g (0.02 mol) of tetramethyl pyromellitate in
no ml of dry DMSO was added. The mixture was stirred at room
temperature for 2h hours. The DMSO was removed under reduced
pressure. To the sticky red residue was added 200 ml of 15%
aq. hydrochloric acid. The initially homogeneous solution
slowly precipitated a yellow solid which was filtered out and
washed with water. It was dried in a vaccuum desiccator and
recrystallized from benzene. Mp 228-23100, yield 3.7 g (hQ%)-
Mass spec- m/e= 378 (7%). 376 (8%). 37h (10%). 31.1 (3%). 339 &7%).
317 (8%). 315 (12%). 145 (100%)-

2 -Bis carboetho -1 - drindacene tetraone dis—
odium salt,(§9). In a flame dried 250 ml reaction apparatus
(flask, condenser, addition funnel, N2 atmosphere), h.3 g

(0.18 mol) of THF-washed sodium hydride was added to 100 ml

103
of dry THF. To this mixture was added 10 g (0.032 mol) of
tetramethyl pyromellitate (which had been oven dried and
stored in a desiccator). The mixture was brought to reflux.
A solution of 7.0 g (0.08 mol) of ethyl acetate and 25 ml of
THF was added slowly while the reaction mixture continued to
reflux. After the addition was complete the reaction mixture
was stirred at reflux for 2h hours. After cooling the reac-
tion mixture was rotary evaporated and the brown residue was
dispersed in ethanol causing considerable foaming due to the
presence of excess sodium hydride. The resulting orange solid
was centrifuged out, diSpersed in water and recentrifuged.
The resulting red solid was dried 13_yggpg_to give 8.2 g

(6u%) of brick red solid. Mp 30000.

1,3,517-Hydrindacene tetraone(19). 5.0 g (12.h mmol) of
finely divided red salt (99) was dispersed in 50 ml of glacial
acetic acid and warmed on a steam bath. A small amount of
con. HCl was added by drOpper until the red suspension turned
light orange. Heating was continued as the mixture darkened,
evolved gas and became homOgeneous. Within a few minutes a
gray precipitate formed. The mixture was cooled in an ice
bath then centrifuged. The resulting silvery solid was washed
once with acetic acid and once with water then vaccuum dried
to give 1.3 g (50%) of solid. This material is quite insolu-
ble in most solvents. It can be sublimed (18500/1 torr).

Mp 300°C. Mass spec. m/e 21b (100%), 186 (12%), 172 (31%),

10h

158 (17%), 1th (27%), 130 (10%), 102(26%). IR: 1750 cm-1(w),
1710 (s). 13u5 (m). 1225 (m). 877 (s). 735 (w). 1355 (m).

216-Dibenzylidene-113,5,7-hydrindacene tetraone(81). The
above procedure was followed, except that an excess of benzal-
dehyde was added before heating. A green-yellow solid is
obtained. After washing with acetic acid and drying, it can
be recrystallized from anisole. Mp 330a332 dec. Mass spec.
m/e: 390 (80%). 389 (100700. 276 (16%). 1911 (1.2%). 129 (35%).
102 (3%).

2,6-Dinitro-1,31517—hydrindacene tetraone(§}). To a
mixture of 10 ml acetic anhydride and 15 ml glacial acetic
acid was added in one ml portions 5 m1 of con. nitric acid.
The temperature was kept below 60°C with an ice bath. After
the addition was complete the temperature was adjusted to 35°
and the mixture was poured over 2.6 g (12.1 mmol) of the gray
tetraketone 12, The mixture was warmed to 35° and maintained
for twenty minutes by alternate warming and cooling. It was
then cooled to 5°C and suction filtered. The dirty yellow

solid was rinsed with ether and dried to give 2.70 g (7u%).

2.6-Dibromo-2l6-dinitro—11345,7-hydrindacene tetraone(§g).
In a 50 m1 Erlenmeyer flask 2.63 g (8.22 mmol) of freshly
prepared5 pyridinium hydrobromide perbromide and 15 ml acetic
acid were heated on a steam bath until all the solid dissolved.
It was then poured rapidly into a solution of 2.5 g (8.22 mmol)

of dinitrotetraone §2 in 25 ml water. The resulting white

105
precipitate was suction filtered, dissolved in ether, dried
(NaZSOA) and evaporated to give 3.0 g (80%) of a white
crystalline solid.

Pyrolysis of 8b. In a large test tube 3.0 g of the white
solid §Q and 10 m1 of o-dichlorobenzene (bp 180°C) were heated
over a small flame for 3 minutes. The solid blackened and
after cooling was intractable.

p:Phenylene bis(3,3'-acrylic acid) (101). A mixture of
15 g (0.112 mol) terephthalaldehyde, 27.0 g (0.260 mol)
malonic acid, h2 ml pyridine and a trace of piperdine were
heated at 100‘3 for 5 hours. The cooled reaction product was
taken up in 300 ml of water and neutralized with 10% aq.

HCl. The resulting white solid was filtered, washed with
water, suspended in hot acetic acid and suction filtered.
Yield 20.0 g (82%). Mp 300°. IR: 1685 cm'1(s), 1630(m),
1330, 1320, 1290, 1230, 985, 950, 830. For lit. values see

Ref. Shb.

Diethyl pephenylene bis(3,3'-acrylate)(102). Method A.
To a solution prepared by adding 2.3 g (0.10 mol) of sodium
to 30 ml of absolute ethanol was added to a solution of h3 g
(0.10 mol) of carboethoxymethyltriphenylphosphonium bromide
in 70 ml of absolute ethanol. After the addition was compl-
ete the solution was allowed to stand for one hour. To the solu-
tion was added 6.7 g (0.05 mol) of terephthaladehyde along

with an additional 50 ml of ethanol. The solution was stir-

106

red for two days at room temperature. The solution was
filtered to remove a small amount of sodium bromide, then
rotary evaporated. The resulting sticky residue was digested
with low boiling ligroin and filtered. This process was
repeated several times until the white solid showed no ethyl
signals in the nmr. The combined ligroin extracts were rotary
evaporated, giving a white semi-solid which is apparently a
mixture of the desired product and triphenylphOSphine oxide.
They can be separated by chromatography on alumina with

chloroform.

Method 3. A solution of 15 g of p-phenylene bis(3,3'-
acrylic acid was heated at reflux with 100 ml of absolute
ethanol in 1 ml of concentrated sulfuric acid. After 2h hours
the mixture had become homogeneous (the starting diacid is
quite insoluble in ethanol). The hot solution was filtered to
remove a small amount of insoluble material. Upon cooling, it
deposited beautiful white needles. Mp 92-930; lit.5ua9S-96°C.
Concentration of the mother liquor led to additional product.
Total yield 17.2 g (91%). IR: 1717 om'1, 16h0, 1330, 1310,
1290, 1210, 1190, 1030, 1000, 880, 835. Pmr:$1.31,t(J=7Hz),
6H;$u.22,q(J=7Hz),hH;$6.39,d(J=1682),2H;f7.h7,s,hs;37.6o,d
(J=8Hz),2H.

piethyl p—phenylene bis(3,3'-progionate)(81). A.mixture
of 10 g (37 mmol) of diethyl p-phenylene bis(3s3'-acrylate)

and 0.1 g of 9% palladium on charcoal in 100 m1 of ethanol

107

was hydrogenated in a Parr hydrogenator at room temperature
for 12 hours. Though the starting material was not completely
dissolved, the product was. The solution was filtered to
remove the catalyst, and the solvent was removed giving 10.1 g
(100%) of white solid, mp 69-7000, lit.71 69°. Pmr: 51.19.t
(J=7Hz),6H; 32.69,m(ae'hh'),8h; 5'h.0h,q(J=7Hz),hH; g‘7.00,s,hH
p:Phenylene bis(3.3'1pr9pionic acid)(8§). A solution of
10 g (36 mmol) of diester 82 in 100 m1 of ethanol was mixed
with a solution of h.S g (80 mmol) of potassium hydroxide in
50 ml of water. The mixture was heated at reflux for 15 hours.
After cooling, the ethanol was removed by rotary evaporation.
The aqueous solution was made acidic with 10% aqueous HCl.
The resulting white precipitate was filtered out and washed
thoroughly with water. Recrystallization from ethanol gives
7.9 g (99%) of white needles. Mp 225-22700; lit.71 223°C.
Isophthalaldehyde.SS In a two liter, three necked round
bottom flask equipped with overhead stirrer, addition funnel,
and thermometer extending to the bottom of the flask were
placed 29 ml (25 g, 0.21. mol) of m-Jq'lene, 928 m1 (1 kg)
acetic anhydride and 382 ml (h00 g) of glacial acetic acid.
The solution was cooled in an ice bath and 83 ml (150 g)
of concentrated sulfuric acid was added through the dropping
funnel at such a rate that the temperature stayed between S
and 10°C. Over a period of 1% hours, 90 g (0.90 mol) of

chromium trioxide was added in small quantities so that the

108

temperature did not rise above 100. After the additiom \ 3
complete, stirring was continued for 3 hours at 50. The
green reaction mixture was poured into two liters of crushl
ice and extracted three times with 250 ml portions of chlori-
form to give two liters of light green solution. It was
concentrated to half its volume under hOO, then steam distil-
led until three liters of distillate were collected. This
was extracted with chloroform, which was dried and rotary
evaporated to give a yellow oil from which crystallized h.h g
of white needles. An additional 2.7 g could be obtained from
the oil on recrystallization from ethanol/water. Total yield
7.1 g (22%). Mp 85w8900; lit.SS 890. Small amounts of 3-
formyl benzoic acid and iSOphthalic acid could be isolated

from the steam distillation still pot.

m-Phenylene bis(3,3'-ac;ylic acid)56(lgé). A mixture of
2.2 g (16.8 mmol) of isophthalaldehyle, 3-h g (33.0 mmol) of
malonic acid and 10 ml of pyridine containing a trace of
piperidine was heated at 1000C overnight. After cooling the
solution was poured into water and made acidic with 10% aq.
HCl. The resulting white solid was filtered out and recrys-
tallized from acetic acid to give 2.9 g (80%) of white solid.

56

Mp 281-28h°c; lit. 280°C.

m:§heny1ene bis(3,3'-prqpionic acid)(2Q). To a mixture
of 2.25 g (10.3 mmol) of m-phenylene bis(3,3'-acrylic acid)

and 15 ml of water was added enough con. ammonium hydroxide

109

to dissolve the solid. To this solution was added 0.05 g of
5% Pd on charcoal. The mixture was shaken under hydrogen at
room temperature for 1% hours (until no more hydrogen was
taken up). The catalyst was filtered out and the solution
was made acidic with 10% hCl. After a moment a flaky white
solid began to appear. After an hour this solid was filtered
out, washed with water and dried giving 1.9 g (85%). Mp 1h8-

56 1h6-1h7°c.

150°C, lit.
6:(2~Carboxyethyl)31-indandone(108). Method A. A mixture
of h.5 g (16.2 mmol) of p—diester 85 and 50 ml of polyphos-
phoric acid was warmed with manual stirring on a steam bath
until the mixture became homogeneous and darkened considerably.
After cooling, the syrupy mixture was poured into water and
stirred until the PPA was completely dissolved. The aqueous
solution was extracted thoroughly with chloroform. The organ-
ic solution was dried (Nazsoh), filtered and evaporated to
give a yellow solid. Fractional crystallization (ethanol/
water) gave 2.5 g (76%) of 192_(mp 1h0-1h2.5°c) and 0.57 g
(13%) of oily 198, Nmr of 192; 32.5-3.3,m,83; S7.2—7.75,m,3H
Nmr of 1985 S1.19,t(J=3.5Hz),3H;§§2.5-3.2,m,8H; gh.03,q(J=3.SHz).
2H; g6.9-7.7,m,3H.
Method B. A mixture of 3.5 g (15.8 mmol) of p-phenylene
bis(3,3'—pr0pionic acid) and 15 ml of thionyl chloride
(freshly distilled from quinoline) was heated on a steam bath
under reflux. After one hour the solid had dissolved and no

fUrther gas was evolved. The mixture was allowed to cool,

110

and the thionyl chloride was removed under reduced pressure.
The resulting yellow solid was dissolved in carbon disulfide,
cooled and stirred while 5.0 g (hO mmol) of aluminum chloride
was added all at once. A c0pious quantity of HCl was released
and the mixture turned red and lumpy. After standing over-
night it was refluxed for an hour. Ice water was added slow-
ly until all the aluminum chloride was hydrolyzed. The mix-
ture was extracted thoroughly with chloroform. The combined
extracts were washed with 9% sodium bicarbonate, dried (NaZSOh),
filtered and evaporated to give a dark solid. This was re-
crystallized from ethanol/water to give 3.0 g (95%) of 102

(mp 139—1h200). Nmr: see above.

Attempted cycligation of 192 with PPA. Warming of 1.0 g

 

of 102 with 15 ml of polyphosphoric acid overnight under N2,
followed by hydrolysis and work-up as above gave quantitative

recovery of starting material.

53(2-carboxyethyl)-l-indanone(109) and 71(2-carboxyethyl)-
1-indanone(110). A mixture of 1.0 g (h.5 mmol) of m—phenylene
bis(3,3'-propionic acid) and 20 ml of polyphosphoric acid was
heated for two hours on a steam bath, then allowed to stand
overnight. Pouring the mixture into ice-cooled water gave 0.6 g
(67%) of white solid (199). Mp 165-166.5°C (lit.S3 165-166)
from acetone. The mother liquor was extracted with chloro—

form which was dried, filtered and evaporated to give 0.3 g

(33%) of a mixture of 109 and 110. hp 93-100, 1h0-150°c rom

111

ethanol/water. Nmr of 199; 32.5-2.9,m,hH; 52.9-3.5,m,hH;; 7.32,
dd(J:8.75,1.25Hz),1H;:57.h5,d(J-1.25Hz),1H;€37.57,d(Jz8.75Hz),
1H. (250 HMz).

142,h,5-Tetrakis(hydroxymethyl)benzene(lgg). The thimble
of a dried Soxhlet apparatus under nitrogen was charged with
5.0 g (16.1 mmol) of tetramethyl pyromellitate. A slurry of
5.0 g (132 mmol) of lithium aluminum hydride and 100 ml of
ether was boiled in the flask. After all the ester was dis~
solved and the reaction mixture cooled, the excess LAB was
destroyed with saturated ammonium sulfate. After the ether
was removed, the white residue was boiled in 200 ml of water.
The hot solution was filtered and on cooling deposited 0.6 g
(195) of white needlts; mp 132-19300 (lit.61 13300). Concen-
tration of the mother liquor gave an additional 0.9 g (28%).

