YL

*PARA»
IN

er
--PARA-MEIHOXYPHEN
1.

 

  

Di
3;

Ifins

 

y .

H
1:
uRE'

 
 

.leT
‘JQUHV

3

NE
RUCT

E 3.
EL;-

 
 

 

 

 

 

 

 

 

F... .
...nun..4mc... T-...& E-
H 8 ME m
.. . . ..l P .L. . ...Du...A :D
. _ I . smv. _ a. A
. . N ......I
. . H II, R U
C.“ H C4 B
4 . m m m w..m.m

. R: . . ,
. , .. W..M.,m . .T. “Wu

 

 

 

 

 

 
 
 

 

 
 

 
 
   

 

 

 

 

 

 

 

       

_. n r. . . ...
. _. . . .
Tm. HH;~J..~:.4: .
......i .

, .
... king}: 0.

    

a fi
. 1 :v
.cw_.r¥a1

$57.,

2,.

        

  

3;
,4:
34,.
....

‘-

             

i
' 2:0
24‘

- ,
'A‘
*1

 

 

3.24.
t .1: (1

5.2.8:...

 

 

 

fEPh;

       

,3
:.:, «1...:

       

                

 

.4.
3.. .3. ”MW.

we .0

 

éLD

 
   

                    

r 5
. ..¢.,w..../v.(.rr 1 4
.l. Iflrffqv J». ...:
3 r 11:: 5.4.11. flea
. a}: .. .1 1:15;

:.. .ffa-

   

      
  
 

If.

 

l

. z .7
(Wirfitfilawiqf mm
’V/ 5...» .A. ..

. . . a. T:
71?“! H 147.91. ...IJ
I. Funk/Mn aft.
3 .7 . r 1..
~.1 (in!) a

                 

‘esns

u. ....
13.55742»,
2'... .f

'*NHCHI

    

 

           
     

      

145.;
.1; .rv. J

  
  

 

 

                   
        

 
 

   

 

  

 

      

 

 

 

 

 

 

 

. , , f
. . ... . :

h .../Mt

., . . . . : ....
. .. .. . . , <44. 3' f
.. .. . . .. . 4 , . 4. . .v... r ftff
. . . . bu . IEv.Ir.s.
. E . 4!..Iu/
I . . if): .
‘ ... r . uvllrlulvartun
. , LEC..C..e.
A v 1 d

 
 

E...

 

  

1-... .fi?

 

; 4.5..11413171. . .

 

  
  

      
    

“Mama
93 5’45 (.1233; IF”

~‘,‘, .43 . This is to certify that the

7 35%
r‘» J-
-'~.,'. ’.'

,. -

«l

' thesis entitled
Part I: Approaches to the Synthesis of 2-para—nitrophenyl—
4,6—di—para-methoxyphenyl—benzene-l,3-quinone
Part II: The Crystal Structure of 2,4—di—para-methoxyphenyl—
cyclobutadiene-l,3—quinone

presented by
William Terry Suggs, Jr.

has been accepted towards fulfillment
of the requirements for

_Eh‘.D..___degree in Jhm

 

Major pr‘ofessoxr

DateM—_ . ‘ f 3

12% ' ° ‘
' .H‘" :

0-7639 l

 

ABSTRACT
PART I
APPROACHES TO THE SYNTHESIS OF 2-PARA-NITROPHENYL-
4,6-DI-PARA-METHOXYPHENYL-BENZENE-l,3-QUINONE
PART II

THE CRYSTAL STRUCTURE OF
Z,4-DI-PARA-METHOXYPHENYL-CYCLOBUTADIENE-l,3-QUINONE

By

William Terry Suggs, Jr.

PART I

Synthetic approaches to 2-p-nitropheny1-4,6-di-p-
methoxyphenyl-benzene-l,3-quinone gl have been investigated.
Benzene-1,3-quinone itself is unknown and is expected to
be unstable, since no uncharged resonance contributors
can be drawn for such a system. However, molecular orbital
calculations indicate that the molecule may indeed be stable
and isolable when the appropriate stabilizing groups are
added, as noted above.

Reaction of ethyl p-methoxyphenylacetate %g with diethyl
oxalate and sodium ethoxide produced the oxaloester ég which
was, in turn, reacted with aqueous formaldehyde and potassium
carbonate to yield ethyl a-p—methoxypheny1propenoate éé.

Base condensation of éé with p-methoxyphenylacetone gave

2 William Terry Suggs, Jr.

4,6-di-p-methoxyphenylcyclohexane-l,3-dione éé. Arylation
of éé in sodium hydride/hexamethylphosphoramide with p-
nitrofluorobenzene afforded 2-p-nitrophenyl-4,6-di-p—
methoxyphenylcyclohexane-l,3-dione SQ. Bromination (Brz/
sodium acetate) and dehydrobromination (HBr, DMF) of ég
yielded 2-p-nitrophenyl-4,6-di-p-methoxyphenylresorcinol
gg. Direct routes to gl by both dehydrogenation (DDQ) of

SQ and oxidation of g% are currently under study.

ArCHzCOZEt ———> ArCIJHCOZEt ——> ArCCOzEt

COCO Et CH

2 2
a
it éé
AT 1‘ Ar 1‘ A AI‘
6—- <___
“O H 0 H 0 0H
Ar' Ar'
23 We éé
Ar Ar
Ar = p-methoxyphenyl
O .—
O Ar'= p—nitrophenyl
Ar'

3 William Terry Suggs, Jr.

PART II

The crystal structure of 2,4-di-p-methoxyphenylcyclo-
butadiene—l,3-quinone has been determined using three—
dimensional X-ray data collected by means of a Picker
FACS—l four-circle diffractometer. The crystalline form
used was one of four modifications, two low density (1.38
g/ml), and two high density (1.54 g/ml) forms obtained
by us depending on crystallization conditions. The space
group is Pl, with unit cell dimensions a = 3.89 R,
b = 10.80 R, c = 9.03 R, a = 77.88°, 0 = 74.12°, y = 85.15°,
and a density of 1.38 g/ml indicates there is one molecule
per unit cell. A least-squares refinement of the positional
and anisotropic thermal parameters gave a final R1 value
of 6.2% for 364 reflections.

The zwitterionic nature of the molecule can be seen
by the partial delocalization of the bonds in the 4-membered
ring (which is square planar) and the bonds joining the p-
phenylene and 4-membered rings. The nearly planar p-
phenylene rings themselves are quinone—like in structure
and are parallel, but not coplanar. The columnar packing
of the molecules resembles that of graphite, and observed
interplanar spacings are 3.46 R. Stereoscopic views of the

molecule and its unit cell orientation are given below.

4 William Terry Suggs, Jr.

   

 

PART I

APPROACHES TO THE SYNTHESIS OF 2-PARA-NITROPHENYL-
4,6-DI-PARA-METHOXYPHENYL-BENZENE-1,3-QUINONE

PART II

THE CRYSTAL STRUCTURE OF
2,4-DI—PARA—METHOXYPHENYL-CYCLOBUTADIENE-1,3-QUINONE

By

William Terry Suggs, Jr.

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

1973

Parents

ii

ACKNOWLEDGMENT

The author would like to express his appreciation to
Professor Donald G. Farnum for his guidance and very
helpful assistance during the course of this research

project, as well as for his sincere friendship and interest.

Special thanks are also due to Dr. Mel Neuman for his
invaluable aid in the crystal structure determination, and
to Mr. Dave Watson for his experimental help in the latter

stages of the benzene—meta—quinone synthesis.

Finally, my life has been made very enjoyable by both

my fellow graduate students, and the art of figure skating.

iii

TABLE OF CONTENTS

Page
PART I
APPROACHES TO THE SYNTHESIS OF 2-PARA-NITROPHENYL-
4,6-DI-PARA~METHOXYPHENYL-BENZENE-1,3-QUINONE

INTRODUCTION . . . . . . . . . . . . . . . . . . . . 2
HISTORICAL . . . . . . . . . . . . . . . . . . . . . 5
RESULTS AND DISCUSSION . . . . . . . . . . . . . . . 16
EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . 53
General . . . . . . . . . . . . . . . . . . . . 53
p-Methoxyphenylacetyl Chloride. . . . . . . . . 53
Sodium p-Methoxyphenylacetate . . . . . . . . . 54
-di—p-Methoxyphenylacetic Anhydride £3. . . 54
Ethyl p-Methoxyphenylacetate gg . . . . . . . . 55

Reaction of Ethyl p- .Methoxyphenylacetate with
Iodoform. . . . . . . . . 55
Diethyl a,a'-di-p—Methoxyphenylsuccinate 3% . . 55

Reaction of Ethyl Phenylacetate with Iodoform.
Diethyl a,o'-Diphenylsuccinate 33 . . . . . . . 56
Ethyl p-Methoxyphenylmalonate 3x. . . . . . . . 57
Ethyl l,3-di-p-Methoxypheny1g1utarate 38. . . . 58
Ineflwl lngiphenylglutarate 4%. . . . . . . . . 59

iv

TABLE OF CONTENTS (COntinued)

1,3-Diphenylglutaric Acid 43. . .
1,3-Diphenylglutaric Anhydride 4% .
Reaction of 1,3-Dipheny1g1utaric Anhydride
with Methylmagnesium Bromide. Monomethyl
Ketone of 1,3-Dipheny1glutaric Acid 45. . .
Reaction of Ethyl Phenylacetate with Diethyl
Oxalate. Oxaloester of Ethyl Phenylacetate
49. . . . . . . . . . . . . . .
Ethyl d-Phenylpropenoate 5Q . .
4,6-Diphenylcyclohexane~l,3-dione 41. . .
p—Nitrophenyliodonium Bromide 51. .

A. p-Nitroiodobenzenedichloride

B. p-Nitroiodosobenzene

C. p-Nitrophenyliodonium Bromide.

2—p—Nitrophenyldimedone 5%. . .

A. Reaction of dimedone with p-Nitro-
phenyliodonium Bromide

B. Reaction of dimedone with p-Nitro-
fluorobenzene.

Oxaloester of Ethyl p-Methoxyphenylacetate 5%

Ethyl a-p—Methoxypheny1propenoate ES. . .

4,6—Di-p-Methoxyphenylcyclohexane-l,3-dione 53.

Page

59

60

61

62
63
64
65
65
65
66
67

67

68
69
69

70

 

TABLE OF CONTENTS (Continued)

2-p-Nitrophenyl- 4 ,6-di-p-methoxypheny1-
cyclohexane- l ,3 dione SQ. .

2- -p- Nitrophenyl-o 4 ,6 diphenylcyclohexane-
1,3-dione 48.

BIBLIOGRAPHY

PART II
THE CRYSTAL STRUCTURE OF
2,4-DI—PARA-METHOXYPHENYL-CYCLOBUTADIENE-l,3-QUINONE
INTRODUCTION .
EXPERIMENTAL
Preparation
Crystallization of the Polymorphs
CRYSTAL DATA .
Solution and Refinement .
RESULTS AND DISCUSSION .
SUMMARY. .
BIBLIOGRAPHY . . . . .

vi

Page

71

73
80

86
89
89
90
94
97
103
116

118

TABLE

LIST OF TABLES

PART I

APPROACHES TO THE SYNTHESIS OF 2-PARA-NITROPHENYL-
4,6—DI~PARA-METHOXYPHENYL-BENZENE-1,3-QUINONE

Reaction of 1 ,3-Dipheny1g1utaric Anhydride
with Methylating Agents. . . . .

Reaction of 4,6-Dipheny1cyclohexane-
1,3-dione with p-Nitrophenyliodonium
Bromide. . . . . . . . . . . . .
Reaction of Dimedone (1.0 eq.) with

p-Nitrohalobenzene (1.0 eq.) in a sodium
hydride-HMPT System. . . .

PART II
THE CRYSTAL STRUCTURE OF

2,4-DI-PARA-METHOXYPHENYL-CYCLOBUTADIENE-1,3-QUINONE

Crystal Data .
Observed and Calculated Structure Factors.

Final Atomic Coordinates and Anisotropic
Thermal Parameters

Hydrogen Atom Positions.

Ifitramolecular Interatomic Distances
( ) and Angles (Deg.).

Bond Lengths for Selected Cyclobutyl
Ring Systems . . . .

Equations for Selected Planes.

vii

Page

76

77

78

95

104

105
106

107

108

112

LIST OF FIGURES

FIGURE Page

PART II

THE CRYSTAL STRUCTURE OF
2,4-DI-PARA-METHOXYPHENYL-CYCLOBUTADIENE-1,3-QUINONE

1 Model of the purple compound . . . . . . . 97
2 Stereosc0pic view of the purple compound . 114
3 Crystal packing of the purple compound . . 115

viii

PART I

APPROACHES TO THE SYNTHESIS OF 2-PARA-NITROPHENYL-
4,6-DI-PARA-METHOXYPHENYL-BENZENE-1,3-QUINONE

2 William Terry Suggs, Jr.

4,6—di—p-methoxyphenylcyclohexane-l,3-dione 53. Arylation
of S3 in sodium hydride/hexamethylphosphoramide with p-
nitrofluorobenzene afforded 2-p—nitrophenyl-4,6-di-p-
methoxyphenylcyclohexane-l,3-dione 86. Direct routes to %1
by both dehydrogenation (DDQ) of 86 and oxidation of g%

(available from work of W. D. Watson) are currently under

study.

ArCHZCOZEt —> ArC'IHCOZEt ——> ArfiCOzEt

COCOZEt CH2
8%
128‘, £38
9 %
Ar' Ar'
8% £8 88
Ar Ar
Ar = p-methoxyphenyl
O -
0 Ar'= p-nitrophcnyl
Ar'

81

Benzene-1,3-quinone, however, is unknown. It is
expected to be unstable, since no uncharged resonance
contributor can be drawn for such a molecule. In fact,
its dipolar contributors are formally related to the 4
w-electron antiaromatic benzene dication as are those of

the isomeric quinones.

Ow ©

Benzene—1,3-quinone may also be represented as a
triplet diradical species, a hypothesis that could easily

be tested by esr investigation.
0 O 0'
@HQ 9 (l
O' ’

One might imagine that, if formed, a benzene-1,3—
quinone would be reactive and should be quite easy to

reduce to its hydroquinone, even when compared with the

1,2- and 1,4-quinones.

Our research efforts were concerned with the synthesis
and characterization of a benzene-1,3-quinone, in particular
the 4,6-di-p-methoxyphenyl-2-p-nitrophenylbenzene-1,3-
quinone, which we felt, from Hfickel molecular orbital
calculations, would be a highly stabilized and hence readily

accessible derivative of the parent compound (see discussion).

HISTORICAL
In the years preceding the molecular orbita13’4’5
approach to the stability of a compound, numerous attempts
at the synthesis of benzene-1,3-quinone were made.
In 1880, Stenhouse6 reported the oxidation of
trichloro-orcinol with alkaline potassium ferricyanide.

The product was a para-quinone, C7H4C1203, which was

()11 () (311

C 1 Cl 1 H C1 1
K3138 (CN)6 > 2

 

CH H 0“ CH3 OH CH H

C1 O OH

reduced to the hydroquinone.

Although he did not describe this work, he reported
that tribromo-orcinol, 2,4,6-trichlororesorcinol, and
2,4,6-tribromoresorcinol gave corresponding results.

Jackson, however, in 1896, reported that tribromo-
resorcinol was unaffected by ferricyanide, but that
alkali destroyed the compound.7

Meyer and Desamari used several different reagents

to try to convert tribromoresorcinol to a meta-quinone,

but met with failure.8 The conditions used for the
preparation of the water sensitive ortho-quinone, silver
oxide in ether with anhydrous sodium sulfate, had no
effect. Neither treating this compound with nitric acid
in cold glacial acetic acid nor boiling it in benzene
with lead dioxide produced any reaction. Treatment with
nitric acid in warm glacial acetic acid converted tribromo-
resorcinol into 2-bromo-4,6-dinitroresorcinol.

Davis and Hill, in 1929, oxidized tribromoresorcinol
with a two phase chromic acid-benzene system, and isolated

two diphenoquinone reaction products.9

OH
Br T H Cr 0

 

Benzene

Br

 

More recently, Musso, Massen, and Bohrmann10 reported
the air oxidation of an alkaline solution of 4,5,6-trimethyl-
resorcinol. The product was said to exist as any of the

following tautomers depending on the environment:

OH O OH 0
CH3 0 2 CH CH CH
.__;> __;>
[ill] ‘7§?r€> ‘<7“' <7“
CH3 .1! CH3 OH CH3 0 CH3
OH 0H
CH3 CH3 H CH3 CH3

These formulations could arise from the addition of
water to a dipolar meta-quinone, similar to the known

spontaneous reaction of ortho-quinone with water.

' OH
CH CH%:[::;]§
H20
______;>
CH 0 CH3 0
3 CH OH
CH3 3
0 0 OH
0' OH OH
______;> §>
/
H on

H OH

 

In light of this interesting result, McCarthy11
approached the synthesis of a meta—quinone through the
cyclohexadienone derivative, 2,4,6-trimethyl-4-chloro—3—

hydroxy-Z,S-cyclohexadienone %.

Since this compound is formally a meta-quinone-hydro-
chloric acid adduct, any reagent (e.g. NaH in glyme) that
would simultaneously extract both a proton and a chloride

ion should form a zwitterionic meta-quinone.

o
+ M+H- —————>
OH 0"

The most direct synthesis of g appeared to be the

Cl

t-butyl hypochlorite chlorination of 2,4,6—trimethyl—

resorcinol, a compound which was first prepared by Knecht.12

OH O

 

t‘BUOCl :
/
H OH

Equivalent amounts of the two compounds did not react
to yield the desired product but, instead, gave 42% of the

starting material and 57% of a dichlorinated derivative 3.

CH

Similar results were obtained with 2,4,6—trichlororesorcinol.

Since hypochlorite had apparently reacted with the enol
group of the desired product in preference to a second
molecule of the resorcinol derivative, McCarthy attempted to
block one of the hydroxyl groups.

Barton and Quinkertl4 had reported the synthesis of
mesorcinol monoacetate S in moderate yield by direct
photolysis of 6—acetoxy—2,4,6-trimethyl-2,4—cyclohexadienone
4 (which in turn is prepared by lead tetraacetate oxidation

of 2,4,6-trimethylphenol).

 

 

OH O OH
Pb(OAc)4 >> AC hv \_
pyrox "
EtZO OAC

84:.
an

13

10

Reaction of g with t-butyl hypochlorite gave the

desired 4—chloro-3-acetoxv-2,4,6—trimethylcyclohexadienone

6.

Ac

Removal of the acetate blocking group with cold
concentrated sulfuric acid, however, removed the desired
chlorine as well, resulting in formation of 4,5,6-trimethyl—
3,4-dihydroxy—2,S—cyclohexadienone 7, which was also
prepared by air oxidation of mesorcinol with aqueous

sodium hydroxide.10

OH

Numerous synthetic attempts at 2 via direct replacement
of the hydroxyl of Z with chlorine unfortunately met with

failure.

11

The trimethyl cyclohexadienone system was abandoned in
favor of the tri-t-butyl system since t-butyl groups had
been used successfully to stabilize another inherently

unstable molecule, the phenoxy radical.16’17

18 of 2,4,6-tri-t-butyl

Lead tetraacetate oxidation
phenol yielded 32% of the ortho-acetoxy dienone 8, which
could only be separated with great difficult from its para—

acetoxy isomer, another reaction product.

OAc

Irradiation of 8, however, gave only 4,6-di-t-butyl-

resorcinol monoacetate 9.

OAc OH
hv \

pyrex
Et 0 OAc

The triphenyl system was next investigated. The great

stability of, for example, the triphenylphenoxyl radical

12

recommended use of the sym-triphenyl nucleus for stabilizing
an otherwise unstable meta-quinone. The bulky phenyl
substituents could also prevent such a molecule from
dimerizing.

2,4,6-Triphenylphenol 10 was prepared from 1,3,5—

triphenylbenzene as indicated below.19

OH
24> HNO____Z__> ‘1’
(DE 3d) (fuming);

IQ

 

Lead tetraacetate oxidation of IQ afforded a 3:1 mixture
of 6-acetoxy-2,4,6-triphenyl-2,4-cyclohexadienone 11 and 4-
acetoxy-Z,4,6-tripheny1-2,S—cyclohexadienone 12. The

isomeric dienones could not be separated, so the mixture was

AcO
¢ Ac ¢

U. 1%

photolyzed (Pyrex, ether). The only photoproduct obtained

was 13, a bicyclohexenone. Further experiments indicated

13

 

R

that 13 arose exclusively from the ortho-dienone 11.
Additional photolysis of 13 gave rise to a complex mixture
of phenolic substances which could not be separated.
McCarthy had also mentioned the possibility of
obtaining 2,4,6-triphenylresorcinol 14, via further
photolysis of 13, followed by careful separation and

hydrolysis. The resorcinol could then be air oxidized

OH

OH
¢ ¢ , ¢
{é h: E \[:::I: hydrglyms:>
OAc OH
¢

to the triphenyl-dihydroxy-cyclohexadienone analogous to

 

7, and this derivative in turn could be directly chlorinated

to the triphenylchlorohydroxycyclohexadienone lg analogous

14

to 2. Removal of HCl from 15 would give 2,4,6-tripheny1-

benzene-1,3‘quinone.

O O 0
'd' ¢ HCI ¢ ¢ - ¢ ¢
1% OX1 ize:> ? E9 __1¥H€> -
7 H H
¢ H ¢ C1 ¢

302%

 

In later unrelated work, 2,4,6-triphenylresorcinol has

been synthesized by the following two sequences.

OH OMe
Me SO Schmidt
2 1%}
(H2804,HN3)

 

 

 

 

COZH CO2H
51.8
OH OMe
HNO , H 0
/ Z 2 Q Q
. <
OH NHz
IC1
I OH
hv
@ Q @ > “1’
(36%)
I on

(75%)

15

OH [:3 0H
1 eq. I

I

h .
H > V § 1,4 (126)
0H 3 68' I2 H ©

anh. alcohol

 

(90%)

The first reaction scheme is lengthy, and the overall
yield of 14 is low but, nevertheless, the desired compound

is produced. The second reaction schemezz’23

is an
improvement on the first, for it uses only two steps,
starting materials are readily available, and the high
yield in step 1 compensates for the low yield in step 2.
Availability of the triphenylresorcinol 14 should

provide a good test of McCarthy's chlorodienone route to

the benzene-1,3-quinone.

RESULTS AND DISCUSSION

A careful examination of past efforts directed toward
the synthesis of a benzene-1,3-quinone led us to the
conclusion that the general synthetic approach toward the
compound, via the appropriate cyclohexadienone precursor,
did not show much promise of success. We also felt that in
addition to finding a new synthetic route to the meta-
quinone, we would aim directly at the preparation of a
highly stabilized derivative of the compound.

Benzene—1,3—quinone could exist as a hybrid of any of

a large number of resonance contributors, notably:

o 0’
R1 R 3 R1 3 R R3
<———> <—-> <—-——>
_ 0' o 0
R2 R2 R2

ii 18 18 88

0’ .0
a + 1
R1 3 R1 + R3 R1 3
<—-—> <———>
.- — 0
R2 R2

16

17

Application of simple molecular orbital theory to
the problem, however, suggests that only 11-20 are important
contributors. This fact becomes evident when one calculates
atomic charge densities for benzene—meta-quinone. The
calculated charges resemble those expected for a
zwitterionic species consisting of a non-interacting allyl
cation and a B-diketone enolate anion. The non-interaction
of the two moieties is apparent when their frontier
molecular orbitals are examined. The lowest unoccupied
molecular orbital (LUMO) of the 2 n-e- allyl cation is a

vacant non-bonding orbital with the following symmetry:

 

The highest occupied molecular orbital (HOMO) of the 6 n-e-
B-enolate anion is a filled non-bonding orbital which is

similar to that of pentadienyl anion, but the presence of

 

electronegative oxygen atoms in place of carbon atoms at
positions 1 and 5 symmetrically decreases the charge density

at carbon atoms 2 and 4. Therefore, while the nodal points

18

at atoms 2 and 4 in pentadienyl anion prevent interaction
with the unsymmetrical orbitals at atoms 1 and 3 in allyl
cation, the change in charge density at atoms 2 and 4 in
the B-enolate anion may cause the nodal points to be
displaced somewhat. Although the wave function may thus
have some amplitude at atoms 2 and 4, there is still no
interaction between the two moieties because (LUMO) of the
allyl cation is antisymmetric, while (HOMO) for the enolate

anion is symmetric, i.e.

It now becomes apparent that electron releasing groups
at R1 and R2, and an electron withdrawing group at R3 would
stabilize the respective positive and negative charges
generated by the allyl cation and the B-enolate anion. If
R1=R2=para—methoxyphenyl and R3=para-nitrophenyl, the
benzene—1,3-quinone should have good stabilization and
indeed might be isolable.

Our route to 2-p-nitrophenyl-4,6—di—p-methoxyphenyl-
benzene-1,3-quinone 21 was based on the idea that simple
oxidation (e.g. Br2 or HgO) of the appropriately substituted

resorcinol 22 should afford the desired product.

19

 

 

We did not attempt to prepare the readily accessible
22,23

2,4,6-triphenylresorcinol and convert it to 2,4,6-

triphenylbenzene~l,3-quinone because of the much lower
stability of the triphenyl ring system compared to the p-
nitrOphenyl-di-p-methoxyphenyl system.

We felt that 22 might be prepared as follows:

CH30 CH CH3O\© © OCHS
3
0
'c! “ ‘
CHzo H X / 131130 BF4 \
/
KF, DMSO

, -H o 0
0 0 '0 2

O O

8% 88‘.

20

V/

8%

 

The first step would involve a Knoevenagel condensation
of methyl formate with a,a'—di-p—methoxyphenylacetic
anhydride to form the intermediate 23%, which should then

self-condense to 24. The choice of an anhydrous alkali

CH3O\ 0 OCH

88%

metal fluoride (KF) for the base in a non-aqueous solvent

26 27

was based on reports by Rand and LeGoff that various

Knoevenagel and Knoevenagel-Michael condensations catalyzed

21

by alkali metal fluorides had proceeded in fair to excellent
yields. The high basicity and low nucleophilicity of
fluoride ion together with its ability to form strong

hydrogen bonds presumably account for its reactivity.

The next step, the reaction of the acid anhydride 2%
with Meerwein's reagent (triethyloxonium fluoroborate),28
is unique. Trialkyl onium salts are among the strongest
known alkylating agents, reacting with ethers, sulfides,
ketones, B-diketones, esters, B-ketoesters, and amides on
oxygen, nitrogen, or sulfur to form onium fluoroborates
that can in turn react with nucleophilic reagents.29’30’31’32
We were hopeful that the anhydride carbonyl groups would be
nucleophilic enough to react with the "free" ethyl cation
generated by this reagent and that we could isolate the
product as its fluoroborate salt.

Step three involves base abstraction of a methyl proton

from weakly acidic (pKa % 25)33 p-nitrotoluene. The anion

formed would then condense with 25 as seen below:

 

EtO Et
+ .-
BF4
CH2
NO2
Ar = p-methoxyphenyl 22

The only precedent for this type of reaction is the
triethylamine catalyzed condensation of nitromethane
(pKa % 10)34 with 2,4,6—triphenylpyrilium tetrafluoroborate

to form 2,4,6-triphenylnitrobenzene.35

Nitromethane is much more acidic than p-nitrotoluene,
but a strong enough base (e.g. dimsyl anion) should be able
to abstract a proton, generating the desired p-nitrotolyl

anion. Furthermore, the chemical resemblance between a

23'

’/ I and a ” I group would allow
\. \
Et OEt
+ _ + _
BF4 BF4

reasonable chance of success for this reaction.

The required a,a'-di~p-methoxyphenylacetic anhydride
was easily obtained from p-methoxyphenylacetyl chloride
and sodium p-methoxyphenylacetate according to a literature

method.24’25

Attempted condensation of 88 with an equivalent of
methyl formate, with anhydrous KF in either dry ethanol
or N,N-dimethylformamide, proved fruitless. Only unreacted
starting materials were obtained, possibly due to the very
low solubility of the KF.

Further investigation of a suitable preparation of 88

37 who

turned up an article by Phalnikar and Nargund,
claimed to have synthesized the unsubstituted analog of 88,
1,3-diphenylglutaconic anhydride 88. Their reaction scheme
involved base condensation of ethyl phenylacetate with
iodoform to produce diethyl 1,3-diphenylglutaconate 88,
followed by saponification to the diacid 88 and dehydration

with acetyl chloride to the anhydride 88.

24

0 G 3
((31 H CH1 /
CH COEt + NaOEt 3 \1

 

 

 

2 /’
COzEt COZEt
£8
aq. NaOH
0 G 9
N
CH3CC1
,/ /’
‘\
£8 C021 COZH
81

A reasonable mechanism for the formation of 88 could
involve insertion of diiodocarbene into the methylene carbon
of the ethyl phenylacetate. Attack by the enolate anion of
a second molecule of ethyl phenylacetate to form the bridged

diester 88 and then dehydroiodination would afford 88.

m @9
[::::L\T 2 CHCOzEt \\
. /
HZCOZEt + .CI2 *———;> CHCOZEt -HI
: \CHco Et

 

 

88 2’ “30+ HOT 2
\ -
-HI ‘\

CHCOZEt

25

Analogous preparation of the p-methoxyphenyl derivative
seemed straightforward. Thus, we prepared the required
ethyl p-methoxyphenylacetate, 88, by the classical
esterification of commercially available p—methoxyphenyl
acetic acid. However, reaction of the ester 88 with iodoform
according to the literature procedure afforded an 11.5% yield
of a,a'-di-p~methoxyphenylsuccinate, 88, mp l43.5-l44.5°,
instead of the desired 88. The mass spectrum parent peak
at m/e 386 (calculated molecular weight for 88 is 386) and
the combustion analysis established a molecular formula
of C22H26O6' The nmr shows no resonance for the vinyl
proton for 33, but instead a two proton singlet at a higher
field (I 5.73) for the methine hydrogens of 88. The
succinate diester 88 may be formed by iodination of the

enolate of 88, followed by nucleophilic displacement of

iodine by another molecule of enolate.

 

 

CH30—<::::>—CH2COZEt CHIS ,, CH30— —CHC02Et
CH3 CH3 CH30—<:::>>—CHCOZEt
r 1 ‘
COZEt COZEt

«"38

26

In view of this result, we reran the reaction with

ethyl phenylacetate as Phalnikar and Nargund had done, and

37

obtained 2% of a white solid (lit. 15%), mp 139-140°

37 140°) plus nearly all unreacted starting material.

(lit.
The data for the product (nmr T (CC14) peaks at 2.5—2.9
(10 H,m), 5.83 (2H,s), 6.23 (4H, q, J=7), 9.15 (6H, t, J=7),
ir absorptions at 1729, 1205, and 1030 cm-1, and a mass
spectral parent peak at m/e 326 (calculated molecular weight
for 88 is 338)) were consistent with ethyl a,a-dipheny1—
succinate 88 (calculated molecular weight: 326) rather than

the glutaconate ester 88.

 

l l

COzEt COZEt

1%»?

We were still hopeful of preparing the substituted
glutaconate ester 88 and Opted to run a base condensation
of ethyl p—methoxyphenylacetate with methyl formate. The
enolic B-ketoester thus formed should react with tosyl
chloride and the resulting tosylate should be easily

displaced by another molecule of the ester enolate, forming

818-

27

base _
CH30—<::::>-CH2COZEt + HCOZMe —————€>' CH3O—<:::;>—fi COZEt

 

 

/C\
H OH
TosCl
V
CH 30 @gHCO 2E 1:
£8 + + 8% <3———————- CH30— —fi-C02Et
./C\
H OTS

Unfortunately, the initial condensation reaction
failed; only unreacted starting material was obtained,
whether KF or NaOEt was employed as the base.

Having quickly abandoned the above reaction sequence,

we devised the following route to 88:

if B
ArCHO + Arccn base > ArCCH=CHAr HCN >

3
84,

 

 

Ar = p-M60©*

28

 

 

0 CN 8N 8N 8N 8N
“ HCN d
- _ ehydratexx =
ArCCHZCHAr —————5>. Ar? CH2 CHAr ,,. ArC CHCHAr
OH
H 0*
£8 88 3
COZH 802H
1, dehydrate Ar ===CHCHAr
784 \

Numerous substituted chalcones (including p-methoxy—
phenyl) of structure 88 as well as their HCN addition

products 88 have been prepared.38

We thought that HCN could
be added to 88 to give the cyanohydrin 88, which could be
dehydrated and then hydrolyzed to the diacid. Dehydration
of the diacid would give 88.

Again, however, we met with failure. Hydrogen cyanide
did not add to 88. Only unreacted starting material was
obtained, probably due to the steric and conjugative effect
on the carbonyl group by the aromatic ring.

Another, more promising route to 88 involved
preparation of its saturated counterpart 88, which would
then be dehydrogenated. Our procedure, a modification of

40

work by Souther39 and Eliel, is outlined below. It

involves base condensation of ethyl p-methoxyphenylacetate

29

CO Et

I 2
1 . NaOEt \
CH30—<::::>—CH2COZEt + EtOZCCOZEt 2. A I/VCH30—<:::>>—?H
2
81

 

CO Et

NaOEt,

CHZI2

 

OCH OCH

 

 

 

 

 

<E_§£§Li£_.
////\\\W anhydride
CO Et CO Et
0/\ 0/ O
88 738
4.8
dehydrogenate
\V
84.

with diethyl oxalate, followed by decarbonylation to ethyl
p-methoxyphenylmalonate 88. Base condensation of 88 with

methylene iodide should yield diethyl 1,3-di-p-methoxy-

30

phenylglutarate éfi which could be saponified to the diacid
ég, then dehydrated to the anhydride gg.

Furthermore, if the anhydride gQ could not be
dehydrogenated to g%, the less reactive diester §§ could
be dehydrogenated to its unsaturated counterpart él which
in turn could be saponified to the unsaturated diacid él,

then dehydrated to form gg.

COWBOOHS 6300.Na013t 3‘3 acetic A 4
—_—“—;> anhydride "%%

COZEt CO ZEt CO 2Et CO 2Et

éfé «75% 4.9L

 

Reaction of equivalent amounts of ethyl p-methoxy-
phenylacetate, diethyl oxalate, and sodium hydride afforded
the oxaloester condensation product which was decarbonylated
by heating at 180° under vacuum for several hours. Vacuum
distillation of the resulting mixture gave 37.5% ethyl
p—methoxyphenylmalonate, bp ll4°/0.0S mm, characterized
by its spectra.