1,2,h,52Tetrakis(bromomethyl)benzene(118). Method A.
A mixture of 3.1 g (15.6 mmol) of 1,2,h,5~tetrakis(hydroxy-
methyl)benzene and 100 ml of glacial acetic acid was stirred
while gaseous hydrogen bromide was bubbled into the mixture.
The temperature increased, the solid slowly dissolved and the
solution became greenish. 0n cooling a light green solid
crystallized and was filtered, washed with water and air dried.
Yield 6.1 g (85%). Mp 157-159°c recrystallized from acetone.

Concentration of the acetic acid solution gave an additional

0.7 s (10%)-

112

Method B. A mixture of 22.8 g (0.170 mol) of finely
ground durene, 125 g (0.702 mol) of N-bromosuccinimide, 1.5 g
(6.0 mmol) of benzoyl peroxide and 100 ml of carbon tetra—
chloride was heated at reflux with mechanical stirring.

After 30 minutes a vigorous reaction commenced causing con—
siderable foaming. After the reaction appeared to have
subsided, the mixture was stirred at reflux for an additional
two hours. The hot mixture was filtered and the white solid
was rinsed with hot carbon tetrachloride. After standing
overnight the combined organic wasnes had deposited 12.5 g
of large plates (mp 139-15hOC) Concentration of the mother
liquor led to an additional 5.9 g. Crude yield 18.h g (25%).
Crystallization from acetone gave 15 g (mp 1h8-152; lit.6
160°C). Nmr: Sh.h0,s,8H; $7.33,s,2H.
2,2',6,6'-Tetrakis(carboethoxylhydrindacene(121). Under an
N2 atmosphersodried 3-necked round bottom flask was equipped
with magnetic stirring bar, addition funnel and condenser.
Enough ethanol was added to 1.0 g (h2 mmol) of sodium metal
in 50 ml of dry THF to react with all the sodium. To this
solution was added 9.0 g (56 mmol) of diethyl malonate followed
as quickly as possible by 6.0 g (13 mmol) of 1,2,h,5—tetrakis
(bromomethyl) benzene (prepared by Method A above) and 50 ml
THF. The reaction mixture was heated at reflux for ten hours.
After cooling most of the solvent was removed by rotary evapor-

ation. Water was added and the waxy white solid was centri—

113

fuged out. After rinsing with water and drying'it was recrystal-
lized from ethanol, giving in two crOps h.7 g(80fl). Mp 159.5-
160.5°c. Nmr: 51.22,t,12H; 83.hh,s,8H;<Sh.12,q,8H, 56.82,s,2H.
Mass spec. m/e: the (9%), 372 (30%), 298 (110%), 153 (5&% .

2,2',6,6'-Tetracarboxyhydrindacene(122). A mixture of
h.0 g (8.96 mmol) of tetraester 1gp, h.0 g (72.h mmol) of
potassium hydroxide, 50 ml of water and enough ethanol to
dissolve the solid was heated at reflux for 20 hours. After
cooling the homogeneous solution was made acidic with 10%
HCl. When precipitation of the white solid was complete, it
was filtered, washed with acetone and dried to give 2.9 g
(95%) of tetraacid 12g.(mp32000). Mass spec. m/e: 2&6 (39%),
200 (148%). 155 (10025). in (947123).

2,3,5,6-Tetrakis(hydroxymethyl)hydrqguinone(12S). A
mixture of 22.0 g (0.20 mol) of hydroquinone and 20 ml of
water was purged with N2 gas. To the mixture was added 80 ml
of NZ-purged 10% aqueous sodium hydroxide (0.20 mol). Under
nitrigen, 3&9 g of freshly Opened aqueous formaldehyde was
slowly added. The homogeneous solution was allowed to stand
under nitrogen at room temperature for 3 days. It was
acidified with Né-purged 10% aqueous hydrochloric acid and
cooled in an ice bath fOr several hours. The light brown
solid was filtered out, washed thoroughly with water and

vaccuum dried to give 32 g (70%) of whitish solid. Mp 200-

11h
210°C in a preheated oil bath, 1it.6h 210-22.

243,5,6—Tetrakisjphloromethyl)—1,hediacetoxybenzene (131).
Excess hydrogen chloride was bubbled into a mixture of 10 g
(243 mmol) of hexaol _1_5_ and 100 m1 of glacial acetic acid.

The solution became warm and darkened as the solid dissolved.
After standing overnight, the solvent was removed under reduced
pressure. To the dark residue was added a solution of 15 ml
of acetic anhydride, 2 ml of concentrated sulfuric acid and

100 ml of glacial acetic acid. The solution was stirred
overnight, then diluted with water and immediately centrifuged.
The light green solid was washed with water and vaccuum dried.
Recrystallization from ethyl acetate gave 7.1 g (h2%) of the
tetrachlorodiacetate. Mp 217.5—220.50C. Nmr: 52.h1,s,6H;
Sh.5h,s,6H. Mass spec. m/e; 390 (0.173), 380 (0.1%), 3L.8 (1%),
31:6 (3%). 32m (2%). 308 (379), 306 (1273). 3% (25:12). 302 (2135).
is (10%).

Cyclopentadiene-benzoquinone adduct(l}§). Cracking of
dicyclOpentadiene was carried out by heating and distilling
the resulting cyclOpentadiene through a column of glass hel-
ices, bp hO-SOOC. To an ice-cooled slurry of 20.3 g (0.172
mol) of benzoquinone in 200 ml of absolute ethanol was added
drapwise 11.h g (0.172 mol) of freshly cracked cy010pentadiene.
After the addition was complete the mixture was allowed to
warm with stirring until the solution became dark red-brown.

The ethanol was distilled off under aspirator pressure. The

115

dark residue was extracted with hot hexane until the hexane

appeared very light yellow. The combined hexane extracts

were hot filtered and concentrated. 0n cooling 16.7 g (53%)

i .0 . 5b 0
of yellow needles formed. Mp lb C; lit. 71-73 C. Nmr:
£10h‘106,m2fl; 53005-3025,:n2fi;5‘301‘4“305,m25; 3"6.0--6.1,m,2H
56.h5,s,2H. For lit. nmr see Ref. 65b.

1,h,5,8-Dimethano-1,h45187tetrahydroe911O-diacetocy—

anthracene(135). A mixture of cyclOpentadiene-benzoquinone

 

adduct (8.5 g, h5.6 mmol) and h5 ml of acetic anhydride was
heated at reflux for two hours then allowed to cool in an ice
bath. From the solution crystallised h.i g (62%) of an
orange-white product. It was filtered out, rinsed with ether
and recrystallized from ethyl acetate (prisms, mp 222-22hoC,
lit.658 250°C). The by-product, 1,h«diacetoxybenzene, can
be isolated from the acetic anhydride mother liquor. Nmr of
132: 82-17.m1.H; SIB-27.8.68;(33.70.111.131; warms. Mass Spec.
m/e: 322 (has), 280 (2372;), 238 (10031,).
2,5-Bis(chloromethyl)-1,hedimethoxybenzene(1h2). A
mixture of b.5 g (32.6 mmol) of 1,h-diuethoxybenzene, 25 g
(0.83 mol) of paraformaldehyde, 80 ml of concentrated hydro-
chloric acid and 100 ml of glacial acetic acid was heated on
a steam bath for 2h hours. The hot mixture was filtered and
after cooling, the mixture was poured over ice, the solid

filtered out, washed with water and vacuum dried to give

6.7h g (88%) of yellowish product. Recrystallization from

116

72 167.50C. Nmr:

acetone gave needles, mp 163.5-1oh.5, lit.
5‘3.79,s,6H;£§h.55,s,hH;116.79,s,2h. Mass spec. m/ez238 (ME),
236 (211%). 23b (38%). 221 (14%, 219 (6%). 201 (32360. 199 (100%).
13b (111:6 .
1,h-Dimethoxy-ZLS—bis(2A2-dicarboethoxyethyllbenzene(lgj).
In a dried apparatus under nitrogen, 0.7 g (30 mmol) of sodium
was dissolved in a minimum of absolute ethanol in 150 ml of
dry THF. To the sodium ethox1de was added quickly h.8 g
(30 mmol) of diethyl malonate followed immediately by 3.3 g
(1h.1 mmol) of 2,5-bis(chloromethyl)-Eh-dimethoxybenzene(1&2).
The reaction mixture was heated at reflux for 20 hours. After
cooling, the solvent was evaporated, the residue taken up in
water and the resulting yellow solid was filtered and washed
with water. After drying, it was recrystallized from petro-
leum ether to give 3.0 g (h5fi) of white needles (mp 99.5-
100°c). Nmr: 51.17,t(J.—.3Hz),12ii;S3.08,d(J=Lmz).hH; 53.57-
3.77, m,2H; $3.70,86H;§h.0h,q(J=3Hz),8H; 56.5h,s,2H.
1LQrDimethoxy—245-bis(2,2-dicarboxyethyl)benzene(lgg).
A mixture of 2.9 g (6.0 mmol) of tetraester 1g3 and 1.h g
(25 mmol) of potassium hydroxide in 10 ml of water was diluted
with enough ethanol to dissolve the solid. The mixture was
stirred at reflux overnight. After cooling to room tempera-
ture the solution was poured into 5% aqueous hydrochloric
acid, then stored in the refrigerator overnight. The re-

sulting white solid was filtered out and air dried giving

117

1.8 g (82%) of off—white powder. An additional 0.h g (18%)
can be obtained by concentrating and chilling the mother
liquor. Mp 2200C dec, resulting material mp 189-19200; lit. 3
226 and 197-199.5.

1,h-Bis(2-carbomethoxyethyl)benzene(116). To a solution
of 0.60 g (2.2 mmol) of N,N,N‘,N'-tetramethyl-1,h-bis(2-
carbamidoethyl)benzene, 29, in 5 m1 of chloroform was added
0.50 ml (0.72 g, h.h mmol) of methyl trifluoromethanesulfonate.
The homOgeneous solution was heated at reflux for 3 hours
after which time a dark oil had separated. To the mixture
was added 5 ml of water and stirring was continued for 2 hours.
The organic layer was separated and the aqueous layer extracted
twice more with 5 ml portions of methylene chloride. The

combined organic layers were dried (MgSO filtered and

14),
evaporated to give an orange oil which eventually solidified.
Recrystallization from methanol gave 0.35 g (6hfl) of white

7h ,0 ,
115 C). Mass spec. m/e: 250

needles, mp 110-11hOC (lit.
(32960. 190 (5573). 176 (38%). 13o (362:3), 117003;). Fmr:
32.5-2.9,m,8H; $3.60,86H;-5 7.0l,s,hh.
JJgggimethoxy-2,5-bis(2~carboxyethyljbenzene(111), In
a small round bottom flask tOpped with an uncooled condenser,
h.0 g (11 mmol) of the his malonic acid 1gp was heated to
2h0°C in an oil bath and the resulting melt was allowed to

stand at 2h0°C for 15 minutes. After cooling to room tem-

perature, the brown SOlid was recrystallized from acetonitrile

118

to give 2.7 g (87%) of white needles, mp 195-19700 (lit.]3

197499.500). Mass spec. m/e; 28?. (10(55), 223 (517%). Pmr:
S2.5-2.8,m,8H; S3.73,s,6H; é‘6.73,s,2H.
1,h—Dimethoxy-2,5-bis(2»dfmethylaminocarboryletgyl)ben~
mug). A solution of 0.9 g (3.2 mmol) of the diacid 113
in 5 ml of thionyl chloride was heated at reflux for 3 hours.
The excess thionyl chloride was removed under reduced pressure
and the resulting solid was dissolved in 10 ml of methylene
chloride and added with stirring to an ice-chilled mixture of
5 ml of 25% aqueous dimethyl amine and 5 ml of methylene
chloride. After stirring overnight at room temperature, the
organic layer was separated and the aqueous layer was extract-
ed twice more with 10 m1 portions of methylene chloride. The
combined organic layers were washed with 10 ml portions of
10% NaOH and water, then dried, filtered and evaporated to
give 1.0 g (91%) of crude solid. Recrystallization from
acetone gives plates, mp 126.5—12700. Mass spec. m/e: 336
(607.), 177 (67%), 72 (10073). Pmr: 32.h-2.9,m,8H;S2.95,s,12E;
33.71,s,6H;.S6.62,s,2H.
1LL-Dimethoxy:2,5gbisj2-carbamidoethyllbenzene(lgg).
A mixture of 2.70 g of the diacid 113_and 50 ml of thionyl
chloride was warmed for h hours. The thionyl chloride was
removed under vaccuum. The residue was dissolved in toluene,
which was in turn removed under reduced pressure. The result-

ing yellow solid was dissolved in 100 ml of ether and to it

119

was added slowly with stirring 12 m1 of concentrated ammonium
hydroxide. A yellow precipitate formed. After 15 minutes
the ether was rotary evaporated and the yellow residue was
washed with water and suction filtered until dry. Recrystal-
lization from ethanol gave 1.75 g (66%) of yellow powder, mp

6 o
9 225-228 C. Mass spec.