The glutarate diester éé, prepared in 50-60% yield
from £1, sodium ethoxide/ethanol, and methylene iodide,
was obtained as a light green viscous oil, bp 224°/0.1 mm,

and was characterized by its spectra. The multiplicities

31

of many of the nmr signals, however, together with the
mass spectrum, indicated that the glutarate ester ég was
contaminated with a small quantity of malonate ester éz.
Although this problem could have been remedied by more
careful fractional distillation, we chose to use the
impure product.
The following scheme was chosen: conversion of the
glutarate ester ég to its unsaturated counterpart él,
by means of bromination at the benzylic position of QQ,
and then dehydrobromination. Employment of the standard
technique41 for bromination of allylic and benzylic carbon
atoms, addition of N-bromosuccinimide with a catalytic
amount of benzoyl peroxide to the ester, afforded a mixture
of two compounds (TLC-Silica Gel), one of which was
unreacted starting material. The nmr of the mixture was
too complex for analysis so, having assumed that the other
compound was the bromo-derivative, we attempted to eliminate
HBr by reaction of the crude mixture with NaOH in dimethyl-
sulfoxide. However, only the saturated diacid was obtained.
At this time, our supply of the glutarate ester éfi
was nearly exhausted and, since it was difficult to
synthesize, we decided to prepare the more easily accessible
analog, diethyl 1,3—diphenylglutarate ég, which could be
used as a model for éfi in the bromination~dehydrobromination

studies.

32

l NaOEt

\
CH ”:7
CO Et

2
COZBt COzEt

ié

 

The glutarate ester %% was obtained in 47% yield

(11t.39

39%) in a manner analogous to that for its
p-methoxy counterpart éé.

Our experience with the dehydrobromination of
glutarate ester fig led us to adapt the procedure of
Collington and Jones,43 who had prepared cyclic a,8-
unsaturated ketones from a—bromoketones by elimination
of HBr with N,N-dimethylformamide.

Unfortunately, repeated attempts at bromination of
the diester %g with NBS resulted in only minute amounts
of the bromoester, which in turn would not eliminate HBr,
even in refluxing DMF. A photochemical bromination of
4% (Pyrex, molecular bromine) was also tried, but resulted
only in recovery of unreacted starting material.

We also attempted to form the ester enolate anion
with sodium hydride, then brominate with either molecular

bromine or pyridinium bromide perbromide, but again,

obtained only unreacted starting material.

33

Finally, a direct dehydrogenation of $& was attempted
(by refluxing the diester in o-dichlorobenzene containing
10% Pd/C), but resulted in quantitative recovery of
starting material.

Our next route to the meta-quinone is given below. It
involves saponification of the glutarate ester gg to the
diacid g3 and dehydration to the anhydride $g. Treatment
of the anhydride with dimethylcadmium (or methyl Grignard)

should given the ketoacid $3, which could be esterified to

saw“ as?

 

 

 

 

CO 2Et CO 2Et CO 2H CO 2H
0
4% 4% 4,4,
MeZCd(MeMgBr)
Base CH2N2
‘\
CO 23CH C02 H
O O
4z CH3 CH3
$3 4%

42> — inone
/, meta qu

 

34

%Q with diazomethane, then base cyclized to 4,6—diphenyl-
cyclohexane-l,3-dione $Z' The very acidic C-Z hydrogen of
4% should be easily removed by base, and the resulting anion
could displace halide ion from an activated p-nitrohalo-
benzene to form the arylated dione 4g. This dione (at
least when containing p—methoxyphenyl substituents at the
4- and 6-positions) is an immediate precursor to the meta-
quinone, needing only to be dehydrogenated to the desired
product.

1,3—Diphenylglutaric acid £3 was prepared by addition
of excess 5% aqueous sodium ethoxide to the diester gg.
The white crystalline diacid was obtained in 50% yield,

39

mp 164—166° (lit. mp 164°) after recrystallization from

water.
The white, crystalline 1,3-diphenylglutaric anhydride

39 mp 142.5-

33 was obtained in 54% yield, mp l42-l45° (lit.
l43.5°) by refluxing the diacid $3 in excess acetic
anhydride.

The formation of ketoacids from the reaction of cyclic
anhydrides with Grignard and organocadmium reagents is well

45,46,47,48

known. In a typical reaction, we prepared

dimethylcadmium by modifying the general procedures of

47 49 (see Experimental - Table 1).

deBenneville and Gilman
The dimethylcadmium reactions were generally
unsuccessful, however. Normally, workup afforded a viscous

oil consisting of some anhydride, much diacid (from

 

35

hydrolysis of the unreacted anhydride during workup), and
another compound (TLC, Silica Gel, CHClS) which was assumed
to be the desired product on the basis of nmr data for the
mixture (the appearance of a broad singlet at T 4.3 to —l.l
(carboxylic acid ~OH) and a sharp singlet at T 8.0 (methyl
ketone)). However, so little ketoacid was obtained by this
method (no doubt due to the many variables involved in
correctly preparing the dimethylcadmium) that it was
abandoned.

The Grignard reactions were generally more successful
(see experimental — Table 1). Although similar product
mixtures consisting of unreacted anhydride, diacid, and
ketoacid were obtained, the relative amount of ketoacid
formed was great enough to allow its isolation by column
chromatography (Silica Gel, benzene-chloroform). Purification
of the chromatographed product by vacuum sublimation afforded
a few percent yield of white tacky crystals of $5, which was
characterized by its spectra.

Rather than try to optimize the yield of the ketoacid,
we felt it would be better to approach the synthesis of the

model 4,6-diphenylcyclohexane-l,3-dione by a different path:

36

NaOEt
+ EtOZCCOZEt
CHZCOZEt lHCOZEt
COCOzEt
HCHO, aq. ch03
ll
CHZCCH3
NaOEt
-CO Et
0 r0 I 2
CH2
u r29

The reaction sequence involves base condensation of
ethyl phenylacetate with diethyl oxalate, and treatment
of the resulting oxaloester 42 with aqueous formaldehyde
and potassium carbonate to form ethyl a-phenylpropenoate
(ethyl atropate) 5Q. This interesting reaction
mechanistically involves removal of the methine proton of
42, resulting in the ester enolate which condenses with
formaldehyde to give the cyclic intermediate 42%. Loss of

the half-ester of oxalic acid affords 5Q.

37

0'9
0
9 ll CH2
CHCOZEt é) CCOZEt ,C\ l
|_ ____> |_ H__11_> C-COZEt
0-0 0-0 I
J: I c=o
0 Et co Et l
2 2 COZEt
%2
V/

 

C=CH2 E /C—C}\iz E [Cu~ \CHZ
GOJC‘fi/O (0%?”
Et OEt 6

éQ O

O 0 42a
H H m
+ EtOC-C-OG

The last reaction, a Michael addition-cyclization of

5Q with phenylacetone, is mechanistically represented below:

 

38

The oxaloester £2 was obtained by a modification of
the procedure of Ames and Davey50 as a yellow oil which
consisted of 80% 4% and 20% ethyl phenylacetate. The
ethyl phenylacetate was very carefully distilled off,39’51

resulting in 92.5% pure oxaloester.

 

39

Reaction of the purified oxaloester with a 10% excess
of formaldehyde and aqueous potassium carbonate afforded
73% ethyl a-phenyl propenoate (92.5% pure - still
contaminated with 7.5% ethyl phenylacetate), bp 58°/0.4 mm

(lit.50

bp 76°/1.2 mm).

4,6-Diphenylcyclohexane-l,3—dione was prepared by
reacting equivalent quantities of phenylacetone, ethyl
d—phenylpropenoate, and sodium ethoxide/ethanol. White
crystals of the dione were obtained in 26% yield by ethanol

recrystallization, mp l73.5-l76.5° (lit.50

mp 161-163°).

The presence of a B-diketone moiety was established by an
nmr of the sample in a DZO-K2C03 solution (sample pH m 9),
in which the signal due to the o-proton disappeared. The
occurrence of a single (a) proton at r 4.30 verifies the
enolic nature of the diketone. The major absorptions in the
infrared spectrum at 1600 and 1553 cm'1 also indicate the
presence of an enolic B-diketone. The mass spectrum showed

a parent peak at m/e 264 (calculated m.w. for the dione is

264) plus major fragments at 236 (P-CO), 193

+
(.GH=CH-CH2©), 179 (@-C=CH©), 160
(Pl::)-CH=CH2), 132 (Pl::]-CH=CH2-CO), and 118 ([::lcn=c=0).

The next step in our synthetic sequence was the
introduction of a p-nitrophenyl group at C-2 of the diketone
Q7. We first chose p-nitrophenyliodonium bromide 51 instead

of a p-nitrohalobenzene as the arylating agent since another

4O

cyclohexane-l,3-dione, dimedone (the 5,5-dimethyl derivative),
had been arylated at C-2 with both phenyl- and o-nitrophenyl-
iodonium halides, with sodium t—butoxide/t-butanol as the

base/solvent system.52’53

The product yields were low
(22% and 29%, respectively), but we felt that since both
dimedone and dione %7 had the same skeleton, and the p—
nitrophenyliodonium salt was probably superior to at least
the phenyliodonium salt as an arylating agent (because of

the e’ withdrawing ability of the p—nitro group),54 at least

an acceptable yield of 2—p-nitr0pheny1-4,6-diphenylcyclo-

 

hexaneel,3~dione should be obtained.
Consequently, p-nitrophenyliodonium bromide was

prepared from p-nitroiodobenzene via the following

sequencezss’56

o N— ——I flu?“ o N— --IC1 NaOH \ o N— ~10
2 O 2 2 aq. K2C03/’ 2
O
n e .

conc.
H2804,
-40°

2) aq. NaBr

 

 

o..@.@

it

41

Reaction of p-nitroiodobenzene with chlorine gas at 0°
produced 84% p-nitroiodobenzene dichloride, a bright yellow
solid, which was immediately reacted with a mixture of
ice-cold aqueous potassium carbonate/sodium hydroxide to

give 98% p-nitroiodosobenzene.55

Addition of p—nitroiodoso—
benzene to a slurry of benzene and concentrated H2804 at
—40°, followed by addition of aqueous sodium bromide to the

56

reaction mixture yielded 67% (lit. 41%) p-nitrophenyl—

iodonium bromide, 51, mp 147.5-150° (lit.56

mp 149°), which
was characterized by spectral and combustion analysis.
We adapted the dimedone arylation procedure of

Beringer53

to the dione 41. Para-nitrophenyliodonium
bromide was added to a suspension of an equivalent amount
of the dione in base. A series of reactions were run, in
which the relative amount and the nature of the base, the
addition temperature, the reaction temperature and time,
the workup procedure, and the relative amount of diaryl-
iodonium salt were varied (see Table 2 in Experimental).
The reactions run in the sodium t-butoxide/t-butanol
base/solvent system were unsuccessful (affording only
unreacted dione, some diaryliodonium salt, and some p—
nitrophenol and iodobenzene as its decomposition products),
apparently due to the low solubility of both dione and
diaryliodonium salt in t-butanol. However, use of sodium
ethoxide/ethanol as the base/solvent system, which did

dissolve the starting materials, gave the same results.

42

The reaction was also run in two equivalents of sodium
t-butoxide/t-butanol since, by the nature of the intermediate
formed in the reaction, two equivalents of base Were necessary
to drive it to completion. Again, however, only unreacted
starting materials were recovered.

At this point, we thought it advisable to turn back to
the arylation of dimedone. Consequently, reaction of
equivalent amounts of dimedone, p-nitrophenyliodonium
bromide, and sodium t-butoxide/t-butanol gave a 10% yield
of yellow crystals of somewhat impure Z-p-nitrophenyldimedone
5%, mp 230-235°, characterized by nmr and ir. Dimedone,
however, is completely soluble in t-butanol, but dione 47
is not. The yield of Z-p—nitrophenyldimedone is also too
low to be considered synthetically acceptable.

In light of these results, we sought a more suitable
base/solvent system for the arylation reaction, and finally
decided on sodium hydride/hexamethylphosphoramide (HMPT).
Sodium hydride was chosen because it is a very powerful, yet
non-nucleophilic, base. HMPT was decided upon because it

57 it is a mobile liquid,

has a number of unique properties:
miscible with water and many polar and non-polar organic
solvents (and will easily dissolve both the diones and the
diaryliodonium salt under study), has a high boiling point
(235° at 760 mm), and is aprotic (with the highest basicity

and lowest dielectric constant of all common aprotic solvents).

HMPT readily solvates cations, but not anions, and is

 

43

remarkably stable to nucleophiles, making it an ideal solvent
for reactions with nucleophilic or basic reactants. Since
reactivity of anions in aprotic solvents is greater than in
protic solvents of similar polarity (assuming the anion is
completely soluble in both types of solvent), reactions
involving anions should be accelerated by such a solvent.

Indeed, numerous alkylations58 and arylation558’59’60’61

(by
use of the respective alkyl and aryl halides) of anions
generated in HMPT on carbon, nitrogen, and sulfur have been
reported.

We again attempted to arylate 4,6-diphenylcyclohexane-
1,3-dione by formation of its enolate anion in NaH/HMPT,
and then by addition of a solution of p-nitrophenyliodonium
bromide in HMPT. Once more we were frustrated in our
attempts (see Table 2 in Experimental), since only unreacted
dione and decomposition products of the diaryliodonium salt
were obtained.

We were still convinced of the merits of the NaH/HMPT
system, and opted to use a different arylating agent for
several reasons: the diaryliodonium salt was difficult to
prepare and purify, its complex nmr spectrum and its
decomposition products made analysis of a reaction difficult,
and the arylation reactions which were run in NaH/HMPT used

58‘61 as the

simple p-nitrohalobenzenes and p—dihalobenzenes
arylating agents rather than the diaryliodonium salt.

Finally, if the arylation reaction is indeed a simple

44

nucleophilic aromatic displacement, the rate of reaction of
halobenzene with the enolate anion from the dione should be
P > C1 > Br > 1, since F has the highest negative inductive
effect and I has the lowest

We first decided to arylate readily available dimedone,
using the p-nitrohalobenzene, NaH/HMPT system. A number of
procedures were tried collaboratively with W. D. Watson?0
and these are summarized in Table 3 (Experimental). One can
see from the results in the table that the order of
reactivity of the p-nitrohalobenzene is indeed F > C1 % Br > I.
The product, Z-p-nitrophenyldimedone, can be obtained in high
purity (mp 236-238°) and in respectable yield (50%). Addition
of a 2nd equivalent of sodium hydride and especially addition
of an equivalent amount of an alkali metal salt, however, were
only detrimental to the reaction.

Having finally found an acceptable arylation procedure,
we decided to prepare and then arylate the dione that we
finally needed, 4,6—di-p—methoxyphenylcyclohexane-1,3-dione

Dione 53 was prepared in a manner analogous to that for
4,6-diphenylcyclohexane-1,3-dione, but with some significant
experimental modifications (see Experimental). The overall

reaction scheme can be seen below.

4S

2

NaOEt
CH30—<::::>—CH2COZEt + EtOZCCOZEt —————§>- CH30—<::::>—CHC02Et

COCO Et

£4.

HCHO,
aq. cho

CH30—<::::>—fiCOZEt
CH2

éé ii

 

 

 

The oxaloester 54 was obtained as a light yellow
powder, mp 50°, in 49.3% yield, and was characterized by its
nmr, ir, and mass spectra.

Reaction of 10% excess formaldehyde and aqueous
potassium carbonate with the oxaloester afforded 70.1%
ethyl q-p-methoxypheny1propenoate 55, bp 105°/0.25 mm,
whose structure was confirmed by its spectra.

4,6-di-p—Methoxyphenylcyclohexane—1,3-dione 53 was
prepared by reaction of equivalent amounts of ester 55,
p-methoxyphenylacetone, and sodium ethoxide/ethanol. White
crystals of the dione, mp 90-95°, were obtained in 25% yield
on ethanol recrystallization. The compound's structure was

confirmed by ir, nmr, mass spectral, and combustion analysis,

46

and was found to exist as the enolic B-diketone.

Reaction of equivalent amounts of dione 55 and p-
nitrofluorobenzene with a 25% excess of sodium hydride/HMPT
afforded 23.1% light yellow crystals of Z-p-nitrophenyl—
4,6-di-p-methoxyphenylcyclohexane-1,3-dione 55, mp 194-
196.5° on recrystallization from benzene. Spectral data
(ir, nmr, mass) plus combustion analysis confirmed the
triaryl dione's structure.

To extend our investigation of the arylation reaction,
we also prepared Z-p-nitrophenyl-4,6-diphenylcyclohexane-
1,3-dione 55 from 4,6-diphenylcyclohexane-l,3-dione 55 in
an analogous manner. Dione 55 was obtained as light yellow
crystals, mp 240-243° (1:1 benzene:methanol) in 28.4% yield.
Its structure was also confirmed by spectral and combustion
analysis.

Conversion of Z—p-nitrophenyl-4,6-di-p-methoxypheny1-
cyclohexane-l,3-dione 55 to the meta-quinone 55 should be
possible by either aromatization (loss of 1 mole H2) of the
dione to the resorcinol 22 which could then be oxidized
(loss of a second mole Hz) to the desired product, or by a
direct dehydrogenation of the dione (loss of 2 moles H2 in

one step without isolation of the intermediate resorcinol).

 

47

 

 

 

48

Work done by W. D. Watson in these laboratories took

adxuantage of the following known reactions.62—66

Br
Br HBr / NaOtB%-
Q0 N 35A? DMF i t - BuO
0 H 0 H HO OH
Br Br

Brr

He was thus able to prepare the resorcinol 2% as follows:

 

 

49

 

NO

This compound was valuable to us as a reference
compound in our attempts to dehydrogenate 55.

In the meantime, we also began exploring pos-
sible paths for the direct dehydrogenation of dione
55 to the meta — quinone. Common dehydrogenation
agents include sulfur, selenium, palladium and plati-

num catalysts, and several quinones. From this

h——WA

 

 

50

list, we decided to use 2,3-dichloro-5,6-dicyano-l,4-
benzoquinone (DDQ), both because of its high oxidizing
power when compared with other quinones,64 and the
relative simplicity and mildness of its reaction conditions
when compared with most of the other methods (e.g.
catalytic, sulfur, or selenium dehydrogenations normally
require very high temperatures).

Consequently, reaction of one equivalent of a dioxane
solution of dione 55 with two equivalents of DDQ for
24 hr. at reflux afforded quantitative yields of both the
DDQ hydroquinone and a black crystalline material, mp
115-125°, which, on TLC (Alumina) examination, gave two
spots, an orange one corresponding to the resorcinol
derivative 22, and a light yellow one at slightly higher
Rf. No starting material, dione 55, was indicated in
the reaction product. However, no pure compound has yet
been isolated from the second Spot. The nmr spectrum of
the crystals exhibited considerable multiplicity in the
aromatic region (12.5H), plus a sharp singlet (0.5H) at
r 4.7, and sharp singlets at T 6.1 and 6.2 (N3H each;
methoxy protons for the resorcinol and some other component),
but no methine or methylene protons. The ir indicated a
moderately strong carbonyl band at 1690 cm-1, which is not
present in either the dione 55 (1640 cm-1) or, of course,
the resorcinol 25. The mass spectrum consisted of peaks at

m/e 443 (P+, resorcinol), 441 (twice as intense as 443) plus

 

51

major fragments for the resorcinol and others. Npreak at
m/e 445 (P+, dione 55) was observed. The 3H:3H ratio of
the methoxy protons in the nmr seems to indicate that the
black crystals are a 50:50 mixture of resorcinol and some
other component.

Simple recrystallization would not separate the
compounds, so separation was attempted by dry—column
chromatography (alumina, ether).67 However, in this
particular case, the separation was not comparable to
that observed on the TLC plates, for the lone wide brown-
green band which moved down the column afforded black
crystals again, on extraction of the alumina with acetonitrile.
These crystals exhibited an nmr spectrum which was virtually
the same as that of the original crystals, but with somewhat
better resolution.

It would be tempting, at this stage, to speculate that
the meta-quinone 25 is the other component in the reaction
product, but a better separation procedure (probably a
modification of the dry—column chromatographic technique,

a preparative TLC) as well as possibly a modification in
reaction conditions will have to be found before we can state
with certainty that we have prepared it.

Nevertheless, the resorcinol, which is a key
intermediate in the meta—quinone synthesis, is available

from Watson's work,70

and we are currently exploring the
alternate oxidation pathway to our desired product. A

great variety of oxidizing agents have

52

been employed in converting 0- and p-dihydroxybenzenes into

68,69

quinones, and we are quite hopeful that we can find

the right one for our resorcinol.

EXPERIMENTAL

General. All melting points were taken on a Thomas—
Hoover apparatus and are uncorrected. Infrared spectra
were recorded either as neat films or as mulls in Nujol on
a Perkin—Elmer 137 spectrophotometer. Nuclear magnetic
resonance spectra were taken on Varian A—60, 56/60, and T-60
spectrometers. All spectra were taken at ambient temperature,
and are recorded in tau (T) values relative to tetramethyl-
silane (TMS, I 10.00). Coupling constants (Av, J) were
recorded in Hertz (Hz). An Hitachi RMU-6 spectrometer was
used to obtain all mass spectra. Elemental analyses were
performed by Spang Microanalytical Laboratory, P. O. Box 1111,
Ann Arbor, Michigan; and Chemalytics, Inc., 2330 S. Industrial

Park Drive, Tempe, Arizona.

p-Methoxyphenylacetyl Chloride. This compound was
prepared according to Farnum.25 Thionyl chloride (19.0 g,
0.16 m) was added to p-methoxyphenylacetic acid (21.0 g,
0.126 m) maintained at 45-50° in an oil bath. The mixture
was heated for 2 hours to give a dark red oil. The product
was vacuum fractionally distilled, and p-methoxyphenylacetyl

chloride (13.3 g, 57%) was collected at 97° (1.0 mm),

53

 

54

(lit.25 bp 125°/9 mm). The light yellow oil had ir (neat)
absorptions at 2940, 1825, 1515, 1470, 1300, 1275, 1190,
1110, 1030, 955, 820, 780, and 720 cm‘l.
Sodium p-Methoxyphenylaeetate. p-Methoxyphenylacetic
acid (25.0 g, 0.15 m) was dissolved in a solution of sodium
hydroxide (6.0 g, 0.15 m) in 50 ml water. The water was
evaporated at reduced pressure on a rotovap, and the white,
moist solid obtained was dried for several hours in a
vacuum oven. The sodium p-methoxyphenylacetate obtained

weighed 25.9 g (93%) and melted at l92-200° with

decomposition.

a,a'-di—p—MethoxyphenyZaeetic Anhydride 25. p-Methoxy-
phenylacetyl chloride (10.0 g, 0.054 m) was slowly added to
a stirred suspension of sodium p-methoxyphenylacetate
(10.2 g, 0.054 m) in 100 m1 dry benzene. After the addition,
the reaction mixture was stirred for 3 hr at room temperature,
then filtered, and the filtrate evaporated on a rotovap.
Evaporation afforded 16.4 g of a light brown crystalline
solid which, on recrystallization from 200 m1 of 95:5
petroleum ether (60—110°): benzene gave white crystals of
a,a'-di-p—methoxypheny1acetic anhydride (9.0 g, 53%),
nu) 74.5—76.5°; ir (Nujol) 1812, 1748, 1258 cm‘l; nmr T
((3c14) 3.25 (8H, A282 q, 1:9), 6.32 (6H, s), 6.5 (4H, s).
O C, 68.79; H, 5.73.

AnaZ. Calcd. for C H

18 15 5‘

FOLunl: (L 68.77; H, 5.80.

55

Ethyl p-Methoxyphenylacetate 25. p-Methoxyphenylacetic
acid (20.0 g, 0.12 m), thiophene-free benzene (34.5 ml),
absolute ethanol (27.6 g, 0.60 m) and concentrated sulfuric
acid (0.5 ml) were mixed in a round-bottom flask to which
was attached a Dean-Stark trap and reflux condenser. The
reaction mixture was refluxed for 24 hr, during which time
the water formed from the reaction separated in the trap.
The benzene was removed by distillation, and the red oil
remaining was allowed to cool to room temperature. The oil
was washed, first, with water, then with 5% aqueous sodium
bicarbonate, and again with water. It was then dried over
anhydrous magnesium sulfate and finally vacuum distilled.
Ethyl p—methoxyphenylacetate (15.9 g, 66%) was collected at
88° (0.4 mm); ir (film) 1739, 1515, 1250, 1163, 1036, 826
cm‘l; nmr T (cc14) 3.27 (4H, A282 q, Av = 14, J=8), 6.07
(2H, q, J=6.5), 6.41 (3H, s), 6.68 (2H, s), 8.85 (3H, t,

J=6.5).

Reaction of Ethyl p-Methoxyphenylacetate with Iodoform.
Diethyl a,a'—di—p—MethoxyphenyZsuocinate 52. Ethyl p-
methoxyphenylacetate (9.7 g, 0.05 m) was added to a cold
stirred solution of sodium ethoxide (prepared by dissolution
Of 3.5 g (0.15 m) sodium in 35 m1 anhydrous ethanol) under
£1 nitrogen atmosphere. Iodoform (9.85 g, 0.025 m) was
salowly added, and the reaction mixture stirred overnight at
IWDom temperature. Finally, it was heated on a steam bath

fO'r 1 l/2 hr, then cooled to 0°, and slowly acidified to

 

 

56

pHS with dilute acetic acid. Ether (100 ml) was added to
the reaction mixture, the sodium iodide formed was filtered
off, and the ether layer washed twice with water, and
dried over Drierite. Evaporation of the ether on a steam
bath afforded an orange semi—solid which, on two
recrystallizations from ethanol, gave white crystals of
ethyl a,a'-di—p-methoxyphenylsuccinate (1.15 g, 11.5%),
mp l43.5—l44.5°; ir (Nujol) no hydroxyl, 1724 cm-1; nmr T
(CC14) 2.89 (8H, AZBZ q, Av = 25, J=9), 5.73 (2H, s), 6.14
(4H, q, J=6.5), 6.23 (6H, s), 9.07 (6H, t, J=6.S); m/e
386, 313, 240, 225, 193.

Anal. Calcd. for C H O ' C, 68.39; H, 6.73.

22 26 6'
Found: C, 68.16; H, 6.80.

Reaction of Ethyl Phenylacetate with Iodoform.

Diethyl a,a’-Diphenylsuccinate 55. Ethyl phenylacetate
(16.42 g, 0.1 m) was added to a cold well-stirred solution
of sodium ethoxide (prepared by dissolution of 3.5 g (0.15 m)
sodium in 35 m1 anhydrous ethanol) under an atmosphere of
nitrogen. The reaction mixture rapidly formed a thick paste
hfliich had to be heated in order to form a homogeneous
ssolution. lodoform (20 g, 0.05 m) was slowly added to the
warm solution. The reaction mixture first turned orange in
color, and then black, as a rather vigorous reaction took
Place, and sodium iodide precipitated out of solution. The
Néaction mixture was stirred overnight at room temperature,

€N1d was finally heated for 1 hr on a steam bath.

57

The mixture was cooled to room temperature and acidified
to pHS with dilute acetic acid. Ether (100 ml) was then
added, the sodium iodide was filtered off, and the ether
layer washed twice with water and dried over Drierite.
Rotary evaporation of the ether afforded 20 m1 of a dark red
liquid which, on standing, slowly began to crystallize. The
semi-solid mass thus formed was recrystallized twice from
absolute ethanol to give white crystals of diethyl a,a'-
diphenylsuccinate (0.8 g, 2.5%), mp 139-l4l°; ir (Nujol) no
hydroxyl, 1729 cm’l; nmr T (0014) 2.5-2.9 (10H, m), 5.83
(2H, s), 6.23 (4H, q, J=7), 9.15 (6H, t, J=7); m/e 326, 281,

252, 207, 179, 163.

Ethyl p-Methoxyphenylmalonate 52. A mixture of ethyl
p—methoxyphenylacetate (9.7 g, 0.05 m) and diethyl oxalate
(7.3 g, 0.05 m) was added, dropwise, under nitrogen, to a
rapidly stirred, reom temperature suspension of sodium
hydride (2.2 g of a 57% mineral oil dispersion, 0.05 m) in
40 m1 dry benzene.

Following the addition, the reaction mixture was
refluxed for 3 1/2 hr, then cooled to room temperature, and
hydrolyzed with water and dilute hydrochloric acid. The
aqueous and organic layers were separated, the aqueous layer
was extracted with benzene (50 m1), and the organic and
belizene extracts were combined and dried over Drierite.

I3Vaporation of the benzene on a rotovap yielded an orange oil,

58

which was then heated at 180°/ 30 mm for 2 1/2 hr to effect
decarbonylation. Vacuum fractional distillation of the new
black oil gave ethyl p—methoxyphenylmalonate (5.0 g, 37.5%),
bp 114°/0.05 mm; ir (film) 2900, 1730, 1626, 1493, 1460,
1361, 1026, 840 cm 1; nmr T (CC14) 3.05 (4H, AZBZ q, Av =

19, J=8), 5.60 (1H, s), 5.90 (4H, q, J=7), 6.33 (3H, s), 8.82
(6H, t, J=7); m/e 266, 193, 180, 165, 148, 121.

Ethyl J,3-di—p—Methoxyphenylglutarate 55. A mixture
of ethyl p-methoxyphenylmalonate (13.3 g, 0.05 m) and
methylene iodide (6.7 g, 0.025 m) was added, dropwise, to a
vigorously stirred solution of sodium ethoxide (0.05 m -
prepared by dissolution of sodium (1.15 g, 0.05 m) in 15 m1
anhydrous ethanol) under nitrogen. After the addition, the
reaction mixture was refluxed for 67 hr. Excess ethanol
was removed by distillation, and the reaction mixture was
cooled to room temperature and slowly hydrolyzed with water.
Ether (25 ml) was added, the organic and aqueous layers
separated, and the aqueous layer extracted with 2 x 25 ml
portions of ether. The organic layer and ether extracts
were combined and dried over Drierite.

Evaporation of the ether on a rotovap gave a yellow
oil which, on vacuum fractional distillation, afforded
ethyl 1,3-di—p—methoxyphenylglutarate (5.0 g, 50%), bp
224°/0.1 mm; ir (film) 2940, 1735, 1613, 1493, 1460, 1370,
1299-1149 (bd), 833, 800 cm-1; nmr T (CC14) 2.70-3.45

 

l'"'

 

59

(8H, m), 6.05 (4H, q, J=6.5), 6.65 (2H, m), 7.52 (2H, m),
8.88 (6H, m, J=6.5 and 8.0); m/e 400, 354, 325, 292, 279,

266, 251, 220, 206, 193.

Diethyl-1,3—diphenylglutarate 52. This compound was

39 A mixture of diethyl

prepared according to Souther.
phenylmalonate (118 g, 0.5 m) and methylene iodide (67 g,
0.25 m) was added, dropwise, to a vigorously stirred
solution of sodium ethoxide (0.5 m - prepared by dissolution
of sodium (11.5 g, 0.5 m) in 150 ml anhydrous ethanol) under
nitrogen. Following addition, the mixture was refluxed for
3 days. Excess ethanol was removed by distillation, and

the reaction mixture worked up in similar fashion to that
for 55. Vacuum fractional distillation yielded diethyl—1,3-

diphenylglutarate (40.0 g, 47% - lit.39

39

32%), bp 145°/0 2 mm
(lit. bp 216°/7 mm); ir (film) 2960, 1730, 1613, 1493,
1449, 1235, 1163, 1099, 1026, 860, 735, 695 cm'l; nmr T

(CC14) 2.9 (10H, 5), 6.05 (4H, q, J=7), 6.65 (2H, t, J=8),

7.51 (2H, d, J=8), 8.91 (6H, t, J=7).

L3—Diphenylglutaria Acid 55. This compound was also
prepared according to Souther.39 Diethyl-l,3-diphenyl~
glutarate (5.0 g, 0.0147 m) was added to a warm solution of
aqueous sodium ethoxide (prepared by dissolving sodium
(1.035 g, 0.045 m) in a mixture of 14.0 ml ordinary ethanol
and 0.7 ml water). Within seconds, the entire reaction

mixture had solidified. The solid mass was washed with ether,

 

 

60

filtered, and dried on a steam bath, yielding the disodium
salt (4.1 g, 80%). The dry salt was dissolved in 30 ml H20,
decolorized with Norite, and filtered. Acidification of

the hot aqueous solution with concentrated HCl afforded a
yellow-white crystalline precipitate of 1,3-diphenylglutaric
acid, which, after recrystallization from H20, weighed 2.0 g

(47%), mp l62-164° (1it.39

164°); ir (Nujol) 2700 (bd.),
1695, 1439, 1282, 950, 730, 695 cm'l; nmr T (c0013) -1.7

(2H, bd s), 2.80 (10H, 5), 6.48 (2H, m), 7.46 (2H, m).

1,3-Diphenylglutaria Anhydride 55. This compound was

40 1,3-Diphenylglutaric acid

prepared according to Eliel.
(1.0 g, 0.0035 m) was dissolved in acetic anhydride (10.0
ml, 0.1 m) and the solution was refluxed for 1 hr. The
excess acetic anhydride was removed under vacuum, and the
dark brown residue was refluxed with 10.0 ml dry benzene

for 1/2 hr. The light yellow clear liquid was filtered and,
on addition of 10.0 ml petroleum ether (30-60°), yellow
crystals rapidly formed. The crystals were filtered and
recrystallized from 1:1 benzenezpetroleum ether (30-60°).
The resulting white crystals of anhydride 55 (0.5 g, 54%)

40

had mp 142-145° (lit. mp 142.5-143.S°); ir (Nujol) 1818,

1770, 1460, 1379, 1042, 775, 700 cm'l; nmr T (00013) 2.70
(10H, 3), 5.97 (2H, t, J=9), 7.53 (2H, t, J=9).

Anal. Calcd. for C H O

17 14
Found: C, 76.73; H, 5.39.

3: C, 76.69; H, 5.26.