225-2270C (turns red on melting), lit.
m/e: 280 (100%). 21.7 (5%). 235 (22%). 222 (11%). 177 (25%).
1,h-Dimethoxy—2,5-bis(2-cyanoethy1)benzene(1h7). A
mixture of 1.50 g (5.36 mmol) of the diamide 1 , 1.0 g
(7.07) mmol) of phosphorus pentoxide and 75 ml of mesitylene
was heated at reflux for two hours. The hot yellow super-
natant was decanted from a black residue. 0n cooling the
mesitylene solution deposited 0.50 g (38%) of yellow leaves
(mp 152.5-15h.5, lit.69 155—156). The mother liquor was con-
centrated under vaccuum to half volume and chilled to give an
additional 0.25 g (19%). Total yield 57% crystallized from
ethanol, mp 15L.-155°0. Pmr: 8‘2.l.-2.9,m,8n; €3.77,s,611;86.63,s,
2H. Mass spec. m c: 21.8 (895), 22-) (693), 2014 (10095.), 189 (129;).
FuSion of 1,h-bis(2-carboxyethyl)benzene(88) in PCl Z
Alglj. A mixture of 0.90 g of the diacid (h.1 mmol), 2.0 g of
phosphorus pentacnloride (10 mmol) and 2.0 g (15 mmol) of
aluminum chloride was heated slowly in an oil bath to the
melting point and allowed to stand for 15 minutes. After
cooling the dark solid mass was treated carefully with water.
After the hydrolysis was complete, a quantity of black tar

remained. The mixture was extracted thoroughly with methylene

120

chloride. The organic solution was then dried, filtered and
evaporated to give 0.25 g (2(6) of a dark oil which by nmr
consisted almost exclusively of 6-(2-carboxyethyl)-1-indanone

(191) -

Fusion of 1,hrdimethoxy—ZJS-bis(2-carboxyethyl)benzene

 

(113) in PCISZA1C13. The same procedure as above was carried
out starting with 1.20 g (p.25 mmol) of the dimethory diacid,
3.0 g (15 mmol) of phosphorus pentachloride and 2.0 g (15
mmol) of aluminum trichloride. The resultant methylene

chloride extract gave only a trace of unidentifiable material.

Part III

REACTIONS OF
moronic POLY(N,N-DD1E.THYL memes)
WITH ELLCTRCPHILBS

121

INTRODUCTION

Our investigation of N,N-dimethyl amides 22_and 13§_as
well as nitrile 1 , specifically their reactions with meth-
ylating agents and other electrophiles, led us to investigate
other examples. Several aromatic poly(N,N-dimethyl)amides
have interesting chemical and physical properties. For
example, all the amides shown in Figure 50 show appreciable
water solubility, but those with amide groups ortho to one
another} 15g and 155, are very soluble in water as well as

most other organic solvents.

CONMe2 CONMe2 Me2 NCO :::: :% CONMe2
[:::]/ [:::]:CONMe: Me2 NCO CONMe2

ONMe2

CONM82

CONMe2

CONMe2

15!; .151

Figure 50. Aromatic poly(N,N-dimethyl)amides.

It was of interest to us to explore the reactions of
diamide 12Q_with electrophiles to determine to what extent,
if any, the neighboring amide group enters into the reaction
of the other. For example, with methylating agents several
structures are pOSSible, listed in Figure 51. It was of
interest to us to determine which of these, or others, is the

correct structure of the salt.

In addition, we desired to investigate the reactions of
15g with chlorine electrophiles such as thionyl chloride,
phosphorus oxychloride and oxalyl chloride. The structure
of the reaction product could be one of the following or
some other. For both types of reaction the salts can be
characterized by 1H-- and 1 Cnnmr, as well as by reaction
with nucleOphiles to give isolable neutral molecules.

The ready availability of amides _§}J l§§_and 121 makes
their use as model systems for the above reactions possible.
In addition, reaction of the tetraamide l§§_with the above
reagents could give insight into the effect of a second iden-

tical reaction center in the same molecule.

It was also of interest to investigate the products, if
any, of reaction of the salts of'lfig with aromatics such as
1,h—dimethoxybenzene and N-methyl pyrrole. It was our intent
to determine whether acylation would occur, and, if so,

whether one or both amide groups would react.

12).;

Also, it was thought that salts of tetraamide 155_could
serve as an entry into the polyketo hydrindacene system,
explored in Part II of this thesis, by reaction with masked
carbonyl anion equivalents such as malonic ester or 1,3-

dithiane anions.

125

Me\N /Me Me :h/Ie
+

 

15.1;
or
.4.
- Me N
X MeBN OHIe Z/YOI‘v‘Ie2
\ ..
O or I O X
/
+
._ + - N\
X Me/N‘Me X Me/ M8

Figure 51. Possible products of methylation of N,N,N',N'-
tetramethylphthalamide 15h.

126

 

-Me\ /Me
C1 + N
Cl
Me\ /Me (:1
N * r
\0 Ma/ \Me We/ \Me
”O I or
z’N‘x
Me Me Me2N 1
O
(:1 "I,N\Me

 

Figure 52. Possible products of reaction of 15h with chlorine
electrophiles.

OMe

(DMe

/’ ' \

\\ l //

(NWe
CONMCQ + /////////’g 0M8
L E
I I
WT/ N
0 Me

-————————)
CONMB2 \\\\\\\\\$ 0
/\
{a} Gib

Figure 53. Possible products of reaction of salts of 15;,
with aromatics.

127

MeZNCO CONMe2
155
MeZNCO CONMe2
MeUTf '
OI‘
80012 Me2

(lumen)2“,b FtOCO

EtOCO

 

<3

N

Me2

 

Figure Sh. Amide 155 as an entry into the polyketohydrindacene
system.

128

RESULTS AND DISCUSSION

Preparation of the amides. The ortho, meta and para
diamides 1 , 159 and 151, as well as N,N-dimethyl benzamide
l5j_were prepared by the reaction of the corresponding acid
chlorides with excess 25% aqueous dimethylamine. (As these
amides were quite soluble in water it was necessary to extract
them from the aqueous reaction mixture with c0pious quantities
of methylene chloride followed by back extracting with a
minimum of aqueous sodium hydroxide to remove a small amount
of dimethylamine hydrochloride which was inevitably extracted
into the organic layer). The tetraamide l55_was prepared in
one step, albeit in fairly poor yields (22%) irom pyromellitic
acid by reaction with dimethylformamide and phosphorus pent-

75

oxide. Alternatively, reaction of pyromellitic anhydride
with dimethylamine gave a nearly quantitative yield of the
diacid-diamide (15§). Treatment of this material with thionyl
chloride followed by dimethylamine gave the desired tetraamide
in 55%1yield. This is shown in Figure 55.

Reactions g£_15hwith electrophiles. The reactions of
‘153 with various electrophiles, P0013, (COCl)2, (CF300)20,

CH 080 CF and (CH3)2$Oh, both neat and in solution were

3 2 3

monitored by proton and carbon-13 nmr. The results are given

in Table h.

129

 

 

 

HOOC CONMe2
Me2NH
—-————-—-.»
Me2NCO COOH
15h
1)80012
2)Me2NH
t
HOOC -\\ (xxui MezNCU CONM82
P205
// -—~———+ ;
HOOC cooa DMF mech CORMez

155

Figure 55. Preparation of l,?,h,5-tetrakis(N,N-dimethylnmino-
carbony1)benzeno, 155.

13o

mma
mm
ma

mmw

4

Q

.—
(\l

D

u

mpmm o>opw mmm

m .mq.m
.DD.wm «moNIJ.N
m pom

. m up .mm.m

H

m: ..pp,mmunmm.a

4 pmm

mm . m up .aw.m

mo .

m up .oo.m

m3 ..Dp.mm .MN.>

adnpommm

.hmoomonpommm

AmHQQMam xmms hhm>

m>HommHemn
soap .mQSp msammOMo
Show mflwpmhuo zoaamm

auruoum an we
auruonm a: om
sumumuoumua an we
muqumue we a
a" umuw gas on

h.co « Gas or

mummmmm o

pmm mm m>Hommfl©mH
:mnp .mmmmonocfi
m pom mm Hammad
mampmhhc soaamm

soapsflom
msomswmoaoz

 

mesmeaoo

 

 

mauom .p.u maoom
maomo .p.n maoom
mHomo .p.H mac:
mawzm

psm>aom macauwucoo noupomfim

Ham an umHOpflnoa mm mmaanmonpooaw :pfi3 mmw Ho msOflpowmm

.4 manna

131

mm .m an .m
mm .m up .5

 

m3 . .Dn.mm . wmé.
mm? .m up.hoo.m
a pmm

ma ..pn.sm .mm.e
mm? .m .mm.m
m.smm

 

 

mace o Hausa
aumumnosmue
mas om

mdomsmMOBon
.wm>do>m mmm mo
undead mSOHQoo

pampmcoo msamswu
new Hum seep
muqumua
"anamapaca
hasoam show
mmflwmwe mean:
Um>ao>m mam mo
unseen Hamnm

nowadaom
maomcmmoson

 

mpsmasoo

A.e.»eoov a magma

m
mooo ho

fi
20(90

m

ZU no

uwfiom onHmh

pgmaa .mmm:

nhao o» cwpmno
Imm>o .mHS am

ma

omo SH .p.H

.mamp Boon as

meo a
made

a
dmwmm

 

psm>aom

H nm>HommHu
ammo zoaamh
owe .mnn am
one ca .».H

 

macapaeeoo

mnaoooV

NAHOOOV

MES
maamm

tenuomam

132

a .o.m1m.e
m .Wwom
m .am.m
m an .sr.m
a .equ

m umm

m4 ..Qfl.mm MN.©
mm .m up ~0.m
mm .m up .om.m
.w.mmw

ma ..np.ms im~.e
mm .m up .om.m
mm .m an .wm.m

m: . m .om.»
mm? .m Hp.hmmum

o pmm

mF am fincm
mm . a .m-e
mma .up .0 m

w name we
a mass m
mumue made 0
haze «
m5 is 2
mdomCmmoaon

Nu
r»
M“

anomQQMoaon

hflco o mhme m
Prue“, Ame H
qua”? an a
mum"? an m
mumumnoumue gas or

mfi3oam Show
mmacmm: wears
.mcaanpsp 0:

 

mpsmaaoo

Zomno

 

pam>aom

A.e.p:oov a magma

Op.“

.9.“

cwaom

SOHHmh .mpmno
|Qm>m .mhn :N

smog .maomo

 

meoapaecou

ommfimmov

onommov

mfiaoouv
mflanm
nonpomam

 

133

m4 ..pn.mm .me.~
mm . m .mo.a

mm. .
mm .m HM .mmum
mmmmmmmw

ms
omcmmoeon
m
Homo

 

unease
o
psm>fiom

A.u.vsoov J maps
8

OP.“

 

 

escapaecoo

m
momomOmmo

 

maasaonpomam

13h

Thonnt no clear pattern emerces from tiese data (except that
these reactions are more complicated the expected), several iten;
can be noted.

1) On treatment with phosohorus oxychloridc the pair of sicnals
normally arisinc from the I-mcthyl groups coalesccs into a singlet.
This coalescence may he due to catalysis of C-h bond rotation by
traces of proton. In any case this phenomenon was observed with all
amides studied in thionyl chloride and phOSphorus oxychloride. The
reaction with phosphorus oxychloride obviously goes through an in-
soluble intermediate in which the four methyl groups appear to be
identical and the aromatic rorion symmetrical. The soluble final
product is unsymmetrical (note the aromatic rerion).

2) The product of reaction with oxalvl chloride is apparently
hydrolyzed by traces of water in triflucroacetic acid as dimethyl

ammonium ion is apparent in the spectrum. (The same spectrum aris s

 

from diamide 19h in TF1.) The produ,t in CDqCN oiffers depending on
whether or not the reaction mixt re is heated, hut the rain set of
peaks is the same in both cases: a broad sinrlct for the methyls
and an apparent sin'lct for the aromatic protons.

3) The diamide 152 does not react at all with trifluoroacetic
anhydride.

h) 't reacts only very slowly with dimethyl sulfate and then gives

a com lex mixture, perhaps due in part to the presence of a srnll

amount of methyl hydrogen sulfate.

135
S) The diamide appears to add one methyl group immediately
(and then remains constant) on reaction with methyl trifluoro—
methanesulfonate, although the aromatic region appears symme-

trical.

Reactions of 15h and related amides with methyl triflate.
As the spectrum of lip with methyl triflate appears to be the
most straightforward, the reactions of this electriphile will
be taken up first. Later, we will explore the reactions of
thionyl chloride and, to a lesser extent, phosphorus oxychloride.
The reactions of aromatic poly(N,N-dimethyl amides) with
methylating agents, can be conveniently followed by 1H- and

13C-nmr, as well as by the known hydrolysis of O—alkylated amides

’7

6
to give esters.‘

O pMe Me
R—</ M———-—9€OTf R .g\ 3129—, R \
NM82 +NMe2 0

Figure 56. Preparation of esters from amides by alkylation/
hydrolysis.

136

Treatment of N,N-dimethy1benzamide with methyl trifluorometh-
anesulfonate clearly gives the salt 122 by pmr (UDClS):fSY.65,
br s,SH(ArH); 53.97,s,3H(OIvle); 3'3.53,s3H(:me); é’.3.23,s,3}{(1me);
and, by cmr: Sl?h.6(ArCON); 5133.5, 8130.1, 5127.9, 612h.O(Arom.

C's); 862.6(Ui‘le); 5'u2.8(m~1e); S39.h(2fi~ie).

 

O OMe
l
- \sv/Me
lee2 + MeOTf —+ N: + Tro"
Me
159

Figure 57. Reaction of N,N-dimcthylbenzamide with methyl
triflate.
Treatment of the salt in chloroform with water gives quanti-

tatively methyl benzoate.

In cloroform, N,N,N',N'-tetramethyl iSOphthalamide reacts
very rapidly with excess methyl triflate to give the dimethyl-
ated salt léQ, This salt immediately oils out of the chloroform
solution, but is soluble in acetonitrile, with which it apparently
does not react, at least not immediately, as shown by nmr. pmr
(CD3CN): 97.92, br, s,hH(Ar—H); $3.95,s,3H(0Me);S 3.h3,s3H(NMe);
3.17,s,3H(NMe). cmrzs172.7(ArCON);S133.8, $132.8, $128.9,
8126.1(Arom. C's); S 63.6(0Me); 813.2, 539.7 (NMe). Treatment

of the oil in acetonitrile with water gave dimethyl isophthalate.

137

 

Jig-39‘ Newt, MeO OMe 160
MeZNC CONMe2 2)H20 ___.
Me N + 4-NMe
2 2_
\
A = -CONMe2 |
//'\\
E = -CQQMe MeOCO COOMe

Figure 58. Reaction of 1%6 with methyl triflate.

N,N,h',N"~tetramethylterephthalamide reacts in chloroform
with excess methyl triflate very rapidly and gives crystals
which make obtaining an nmr Spectrum difficult. An initial
Spectrum, taken before crystallization becomes heavy, shows
(besides the sharp singlet até§7.27 and the broad signal at
3'2.98 arising from the starting material) a clear aa'bb'
multiplet atS7.SS and sharp singlets at 53.90, 3‘3.L;7, $3.20.
Clearly one of the amide functions is becoming methylated.