 

 

61

Reaction of 1,3-DiphenngZutaric Anhydride with
Methyl—magnesium Bromide. Monomethyl Ketone of 1,3-Diphenyl-
glutaric Acid gé. In a typical reaction, methylmagnesium
bromide (3.4 m1 of a 2.95 M solution in ethyl ether - Alfa
Inorganics, no. 87324, 0.01 m) was added, dropwise, under
nitrogen, to a rapidly stirred solution of 1,3-diphenyl-
glutaric anhydride (2.22 g, 0.0083 m) in 20 ml dry
methylene chloride maintained at -10° (acetone-ice bath).

After addition, the reaction mixture was stirred at
-10° for 2 hr. Hydrolysis was accomplished at 0° by
addition of H20 and a slight excess of 10% H2804. After
separation of solid magnesium salts, the organic layer
was separated, and the aqueous layer was extracted with
ether. The organic layer and ether extracts were combined,
extracted with excess 10% NaOH (to remove the keto acid),
and the NaOH extract slowly acidified (0°) with 25% H2804.
The acidified extract stood overnight in a refrigerator,
and a small amount of orange solid precipitated out. The
solution was then saturated with sodium chloride, extracted
with ether (3 x 200 m1), and the combined ether layers washed
with water, dried over Drierite, and evaporated, yielding a
slightly pink viscous liquid. Chromatography of the liquid
(Silica Gel, 95—200 mesh, 1.5 x 21 inches), beginning with
benzene as the eluting solvent and gradually increasing
Polarity to ethyl ether yielded (fraction 7), a green viscous

oil which, on standing at room temperature, began to

62

crystallize. Sublimation (130°/O.6 mm, 15 1/2 hr) of the
oil yielded m20 mg of sticky white crystals of 45; ir
(Nujol) 2700, 1724, 1460, 1380 cm’l; nmr r (CDC13) 0.4

(1H, bd s), 2.77 (10H, m), 6.55 (2H, m), 8.05 (3H, s), 8.68

(2H, m); m/e 282, 264, 236, 22, 194, 149, 139, 91.

Reaction of Ethyl Phenylacetate with Diethyl Oxalate.
Oxaloester of Ethyl Phenylacetate £2. Diethyl oxalate
(27.38 g, 0.1875 m) was added dropwise to a magnetically
stirred solution of sodium ethoxide (0.1875 m - prepared by
dissolving sodium (4.35 g, 0.1875 m) in 54.25 m1 anhydrous
ethanol) under nitrogen. Ethyl phenylacetate (40.725 g,
0.25 m) was then added dropwise to the reaction mixture.
The entire solution solidified into a light yellow mass
within 15 minutes after the ethyl phenylacetate addition.
The solid mass was vacuum filtered, washed several times
with ether, and air-dried, affording 41.0 g (76% of the
sodium salt) light yellow powder. This solid was added to
an ice~cold solution of 111 ml 5% HCl (0.16 m HCl). The
suspension of solid in the dilute acid stood overnight at
room temperature, causing formation of an orange-red
heavier-than6water oil, which was separated. The aqueous
layer was extracted with ether, and the ether layer and oil
were combined, filtered, washed with H20, and dried over
Drierite. Evaporation of the ether gave a yellow oil which

consisted of 80% oxaloester and 20% ethyl phenylacetate

63

(determined by comparison, in the nmr, of the relative areas
of the methine proton (r 4.80) of the oxaloester and the
methylene protons (r 6.55) of ethyl phenylacetate). Careful
vacuum distillation of the ethyl phenylacetate (70°/0.5 mm)
resulted in 27.2 g (55%) of 92.5% pure (determined as before)
oxaloester, ir (film) 2940, 1725, 1650, 1370, 1266, 1064,
1031, 700 cm‘l; nmr T (cc14) 2.70 (5H, s), 4.80 (1H, s),

5.90 (4H, m), 8.82 (6H, m); m/e 264 (M+), 236, 190, 164, 145,

136, 118, 105, 91, 77.

Ethyl-u-phenylpropenoate 50. The purified oxaloester
49 was added to a solution of 38% aqueous formaldehyde
(2.9 ml, 0.039 m, 10% excess) in 9.6 m1 H20, and the
resulting emulsion stirred vigorously at 10-15°. A solution
of potassium carbonate (5.02 g, 0.039 m) in 9.3 ml H20 was
added, dropwise, to the emulsion. After the addition, the
mixture was stirred 2 hr at 10-15°. When stirring was
stOpped, two layers had formed, the lower, aqueous layer
having pH m7. A few drops of 5% HCl were added, followed by
ether (50 ml), and the 2 layers were separated. The aqueous
layer was extracted with 50 m1 ether, and the ether layers
combined and dried over Drierite. Evaporation of the ether
gave 5.8 g clear, light yellow oil which, on vacuum
distillation, afforded ethyl-a-phenylpropenoate (4.2 g, 73%,

92.5% pure* based on comparison, in the nmr, of the relative

 

*92.5% ethyl-a—phenylpropenoate, 7.5% ethyl phenylacetate

EL

 

64

areas of the vinyl protons of 50 (r 3.75 and 4.20) and the
methylene protons (r 6.55) of ethyl phenylacetate), bp
58°/0.4 mm; ir (film) 2985, 1724, 1370, 1333, 1190, 1100,
1030, 780, 700 cm‘l; nmr T (cc14) 2.70 (5H, s), 3.75

(1H, d,J=2), 4.20 (1H, d, J=2), 5.80 (2H, q, J=6), 8.75
(3H, t, J=6); m/e 176 (M+), 164, 148, 132, 103, 91, 77.

4,6-Diphenylcyelohexane-l,3-dione 32. A mixture of
ethyl—o-phenylpropenoate (4.0 g, 0.0228 m) and phenylacetone
(3.05 g, 0.0228 m) was added, drOpwise, under nitrogen, to a
warm solution of sodium ethoxide (prepared from dissolution
of sodium (0.52 g,0.0228 m) in anhydrous ethanol (14.8 m1)).
After the addition, the now bright yellow solution was
refluxed for 3 hr, then allowed to stand at room temperature
overnight.

Approximately 30 ml H20 was added to the reaction
mixture, and the now cloudy suspension (pH ~10) was cooled
to 0° and slowly acidified with 10% HCl to pH 1. On
acidification, a yellow oily solid precipitated out. The
solid was vacuum filtered, washed with H20, and air dried,
weight: 6.0 g. It was then stirred in 30 ml ethyl ether,
whereupon the ether became yellow in color and an off-white
powder remained insoluble. This powder was vacuum filtered,
washed with several portions of ether, and recrystallized
from ethanol to give white crystals of 4,6~diphenylcyclo-

hexane-1,3-dione (1.55 g, 26%), mp 173.5-176.5° (lit.50

 

 

65

mp 161-163°); ir (Nujol) 1600, 1553, 1176, 850, 760, 700
cm’l; nmr T (DMSO—dé) 2.70 (10H, 3), 4.30 (1H, s), 6.40
(2H, m), 7.50 (2H, m); m/e 264 (M+), 236, 193, 179, 160,

132, 118, 103, 90, 77.

p-Nitrophenylphenyliodonium Bromide 5%.

A. p-Nitroiodobenzenedichloride. p-Nitro-
iodobenzene (5.0 g, 0.02 m) was suspended in dry chloroform
(60 ml) maintained at 0°. The suspension was stirred very
vigorously (Hershberg mechanical stirrer), and a steady
stream of dry chlorine gas was passed over the suSpension
for 3 hr. The resulting bright yellow solid was vacuum

filtered, washed with a small volume of CHCl and air-dried,

3’
weight: 5.36 g (84%).

B. p-Nitroiodosobenzene. The p-nitroiodobenzene-
dichloride (5.36 g, 0.0167 m) was immediately and thoroughly
ground with anhydrous sodium carbonate (5.0 g) and 10 g
crushed ice in a mortar which was chilled in an ice bath.

To the resulting thick paste, 14 ml of 5M aqueous sodium
hydroxide was added in small portions, with repeated
trituration after each addition. Finally, 10 ml H20 was
added to make the mixture more fluid, and the bright yellow
suspension was refrigerated overnight. The suspension had
turned yellow-brown in color, and the solid was vacuum

filtered, washed with 4 x 30 m1 portions of H20, air dried,

66

and then washed with a small amount of acetone, weight:

4.36 g (98%), mp 87° (eXp1.).

C. p-Nitrophenylphenyliodonium bromide. p-
Nitroiodosobenzene (4.0 g, 0.015 m) was very slowly added
(25 min) to a mechanically stirred slurry of concentrated
sulfuric acid (22.5 ml) and benzene (2.34 g, 0.030 m),
which was maintained at ~35 to -40° (Dry Ice-chlorobenzene
bath). The reaction mixture turned blue-green in color

during the addition, and was stirred an additional hr at

 

—35 to -40°. The slurry was poured onto 150 g ice, the
resulting precipitate filtered off, and the clear yellow
filtrate was shaken with 0.5 g Norite, filtered, and treated
with a solution of sodium bromide (3.0 g, 0.030 m) in water
(15 m1). A light yellow precipitate immediately formed,
which was vacuum filtered, washed with H20, then ether, and
air dried to give 4.1 g (67%) of p-nitrophenylphenyliodonium

53

bromide, mp 147.5—150° (lit. mp 149°); ir (Nujol) 1613,

1567, 1527, 1351, 1010, 995, 855, 735 cm'1

; nmr r (DMSO-dé)
1.70 (m, 6H), 2.40 (m, 3H); m/e 408 (P + 2), 281, 249, 219,
204.

Anal. Calcd. for ClegBrINOZ: C, 35.46; H, 2.22.

Found: C, 35.41; H, 2.28.

67

Z-p—Nitrophenyldimedone 5%.

A. Reaction of dimedone with p-nitrophenylphenyl-
iodonium bromide. Dimedone (1.40 g, 0.01 m) was slowly
added to a well-stirred solution, under nitrogen, of sodium
t—butoxide (0.01 m - prepared by dissolving sodium (0.23 g,
0.01 m) in dry t-butanol (50 ml)) at room temperature.
After gas evolution (H2) had ceased, the temperature was
raised to 65°, and solid p-nitr0phenylphenyliodonium
bromide (4.06 g, 0.01 m) was slowly added. The reaction
mixture was stirred at 65° for 4 hr, then cooled to room
temperature. The resultant orange suspension was hydrolyzed
with H20 (150 m1), then made basic (to remove acidic
unreacted dimedone and product) by addition of a solution
of potassium carbonate (1.38 g, 0.01 m) in 10 m1 H20. The
suspension was filtered, and the filtrate extracted with
ethyl acetate (2 x 200 ml) - to remove organic products
from the decomposition of the iodonium salt. The extracted
aqueous filtrate (pH 9) was slowly acidified with 10% HCl to
pH 1, causing formation of a small amount of yellow solid.
The solid was vacuum filtered, washed with H20, and air-dried.
Recrystallization from 1:1 benzenezmethanol yielded impure
2-p-nitrophenyldimedone (0.27 g, 10%), mp 220-235°; ir (Nujol)
2600, 1582, 1515, 1340, 1031, 847, 758, 730, 700 cm‘l; nmr T
(DMSO-dé) 2.15 (4H, A2B2 q, Av = 42, J=9), 7.60 (4H, s), 8.90

(6H, s).

 

 

68

B. Reaction of dimedone with p-nitrofluorobenzene.
Sodium hydride (0.58 g, 0.024 m, 1.01 g 57% mineral oil
dispersion) was added to 20 m1 anhydrous (vacuum distilled
from calcium hydride) hexamethylphosphoric triamide (HMPT)
under nitrogen atmosphere at room temperature. After 1/2 hr
of stirring, the gray suspension was cooled to 15°, and
dimedone (2.8 g, 0.02 m) was slowly added. Considerable
bubbling (H2) occurred, and all of the dimedone had
dissolved within 15 min.

After the reaction mixture had stirred for 40 additional

 

min. and had warmed to 20°, a solution of p-nitrofluoro-
benzene (2.83 g, 0.02 m) in 15 ml HMPT was added dropwise.
Immediately upon addition, the color of the reaction mixture
turned deep red. Stirring was continued for 1/2 hr at room
temperature, then at 80° for 9 hr.

The solution was cooled to 0° and slowly hydrolyzed
with H20 until the total solution volume was 450 ml, pH 7.
Slow acidification to pH 1 with 10% HCl caused formation
of a considerable quantity of yellow precipitate. The
precipitate was refrigerated for 1 hr, then vacuum filtered,
washed with H20 (200 m1) and air dried. It was then dried
for several hours at 60° under vacuum and weighed 3.66 g.
Recrystallization from 1:1 benzenezmethanol afforded light
yellow crystals of 2-p-nitrophenyldimedone (2.59 g, 50%),
mp 236-238°; ir (Nujol) 2632, 1613, 1580, 1524, 1344, 1266,

1

1031, 760, 700 cm“ ; nmr T (DMSO-d6) 2.30 (4H, A282 q,

 

 

69

Av = 42, J=8), 7.65 (4H, s), 9.05 (6H, s); m/e 261 (M+),
246, 233, 215, 205, 177.

Oxaloester of Ethyl-p-methoxyphenylacetate £3. Diethyl
oxalate (54.76 g, 0.375 m) was added, dropwise, to a well-
stirred solution, under nitrogen, of dry benzene (150 m1)
and sodium ethoxide (0.375 m - prepared by dissolving
sodium (8.7 g, 0.375 m) in anhydrous ethanol (108.5 m1)).
Ethyl-p-methoxyphenylacetate (72.76 g, 0.375 m) was then
added dropwise. The now dark orange solution was stirred
at room temperature and, after 4 hr, had completely
solidified, The solid was vacuum filtered, washed with
ether (800 ml) and air-dried. The sodium salt (60.2 g) was
dissolved in ice water (480 ml), and the light green
solution slowly acidified with 5% HCl to pH 2, causing
formation of a considerable quantity of white crystalline
solid. The solid was vacuum filtered, washed several
times with H20, and air-dried. The oxaloester weighed
54,0 g (49.3%), mp 50°; ir (Nujol) 1770, 1739, 1299, 1242,
860 cm'l; nmr T (00013) 2.85 (4H, A282 q, Av = 13, J=9),
4.65 (1H, s), 5.7 (4H, m, J=4), 6.2 (3H, s), 8.7 (6H, m,
J=4); m/e 294 (M*), 248, 220, 193, 165, 148, 135, 121, 107.

Ethyl-a-p-methoxyphenylpropenoate £5. The oxaloester 5%
(50.0 g, 0.17 m) was added to a solution of 38% aqueous

formaldehyde (10.4 g, 0.34 m) in 91 ml H20 contained in an

 

70

Erlenmeyer flask. The suspension was cooled to 15°, stirred
vigorously, and a solution of potassium carbonate (24.6 g,
0.1785 m) in 45.5 ml H20 was added, dropwise. After the
addition, stirring was continued at 15° for 3 hr. Ethyl
ether (200 ml) was added to the 2-1ayer suspension and the
organic (upper) layer separated. The aqueous layer was
extracted with ether (200 m1), and the ether layers were
combined and dried over Drierite. Evaporation of the ether
afforded 31.0 g of a cloudy yellow oil. Vacuum distillation
of the oil at 105°/0.25 mm yielded ethyl-a-p-methoxyphenyl-
propenoate (24.6 g, 70.1%); ir (film) 3030, 1730, 1626,
1621, 1515, 1290, 1256, 1190, 1093, 1036, 840 cm—1; nmr r
(CDClS) 2.8 (4H, AZBZ q, Av = 22, J=9), 3.7 (1H, d, J=2),
4.1 (1H, d, J=2), 5.65 (2H, q, J=8), 6.2 (3H, s), 8.65 (3H,
t, J=8); m/e 206 (M+), 194, 179, 162, 133, 121.

4,6—di-p-Methoxyphenylcyclohexane-J,3—dione éé- A
solution of p—methoxyphenylacetone (4.43 g, 0.027 m) and
ethyl—a-p-methoxyphenylpropenoate (5.6 g, 0.027 m) was
added, dropwise, to a well-stirred room temperature solution
of sodium ethoxide (0.03 m - prepared by dissolution of
sodium (0.69 g, 0.03 m) in 19.5 ml anhydrous ethanol) under
nitrogen. Within 30 min. after the addition, the reaction
mixture had solidified.

The mixture was cooled to 0°, 50 ml H20 added, any

insoluble material filtered off, and the ice—cold filtrate

 

 

71

(pH 10) acidified slowly to pH 4 with 10% NC]. On
acidification, a yellow sticky solid precipitated out, and
was washed several times with H20 and air-dried. The solid
was then dissolved in ethyl ether (75 m1), and the ether
was extracted with 10% aqueous potassium carbonate (50 ml).
The carbonate layer was separated, cooled to 0°, and
acidified with 10% HCl to pH 5, affording another sticky
yellow solid. This material was also washed with H20 and
air-dried, weighing 7.1 g. Two absolute ethanol
recrystallizations gave white crystals of 4,6-di-p-
methoxyphenylcyclohexane-1,3-dione (2.2 g, 25%), mp 90-95°;
iT (Nujol) 1629, 1613, 1527, 1250, 1190, 1036, 840 cm‘l;
nmr T (acetone-d6) 3.0 (8H, q, J=9), 4.35 (1H, s), 6.2
(6H, s) overlapping 6.35 (2H, t, J=6), 7.6 (2H, t, J=6);
m/e 324 (M+), 296, 253, 227, 190, 175, 148, 134, 121, 91,
77. This material was adequate for preparations, but required
extensive purification to remove a persistent small purity
before acceptable combustion analyses could be obtained.