The signal arising from the N-methyl protons on the unaffected
amide group are obscured by the corresponding protons in the
starting material. Hydrolysis gives a 6:h mixture of p-

carbomethoxy—N,N-dimethylbenzamide and dimethylterephthalate.

138

Apparently the diamide can be dimethylated but in the reaction
mixture rapid crystallization prevents complete reaction.

Similarly, 1,2,h,S-tetrakis(N,N-dimethylaminocarbonyl)
benzene reacts rapidly with methyl triflate to give white
crystals. Hydrolysis gives a white crystalline compound
having the correct molecular ion and pmr for a diester-diamide.
The question then arises as to whether the structure of the
hydrolysis product is lél, arising from the m-dimethylated
tetraamide, or l§g_from the p-dimethylated isomer, or a mix-

ture of the two. Careful crystallization from methanol gives

 

A
15%;
E
A 1'1
122eq. MeOTfi +
A A 2 v y r \ 7
) H80
E A
-1
.52 85%
A E
A :~CONM8
2
E = -002Me 162

Figure 59. Reaction of tetrakis amide 155 with methyl triflate
followed by hydrolysis.

beautiful prisms with a fairly sharp melting point (210.5-

213.0°C). The proton nmr (CD013) shows sharp singlets at

139
37.80 (as), 5‘3.83(6H), :,‘3.1o(.61~i) and 5'2.77(6H). This material
amounts to about half of the isolated product. Based on the
unique sharp singlet in the aromatic region, it is assumed
that the structure of this material is lég which has only
one kind of aromatic proton. Evaporation of the methanol
mother liquor gave a material with a broader melting point
range (whose pmr showed it to consist of the above compound
1gg) and a material shown to have the structure j§fl_in.a
ratio of about 3:1, based on the pmr. The structure assign-
ment as lél is based on the appearance in the pmr of two
singlets in the aromatic region as expected for l§l_which
has two different aromatic protons. The chemical shifts of
these two protons are surprisingly different (7.10 ppm and
8.50 ppm). However, they can be confirmed by comparing these
chemical shifts with those of the aromatic protons in the
meta diamide, N,N,N'N'-tetramethylisophthalamide and in the
meta diester, dimethyl isophthalate. In the former all the
aromatic protons appear at 7.3h ppm, while in the latter one
aromatic proton (assigned by Sadtler77 as the proton between
the ester groups) appears at 8.63 ppm.

A130-nmr spectrum of the original product (mp 175-20500)
is consistent with a mixture of the two isomers. It shows
two amide carbonyl resonances ( $168.89 and 168.71, one for
each of the isomers. However, there is only one ester carbon-

yl resonance (‘316h.19) apparently due to coincidental overlap

1&0

13
of the very similar carbons. The spectrum also shows seven

aromatic resonances. The two isomers have three and four kinds
of aromatic carbons, respectively. Thus, lég outweighs l§l_by
ahmt7fl.

This result is surprising since, after the first amide
group is methylated, the resultant positive charge is expected
to be delocalized into the aromatic ring at the para and ortho
positions. Thus, a second methylation of the amide group
para to the previously methylated group would appear not to be
favored, relative to a methylation on the amide group meta to
the previously methylated group. Therefore, it would be pre-
dicted that the two methylated groups would be meta to each

other rather than primarily para, as was found experimentally.

 

NMe
i

2

 

 

H20
1
1‘: \‘\ ‘ j,
A / / \
1&1
Fugmma60.

fit. .8\\/2
Med ‘ '/
+ .
L A/ {/ki

1‘ pl b r)
(-

l

\‘OMe

 

J

 

Dimethylation of 155; Mechanism 1.

‘-——+

 

others

 

 

1142
One possible eXpJanation involves participation by the
amide group ortho to the initially methylated group. Should
the ortho group participate, the positive charge would be
delocalized into the ring at the positions meta and ipso to
the initially methylated amide group. Thus, it might be

expected that the second methylation would occur at the amide

group para to the initially methylated group, as was found

 

experimentally.
TIA
A / A
1
r’ ' a
Me2N OMe
A
0 +‘* +—+ others
A
+ hMe2 ' NMe2

 

 

 

 

 

 

1.6.1. as.

Figure 61. Dimethylation of 155: Mechanism 2.

1113

Two factors related to this discussion should be noted.
First, the idea that the ortho amide group participates in
the initially formed cation is supported by the discussion
of N,N,N',N'-tetramethylphthalamide found below. It is also
supported (though not proved) by the fact that no dimethyl-
h,S-bis(N,N-dimethylaminocarbonyl)phthalate was detected.
This could be because the ortho group is "protected" by
participation with the initially formed cation. 0n the other
hand, it could simply be due to a steric interaction.

Secondly, it should be noted that the idea of participa-
tion by the ortho group in the initial methylation product
as an explanation for the product distribution depends upon
methylation and neighboring group participation occuring in
two separate steps rather than simultaneously. Were the two
steps to occur simultaneously, the second methylation would

give the wrong isomer.

 

MBZN (Wk?
A X ,.
E: 0 Naomi}
\ \'
A 1
NMen
c’.
E E
A A
lél

Figure 62. Dimethylation Of 155: Mechanism 3.

1th

Of the diamides studied the most interesting is the
ortho diamide. Unlike the others, it is only monomethylated,
even in the presence of excess methylating agent. The pmr
spectra in the presence of 1 equivalent of methyl triflate
and in the presence of excess methyl triflate are identical
(except for the signal arising from the methylating agent
itseli): 57¢75,br s,hH; $3.97,s,3a; Q3.h5,s,3H; $3.17,br s,9H.
(CDClB). The resonances at‘53.97, S3.h5 and 53.17 correspond
with those arising from other methylated N,N-dimethyl amides

studied (see Table 5), thus supporting structure A, Figure 63.

M8 \ /Me

+ TfO-
\\‘ (flhe

Me\N/Me //

MeOTf

N'+ (FfO

Figure 63. Reaction of 15h with methyl triflate.

The large signal at $3.17 does not correspond exactly with the

1&5

typical N-methyl signal of N,N—dimethylamides (3" 2.87-3.00) as
would be necessary for structure A. This difference will be
discussed later. On the other hand, structure 3 requires a
signal at about,82.2 for the amino methyl groups,78 which is

not present in the spectrum.

Table 5. Methyl pmr resonances in N,N-dimethyl amides and
O-methylated N,N-dimethyl amides.

ONMe
2 l +
MeO,
((j 2 .97 T;N\Me
3 97

[::j 3:33:
3.23

| Me

ONMe2 /N\Me

CONMe2

/’

/' 2.98

5..

\

3.90.
3047’
CONMe2 3.20
[j:1\ 3.00 CONMe2
CONMez OMe 3.95,
I + 3.439
MeO -’N\Me 3.17
Me
ow 2»;
1e
come2 2'87 OMe

All chemical shifts are in parts per million downfield from
internal TMS in CD01,

v
/

1&6

All chemical shifts are in parts per million downfield from
internal TMS in CD1,.

Table 6. Chemical shifts predicted for possible products
of methylation of 159.

predicted for

 

; 3.9a, 3.u7, 3.20, 3.00(x2).

found:

3-97. B-LS. 3.170(3)-

predicted for Jz

: 3.95. 3.147. 3.2(01~:e)78. 2.2013097”

1 _
The 3C—nmr spectrum of the methylated N,N,N',N'-tetra-

methylphthalamide shows two carbonyl resonances (5 175.2,
5157.1), six aromatic resonances (513h.1, $132.6, $131.9,
$128.9, S128.6,S12h.5) and five methyl resonances, one

O-methyl (E 62.14) and four different N—methyls (S 142-3:

539-7, S3951, $35.9). The fact that there are two carbonyls

1h"!

and five methyls corresponds with structure A, but not structure

B.

The chemical shifts of the carbonyl signals correspond

fairly well with one amide and one O-methylated amide, as shown

by comparison with the chemical shifts of the carbonyl carbons

given below.

The slight difference between predicted and actual

chemical shifts will be discussed later.

1
3C chemical shifts

Table 7.

N,N-dimethyl amides
amides.

CONMe2

i) 170.879
CONMe)
CONMeq
ONMe2

1:1.O(CDBCN)
COMM< (,

170.5

@001” $192
CONMe2

of the carbonyl carbons of
and O-methylated N,N-dimethyl

Me
.IMe
1f+ 174.6
Me
MeO
Me
/ N’
\‘1 Ih+ 173.8
e
,Me (CDBCN)
MeO M
Me

1&8

All chemical shifts are in parts per million downfield from
unless otherwise noted.

internal TMS in CDCl3

Table 8. Methyl 13C- chemical shifts for amides and 0-
methylated amides(O-methyl chemical shifts)

CONMeQ

(,4,
I

I
CONMe2

CONMe

4.. /
\/

VI
I

2

A’\
CONMQW

..

‘ ,CONMeM
m ‘-
1 it.
x / -\

CONtag

2:.0,

‘1
,'./o‘*

MeZNCO\” \/C0NMe2
Me000'/ /' coome
Me '
MeO\~ N +
rf‘ Me
Me
\\. +
L\,jf \f’k“Me
OMe
0:1
\/1: «1'19
l +

35.1.
38. 8,
53. 2(0Me)

59.3,
42.8,
'2.6(0Me)

1L9

1
Table 9. 30 chemical shifts predicted for possible products
of methylation of 15h.

 

predicted for Ode ofiflg. +gflg. 0959- +_:Q
63.3 h3.2 39.h 171.3 17h.2
\ CONMe2 39-7 35.3
+
/’ v/NMeZ
(DMe
60- 38- 35-
Meg)(mvle2 65 118 1,0
\ ”z
I 0
./’
Ni
/
Me Me

62.h h2.3 33.1 167.1 175.2

found 39.7 35.9

Note that the signals in the range of 3S-h5 ppm could
arise from the N,N-dimethylamino group of structure 3
(typical chemical shifts of N,N-dimethyl tertiary amines
are 38-h8 ppm)80 or from both types of N-methyl groups of
structure A (see Table 8 for typical N-methyl chemical shifts).

Thus, from all the evidence it must be concluded that
the correct structure for the methylated diamide is structure
A. However, several questions must be answered. First, why
is the omdflumide only monomethylated while under the same
conditions, the m-and p-diamides are dimethylated? Secondly,

why do the methyl protons of the unaffected amide function

have chemical shifts of 3.17 ppm rather than 3.00 as do other

150

amide methyl groups? Finally, why are the chemical shifts of
the carbonyl carbons (particularly that of the neutral amide)
shifted from 171.3 and 17h.2 as predicted to 167.1 and 175.2

as found? All of these questions can be answered by invoking

a small contribution by structure B, thus involving the second
amide group to a considerable extent. This type cf neighboring
group participation was invoked earlier in the 1,2,h,S-tetrakis

amide case.

 

Figure 6b. Preposed "hybrid" structure of methylated 159.

Reactions of nitriles with methylating agents. As an
analogy to the amides just discussed, and because they are
germane to a mechanistic discussion to come later, the reac-
tions of several nitriles with methyl triflate were explored.
Neither benzonitrile, nor phthalonitrile react with methyl
triflate in chloroform solution, as evidenced by nmr spectra
of the reaction mixtures. These show only signals arising

from the original nitriles and from methyl triflate. However,

1S1
methyl triflate does react with acetonitrile as evidenced by
the change in the nmr spectrum of methyl triflate in aceto-
nitrile-d3. Within seven hours, the methyl singlet atigh.30
is replaced oy a group of three singlets of equal magnitude
(J=1.SHz) centered at 3.77. The multiplicity of this signal
is apparently due to coupling of the methyl protons with the

1h

N nucleus in the N—methylatcd acetonitrile.

.. + -
CD3-C=N: + MeOTf ~—-+v CDB-CEN-CH3 OTf

Figure 65. Reaction of acetonitrile—d3 with methyl triflate.

Anticipating an increase in the methylating power of
methyl triflate on addition of antimony pentahalides, we
investigated the reaction of methyl triflate with antimony
pentachloride, and the reactions of that combination of
reagents with benzonitrile and phthalonitrile. The pmr
Spectrum of a mixture of one equivalent each of methyl triflate
and antimony pentachloride in chloroform-d is identical
to that of methyl triflate alone, suggesting that the alkyla-
ting agent is unaffected by antimony pentachloride. Further-
more, when one equivalent of phthalonitrile is added and the

homogeneous solution allowed to stand at room temperature for

152

several minutes, an off-white crystalline solid separates.

A pmr Spectrum of the mother liquor shows only the singlet

due to methyl triflate. Thus, the phthalonitrile is appar~
ently involved in the crystalline solid, but without includ-
ing the methyl triflate. This suggests formation of an
insoluble phthalonitrile-antimony pentachloride complex.
Further evidence for such a complex is suggested by hydrolysis.
After removal of the supernatant by pipette, the crystals were
rinsed with chloroform to remove traces of antimony penta-
chloride and methyl triflate. Addition of water to the
crystals gave a white powdery solid, insoluble in water or
chloroform, assumed to be antimony pentoxide (Sb205). Extrac-
tion of the aqueous solution with chloroform gave quantitatively
phthalonitrile. This proves that no methylation occurred as
hydrolysis of alkylated nitriles is known to give N-methyl
amides.

Similarly, treatment of benzonitrile gives a purple solu-
tion which soon deposits white crystals. Hydrolysis of the
crystalline material gives benzonitrile and antimony pentoxide.

Reactions of polyLfl,N-dimethylamides) with thionyl chlor-
ides. The pmr spectra of various polyamides were taken in
thionyl chloride as solvent. N,N,N',N'-tetramethylterephthal—
amide initially shows a singlet at 2.97 arising from the

N-methyl signals and a singlet at 7.33 arising from the

153

aromatic protons. The single methyl signal is due to the
coalescence noted earlier. Over a period of about one hour
the signal at 7.33 decreases and is replaced by a pair of
doublets (szhz) at 7.uo and 7.73. The signal at 2.97 de-
creases to half its size while singlets at h.OO and 3.87
increase until together they equal the size of the singlet
at 2.97. Clearly, one of the amide groups first reacts to
give a chloriminum chloride (or the corresponding chloro-

sulfite ester).

Me
'+
O NMGZ MeO /N\Me
SOCl2
——————+
OChMe2 o NMe2

Figure 66. Reaction of 167 with thionyl chloride.

Soon after the complete conversion of the material to
the above, the second amide reacts as evidenced by the forma-
tion of crystals in the nmr tube, and the appearance of a

singlet at 8.07 ppm.

The white crystalline material which separates from the

reaction of the his amide with thionyl chloride is apparently

15h

the his chloroiminium salt 1;}, When this material is heated
at reflux with a solution of p-chloroaniline in chloroform
there is obtained the his amidine lég, Hydrolysis of the
white crystalline salt, 163, gave quantitatively starting

material.

Cl

Me Me M [:::]
l + f
’,N

Me/Nv 0 Me /N\ Cl Me /N
8001
2 -01 NH
_________, p (Z 2 ‘
‘ ' "Ego“ 7
/ /M€‘ \If/fle N/ ,Me
I
Me

I
O h Cl
Me Me

 

Cl

.151 1.6.1 léll

Figure 67. Formation of bis-amidine 16“.

Similarly, N,N,N',N'-tetramethyl iSOphthalamide reacts
to give the meta bis chloroiminium chloride salt as evidenced
by the disappearance of the singlet at 33.00 and the broad
singlet at 7.33 and the appearance of a pair of singlets at

3.07 and 3.87 and a multiplet at 7.2-7.8 ppm.

N,N-dimethylbenzamide reacts in a similar manner, a

singlet at 2.97 ppm being replaced by two singlets at.§3.98

155

and $3.83 and a singlet at 7.30 being replaced by a multiplet
(LS7kh-8.0). However, the reaction is much slower being com-
plete only after two weeks at room temperature. However, pro-
longed heating of N,N-dimethylbenzamide in thionyl chloride
results in 2,h,6-triphenyl triazine, a reaction which will be

further discussed later.

As we have seen N,N,N',N'-tetramethylphthalamide does
not behave like other amides on reaction with methyl triflate.
The same is true on reaction with thionyl chloride. In thionyl
chloride as solvent the amide reacts within five minutes to
give a material with pmr signals at $3.63 (two closely spaced
singlets) and an aa'bb' multiplet at $7.83. Within several
hours a yellow precipitate forms in the tube making further
nmr measurements difficult. However, this material eventually
redissolves. The nmr spectrum changes to a very broad methyl
signal at 3.07 ppm, two complex aromatic multiplets at $8.20-
8.35 and 7.h7-7.93, in a ratio of 1:3, and together amounting
to one third (h:12) of the methyl signal. At this point, no
further chanre is noted in the spectrum even after refluxing

for three days.

The 13C-nmr spectrum of the reaction product of N,N,N',N'-
tetramethylphthalamide with thionyl chloride in deuterochloro—
form was obtained after formation and redissolution of the

yellow solid. The Spectrum consists of one carbonyl resonance

(167.7 ppm), six aromatic resonances (136.8, 13h.2, 133.1,

1.56
131.1, 130.8, 129.7 ppm) and one large methyl signal (h3.7 ppm).
The signals at 3129.7 and 131.1 are about half the size of the
other four aromatic signals. These apparently arise from the
two substituted carbons. When the thionyl chloride is removed
and the solid residue is dissolved in water, there is isolated

mostly starting material and a small amount of phthalic acid.

A chloroform solution of one equivalent each of the his
amide and thionyl chloride shows no change in its pmr, other
than the coalescence of the N-methyl signals. The solution
deposits yellow crystals only when a minimum of chloroform
is used ta dissolve the reagents. At higher concentrations,
the mixture remains a homogeneous solution. In any case,
the pmr spectrum of the mother liquor remains unchanged. The
crystals can be redissolved in greater amounts of chloroform
and the pmr spectrum of this solution is identical to that
of the mother liquor and to that of the starting materials.
Treatment of the yellow crystals with either water or absol-
ute ethanol gives back the starting bis-amide.

This yellow crystalline materia apparently is a loose
complex of the diamide and thionyl chloride present to a
small degree (and hence not apparent in the nmr) in equil-
ibrium with the starting materials. As it is less soluble
in chloroform than the starting materials, it is removed from
the equilibrium at high concentration, by precipitation from

solution.

157

When N,N,N',N'-tetramethylphthalamide is allowed to
react with excess thionyl chloride, the initially formed
yellow crystals redissolve, and the reaction is complete as
indicated by pmr after h days at room temperature or 12 hours
at reflux. No further change results from longer reaction
times. The product is the same as that from the reaction in

thionyl chloride as solvent.

Evaporation of the solvent from this material gives a
light yellow residue, 165, which on hydrolysis gives mostly
starting material along with a small amount of phthalic acid.
This shows that the dimethylamino groups are not cleaved after

the reaction with thionyl chloride is complete.

+
Mew/Me Me‘N,Me
\ Cl
0 UUCl2
/ O “‘3 .O
L
Me Me Me’N\Me
1 4 1?
M82N Cl
+
Me/N‘Me

Figure 68. Reaction of 1Sh with thionyl chloride.

158

When the thionyl chloride product, 152, is treated with anti-
mony pentachloride, in chloroform solution, a green solution is
obtained. From this solution soon separates light blue crystals.
The supernatant was drawn off and replaced with fresh chloroform.
This process was repeated several times, each time waiting for the
mixture to come to equilibrium as indicated by no further color-
ation of the solvent. The individual washings were kept separate.
Each successive washing became less green and more blue, suggesting
the presence of a yellow chloroform soluble impurity. It is assumed
that the blue material has the structure shown in the second re-

action in Figure 69.

159

Me2N Cl
CONMe2
H 0
Cl- ——g—-e
CONMe2
/N‘Me
4.
Me
165
Me
\N’LMe _
[/g SbCl6
SbClS \o
1&2 SbCl6-
N‘Me
Me/+

NH

Figure 69. Reactions of thionyl chloride product, 165.

160

 

o zlbw
w ..e.
, ow
0 2mm 0
0023mm
ivmooem 1mmo
: . o
more? 1 see: 11 . \
m m >8 m 0H >H
m
Abel u enommmowv 0 2mm

0 2mm B
as

mwmcsm no. mommwcwo massechmm om «Um encased on mam man vsoswouowswwwbm.

 

161

When the thionyl chloride product 1§§_was treated with
p-chloroaniline, a brilliant yellow material is obtained. Crystal—
lization from petroleum ether gave as beautiful lemon yellow needles
a compound which was pure by TLC and melting point (range 1.00C).

It was highly soluble in chloroform, less so in ethanol. It had

only aromatic protons as shown by pmr. Thus, it is clear that both
dimethylamine groups from the original bis amide were cleaved. The
mass spectrum of the yellow material 166 showed the isotope pattern
expected for a molecule containing two chlorine atoms. The molecular
ion was m/e 366. This corresponds to a molecule obtained by sub-
stituting both dimethylamino groups in the original bis amide with
p-chloroaniline followed by dehydration. This could arise formally
in several ways, depending on whether one assumes N-alkylation, C-

alkylation or both.

Structure B would appear to be eliminated based on the unsym-
metrical pmr and on the fact that the cmr of the yellow material
shows at least 13 and perhaps as many as 16 different aromatic
carbons. This 13C evidence also apparently eliminates structure
D which would require 18 different aromatic carbons. Thus, it
appears that structure A or C is correct for the yellow material.
There was some evidence for structure C, which is a substituted

11H-dibenz(b,e)azepin—11-one.

162

\ /
'_'N

Figure 71. 11H-dibenz(b,e)azepin-11-one.

This basic ring system was characterized in 1972 by Cooke and
Russell.81 It was recrystallized from pet. ether and was described
as "yellow". Cooke and Russell give ultraviolet data as

(ethanol) 2hh"nm",312 nm(log13 h.h8,3.75). The yellow dichloro
compound has a UV maximum at 255 nm and a much smaller one at 305nm.
Although Cooke and Russell gave no infrared data, the yellow dich-
loro compound gave IR data which were not inconsistent with
structure C .

0n the other hand, there is Considerable chemical evidence,
as will be shown below, that l§§_does not have structure C (orD),
but rather structure A.

.The N,N,N',N'-tetramethylphthalamide/thionyl chloride, lég,
was also treated with p-toluidine to give another bright yellow
crystalline material, 161, Its spectral prOperties are almost
identical to those of the dichloro compound. In addition, the pmr
has a broad singlet at 2.33 ppm and the cmr has two very closely
spaced signals at 20.66 and 20.9h ppm. These resonances correSpond

to the methyl atoms.

163

When this yellow tolyl product was heated with excess KGB in
aqueous dioxane, a trace of p—toluidine was obtained. The main
portion of the product, however, was a white insoluble material,
lég, having the same molecular ion as the starting material. In
the presence of aqueous acid, this white material dissolves in
chloroform from which there is subsequently isolated a white
compound crystallized from chilled ethyl acetate/pat. ether (mp
198-198.5) which was shown by chemical ionization mass spectrometry
to be mostly N-(p-tolyl)phthalamide (352), MW 237, mp82 206°C with
small amounts of N,N'-bis(p~tolyl)phthalamide, MW 3hh, (119) and a
compound having molecular weight 326, either the starting material

or an isomer.

16h

—L
'01
\3

aq. KOH

dicxane

H3
5U

/\
0

CH

chloroform

3

 

O

C
169 1

Figure 72. Reactions of yellow 16].

0“}
N . H
{is + +
”We
0 H
70 3

.J.
O'\
Q

l

.3
m0
(DH

165

In an attempt to confirm the structure of the p-chloro-
aniline produot as an 11H-dibenz(b,e)azepin-11-one, it was treated
with 2,h-dinitrophenylhydrazine in ethanol. There were obtained
beautiful orange crystals (mp 193.5-195, lit.83 19h-19S). The mass
spectrum and pmr showeed it to be N-(p-chlorophenyl)phthalamide 111,
The orange color may apparently arise from a small amount of impurity,
specifically 2,h-DNP or a reaction product of 1§6_with 2,h-DNP, for
which some evidence is given below. The formation of this product
proves the 11H-dibenz(b,e)azepin-11-one structure is n21 that of

166 because such a ketone would not be expected to be hydrolyzed.

J:jfil
/“ ' g

:2\ ;;
‘ 2 4-DNP N Cl
{Ly—C1 ’ > \ Q
\‘g aq. H SO \{
O o
1

 

2 4
O EtOH

Figure 73. Reaction of 166 with 2,h-dinitrophenylhydrazone.

166
Similar 2,h-DNP treatment of the yellow tolyl compound 1§1_gave
the same type of product, n-(p—tolyl)phthalamide 162,82 A second
compound 1lg_was detected in the product when it appeared in the
mass spectrum at a higher temperature. Its mass is h17 which
corresponds to a molecule in which a p-toluidine moiety in the

starting material was replaced with a 2,h—DNP moiety.

0

I
\\«’*<::>—CH3

O

 

N-Ar
1.6.9.
N-Ar 2'4‘DNP . + .
aq. H 8047 OZN N02
0 Et H
167 N-NH
— 1V4 ;.
N‘ CH
/\‘( \ / 3
Ar- = p-tolyl O
31?.

Figure 7h. Reaction of 161 with 2,h—dinitrOphenylhydrazone.

167
When the yellow compound was treated with aqueous acidic ethanol in
the absence of 2,h-DNP there was obtained N-(p-toly1)phthalamide 162
characterized by nmr and melting point (199.5-200.S, lit.82 2060C).

The melting point is low due to a small amount of yellow impurity.

It is clear from these acid catalyzed hydrolyses that the yellow
substances 166_and 161 cannot have the 11H-dibenz(b,e) azepin-11-one
structure. It seems most likely that the correct structures are as
shown in Figures 72-7h. One piece of evidence against such a
structure is the melting point of 161_reported an by Islam et al
(207°C). The melting point of our yellow dimethyl compound is
11h.5-116°c. The proximity of Islam's melting point to that reported
by Porai-Koschiz82 for N—(p-tolyl)phthalamide 162.(206OC) leads us to
believe that Islam may have confused the two compounds. This suspi-
cion is enhanced by Islam's report that the work-up of his compound
included treatment with 10% aqueous H01, a procedure which had been
shown85 (and confirmed by us) to hydrolyze similar compounds to
phthalamides. However, Islam reports as microanalytical data
numbers which agree with structure 161, A second piece of evidence
against structures 166 and 161_is the lack of any recognizable
aa'bb' coupling pattern in the pur. This however, may not be nec-
essary for the prOposed structure. For example, the spectra of
N-(p-tolyl)phthalaimde 162_and N-(p-chlorOphenyl)phthalamide.111
both include singlets for the p-disubstituted aromatic protons.

Their chemical shifts are 7.17 and 7.33 ppm, respectively. The

yellow ditolyl compound 161 and dichloro compound 166 have broad

168
singlets at 7.30 and 7.33 ppm, reapectively.

A solution at the yellow ditolyl compound 161_in CD013 was
treated with two equivalents of methyl triflouromethane-sulfonate.
One methyl group is added.(Ԥ3.30) and the second equivalent of
methyl triflate is unchanged (:ih.1o). The broad singlet (3 2.33)
arising from the aromatic methyl groups in the starting material is
replaced by two singlets (§§2.h3 and 2.38). The aromatic region
remains relatively unchanged (S'7.0-7.8, complex multiplet). These
data are completely consistent with the prOposed structure of the

starting material.

// ,(Hl
,L;::B 3 +/[:::]Xfli3
N CH3\N
N"</__—\\"x CH.5 MeOTf QCHB
O

(HHS 3

Figure 75. Reaction of 161 with methyl triflate.

As a model for 16 , (the product of N,N,N',N'-tetramethyl«
phthalamide and thionyl chloride) phthaloyl chloride was allowed

to react with p-chloroaniline. The white solid obtained (which

169

was insoluble in organic solvents and in aqueous acid or base) was
shown to be N'Nlbis(p-ch]orophenyl)phthalamide, 173, as expected.

, 8
Its melting point was 21S-21BOC (lit. 3

213-21hOC). The mass spectrum
shows a strong peak at m/e 257 which corresponds to N-(p-chlorophenyl)

phthalimide, 11 . The molecular ion

   

Figure 76. Reaction of phthalyl chloride with p-chloroaniline.

(m/e 38h) is a very small peak'(1% of the peak at m/e 257). This
fact suggests facile fragmentation with loss of p-chloroaniline.
The pmr of this material in DMSO-d6 shows an aa'bb' doublet of
doublets centered at 7.hS ppm and a broad singlet at 7.57 ppm in
a ratio of 2:1.