Anal. Calcd. for C20H2004: C, 74.05; H, 6.22.
Found: C, 73.92; H, 6.29.

2-p-Nitrophenyl-4,6—di—p-methoxyphenylcyclohexane—
1,3-dione 88°70 Sodium hydride (0.0060 g, 0.0025 m, 0.104 g
57% mineral oil dispersion) was added to 5 ml anhydrous
HMPT at room temperature under nitrogen. The light yellow
suspension was stirred for 25 min,, cooled to 10°, and

dione 53 (0.6495 g, 0.0020 m) was slowly added (washed in

with 10 m1 anhydrous HMPT). There was considerable gas

   

72

evolution, and within five min., all the dione had dissolved.
This solution was allowed to warm to room temperature,
whereupon a solution of p-nitrofluorobenzene (0.2845 g,
0.0020 m) in HMPT (1.5 ml) was slowly added. The now
orange solution was heated to 84°, and within 15 min., had
turned deep reddish-purple in color. Heating was continued
for 14 hr at this temperature.

The reaction mixture was cooled to room temperature
and hydrolyzed with 75 ml ice water to pH 8. The light
orange solution was acidified to pH 1 with concentrated
HCl, refrigerated for 1/2 hr, and filtered. Saturated
sodium bicarbonate (20 m1) and H20 (10 ml) were added to the
the filtrate, and the mixture warmed on a steam bath for
10 min., cooled to room temperature, and filtered again. The
filtrate was acidified to pH 1 with concentrated HCl, and
the resulting yellow crystals were vacuum filtered, washed
with H20, and air-dried for several hours. Recrystallization
from benzene afforded 0.2055 g (23.1%) 2-p-nitrophenyl-4,6-
di-p-methoxyphenylcyclohexane-1,3-dione, mp 194-196.5°;
ir (Nujol) 1639, 1327, 1350, 1258, 1190, 1115, 1040, 980,
840, 725, 685 cm-1; nmr r (acetone-d6) 2.05 (4H, AZBZ q,
Av = 28, J=9), 2.9 (8H, AZBZ q, Av = 15, J=9), 5.8 (2H, t,
J=8), 6.1 (6H, s), 7.5 (2H, t, J=8); m/e 445 (M+), 415,
324, 311, 296, 281, 255, 165, 148, 134, 121.

Anal. Calcd. for C H NO ' C, 70.10; H, 5.20.

26 23 6’
Found: C, 68.69; H, 5.17.

73,76“

2-p-Nitrophenyl—4,6—diphonylcyclohexane-1,3-dionv $8.
This compound was prepared in an analogous manner to that
of its p-methoxyphenyl counterpart, 56. Reaction of sodium
hydride (0.288 g, 0.012 m, 0.506 g 57% mineral oil
dispersion) and 4,6-dipheny1cyclohexane-1,3—dione 47
(2.6422 g, 0.0100 m) in 10 m1 anhydrous HMPT with a solution
of p-nitrofluorobenzene (1.4150 g, 0.0100 m) in 5 ml
anhydrous HMPT at 80° for 14 hr followed by workup yielded,
on recrystallization from 1:1 benzenezmethanol, 1.093 g
(28.4%) yellow crystals of Z-p-nitropheny1-4,6-diphenyl-
cyclohexane-1,3-dione 48, mp 240-243°; ir (Nujol) 1667, 1639,
1525, 1342, 1305, 1274, 1064, 980, 862, 790, 758, 746, 735,
704 cm 1; nmr r (acetone-dG/DMSO-dé) 2.05 (4H, AZBZ q,
Av = 28, J=9), 2.6 (10H, 5), 5.85 (2H, q, J=8), 7.4 (2H,
t, J=8); m/e 385 (M+), 355, 281, 264, 235, 193, 178, 139,
103.

Anal. Calcd. for C H NO : C, 75.00; H, 4.68.

24 18 4
Found: C, 72.80; H, 4.85.

   

 

 

 

76

A.HH ..OH ..m :oHuumouv momz no né.:oHuoonV oumconnwo :qu pome Honuo uumhuxm

.vHom ouox uso ouwuHmHooum ou »MHvHum can

"nofiuowuuxo owmmm

.Hoxma Hague on» scum vfiom ouox ecu mo cowuwHoww o>HuomHuxo omen-uwum-:o: one :ofiuomuuxo

uozuo xn vozofiHom .ousuxfie coHuomop wHoo on» mo aowumuHmeHum ewom ouaawn

nnsxnoz pudendumeec

.ovwnwxgcm one mo nowusHOm m on woven ma: wumnwwuo
Hxnuos on» ..NH-.o coauomou :H “55H5vmo nguosfiw on» on woven mm3 owwuvxncm ogu ..m-.H macauomoh :Hua

.owwHOHno ssfisvmo msopvxncm wan AmowcmmpocH mmH<V whmcmwuo Hzapoz scum venomoum moz sawsvmo axnuoEwoa

 

 

 

 

: OoNN mN WN : .- OoN :: : oNH
wumvamum o.oH mm o : : : z : .HH
: o.NH mm mm : : : : : .OH
: OoOH mN MN .. z z z : om
:owuomhuxm
Uflmmm O . V O O 2 2 Z .HOSHO : . w
: O.N OH: OH- I : I HXAHQ
cH z mm.~ : .n
addendum o.H ow- ow- HHomzo soaps H.H amH< smmZHmo .o
«.owwuwxcc<
How
uco>Hom
vumvcmum m.H stmop o : Hague o.H 6mm ..h: m.o : .m
cowuomuuxo o.~ mN
owmmm o.m stmoH o vHHom Hague o.H 6mm ..H; m.o : .v
o.m mm are dab
: o.m onwop ston : -ocoucon -oaouzon H.H ston..u: m.H : .m
ocoucon
N N con» oOm ..h: m.o
: o.~ ston o : Hu :u .Hoguo H.H 0mm ..Hn H : .N
: o.N mm N m
enmecapm o.H stmms o .cHom amp axe .em H.H comm ..s: H up A moo .H
ouzvoooum n.H:v oEHH moo“ mama Away msou «scowufloo< uco>Hom ucsoe< «aoflumpmmopm unow<
msxuoz coHuomom cofluumox :oHque< ovfiuoxnc< o>HumHom mo vogue: mcfiuwaxnuoz
mo vogue:

mucow< maflumaxzuoz :uwz owflpvxnc< oHHmHsamfixcocdHQ-m.H mo coHuomom

.H pHnm

H

77

oumconnmu may

.uosvopm and use oumuwmflooud op wowwwwwom can» mm:

.Home gonna on» mo ”.mv :owuumuuxo oumconpmo.v:m .m.m ..m ..nV Hoxma

mucoscm one mo cofiuoauuxo Hosuo .m.m ..w muoHuomoHv H mm on :owumuwwwwwom .Houmz
mo mmooxo wHOMToH a mo :ofluflvem co>Ho>sfi fi.m ..w ..nv mcofiuomou Buzz on» :H msxhoz

.ommn zufiz wouomuuxo no: mm: Home oumuoom Hxnuo esp ..e one .N meowuomou
:H .Home oumconumo xmwwwom wow anaconpmo mmooxo now; Home oumuoom Hxnuo uumuuxm
.oumuoow H»:uo AHH3 ouzuxwe :oHuowop mH mo ouv woanonxc uoauuxm "asxuoz vumvcmumcc

.Eoumxm uco>Hom-ommp on» :H ocowv 0:» mo Amosm-u-smuomz cw cowmcommamv :owusHom
wouufium m o» aaoflusHom 9&2: am me How vHHom m we venom mmz uHmm EswcoonHxanw ones

 

cofiuumuuxo
oumconhmo msaa

 

 

: Woo : o. : Oom WoH : ca
:oHuomhuxo Honum
-smHeHd<
IHOHG3 WWOUXM O.N mN I I O.N I I .m
coHHUMpro Honum Ahmzzv
IHQHN3 MMOUKW O.©H cm I 362 H.H I I .h
: o.n om mm : : : : .o
UHGUCNHW I I I I I I I .m
GOHHUMHuxo
UHWGD OZ O.m I I Smpomz O.N I I .V
AIOHMV
vkwwgwym m o? z : “mowz : : .. om
:ofiuomhuxo
Uflmmfl 02 I I I I I I I .N
Hzosm-uc
pudendum o.s oo oo smuosz .ep o.H .eo o.H .co o.H .H
mssemuosd H.ssH.meHe Home made more game Hp=6>HomV pmdm «uHam ansopoH pdon
«emsxhoz ommm

mcofiuflvcou coHuomom

 

mucsoa< o>fiumaom

opHEoum Esflcowofifixqogmowuflz-u guflz ocoflw-m.H-ocwxozofioxoaxcocdflm-o.v mo cowuomom .N ofinmh

 

 

78

Table 3. Reaction of Dimedone (1.0 eq.) with
p-Nitrohalobenzene (1.0 eq.) in a Sodium
Hydride-HMPT System

 

 

 

Reaction Amt. NaH (eq.) Halobenzene Time (hr) Temp (°C) Results

1 1.2 OZN—<::>-Br 0.5 25 31% product*
95.0 60 mp 237-240°c
27.5 25

2 2.0 " 0.75 25 13%
96.25 80 mp 227-232°
1.25 25

3 2.0 OZN—<::>—I 0.75 25 8.4%
72.25 80
24.0 25

4 2.0 OZN—<::>—C1 1.0 25 10.8%
47.0 80 mp 232-236°
10.0 25

S 1.2** ” 1.0 25 no reaction
47.0 80
26.5 25

6 1.2 ” 1.0 25 23.0%
12.5 80 mp 236—239°
12.0 25

7 1.2 OZN—<::>—F 0.5 25 50%
9.0 80 mp 236-238°

 

*Product is 2 p—nitrophenyldimedone
**1.2 eq. LiCl (anh.) also added

PART I

BIBLIOGRAPHY

79

 

 

10.

11.
12.

BIBLIOGRAPHY

L. Fieser and M. Fieser, "Advanced Organic Chemistry",
Reinhold, New York, N. Y. (1961), p. 845.

J. D. Roberts and M. Caserio, "Organic Chemistry",

W. A. Benjamin and Co., New York, N. Y. (1964), p. 923.
B. Hfickel, Zeit. Physik., 70, 204 (1931); 76, 628 (1932).
A. Streitwieser, "Molecular Orbital Theory for Organic
Chemists", John Wiley and Sons, New York, N. Y. (1961),
p. 256.

M. J. Dewar, "The Molecular Orbital Theory of Organic
Chemistry", McGraw-Hill, New York, N. Y. (1969), p. 153.
J. Stenhouse, Chem. Ber., 13, 1305 (1880).

C. L. Jackson and F. Dunlap, Amer. Chem. J., 18, 117
(1896).

R. Meyer and K. Desamari, Chem. Ber., 42, 2814 (1909).

T. Davis and J. Hill, J. Amer. Chem. Soc., 51, 493 (1929);
T. Davis and V. Harrington, ibid., 56, 129 (1934).

H. Musso, D. Massen, and D. Bohrmann, Chem. Ber., 95,
2837 (1962).

N. J. McCarthy, Ph.D. Thesis, Cornell University, 1968.

E. Knecht, Ann., 215, 83 (1882).

80

 

 

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

81

T. Zincke, Chem. Ber., 23, 3766 (1890).

D. Barton and G. Quinkert, J. Chem. Soc., 1 (1960).

F. Wessely and E. Schinzel, Monats., 84, 425 (1953).
E. Muller and K. Ley, Chem. Ber., 87, 922 (1954).

C. Cook and R. Woodworth, J. Amer. Chem. Soc., 75,
6242 (1953).

G. Bogdanov and V. Ershov, Izv. A. N. SSSR., Otd. Khim.
Nauk, 2145 (1962).

K. Dimroth, G. Brauninger, G. Neubauer, Chem. Ber., 90,
1634 (1957); K. Dimroth, F. Kalk, and G. Neubauer,
ibid., 90, 2058 (1957).

C. Graebe and P. Schestakow, Ann., 284, 316 (1894).

H. Budzikiewicz and W. Meslesics, J. Org. Chem., 24,
1125 (1959).

H. Gfisten, G. Kirsche, and D. Schulte-Frohlinde, Tet.,
24, 4393 (1968).

P. Charbrier, J. Seyden-Penne, and A. M. Fouace,
Compt. Rend., 245, 174 (1957).

J. King and F. McMillan, J. Amer. Chem. Soc., 73,

4914 (1951).

D. G. Farnum et al., J. Amer. Chem. Soc., 87, 5191
(1965).

L. Rand, J. Swisher, and C. Cronin, J. Org. Chem., 27,
3505 (1962).

E. LeGoff, J. Amer. Chem. Soc., 84, 3975 (1962).

28.
29.

30.

31.
32.

33.

34.

35.
36.

37.

38.
39.
40.

41.

42.
43.

44.

82

Meerwein, Org. Syn., 46, 113 (1966).
Meerwein et al., J. Prakt., 147, 257 (1937);

Meerwein, Chem. Ber., 89, 209, 2060 (1956).

>332:

Eschenmoser et al., Angew. Chem. Int. Ed., 3, 490
(1964).

L. Paquette, J. Amer. Chem. Soc., 86, 4096 (1964). .
G. Heiszwolf and H. Kloosterziel, Chem. Comm., 51 (1966).
C. H. Rochester, "Acidity Functions", Academic Press,

New York, N. Y. (1970), p. 26. _
D. J. Cram, ”Fundamentals of Carbanion Chemistry",
Academic Press, New York, N. Y. (1965).

K. Dimroth et al., Org. Syn., 49, 114 (1969).

J. Miller and H. Pobiner, Anal. Chem., 36, 238 (1964).

N. L. Phalnikar and K. Nargund, J. Univ. Bombay, 7,

203 (1938).

W. Davey and D. Tivey, J. Chem. Soc., 1230 (1958).

B. Souther, J. Amer. Chem. Soc., 46, 1301 (1924).

E. Eliel et al., J. Amer. Chem. Soc., 75, 4291 (1953).

L. Fieser and M. Fieser, ”Reagents for Organic
Synthesis”, Vol. 1, John Wiley and Sons, New York,

N. Y. (1967), p. 78.

A. J. Mostashari, private communication.

E. W. Collington and G. Jones, J. Chem. Soc. (C),

2656 (1969).

M. W. Rathke and A. Lindert, J. Amer. Chem. Soc., 93,
2318 (1971); M. W. Rathke and A. Lindert, Tet. Lett.,

3995 (1971).

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

S6.
57.
58.

83

C. Weizmann, E. Bergmann, and F. Bergmann, J. Chem.
Soc., 1367, 1370 (1935).

J. Cason and F. Prout, J. Amer. Chem. Soc., 66, 46
(1944); J. Cason et al., ibid., 1764 (1944); J. Cason,
ibid., 68, 2078 (1946).

P. deBenneville, J. Org. Chem., 6, 462 (1941).

P. Jones and S. Congdon, J. Amer. Chem. Soc., 81,

4291 (1959).

H. Gilman and J. Nelson, Rec. trav. Chim., 55, 518
(1936).

G. Ames and W. Davey, J. Chem. Soc., 1794 (1958).
"Handbook of Chemistry and Physics", 43rd Edition,
Chemical Rubber Co. (1961), p. 1254.

F. Beringer et al., J. Amer. Chem. Soc., 75, 2705
(1953).

F. Beringer, M. Forgione, and M. Yudis, Tetrahedron,
8, 49 (1960).

S. D. Ross, Progr. Phys. Org. Chem., 1, 38-66 (1963).
H. J. Lucas and E. R. Kennedy, "Organic Syntheses,
Coll. Vol. III”, John Wiley and Sons, New York, N. Y.
(1955), pp. 482-3.

F. Beringer et al., J. Amer. Chem. Soc., 81, 342 (1959).
H. Normant, Angew. Chem. Inter. Ed., 6, 1046 (1967).
H. Normant and T. Cuvigny, Bull. Soc. Chim. France,

1866, 1972, 1881 (1965).

 

59.
60.

61.

62..

63{

64.

65.
66.

67.

68.

69.

70.

84

J. R. Campbelle, J. Org. Chem., 29, 1830 (1964).
U. S. Rubber Co. Brit. Pat. 839420 (6/29/60);
Chem. Abstr., 54, 24549g (1960).

J. Bourdais and C. Mahieu, C. R. Acad. Sci. Paris,
Ser. C, 84, 263 (1966).

W. Foerst, "Newer Methods in Preparative Organic
Chemistry", Interscience, New York, N. Y. (1948), 1, 21.
L. M. Jackman, Adv. Org. Chem., 2, 329 (1960).

D. Walker and J. Hiebert, Chem. Rev., 67, 153 (1967).
F. Galinovsky, Monatsh., 80, 288 (1949).

N. Schamp and H. DePooter, Bull. Soc. Chim. Belges,
75, 391 (1966); N. Schamp and H. DePooter, ibid.,

77, 377 (1968).

B. Loev and M. Goodman, Chem. and Ind., 2026 (1967).
C. Buehler and D. Pearson, ”Survey of Organic Syntheses”,
Wiley-Interscience, New York, N. Y. (1970), pp. 730-733.

R. Brockhaus, Ann., 712, 214 (1968).

W. D. Watson, private communication of work done for

preparation of M. S. Thesis, Michigan State University.

 

 

 

PART II

THE CRYSTAL STRUCTURE OF
2,4-D1-PARA-METHOXYPHENYL-CYCLOBUTADIENE-1,3-QUINONE

85

 

 

INTRODUCTION

Cyclobutadiene, like benzene, has its series of
quinone isomers, the 1,2(ortho)-quinone l and the 1,3(meta)-

quinone 2.

 

 

 

 

 

 

 

 

 

 

 

 

 

0‘ R
<—-> 1 * - +/
I + _ ‘9‘€> l +
R ’ *0 R \0 " R “o \P

36 2

Numerous derivatives of each type have been

synthesized}-11

although the parent systems (R=H) have not
yet been prepared.

Both types of quinone have considerable resonance
contribution from dipolar structures, for these structures

resemble a cyclobutadiene dication 3, a closed shell 2n—

D

8

 

electron aromatic molecule.

86

87

The cyclobutadiene-1,3-quinone is related to the
benzene-1,3-quinone in that no simple uncharged structure
can be drawn for it. However, the dipolar resonance
contributors as in 2 (which also imply a singlet ground
state), are not the only possible structures which can be
written for the 1,3-quinone. The molecule might actually

have a triplet ground state, as in 4 or a bicyclobutane-

 

 

 

 

m,
dione framework, as in 5.
ll /0 R O R 0
O
0’ R .0 R O R

4. 8

The synthesis and characterization of 2,4-di-p-
methoxyphenyl cyclobutadiene-1,3-quinone 68 (known in our
laboratory as the ”purple compound") helped to shed
considerable light on the 1,3-quinone's true structure.

Although crystal structures of various cyclobutadiene—

12’13 no structure

1,2—quinones had been reported,
determination of a cyclobutadiene-l,3-quinone derivative
had been carried out. The crystal structure of the

14 is also

dipotassium salt of the squaric acid dianion 7
known. This compound is related to both cyclobutadiene-

l,2- and 1,3—quinones.

88

 

0 OK

We decided to investigate this unusual type of compound
in order to establish bond angles and bond lengths, and to
determine the effect of such a charge distribution on the
syrstem. A single crystal X-ray structure determination and
a laser-Raman spectrum (which gave inconclusive data due
to absorption of incident light by the sample) were the
methods of choice and, indeed, a final structure was
determined, though not without the appearance of a number
of interesting developments. Four different crystal
modifications were discovered, two high density and two
low density forms. The high density forms were produced
at high recrystallization temperatures, and the low
density forms were produced at lower recrystallization
temPeratures (vide infra).

11115 work is concerned with the complete X-ray

StTUCtHJre analysis of one of the low density forms.

 

 

EXPERIMENTAL

Preparation

Compound 6 was prepared by bromine oxidation of the
dihydro compound 8 (R = p-methoxyphenyl) which, in turn,
was synthesized from p—methoxyphenylacetyl chloride and

triethylamine.8

1) (c H ) N R 0 Br CH c1
CH30—<C:)>-CH2COC1 __;§_3_> m .3—2—3—9
.. <—_——.—_—.
2) ”20’ OH HO R H SnClZ, HCl
F3CCOZH

The oxidation product, 6, is a highly insoluble,
high melting, bromine—free, purple, crystalline solid,
obtained in 20-50% yield.

The purple compound could be recrystallized from
boiling benzonitrile-acetonitrile to give delicate purple
needles, mp 212-214° (dec.), which gave satisfactory
combustion analyses. The ir spectrum (KBr wafer) showed
no carbonyl absorption higher than 1650 cm 1, but had

strong, broad absorptions at 1640 and 1590 cm 1, which is

consistent with a conjugated carbonyl group. The nmr

89

80‘

 

90

spectrum (run in CF COZD—CDCIS) exhibited one sharp peak

3
at r 5.97 (area = 3.0) for the methoxy protons, and a
sharp AZB2 quartet at T 2.16 (Av = 1.35 ppm, J = 9 cps,
area = 4.1) for the phenylene protons, which also established
the magnetic equivalence of the p-methoxyphenyl groups. The
uv—visible spectrum (run in CHZCIZ) indicated xmax at 536 mu
(s = 1.41 x 105), 500 mu (6 = 4.15 x 104), and 348 mu (6 =
9.73 x 103), which is indicative of a highly conjugated
n—electron system undergoing a facile n-n* transition (as
seen by the magnitude of e). The four—membered ring's
integrity was demonstrated by the stannous chloride/CFSCOZH-
HCl reduction of Q to the dihydro compound 8 in 15—20%
yield. Compound 6 also does not exhibit an esr signal,
ruling out 4 (R = p-methoxyphenyl) as a possible structure.
The bicyclobutanedione form, 5, is eliminated since it is
not sufficiently conjugated to give rise to a band at 536 mu
in the uv—visible, and the ir absorptions at 1640 and 1590
cm'1 are at too long a wavelength to be a cyclopropanone.
Therefore, the purple compound 6 appears to be a
dipolar structure analogous to the 1,3-quinone 2 (R = p
methoxyphenyl). This conjecture is confirmed by a single

crystal X-ray structure determination.

Crystallization of the Polymorphs
Initially, our task appeared somewhat easier since unit
cell parameters for a crystal of the purple compound had

been determined via Weissenberg photographs.15 Unfortunately,

 

91

the original crystal as well as a description of the
crystallization method were, apparently, lost.

We thus decided to grow several crystals from a
benzonitrile—acetonitrile solvent system. Our procedure
involved slurrying crude, finely powdered purple compound
in boiling (82°) acetonitrile. Next, an equal volume of
boiling (190°) benzonitrile was quickly added. Due to
the large difference in boiling points between the two
solvents, there was spectacular splattering of the solution

on mixing, but in spite of this, the mixture was boiled and

 

stirred for about 30 seconds, then carefully decanted into
a crystallizing dish so that the undissolved residue
remained behind. After the purple solution stood several
hours at room temperature, small crystals deposited which
were vacuum filtered, washed with a small amount of
methylene chloride, and air—dried.

The crystals were examined under a polarizing micro—
scope and found to have two distinct forms, needles and
plates. A needle—shaped single crystal of correct size
for structure determination was found and mounted on a
goneometer head for both approximate alignment and
confirmation of the space group and unit cell parameters.15
A precession camera was chosen instead of a Weissenberg
for these determinations because data interpretation was

more straightforward and the actual operation of the camera

92

was much simpler (especially since a Polaroid cassette
and film were available).

Vinson15 determined that the purple compound
crystallized in a triclinic unit cell, P1 or P1, and had
cell parameters of a = 3.86 R, b = 8.76 R, c = 10.69 R,
and a = 100.02°, B = 95.31°, and y = 99.41°. The resulting
unit cell volume was 348.40 3,3 and the density 1.35 gm/ml,
implying 1 molecule/unit cell.

Precession photographs of the crystal of the purple
compound we had grown via the high-temperature method
revealed that it indeed crystallized in a triclinic unit
cell, but had unit cell parameters of a = 15.34 R, b =
9.99 R, c = 10.42 R, and d = 64.28°, 8 = 56.80°, and y =
76.80°. The resultant unit cell volume was 1203.78 3,3
which is 3.5 times larger than that of the original crystal!
The crystal density was experimentally determined to be

1.54 gm/ml (d 1.52), implying that there are 4 molecules

calc.
per unit cell. To determine the composition of these
crystals, we ran a mass spectrum and observed a parent peak
at m/e = 294 (calculated m.w. for Q is 294), plus major
fragments at 238(P-2CO) and 223 (P-ZCO-CHS). No high
molecular weight peaks (i.e. between 600 and 294 A.M.U.)

were observed, which effectively rules out dimer formation

or solvent system inclusion in the crystal. As a final
check, we ran an nmr (in CDClS-CFzCOZH) of the "high density"

compound, and the Spectrum was virtually identical with that

already reported for the purple compound.

 

93

We modified our recrystallization procedure by adding
hot (90-100°) benzonitrile to the boiling slurry of purple
compound in acetonitrile. No violent reaction occurred
this time when the solvents were mixed, and the crystals
were formed and collected as before. A suitable crystal
for X-ray was found, mounted, and precession photographs
taken.

The photographs revealed that the unit cell is
triclinic with the parameters a = 3.89 R, b = 10.80 R,

c = 9.03 R, and a = 77.88°, 8 = 74.12°, y = 85.15°, and a
volume of 356.61 3.3 The compound's density was
experimentally determined to be 1.38 gm/ml (dcalc. 1.35
gm/ml), implying 1 molecule/unit cell. Both the mass
spectrum and nmr spectrum of the "low density" form of the
purple compound were identical with those of the "high
density” form.

Therefore, a ”high temperature” recrystallization of
6 produces a ”high density" crystal (4 molecules/unit cell),
whereas a ”low temperature" recrystallization of the same
material using the same solvent system produces a ”low
density” crystal (1 molecule/unit cell). The two forms of
the crystal are identical both chemically and, in solution,
spectroscopically. Our collection of a partial data set
for the high density form confirms the discreteness of this

polymorph.

 

CRYSTAL DATA

The needle shaped single crystal of a low temperature,
low density form of the purple compound (mounted on a
glass fiber so that the needle axis (a of the triclinic
cell) was perpendicular to it) was mounted on a Picker
4-circ1e FACS—l diffractometer for precise alignment and
determination of the lattice parameters. Twelve reflections
were measured, and from these, the standard Picker least-
squares program calculated the lattice parameters. The
crystal data are listed in Table l.

Intensity data for the crystal were obtained by use
of the omega—scan technique of the Picker diffractometer,
using Mo Ka radiation (0.70926 R) which was monochromated
with the 002 reflections of a highly oriented graphite
crystal. The X—ray tube was at a 3.0° take-off angle, and
incident and exit beam collimators 1.0 mm in diameter were
used to restrict stray radiation. The scan speed was O.5°
per minute, and attenuators were used to prevent over-
loading of the counter. Stationary-crystal stationary—
counter background counts of 10 sec. were taken at each

end of the scan. Data were collected to 20 = 60°. A set

94

95

Table 1. Crystal Data

Molecular Formula C18H14O4
Molecular Weight 294

Crystal Habit Needle
Crystal Size, mm 0.5 x 0.1 x 0.1
u, mm.1 0.1

Space Group P1

a, R 3.89 (.01)
b, R 10.80 (.02)
c, R 9.03 (.02)
a, Deg. 77.88 (.10)
8, Deg. 74.12 (.10)
y, Deg. 85.15 (.10)
Z l

v, R3 356.61

dexp , g/mia 1.38
dcalc.’ g/ml 1.35

8Measured by flotation

96

of three reflections (0, ~2, 3; 0, -1, l; and 1, l, 1) were
monitored every one hundred reflections to test the crystal
integrity.

6 weighing scheme was applied to the

A Hughes-type1
reflections (i.e. statistically, the best weight (w) given
to a reflection equals the reciprocal of the square of the
standard deviation (0) of the intensity of that reflection
(i.e. w = l/oz); o = 0.05F (where F = the structure factor17

for a particular reflection) when F > 4Fm and o =

in.’
O‘ZOFmin. when P < 4Fmin.)’ The raw data were corrected
for Lorentz17 and polarization17 effects (program DATCOR,
CDC 6500 computer). No absorption corrections were made.

953 reflections were examined and, of these, 412 were
observed at the 20 level based on counting statistics
(i.e. the data set as a whole is a weak one, and only 412 of
the 953 reflections satisfied the statistical criterion for
observability* of I/oI > 2; I = intensity of the reflection,
and 01 (standard deviation in the count) arises from the

l 2 1/2

expression: oI = {10[(0.5 + I) + 4k (IB + 1B + l)]}
1 2

where 10 = scale factor; 0.5 = a statistical probability of
the true intensity, 1; IB + IB are the two background
1
counts; k = (scan time/background time)).
* Unobserved reflections generally arise from accidental

cancellation of opposing wave functions — there are no
systematic absencesdue to symmetry in a triclinic system.

 

 

97

For the final refinement, these 412 reflections were
further reduced to 364 for use in least-squares refinement
by setting the condition that there be a cutoff of 3e_/unit
cell (absolute basis). This cutoff value was chosen as a
recognition of an inability to fit the weakest reflections
in our calculations; e.g., a 0.75 e /unit cell error for
these reflections represents a minimum 25% error for 48
(i.e. 412—364) reflections - these reflections are 15% of

the total observed data.

Solution and Refinement.

A reasonable guess of the molecular structure could
be made (except for the relative positions of the methyl
groups), so we decided to utilize repulsion-only packing

18 based on only the heavy (carbon and oxygen)

analysis
atoms.19 A completely planar model was constructed and
oriented in the unit cell such that the long axis of the
molecule was parallel to the long diagonal of the unit cell.
The structure of the model, complete with estimated20 bond

lengths and angles, is given in Figure l. The molecule is

assumed to be centrosymmetric.

 

Figure 1. Model of the purple compound.

 

 

98

Basically, packing analysis in this case involves
only an orientation of a model compound in the unit cell
in such a way as to minimize electronic repulsions
(i.e. minimize repulsion energy) between it and
neighboring molecules in adjacent unit cells. The input

18 for the calculations

to the computer program, PACK 5,
consists of the coordinates of each atom in the rigid
molecule to be considered, as well as the unit cell
coordinates and an arbitrarily determined rotation matrix.
If the model is a good one, and the rotation matrix is
correct, least-squares refinement of the data will

converge to a minimum value.

After trial—and—error (which mainly consisted of
trying to decide how to adjust the methyl carbon),
least—squares refinement to a minimum repulsion structure
did converge to a reasonable value (4.2, based on Williams'
parametersl8) for the planar model. As a check on the
correctness of the generated coordinates for our model
compound, a Patterson map17 was constructed (program
CONNIE,22a CDC 6500 computer), utilizing data from the 412
observed reflections as well as unit cell coordinates as
input. The fractional coordinates of the strongest vector
intensities were manually estimated, and were used as

input in the determination (program PACK 5 — assumes each

intense vector is an atom) of the distances (R) between

 

 

 

99

the atoms in the molecule. Some improvement in the atomic
coordinates was forthcoming from this treatment.

We then calculated the structure factors (F) for the
molecule, based on 10 heavy atoms (omitting the methyl

carbon) with individual isotropic17

temperature factors

(B = 3.5). Input for the structure factor program

(FOROl) consisted of the lattice parameters, the fractional
coordinates (from PACK 5 and Patterson analyses) for one—
half the molecule (since it was assumed to be centro—

21’28 for carbon, hydrogen,

symmetric), scattering factors

and oxygen, and the 412 reflections (i.e. the Fobs. values).
Comparison of the calculated and observed structure

factors (both in magnitude and in sign) for the individual

reflections initially gave an agreement of R1 = 35%

(where R1 is the unweighted reliability factor, equal to

F

ZIIF ). The structure factor results were

obs.l_I calc.H

le

 

obs.I

used as the basis for a Fourier (true electron density)

map. This plot revealed to us that the methyl carbon had
indeed been misplaced in the model. The atomic coordinates
were refined by least—squares22C so that hydrogen atoms

could be located by rerunning the structure factor—difference
Fourier analysis, using as the basis the set of fractional
coordinates obtained from the least-squares refinement. The
reliability, R1, now converged to a reasonable 8.3%, and

R2 (weighted), converged to 10.2% for use of the individual

isotropic thermal parameters.

 

 

100

Finally, thermal parameters of the oxygens and carbons
were converted to anisotropic17 form (i.e. the assumption
of spherical symmetry in the isotropic case was abandoned,
and the single atomic thermal parameters were replaced by

six parameters21

(B's - represented by a tensor) which in
effect describe the thermal motion of each individual
atom as a vibration ellipsoid, i.e. motion along three
principal axes), and isotropic thermal parameters of

5.0 32 included for the hydrogens. Further structure

factor least-squares analysis of the data reduced R to

1

 

6.2% and R to 6.3%.

2
Up to this point, refinement appeared to be normal
except for a low data-to-variable ratio of 3.0 to 1
(i.e. there are 364 reflections and 121 variables - the
number of variables arising from the fact that there are
11 carbon and oxygen atoms, each with 6 anisotropic thermal
parameters and 3 coordinates (x, y, z), or 99 variables;
addition of 1 scale factor plus 7 hydrogens with 3
coordinates per hydrogen gives an additional 22 variables,
for a total of 121).
Attempted analysis of the thermal parameters by

22b) does not lead to a rigid

least-squares (program TLS
body description of the centrosymmetric molecule. This

appears to indicate that the molecule is either not behaving
as a rigid body, an assumption made in the TLS determination,

or that the thermal parameters are affected by correlation

of parameters.

101

At this point, we examined the correlation matrix17
of the structure factor equations, and found that there
was a relatively high correlation (interdependence) between
coordinates and/or thermal parameters of individual atoms,
a factor which could seriously affect refinement. The-

x, y, and z coordinates for the atoms were correlated Up

to coefficients of 0.5. This we attribute to the unusual
condition that for most atoms, (x + 2y + z) approaches zero;
this is indicated by the fact that the 121, 242, and 363
reflections are unusually strong.

We then decided to attempt refinement of the structure
factors by use of "best overall anisotropic thermal
parameters”. The best overall values were obtained by
using the values of uij (equal to Bij/8n2a:a§)l7 obtained
in the TLS analysis and converting to Bij values. Least-
squares refinement of atomic coordinates enabled Rl to

converge to 6.8% and R to 6.9% for the 364 reflections.