Having considered the spectral and chemical evidence, let us

now turn to the determination of the structure of the product of

170
.N,N,N',N'-tetramethylphthalamide with thionyl chloride, 166, The
Ivar data strongly suggest an unsymmetrical molecule. This is not
surprising, given the evidence of the analogous reaction with methyl
triflate explored earlier. Both hydrolysis and treatment with
antimony pentachloride followed by hydrolysis lead to nearly quantit-
ative recovery of N,N,N'N'-tetramethylphthalamide, suggesting that
both cimethylamino groups remain bonded to their respective carbons
throughout the reactions. Given these results, let us restrict

ourselves to consideration of the two structures below (see also

Figure 68).

Me\\ /AWe
N-+ -
Cl M632T 1
C1 ’
D
N\M
N + e
Me/ \Me Me/ Cl-
A E

Figure 77. Possible structures of 165.

Reaction of 165 with anilines to give imin0phthalimides 166 and
161 (Figure 69) sugggsts (but by no means proves) structure B,

because B has two electron deficient centers, both of which can

171
easily undergo nucleophilic attack. To substitute the neutral amide
structure of structure A requires protonation or participation by

the positive center.

In addition, the fact that yellow crystals form and then re-
dissolve before formation of 165 suggests (but again does not prove)
that structure B is the structure of 166_and that A is the structure
of the initially formed crystalline material. This suggestion
requires a surprisingly slow ring closure from A to B. That this
ring closure is slow is supported by the relatively slow second ring
closure in the dimethylation of 1,2,h,S-tetrakis(N,X-dimethylaminocar—
bonyl)benzene 166_with methyl triflate (Figure 61). In addition,
when considering the geometry of such a ring closure it becomes
obvious that the chloroiminium group must twist out of the plane
of the aromatic ring, and thus out of conjugation, to allow nucleo-
philic attack by the neutral amide oxygen. This clearly raises the
energy of activation relative to the level that otherwise might be

expected.

:\ , .——-—— \+

/
\—-——/‘ H
__ /l“l€3 Me
MEZN \ N “A M8 N \

Cl +\v1e

Figure 78. Conformation of ring closure of 165.

172

Finally, lot us compare the reactions of lSh with methyl

triflate and with thionyl. c‘r'xlorl',dc.

that is, the extent of participation t; t?»

group, would be expected to be greater in

‘3

case than in the methyl triflat, e se. In
case it was shown earlier (figure 6h) that

able participation.

 

The extent of cyclization,

neighboring amide

he thionyl chloride

the methoxy iminium

there is consider-

Therefore, the fact that participation

appears to be complete, or nearly so, in the chloro iminium

case is not surprising.

Mex../Me
N +

CI ‘

Me Me

Figure 79.

Me

_+
/1\‘ ‘Me
Me

M92N OMe

N~4We
Me

Comparison of the extent of amide group participation

in the chloro dimethyliminium ion and in the methoxy

dimethyliminium ion.

173

Addendum. In the course of this work, two unusual, and not
unrelated, reactions were discovered. Because these reactions did
not directly relate to earlier discussions, but merit some attention

in and of themselves, they will be discussed now.
When N,N,N', N'-tetramethylphthalamide, 1§h, was heated in
phosphorus oxychloride there was obtained, along with phthalic acid,

a 10% yield of N~methylphthalimide, 11h.

 

Figure 80. Formation of N-methylphthalimide from N,N,N',N'-
tetramethyl phthalamide or N,N-dimethyl phthala-
moyl Chloride.

1711

Its structure was suggested by mass spec, and pmr (although its pmr
was not identical to that given by SadtlergY) and confirmed by
independent synthesis from potassium phthalimide and methyl iodide.

The formation of N-methylphthalimide in this reaction is some-
what surprising as it involves dealkylation of an amide, a process
for which there is little precedent. Some Czech chemists eXplored
the range and limitations of a similar reaction.86 Pyrolysis of an
N,.N-dialkyl phthalamoyl chloride gives an N-alkylphthalimide and
an alkyl chloride. It is suggested that the mechanisms of these
reactions are similar, that is, involving nucleophilic displacement

by chloride of an acyl quaternary nitrogen.

175

 

O
I
NR2 R2
Cl
T/ NR2
0 0
O
I
l ‘\\ NR2
+ Cl
L\<5/ JR”
C.
Cl

0
)2 / .1
,H \\ ,R
+N‘R f / N\R
0 01‘ 01 R2
0
/j<
N-R
0

Figure 81. Suggested mechanfllm of cleavage of per-N-alkyl
phthalamide or phthalamoyl chloride.

176

Note that either of these reactions constitutes a method for
cleaving a secondary amine to a primary amine, a process for which
there is a surprising paucity of methods. Attempts to increase yield

of N-methylphthalimide failed.

A similar reaction occurs when N,N-dimethylbenzamide is heated
at reflux in thionyl chloride. From this reaction there was obtained
2,h,6-triphenyl triazine as an insoluble white solid identified by
melting point, mass spectrum and solubility preperties. A8 a small
amount of benzonitrile was also formed in the reaction as evidenced
by the odor, it was hypothesized that the nitrile was an intermediate
in the formation of the triphenyl triazine. However, a refluxing
solution of benzonitrile in thionyl chloride gave no trace of

2,h,6-triphenyl triazine.

Figure 82. Formation of 2 h 6-triphenyl triazine from N N-
dimethylbenzamide. ’

177
It is not difficult to prOpose a mechanism for this reaction consist—
ent with the available data. Such a mechanism is shown in Figure 85.
However, no evidence was obtained which would support this mechanism
over others. As was discussed earlier, (Figure 65) the product of
acetonitrile and methyl triflate is N-methylacetonitrilium triflate.
This salt shows no sign of forming 2,h,6-trimethyltriazine even in
the presence of excess acetonitrile as might be expected from a
mechanism like that outlined in Figure 83. However, that salt was not
heated. Furthermore, this mechanism (Figure 83) includes N-methyl~
benzonitrilium rather than N-methylacetonitrilium. As was noted
earlier, benzonitrile did not react with methyl triflate even at
elevated temperatures. Thus, it may be that N~methylbenzonitrilium

is less stable and hence more reactive than N-methylacetonitrilium.

m
0 \\2+
\zm / \3m a H. / .oz
/ z 88 2+ . o _
m . 8 IIII» III...
\ if “I”... 3m Aqmll sass”: \
Emmy

178

 

wHaCHm mu. mnemomma Bmowwbwms mos dam modsmaeo: ow m.t.mudsevsmskw aHHmNHSm macs 2.21mesmdswwdmbumawam.

179

EXPERIMENTAL

Instrumentation, solvents, etc. See experimental section for
Part 1.

N,N.N‘JLN'-tetrmnetiwlpllthalamide(15g). A solution of 36 ml
(50.5 g, 0.25 mol) of phthaloyl chloride in 150 ml of methylene
chloride was added to an ice cooled mixture of 250 ml (62 g, 1.L mol)
of 25% aqueous dimethylamine and 50 ml of methylene chloride. The
mixture was stirred for one hour. The organic layer was separated
and the aqueous layer extracted with two portions of methylene
chloride. The combined organic layers were dried (NaQSOh), filtered
and evaporated to give a white solid which was vacuum dried giving
h7.0 g (85% of off—white solid. {ecrystallization from heptane
gives beautiful white needles, mp 120-12100; lit75b 122-123°c. Nmr:
3‘2.89,br 8,6H; S 302,131: s,6H; 37.23,m(aa'bb'),lm. Mass spec.

m/e: 220 (70.), 219 (21%), 176 (10073), 1148 (25%), 133 (50%.).

NAN-dimethylbenzamldo (151). A solution of 23.2 ml (28.1 g,
0.200 mol) of benzoyl chloride in 100 ml of methylene chloride was
added drOpwise to an ice-chilled mixture of 100 ml of 25% aqueous
dimethylamine (0.55 mol) and 2.5 ml methylene chloride. The mixture
was stirred for an hour. The organic layer was removed and the
aqueous layer was extracted twice with 50 ml portions of methylene
chloride. The combined organic layers were washed once with a small

portion of chilled 20% NaOH to remove dimethylamine hydrochloride.

180

The solution was dried (HaZSOh)' filtered and evaporated to give a
clear colorless liquid. Vacuum distillation gives 25 g (83%) of
colorless liquid. Though h,H«dimethylbenzamide is a solid
(mp hh-h500)95, this colorless liquid has the correct 1H- and
13C-nmr and behaves normally on reaction with methyl triflate or
thionyl chloride. Pmr: $2.97,br 8,6H; $7.27,s,5H.
HLNLN',N'-tetramethylisqphthalamide(156). A solution of 25 g
(0.151) mo1) of iSOphthaloyl dichloride in 100 ml of methylene
chloride was added dropwise to a chilled mixture of 120 ml of 25%
aqueous dimethylamine (30 g, 0.67 mol) and 20 ml of methylene
chloride. After the addition was complete, the solution was allowed
to warm to room temperature and stirred overnight. The organic
layer was separated and the aqueous layer extracted three more times
with 50 ml portions of methylene chloride. The combined organic
layers were washed once more with 50 ml of chilled 20% aqueous sodium
hydroxide, dried (MgSOh), filtered and rotary evaporated to give a
white solid. Recrystallization from petroleum ether gives 30.1 g

87 137-138. Pmrt 83.03,br

(91%) of white needles, mp 13h-135.5; lit.
15,123; $736,151,113. Mass spec. m/e: 220 (25%), 219 (18%), 176 (100%).
N,N,N[.N'-tetr§mstnxlterephthslgmide(151). A mixture of 16.6 g
(0.100 mol) of terephthalic acid and 50 ml of freshly distilled(from
quinoline and cottonseed oil) thionyl chloride was heated at reflux
until the solution became homogeneous (about 2h hours). The thionyl

chloride was removed under reduced pressure and the residue taken up

in toluene, which was in turn removed under reduced pressure. (This

181

is to remove the last traces of thionyl chloride, sulfur dioxide and
hydrochloric acid.) The residue was dissolved in 50 m1 of methyl—
ene chloride and added drOpwise to a chilled mixture of 100 ml of
25% aqueous dimethylamine and 20 ml methylene chloride. After
stirring overnight, the layers were separated and the aqueous layer
extracted twice more with 20 ml portions of methylene chloride. The
combined organic layers were washed with 20 ml each of 10% NaOH and
brine. After drying (MgSOh), filtering and rotary evaporation, a
white solid was obtained. Recrystallization from a minimum of DMF
gave 15.h g (70%) of white prisms, mp 192—197; lit. 207-20800,75a
2020088. Mass Spec. m/e: 220 (63%), 219 (79%), 176 (100%).

1,Q-Dicarboxy-2,5-bis(dimethylaminocgrbonyl)benzene(l§§).
To a heterogeneous mixture of 21.8 g of pyromellitic anhydride
(0.100 mol) and 100 ml of methylene chloride was added with stirring
80 ml (20 g, 0.hh mol) of 25% aqueous cimethylamine. The mixture
was vigorously stirred at room temperature for 5 hours. The organic
layer was removed and the aqueous layer washed once with methylene
chloride. The basic aqueous solution was acidified with 5%.HCl
causing the precipitation of a white solid. It was washed thorough-
ly with water and dried to give 30.0 g (97%) of the diacid-diamide, mp
275. This material was used in the next step without further
purification. Mass spec. m/e: 262 (113%), 219 (58%), 1711 (100%), 1117
(32%)..

1,2,g,S-Tetrghis(ggmethylaminocarbonyl)benzene(155).

Method A. A mixture of 5.00 g (16.2 mmol) of 2,5-dicarboxy-

182

1,h-bis(N,N-dimethylaminocarbonyl)benzene and hO ml of thionyl
chloride was stirred at reflux for four hours. The initially
heterogeneous colorless mixture became homogeneous and golden

yellow in color. After cooling, the excess thionyl chloride was
removed under reduced pressure. The light yellow solid residue was
suspended in toluene and this was then removed under reduced pressure
leaving a white solid. This was taken up in 50 ml of methylene
chloride, in which it is partially soluble, and added slowly to an
ice-chilled mixture of 25 ml of 25%iaqueous dimethylamine (6.25 g,
0.1h mol) and 10 m1 of methylene chloride. After the addition was
complete, the reaction mixture was stirred overnight at room temper-
ature. The organic layer was separated and the aqueous layer ex-
tracted with two more 30 ml portions of methylene chloride. The
combined organic layers were washed with 30 ml of 15%1NaOH and brine,

dried (M580 filtered and evaporated to give 3.2 g (55%) of yellow

h).

solid. Recrystallization from acetonitrile gave white prisms, mp

189-193; 11t.75819u-196.

Method B. To a mixture of 12.7 g (0.05 mol) of pyromellitic
acid and 100 ml of mm was added 111.2 g (0.10 mol) of phosphorus
pentoxide. The temperature rose slightly. The mixture was stirred
at reflux for 20 hours. While still hot, the golden yellow solution
was decanted from a white sticky solid. This was filtered and air
dried to give 3.35 g. Recrystallization from a minimum of DMF gives
3.0 g of prisms, mp 189.58T92.S; lit.758 19h-196. An additional

1.0 g can be isolated from the reaction mixture by partial evapor-

183
ation. Total yield 14.0 {3 (22,76). Pmr: ;a‘2.98,hr s,12H; 33.01,br s,
12H:157.13,s,2H. Mass spec. m/e: 362 (5%), 361 (15%), 318 (100%),
317 (214%). 2714 (2271»). 273 (17:3)-

Observation of nmr spectra over time. When 1H-nmr spectra were
recorded over time (as those in Table h), the reagents and solvents
(if any) were mixed in recorded amounts and at a recorded time in a
clean, dry nmr tube. The tube was then sealed. When each spectrum
was run, the time was recorded and the time of reaction was recorded.
The results of each of these spectra are recorded in Table h or in the
text. These include reactions of each of the amides studied with
excess methyl triflate in chloroform-d, and reactions of each with
excess thionyl chloride in chloroformud. Also included are react-
ions of N,N,N',N'~tetramethylphthalamide 159 with one equivalent of
methyl triflate in CDC13, with one equivalent of thionyl chloride in
CD013 and with thionyl chloride as solvent, as well as the reactions
of benzonitrile and phthalonitrile with methyl triflate and/or
antimony pentachloride in CDClB. Methyl triflate and thionyl chloride
were measured in clean, dry syrin,:s.

Hydrolysis of Oumethyl salt; L. p0ly(R,N-dimethylaninocarbonyl)
benzenes. To a solution of h.h2 mmol of N,N-dimethylbenzamide, the
three isomeric N,N,N',N'-tetramethy1puthalamides of 1,2,h,5—tetrakis
(dimethylaminocarbonyl)benzene in 10 m of chloroform was added one
equivalent per amide group of methyl trit‘uoromethanesulfonate
(i.e., 0.50 ml, 0.72 g, 17.142 mmol for 11,11. 1 imethylbenzamide; 1.00 ml,

1h.5 g, 8.82 mmol for N,N,N',N'-tetramethyliithalamide). The

18h

homOgeneous solutions were allowed to stand for 2h hours. To each

was added 10 ml of water and the mixtures were stirred vigorously

for three hours. The organic layers were separated, the aqueous
layers extracted twice more with 10 ml portions of methylene chloride.
The combined organic layers from each reaction were dried (MgSOh),
filtered and evaporated. N,N-dimethylbenzamide gave 0.60 g (100%)

of methyl benzoate (not further purified). N,N,N',N‘- tetramethyl-
phthalamide gave N,N -dimethyl—2- carbomethoxybenzamide as a colorless
oil. Heating the oil under petroleum ether followed by decanting and

89 91-93°c. Total yield

cooling gave white needles, mp 89-9100; lit.
0.81 g (89%). The salt of N,N,N',N'-tetramethyliSOphthalamide came
out of chloroform solution as a dark oil. Hydrolysis gave 0.80 g
(9am) of dimethyl isophthalate, purified by filtration through a
short silica gel column with methylene chloride, followed by crystal-

lization from cool petroleum ether. Mp 66-67.5; lit.9O 67.8-68.300.

The salt of N,N,N',N'-tetramethylterephthalamide came out of
chloroform solution as white needles. Hydrolysis, followed by work-
up gave 0.87 g of white solid which was shown by nmr to be a 58zh2
(mole ratio) mixture of monoamide-monoester and diester
This amounts to a quantitative yield of the mixture. The two
substances were separated by filtration through a short silica gel
column. Methylene chloride eluted the dimethyl terephthalate
(mp 136-138°c; lit.91 1h1.0-1h1.8°c). Ethyl acetate eluted the
p-carbomethoxy-N,N—dimethylbenzamide (mp 103.5-105.5; lit.92

106-107).

185

On reaction with methyl triflate the tetracarboxamide 155_also
forms white crySLals in chlorcform solution. Filtration and hydroly-
sis followed by normal worn-up and careful recrystallization from
methanol gave white prisms (mp 210.5-213 C) of A,N,N'N'~tetramethyl-
2,5-dicarbomethoxy-1,h-benzenedicarboxamide (16;). Evaporation of
the mother liquor gave a 3:1 mixture (by nmr) of the above p-diamide-
diester lég_and N,N,N'N'~tetramethyl-h,6—dicarbomethoxy-l,3-benzenedi~
carboxamide (161). The overall yield of these compounds was 65% and,
as the crystalline l§g_accounted for half of the isolated material,
the overall ratio of lég;l§l_is 7:1. N,N-dimethyl-2—carbomethoxyben—
samide pmr: 32.63,s,3}i; 3‘3.09,s,3h; 33.82,s,3H; S7.05-7.50,m,3H;
S7.77-7.97,m,1H. Mass spec. m/e: 206 (1,256), 176 (1273), 163 (100%).
Dimethyl isophthalate pmr: 3‘3.91,s,os;5'7.37,t(J=7Hz),1H;88.10,dd
(J22Hz,7Hz),2H; $8.53,t(J:2Hz),1H. N,I—dimethyl-h-carbomethoxybenz-
amide pmr: 53.00,br s, 6H; 3-3.90,s,3h; 37.33,d(J=8Hz),2H; $7.93,
d(J=8H),2H. Dimethyl terephthalate pmr: 53.91,s,6s; 3‘7.99,s,m:.
N,N,N',N'-tetramethyl—2,5-dicarbomethoxy-1,hebenzenedicarboxamide
pmr: S2.77,s,6H; S3.10,s,6H; 33.83,s6H; S7.80,s,2H. N,N,N',N'-
tetramethyl-h,6-dicarbomethoxy-1,3-benzenedicarboxamide pmr: 32.7o,s,
6H; S3.09,s,6H;53.8h,s6H;.S7.10,s,1H;‘S8.50,s,1H. Mass spec of
mixture m/e: 336 (13°23). 335 (31%), 305 (15%). 292 (100%). 162 (17%)-

N,N,N',N' -tet;aggthyl-1 , h:benzene bis (chloro formiminiumj
dichloride(163). A solution of 2.20 g (10.0 mmol) of N,N,N',N'-
tetramethylterephthalamide in 20 m1 of freshly distilled thionyl

chloride was allowed to stand at room temperature for 2h hours.

186

The supernatant was decanted from the beautiful white crystals and

they were washed several times with chloroform.

fiydrolysis of 163. The white crystals formed in the

 

above reaction were dissolved in 30 ml of water and allowed to stand
at room temperature for one hour. The solution was extracted three
times with 20 ml portions of chloroform. The combined organic
extracts were dried (MgSOh), filtered and evaporated to give 2.00 g
(91%) of N,N,N',N'-tetramethylteraphthalamide.

1,L:benzene bis(N,N-dimethyl-N'-(h-chlorophenyl)carboxamidins)
(1_Q). The white crystalline material l§}_(see above) prepared from
2.20 g (10.0 mmol) of N,N,N',N'-tetramethylterephthalamide was taken
up in 25 ml methylene chloride and added to a solution of 5.10 g
(no.0 mmol) of p-chloroaniline in 50 ml of methylene chloride. After
stirring overnight, a yellow solid had formed. The mixture was
chilled in an ice bath and to it was slowly added a chilled solution
of 1.6 g (h0.0 mmol) of sodium hydroxide in 15 ml of water. The
organic layer was separated and the aqueous layer extracted twice
more with 20 ml each of methylene chloride. The combined organic
layers were dried (MgSOh), filtered and evaporated. The residue
was chromatographed on silica gel. Methylene chloride eluted the
excess p-chloroaniline.The biséamidine was eluted with 10% ethyl
acetate/methylene chloride and recrystallized from ethyl acetate,
mp 220-22100. Yield 1.h g (30%). The yield could probably be
improved by further elution of the column with larger preportions

of ethyl acetate. Mass spec. m/e: 390, 392, 39h- Pmr (CD013):

187

56.83,s,hH; S6.76,d(J=8Hz),hH; b6.27,d(J:8Hz),LH; 52.87,s,12H. Anal.
calc'd. for CZEHBUCIQNM: C, 65.61%; H, 5.51%. Found C, 65.77%; h,
5.50%.

2-(h-chloroohenyl)-3-(h-chlorgphenylimino)—2,3ydihydro-1H—
isoindol-1-one(166). A solution of 1.1 g (5.0 mmol) of N,N,N',N'-
tetramethylphthalamide in 10 ml of thionyl chloride was allowed to
stand at room temperature. The yellow solution deposited crystals
which slowly redissolved. After the reaction was complete as monit-
ored by nmr, the thionyl chloride was removed under reduced pressure
leaving a solid yellow residue. This was taken up in 10 ml of
methylene chloride and added by pipette to a solution of 2.55 g
(20.0 mmol) of p-chloroaniline in 15 ml of methylene chloride. A
yellow solid immediately began to appear and the reaction mixture
warmed slightly. After stirring overnight at room temperature, the
mixture was chilled and, with vigorous stirring, enough chilled
dilute sodium hydroxide was added to dissolve the solid and make the
aqueous layer slightly alkaline. The yellow organic phase was
separated and the aqueous phase extracted twice with 10 m1 portions
of methylene chloride. The combined organic layers were dried
(MgSOb), filtered and evaporated to give a yellow residue. Chromato-
graphy on silica gel with methylene chloride as the eluting solvent
gave an initial yellow band followed closely by the starting p-chloro-
aniline. The yellow material was recrystallized from petroleum
ether to give 1.3 g (72%) of fine yellow needles, mp 1h8.5-1h9.s°c.

Mass spec. m/e: 366, 368, 377. Pmr: complex multiplet between 6.5

188

and 8.0 ppm with a large broad singlet at 7.33 ppm. Anal. calc'd.
for C2OH12019N20: c, 65.L1%; H, 3.29%. Found: 0, 65.32%; H, 3.19%.
N,N'-bis(p-chlorgphenyl)phthalamide(113). To a chilled solution
of 1.u1 ml (2.03 g, 10 mmol) of phthaloyl chloride in 35 m1 of
methylene chloride was added a solution of 5.10 g (ho mmol) of
p-chloroaniline. The mixture was allowed to stir overnight at room
temperature. To the mixture, in which some white solid had formed,
was added enough dilute aqueous sodium hydroxide to make the organic
layer slightly alkaline. As the white solid did not dissolve, it was
filtered out, washed with water, ethanol and methylene chloride and
dried to give 3.2 g (83%), mp 215—218; lit.9h 213-21hoC. Mass spec.
m/e: 386, 257, 127. Pmr (DMSO-d6): 37.62,d(J=9hz).hH; 87.57,br s,
lmh $7.28,d(J-9Hz),hH. Anal calc'd. for 020s1h012N202: c, 62.35%;
H, 3.66%. Found: c, 62.h6%; H, 3.62%.
2.3-dihydro-2-(h-methylnhenle-3—(u—methylphenylimino)-
1§3isoindol-1-one(161). See the above procedure for the preparation
of 166, The yellow material was recrystallized from petroleum ether
to give an 81% yield of yellow needles, mp 11h.5-116. Mass spec. m/e:
326. Pmr: 56.5-8.0,m,12H;752.33, 9,63. Anal. calc'd. for 022318
N20: 0, 80.96%; H, 5.56%. Found: c, 80.69%; H, 5.76%.
1,i-bisfg-methylphenylimino)-1,3ydihydroisobenzofu£§g(l§§). A
mixture of 0.25 g (0.77 mmol) of the yellow crystalline 16 , 0.20 g
(2.60 mmol) of potassium hydroxide, five ml of water and enough
dioxane to dissolve the solid was heated on a steam bath overnight.

After cooling, the resulting white solid was filtered, washed with

189

boiling ethanol and air dried to give 0.25 g (100%) of white powder,
mp 1h7-1h8.500. Mass spec. u/e: 326 (85%), 325 (100%), 237 (13%),
193 (10%). 192 (13%). 165 (1076).

Acid catalyzed hydrolysis of 168. The white powder from the
above reaction (0.20 g, 0.65 mmol) was stirred with a mixture of
5 ml of 5%»aqueous hydrochloric acid and 10 ml of chloroform for
30 minutes. The chloroform layer was separated and the aqueous
layer extracted with 5 ml of chloroform. The combined chloroform
extracts were dried (Nazsoh), filtered and evaporated to give a white
solid. Crystallization from chilled ethyl acetate/petroleum ether
gave white needles, mp 198-198.SOC; lit. for N—(p-tolyl)phthala~
mide82, 20600. Mass spec.(chemical ionization) m/ : 238 (100%), N-p-
tolyl) phthalamide+1; m/e: 327 (1.0%, 166 or 168+1; m/e: 3h5 (0.8%),

173+1.

Acid catalyzed hydrolysis of 167. A small amount of the yellow

 

material was dissolved in 5 m1 of ethanol by warming on a steam bath.
To the mixture was added 1 ml of 10% aqueous sulfuric acid. The
solution was allowed to stand overnight, after which time it had
deposited yellow crystals of N-(p-tolyl) phthalamide, mp 199.5-

82 206OC- NEII‘: S70709m(aa'bb')9l-¢H; 37020:S!’+H; 52'}?!

200.500; lit.
8,3H.

Acid catalyzegfhydrolysis of 166. The above procedure gave
orange crystals of N-(p-chlorophenyl)phthalamide, mp 199-20200;
lit.83 1924-195°c. Nmr: S7.73,m(aa'bb'),hll; 5'7.33,e, hll.

249,6-Triphenyl-1,3,S-triazine. A solution of 2.50 g (16.8

190
mmol) of N,N-dimethy1benzamide in 15 ml of thionyl chloride was
heated at reflux for five days. A white solid had formed in the
reaction mixture. After cooling, the white solid was filtered out
and rinsed successively with chloroform, ethanol and water. It was
recrystallized from a large amount of ethanol to give 0.71 g (h1%) of
white needles, mp 233.2-233.SOC; lit.93 23h-23h.500. Mass spec. m/e:
309.

N-methylphthalimide. Method A. N,N,N'N'-tetramethyl-
phthalimide (15;) (1.5 g, 6.9 mmol) was dissolved in 20 ml of
phosphorus oxychloride and heated at reflux for 6 hours. After
cooling, the mixture was poured over enough crushed ice to keep the
mixture cold. The aqueous solution was extracted three times with
20 m1 portions of methylene chloride. The combined organic layers
were dried (Na2SOh), filtered and evaporated to give 0.h g of dark

green oily solid. Chromatography (CH201 on silica gel) gave, after

2
a trace of purple material, 0.30 g (30%) of white needles, mp 129-
131, lit.96 131400. Pmr (CF3002H for comparison with lit.97): 53.15,
s,3H; 7.65,m,hH. Note: Sadtler97 shows a singlet for the aromatic
protons. Mass spec.(chemical ionization), m/e: 161 (100%). Mixed
mp with independently prepared N-methylphthalamide (Method B) showed
no depression.

Method B. In a sealed tube a mixture of 2.0 g (11 mmol) of
potassium phthalamide and 10 ml of freshly distilled methyl iodide

was heated at 130°C in an oil bath for 10 hours. After cooling,

the excess methyl iodide was removed by evaporation and the solid

191

residue was taken up in 20 ml of water. Extraction with three 20 ml
portions of methylene chloride, followed by drying, filtering and
evaporation of the organic extracts gave 1.65 g (95%) of yellow
solid. Recrystallization from ethanol/water followed by chromato-
graphy (CH2C12 on silica gel) gave off—white crystals, mp 130—131.5°

C; lit.96 13hOC. Pmr: same as above.

LIST OF REFERENCES

63.

6.1;.

65.

66.
67.
68.

760

77-
78.

192

A. S. Kende, R. 0. Greenhouse and John A. Hill, Tetrahedron
Lett., 2867 (19 79).

H. v. Euler, E. Adler and G. J. Gie, Arkiv. Kemi. Min. Geol.,
1hB (19h0); CA 3h, 7877 (19h0); h. v. Euler, E. Adler, H.
Hasselquist and M. Lundin, Arkiv. Ksmi. Min. Geol., 18A (19hh);
9.1}. 38. 3875 (19th)

0. Diels and K. Adler, Ber,,62, 2337 (1929); J. A. Landgrebe,
"Theory and Practice in the Organic Laboratory," D. C. Heath
and Company, Lexington, Mass. (1977), p. 338.

G. Fodor and s. Nagubandi, Tetrahedron,,36, 1279 (1980).
D. N. Kevill and G. M. L. Lin, Tetrahedron Lett., 9L9 (1978).
H. J. Barber, L. Bretherick, E. M. Eldridge, S. J. Holt and

W. R. Wragg, J. Soc. Chem. Ind. (London),,62, 82 (1950);
G. T. Morgan and L. P. walls, J. Chem. Soc., 2th? (1931).

 

D. B. Baird. I. Baxter, D. R. Cameron and w. R. Phillips,
J. C. S. Perkin I, 1923, 832.

U. Rahman and N. M. Rodriguez, C. R. Acad. Sc. Paris 0;,
2:29. 230 (1970)

H. Pohl, J. Prak., 1&1, uh (193a).
J. H. woods and R. 3. Gibson, J. Am. Chem. Soc., 71, 393 (19L9).

J. 0. Her t ahl and J. L. Gresham, J. Am. Chem. Soc.,416, 1076
(19511).

F. s. Kipping, J. Chem. soc., 53, 39 (1888).

 

(a ) H. Schindlbauer, Monat. Chem., 99, 1799 (1968).
b H. Schindlbauer and S. Fischer, Synthesis, 63h (1972).

H. Perst, "Oxonium Ions in Organic Chemistry", Academic Press,
New York, 1971, p. 132.

Sadtler NMR # 10550.

R. M. Silverstein, G. C. Clayton and T. C. Morrill, "Spectro-
metric Identification of Organic Compounds, 3rd Ed.", John
Wiley and Sons, Inc., New York, 197h, p. 213.

G. C. Levy et al., "Carbon-13 Nuclear Magnetic Resonance Spectro~
scopy, 2nd Ed.", Wiley Interscience, New York, 1980, p. 123.

h6.

h7-

h8.

h9.

50.

S10

52.

53-

5h.

60.

61.

62.

193

R. west and H. Y. Niu, in J. P. Snyder, "Non-benzenoid
Aromatics", I, Academic Press, New York (1969, p. 311; R.
West and J. Niu, p. 2hh1, Wiley (lnterscience), New York,
1970; R. west, Isr. J. Chem., in press (1980).

R. Gleiter and P. Schanz, Ang. Chem. Int. Ed., 19, 715 (1980).
M. E. Peover, Trans. Faraday Soc., 58! 2370 (1962).

L. F. Fieser and M. Fieser, Reagents for Organic SynthesisL IL
John Wiley and Sons, New York (1967), p. 311; H. D. Becker and
G. A. Russell, J. Org. Chem., 28, 1869 (1963).

C. R. Johnson and W. G. Phillips, J. Am. Chem. Soc.,,21, 682
1969). ‘ ""“"

0. Neilands and M. Vavere, J. Gen. Chem. USSB,‘33, 1032
1963 .

See for example T. Yamazaki, T. Oohama, T. Doiuchi and T.
Takizawa, Chem. Pharm. Bull., 20, 238 (1972).

L. F. Feiser and K. L. Williamson, 0r anic Experiments, D. C.
Heath and Company, Lexington, Mass. (1975), p. 188.

a) P3 Ruggli and w. Theilheimer, Helv. Chim. Acta,‘gh, 899
19L1 .

b) M. Masagawa, F. Suzuki, H. Nakamishi, Y. Suzuki, J. Polym.
Sci. B.,,6, 293 (1968).

Prepared by oxidation of m-xylene with chromic acid. N. W.
Johnson and J. L. R. Williams, J. Am. Chem. Soc.,,69, 2065

(191.7)-
P. Ruggli and P. Bucheler, Helv. Chim. Acta,.30, 20h5 (19h7).

E. Ota, S. Inoue, M. Horiguchi and S. Otani, Bull. Chim.
Soc. Jap.,.52) 3h00 (1979).

Personal communication.

H. Perst, "Oxonium Ions in Organic Chemistry", Academic Press,
New York, 1971, p. 132.

W. Reid and H. Boden, Bg£,,89, 2328 (1956).

M. H. Hubacher and S. Doernberg, J. Pharm. Sci.,’53, 1067
(196A); g; 62, 5250.

P. S. Braterman, B. S. walker and T. H. Robertson, J. C. S.
Chen. Comm., 651 (1977). """"'

38.
39-

h0.

h1.
h2.

19b
D. H. Reid, Quart. Rev., 19, 27h (1965).

J. C. N rtin and R. G. Smith, J. Am. Chem. Soc.,,fié, 2262
(196A)-

M. J. Sabacky, C. S. Johnson, Jr., R. G. Smith. R. S. Gutkowski
and J. C. Martin, J. Am. Chem. Soc., 89, 205h (1967).

E. Muller, A. Moosmayer, A. Rieker and K. Scheffler, Tetrahedron
Lett., 3877 (1967).

N. C. Yang and A. J. Castro, J. Am. Chem. Soc., 829 6308
(1960).

R. Kreilick, JifChem. Phys., Q3, 308 (1965).

D. Kearns and s. Ehrenson, J. Am. Chem. Soc., 8i, 739 (1962).

 

E. A. Chondross, J. Am. Chem. Soc., 86, 1263 (196h).

R. M. Silverstein, G. C. Bassler and T. C. Merrill, "Spectro-
metric Identification of Organic Compounds", 3rd Ed., John
Wiley and Sons, Inc., New York, 197L, p. 227.

M. D. Taylor and L. R. Grant, J. Chem. in., 39 (1955).

M. Jarcho, J. Am. Chem. Soc.,_90, h6hh (1968).

B. A. Stoochnoff and N. L. Bentoin, Tetrahedron Lett., 21-2h
(1973).

s. Holgstrand, Acta Chem. Scand., 18, 1616 (196a).

"Aldrich Catalog Handbook of Fine Chemicals", Aldrich Chemical
Company, Milwaukee, 1981, p. 927.

"Elsevier's Encyclopedia of Organic Chemistry", Series lII,
Elsevier Publishing Co., New York, 1962, p. 12B.

Sadtler pmr 69h3.

K. Hafner, K. H. Hafner, C. Jonig, M. Kreuder, G. Ploss, G.

Schultz, E. Sturm and K. H. Vopel, gag. Chem. Int. Ed.,’g, 123
1963 .

K. Hafner and H. Suss, ADE; Chem. Int. Ed.,‘g, 575 (1973).
E. LeGoff, J._Am. Chem. Soc., 8Q, 3975 (1962).

R. West, "Oxocarbons", Academic Press, New York, 1980.

1h.

1!.”

j.

16.

17.

18.

19.

20.

21.

195

D. Adler and H. Brooks, Phys. Rev., 155, 826 (1967).

B. T. Matthias, Scientific American, 92 (1957); E. B. Yagubskii
and M. L. Khidekel, Russ. Chem. Rev., Q1, 1011 (1972).

K. Bechgaard, C. S. Jacobsen, K.‘M0rtensen, H. J. Pedersen,
N. Thorup, Solid State Commun., 33, 1119 (1980).

(a) D. Jerome, A. Mazaud, M. Ribault, K. Bechgaard, J. Phys.
Lett. (Orsay, Fr.), Al» L-95 (1980). M. Ribault. c. Benedek,
D. Jerome, K. Bechgaard, J. Phys. Lett. (0rsay,_Pr.), 11,

L—397 (1980). R. L. Greene, E. M. Engler, Phys. Rev. Lett.,
Q5, 1587 (1980). K. Andres, F. Wudl, D. B. McWhan, C. Thomas,
D. Nalewajek, A. L. Stevens, Phys. Rev. Lett., 55, 1AA9 (1980).
(b) M. Ribault, J. P. Pouget, D. Jerome, K. Bechgaard, J. ng .
Lett. (Orsay, Fr.),. in press.

(6) F. Wudl, J. Am. Chem. Soc., 193, 706h (1981).

(d) S. Parkin, M. Ribault, D. Jerome, K. Bechgaard, J. Phys.
9;, submitted for publication.

 

K. Bechgaard, K. Carneiro, F. B. Rasmussen, M. Olsen, J. Am.
Chem. Soc., 103, 2ALO (1981).

T. J. Kistenmacher, T. E. Phillips and D. 0. Cowan, Act§70ryst.,
B30, 763 (197A)-

R. C. Haddon, Aust. J. Chem., 28, 23A3 (1975); Nature, 256,
39A (1975)-

A. F. Garito and A. J. Heeger, Apcts. Chem. Res., 1, 232 (197a).

 

A. W. Block, in "Energy and Charge Transfer in Organic Semi-
conductors", K. Masuda and M. Solver, Eds., Plenum Press, New
York, 197b, pp. 159-17h; D. B. Tanner, C. S. Jacobsen, A. F.
Carito and A. J. Heeger, Phys. Rev. Lett., 32, 1301 (197A;

M. J. Cohen, L. B. Coleman, A. F. Garito and A. J. Hee‘er,
ghys. Rev. 3.,719, 1298 (197A).

J. B. Torrance, Accts. Chem. Res., 1g, 79 (1979).

R. C. Wheland, J. Amer. Chem. Soc., 98, 3926 (1976).

E. Phillips, T. J. Kistenmacher, J. P. Ferraris and D. 0.
Cowan, J. C. 3. Chem. Comm., h71 (1973); F. R. Hervstein in
"Perspectives in Structural Chemistry", J. D. Dunitz and J.

A. Ibers, Eds., Vol. h, p. 166, John Wiley, New York, 1971.

R. M. Ietzger, J. Chem. Phys.,’51} 1870, 1876, 2218 (1972).

AK}

AL}.

11.

12.

13.

LIST OF REFERENCES
W. L. Masterton and E. J. Slowinski, "Chemical Principles",
hth Ed., W. B. Saunders Co., Philadelphia, 1977, pp.159-160.

D. S. Acker and W. R. Hertler, J. Am. Chem. Soc.,,éh, 3370,
337A (1962)-

J. Ferraris, D. 0. Cowan, V. Walatka, Jr. and J. H. Perlstein,
J. Am; Cham. Soc., 95, 9A8 (1973).

E. M. Engler, Chemtech, 1976, 27h

"CRC Handbook of Chemistry and Physics, 51st Ed.", Robert
weast, Ed., Chemical Rubber Co., Cleveland (1971).

J. H. Perlstein, Angew. Chem. Int. Ed., 8, 519 (1977).
R. C. Wheland and E. L. Martin, J: Org. Chem., L9, 3101

(1975); G. Saito and J. P. Ferraris, J. C. S. Chem. Comm.,
1027 (1979); J. R. Anderson and O. Jorgenson, J. C. S.

 

Perkin I, 3095 (1979).

D. J. Sandman and A. F. Garito, J. Org. Chem., 39,1165
(197h); A. W. Addison, N. S. Dalal, Y. Hoyano, S. Huizingo
and L. Weiler, Can. J. Chem., 55, A191 (1977); M. Maxfield,
D. 0. Cowan, A. Bloch, T. 0. Poehler, Nouveau J. Chem., 3,

6A7 (1979)-

N. F. Haley, J. C. S. Chem. Comm., 1030 (1979); K. F. Haley,
J. C. 3. Chem. Comm., 207 (1977); F. Wudl, M. L. Kaplan,
B. K. Teo and J. Marshall, J. org. Chem., 5;, 1666 (1977).

H. J. Keller, ”Chemistry and Physics of One-dimensional
Metals", Plenum Press, New York, 1977.

L. L. Klein, "Studies Toward the Synthesis of Organic Conductors",
Ph. D. Thesis, Michigan State University, 1980.

I. B. Coleman, M. J. Cohen, D. J. Sandman, F. G. Yamagishi,
A. J. Garito, A. F. Heeger, Solid State Comm., 12, 1125

(1973).

R. E. Peierls, "Quantum Theory of Solids", Oxford University
Press, London, 1955.

196

80.

81.
82.
83.

8h.

85.
86.

87.

88.
89.
90.
91.
92.
93-

95-

96.
97-

197
J. E. Sarnecki, H. L. Suprenant, F. K. Molen and C. N. Reilley,
Anal. Chem., , 2116 (197 5); H. Eggert and C. Djerassi, J. Am.
Chem. Soc., ,3710 (1973).
R. G. Cooke and I. M. Rus. ell, Aust. J. Chem.,,g5, 2b21 (1972).
B. A. Porai-Koschiz, §;f0bsc. Chi.,’1, 60A (1937).

A. F. M. Fahmy, N. F. Aly, and M. 0. Orabi, Bull. Chem. Soc.
Japan,’51, 21h8 (1978).

A. M. Islam, I. B. Hannout and A. M. El-Sharief, Ind. J. Chem.
Sect. 3., 15, 61 (1977).

w. E. Parham and C. D. James, J. Org. Chem., L1, 1187 (1976).

v. Horak and L. Novotony, Chem. Listy, L6, 357 9 (1952);
v. Horak and L. Novotony, Coll. Czech. Chem. Comm., 18, 80 (1953).

R. B.)Kostenbauder and T. Higuchi, J. Am. Pharm. Assoc., ~5, 518
1956

C. w. Fong and C. R. Hameister, Org. Mass Spec., 535 711 (1978)
C. Mannich and F. Schmitt, A3, 1928, 78.

See Reference 5, p. C-167.

See Reference 5, p. G-166.

Y. Kigugawa, Chem. Pharm. Bull.,'gg, 1059 (1976).

J. H. Bayer and P. S. Ellis, JCS Peggin I,‘1919, h83.

R. D. Reynolds, M. J. Zeigker, D. J. Heuck and C. Johnson,
J. Chem. Eng. Data,_13, 558 (1968).

P. Oxley, M. W. Partridge, T. D. Robson and W. F. Short, J. Chem.
Soc., 19h6, 763.

See Reference 5, p. C-h31.

Sadtler NMR # 19110.