2
Parameter shifts in the final cycle of least-squares were
less than 0.6 times the error.

Considering our TLS results, we envisioned that the
structure is possibly acentric and that the molecule is
folded about the O-C--'C-O line. Unfortunately, there is
both insufficient data available and too high a correlation
between atomic coordinates and thermal parameters to enable

us to carry out the refinement. The data-to-variable ratio

for such a refinement would be 1.5 to 1 - clearly

 

102

unacceptable in light of the large number of weak and
unobserved reflections of the highly symmetric molecule
in the lowest symmetry (Pl) space group.

We made several attempts to further test our results.
All of the data on the crystal were retaken in an
attempt to improve our results, but least-squares l
calculations produced essentially the same results. We
also calculated packing of the resultant molecule and

found no rotational change greater than 1-2°.

 

RESULTS AND DISCUSSION

Our structure factor Table (2), final atomic
coordinates and anisotropic thermal parameters (Table 3),
hydrogen atom positions (Table 4), and intramolecular
distances and angles (Table 5) are listed.

The bond distances which we obtained indicate that
the C—C bonds in the four—membered ring are partially
delocalized, so that one may represent the purple compound

by four equivalent resonance structures, namely:

0 ' ‘ 0 o
+ + + +
Ar<kAr «A» Ar=< >1Ar +4 Ar—< >=Ar ++ Ar=< >-Ar
0 o o‘ 0'

This delocalization can further be seen when one
compares these bond lengths with these of a number of
conjugated and unconjugated cyclobutenedione, cyclobutene,
and cyclobutanedione systems (Table 6).

The data indicate that carbon—carbon double (type 2)
bonds (spZ—spz) in cyclobutane ring systems fall in the

range of 1.325—1.395 R, and carbon—carbon single (type 3)

103

 

 

 

Table 2. Observed and Calculated Structure Factors

104

 

 

104a

'ClL I 1. mus run. I 1. (on In; I I. Mus run. I I. “II 'cAL

       
     

   

   

 

       

L tons vc-L - L was 'Cll. l L runs an I L recs “I, . L rm "... . L n“ "u. ‘ L n“ "I I ‘ m“
‘ I a s . I». Ian -I - a I I Is a 2 1111 1121
" ' ' " 7 ? -) -s . «A m -I A s 4 4 $11 In I su ‘M
'5 '? 3 -0 ~ In a a su In -a a .4 s

0 non 1m -s 2 , _, _. , Q I 1 g s

" “’° ““ ‘ I l -v u n o "5012“ I -| 1 s

' ’2‘ "7 ‘* ? -) o s o y In 2 I o A

° 33‘? 1"“ ’ ‘7 -) 1 s n -v a I o —4

° " “ " " a -1 I I r J -1 - I

" "‘ '? a z s u: 5.3 .1 v 3 l -I -A

I -I I 'I '1 '1 ll) 2 I -‘ ’I 1 —fi
" " ? ) 3-5 1110 nu. -1- 9-1 1 A

' "V ' -J -: s n - -s -| ) s a

‘ 5 ‘ -\ -) A I -I 1 -1 1

I I l I § '9 I J I V
'1 '? ‘ 1 ~ I .1 I 1 I n

‘ ‘ ’ ) a o; -n -1 3 r I

1 1m: us- -Q -\ , -5 , ' -: ‘ , _

‘ "I “1 “ ’ ) t -9 n 2 4 1 a

1 ma ion -9 .1 -) v -s I 2 § 3 u

-1 -d x 1 1 s I -1 s t v

I I I A 7 -S I f 9 VP!

' “ ’ n -~ -5 z -r a «7 a n

' ’ ‘ ~ 0 I 1 z 4 nos um

'I -I h I O I '. -§ '2 § All ‘0'! -) I

I -I I. a -| a ~ I a

7 3 ‘ A l A a I -~

2 ) . ‘ -| -v r a

’ ‘ ‘ a t a -v -r t

-r us: 2 -d I _ , .. ,, 1 5

7 no. inn .9 .. a nu nn .0 a a a 5 AM In?

   

 

If

105

HNVHoo.o

HHvHoo.o-
HHVNoo.o

HNVooo.o-
Hanoo.o-
Hmcmoo.o-
HHvHoo.o-
HHvHoo.o

HNVooo.o-
HHvHoo.o-

mHVmoo.o-
mNm

HonNo.o-
HavaHo.o-
Haemoo.o

Hmvhoo.o-
HecaHo.o-
HOLHHo.o-
Haywoo.o

Hmvmoo.o-
anmHo.o-
HmcHHo.o-

HmvHoo.o
mHm

«whouoEmHmd HmEHoce UHQOHHOch< new moumcHUHoou UHEou< chHm

HmcHoo.o-
Hmvao.o
HHSOHo.o
Hmvnoo.o
Hmcmoo.o
Hmcwoo.o
Haemoo.o
Haemoo.o-
HavNoo.o
Havaoo.o

HmUmHo.o
HHm

HmcmHo.o
HNVmHo.o
Hmcwoo.o
HNVHHo.o
HmuhHo.o
HHVHHo.o
HHUNHo.o
HHVHHo.o
Hmvomo.o
HNLOHo.o

HNvNHo.o
mmm

HNcmoo.o
HHvooo.o
HNSmoo.o
HNVOHo.o
HNcwoo.o
Hmcmoo.o
HNUHoo.o
HHvsoo.o
Hmvooo.o
HmvHHo.o

Hchoo.o

HHNVaNH.o
HNHVOOH.O
HHHmeo.o
HmHVoao.o
HHHvao.o
HHHVNco.o
HHHmeo.o
HVHVHH0.0
HchHoo.o
HHHVHHO.O

HHHcamo.o
HHm

nNNVowwn.o

HHHVmNHo.o

HHHvoomm.o

HchHmNm.o

mnHvNon.o

HOHVmHAH.o

HHHSASHH.O

HHHVHHHH.9

HchwNOH.o

HmHvHooo.o-

HHvaHsH.o-

mm

NH mm HHmeme 4 HamH
+ x: mN + NH u + NHNNm + NaHHmU-mee “Egom ecu we age whouoEmHmm Hwehonu UHQOHuochm 0:94

 

HOHvaAm.o-

mmvnomm.o-

HMHSSHHH.O-

HmHvamH.o-

HmHVonm.o-

HmHvamm.o-

HHHVmme.o-

HQHVmHsH.o-

HHHUommo.o-

HHHSNHoo.o-

HaLHmMH.o-

aux

.m oHan

mevNHwo.o-

HomvmoHH.o

HHHVmHHH.o-

Hmacmmac.o-

quymmmo.o
HHHVmHHN.o
HNHSmmmN.o
Hmmvmnmo.o
HNHvomNo.o
HmmvamH.o

wacwmam.o

mum

mu

co

mo

50

00

mo

vu

MO

ND

Ho

Ho

mEOu<

Table 4. Hydrogen Atom Positions (B

A£2fl+
H4 0.
H5 0.
H7 -0.
H8 -0.
HMl -0.
HM2 0.
HM3 -0.

+The atoms are labelled as:

did

410

239
244
142
057

383

106

HMl, 2, 3 are attached to C9.

.419

.156

.033

.424

.290

.369

=6RZ)

ELS

0.110

H4 attached to C4, etc.

107

Table 5. Intramolecular Interatomic Distances (R)

and Angles (Deg.)a

Distances

Ol-Cl 1
C1-C2 l.
Cl-CZ' l.
C2-C3 l.
C3-C4 l
C3-C8 l.
C4-C5 1.
C7-C8 1.
C5-C6 1.
C6-C7 1
C6—06 l.
06-C9 1.

a
Errors referred
parentheses.

Angles

.23(l) 01—C1—C2 134.
48(2) C1-C2-C3 134.
48(2) C2—Cl—C2' 88.
41(2) C3—C2-Cl' 134.
.43(1) C2—C3—C8 121.
40(1) C2-C3-C4 119.
36(2) C3-C4—C5 119.
36(2) C4—C5—C6 120.
39(2) C5-C6—C7 121.
.39(2) C6—C7-C8 119.
35(2) C3—C8-C7 121.
38(2) C5-C6-06 114.

C7-C6-06 124.

C1-C2-C1' 91.

Ol-Cl—CZ' 136.

to the last significant digit are in

7(13)
6(14)
9(11)
3(14)
5(12)
9(12)
7(12)
2(13)
1(14)
1(14)
4(12)
9(14)
0(13)
1(11)
4(12)

108

Table 6. Bond Lengths for Selected Cyclobutyl Ring Systems

Bond

Compound Bond Length (X) Type* Reference
Purple Cl-C2 1.464 1 8
Squarate Cl- 1.469 1 l4
dianion CZ-C3 1.444 1
3-cyclohexenyl- C -C =0 1.486 1 12
cyclobutene-l,2— CZ-C3=0 1.480 1
dione C =C 1.395 2
3—phenyl— C1-C4=0 1.463 1 13
cyclobutene-l,2- CZ-C3=0 1.507 1
dione C1=C2 1.358 2
Cyclobutene C —C4 1.537 3 23

C —C3 1.537 3

C =C2 1.325 2
Tetramethyl- Cl-CZ 1.58 3 24
cyclobutane— C1-C2' 1.54 3

1,3—dione

* The type refers to the amount of delocalization described
in the text.

109

bonds (sp3-sp3) in cyclobutane ring systems are in the
1.54-1.58 R range. There is another type of bond in the
cyclobutane ring system, however, the delocalized (type 1)
or conjugated (with an aromatic or olefinic species)
bond, which is 1.444-1.507 R long. It is, in effect, an
spZ-sp2 carbon—carbon single bond with some double bond
character (increased e' density) due to interaction of the
p—lobes on the sp2 carbons. The carbon-carbon bonds in
the purple compound fall into this latter category, and
it can be seen from the resonance structures that each bond
in the central ring is, in effect, 1.25 bonds, a fair
representation for the observed bond length.

The fact that the 4—membered ring is a square can
effectively rule out any resonance contributors having

a bicyclo[l.l.0]butane central ring,

for the central ring would be distorted well out of
planarity in such a structure.26
A distortion of the nearly planar phenyl ring to a
quinone—like structure with bond lengths of 1.365 and
1.400 R is found. The shortening of the carbon-carbon

bond (spZ—spz) connecting the central ring and the phenyl

ring is as expected from the resonance forms.

110

The phenyl to methoxy oxygen bond (spo-O) is not
shortened and compares with the spZC-O bond in 1,4-di-
methoxybenzene: 1.36 3.27 It is a normal aryl-alkyl
ether bond. This would seem to indicate that the positive
charge on the aromatic system resides primarily in the ring

and not on the oxygen, i.e.

   

are (some of the equivalent) resonance contributors
to the structure (note that the cyclobutyl-to-phenyl bond

can be considered to be 1.5 bonds), but the following

 

appears to be a minor contributor.

111

Normal20 carbon-carbon spz-sp2 single bonds are
1.48 8 long, and normal20 carbon-carbon spz-sp2 double
bonds are 1.34 3 long. The average of these two values
is 1.41 3 (corresponding to a carbon-carbon bond of
order 1.5), which is equal to the observed value (1.41 R)
for the cyclobutane-to-phenyl bond in the purple compound.

25

The carbonyl bond length is normal for a ketone

carbonyl (carbonyl bond lengths are relatively insensitive
to structural changes, anyway - compare with 1.17 R (C1=O)
and 1.226 R (CZ=O) in 3—cyclohexenyl-cyclobutene-l,2—

12 1.217 R (c1=0) and 1.193 8 (02:0) in 3-phenyl-

13

dione,
cyclobutene—1,2-dione, and 1.20 R in tetramethylcyclo-
butane-1,3-dione24).
The columnar packing of the molecules along the short
a axis resembles that of graphite. At the observed inter-
planar spacing of 3.46 R, the carbons are arranged so that
C3 and C7, via the unit cell translation (1 + x, y, z), and
Cl (l-x, -y, -z), very nearly project onto C4, C6, and C2,
respectively. As a consequence, C8 (1 + x, y, z) is nearly
equidistant from the six phenylene carbons in the asymmetric
unit - the range of contacts being 3.60 to 3.89 R. An
analysis of the graphite structure energetics has shown this
packing to be a minimum energy conformation.19
The non-coplanarity of the two phenylenes in the

molecule (planes equations are given in Table 7) appears to

result from intermolecular forces. The distance along the

'5

 

 

 

112

mN.o- NN.o- “H.o- mH.o- mo.o-
mu nu cu mu «0
MH.o «0.0 mH.o
Ho Nu Ho

 

“my mEoum ocmHo-wo-u:o wouooHom

mo.o-
mo

.N use x ow HeHsoncoaHon
>HHmsu3E mH » use .«u 0» HoHHMHmm mH N .6 on HoHHmpmm mH x
nmonHow mm .moxm oHcHHUHHp can on woumHoH eke moxm Hmcowonuho one+

omHo.o- n NHouo.o - NABNomé + Xmmmw.o Hc-No-Ho

H_No..Ho.Hu.Ho0

0000.0 u Nmm00.0 - sumam.0 + xsmmw.0 mch epsmseps-ssod

H00.Hu.ou.mo.su.muv

~0HH.0 " N200.0 + sg0mm.0 + x0500.0 mch pdescpzd
:oHumSmm oanm

+mocmHm vouuoHem How mcoHumsvm .5 oHpmh

113

plane normal to the (l, 0, 0) origin is ~3.35 R, to the
(-1, 0, 0) origin is 3.57 R, and to the methyl carbon is
0.22 R. Since the methyl carbon occurs in the larger
space, we infer that this placement of the methyl group
influences the packing. The shortest contact for C9
is to 01 (x, y, 1 + 2) at 3.45 R.

Stereoscopic views of the purple compound (program
ORTEP 11,22d CDC 6500) and the unit cell of the purple

compound (looking down the normal to a phenylene plane)

are given in Figures 2 and 3, respectively.

 

Figure 2. Stereoscopic view of the purple compound.

114

Figure 3. Crystal packing of the purple compound.

 

115a

 

 

SUMMARY

The conformation of this molecule may very well be
of a flexible nature, since we have been able to isolate,
at present, four different crystalline forms of the
material, two low density (ours and Vinson's), and two
high density (plates and needles), one of which has been
indexed by us.

One can, of course, speculate on the differences and
similarities between the various forms. The two low
density forms probably differ in placement of the methyl

group relative to the plane of the molecule, while the

 

high density form may arise from some sort of intermolecular

interaction.

116

 

PART II

BIBLIOGRAPHY

117

6.

IO.

11.

12.

BIBLIOGRAPHY

E. Smutny and J. D. Roberts, J. Amer. Chem. Soc., 77,
3420 (1955); E. Smutny, M. Caserio, and J. D. Roberts,
ibid., 82, 1793 (1960).

S. Cohen, J. Lacher, and J. Park, ibid., 81, 3480 (1959).
A. Treibs and K. Jacob, Angew. Chem. Int. Ed., 4, 694
(1965).

A. Treibs and K. Jacob, Ann., 699, 153 (1966).

H. Sprenger and W. Ziegenbein, Angew. Chem. Int. Ed.,
5, 893, 894 (1966).

H. Sprenger and W. Ziegenbein, ibid., 6, 553 (1967).
H. Sprenger and W. Ziegenbein, ibid., 7, 530 (1968).
D. G. Farnum et al., J. Amer. Chem. Soc., 87, 5191
(1965); D. G. Farnum, B. Webster, and A. D. Wolf,
Tet. Lett., 5003 (1968).

A. Treibs and K. Jacob, Ann., 712, 123 (1968).

A. Treibs and K. Jacob, Ann., 74], 101 (1970).

S. Hanig and H. Pfitter, Angew. Chem. Int. Ed., 11,
431, 433 (1972).

I. L. Karle, K. Britts, and S. Brenner, Acta Cryst.,

17, 1506 (1964).

118

 

 

14.

15.

16.

17.

18.

19.

20.

21.

22.

119

C. Wong, R. Marsh, and V. Schomaker, Acta Cryst.,

17, 131 (1964).

W. MacIntyre and M. Werkema, J. Chem. Phys., 42,

3563 (1964).

J. Vinson, unpublished work, B. S. Research Project,

Michigan State University, June, 1968.

E. W. Hughes, J. Amer. Chem. Soc., 63, 1737 (1941).

G. Stout and L. Jensen, ”X-ray Structure Determination:

A Practical Guide", MacMillan, New York, N. Y. (1968).

D. E. Williams, Acta Cryst., A25, 464 (1969).

M. A. Neuman, Proc. Symp. Intermol. Forces and Packing

in Crystals, Tulane Univ., New Orleans, La., Mar. 2—3

(1970), p. 111.

J. March, ”Advanced Organic Chemistry”, McGraw-Hill,

New York, N. Y. (1968), pp. 22-24.

"International Tables for X-Ray Crystallography",

I.U.C., Birmingham, England, Vol. III (1968), p. 201.

We list the major computer programs used in the analysis.

(a) M. A. Neuman, CONNIE, a Fortran Fourier summation
program; (b) TLS, V. Schomaker and K. Trueblood,
Acta Cryst., 21, A247 (1966); (c) WRFLS, a Least-

squares refinement program based on ORFLS of

W. Busing, K. Martin, and H. Levy; (d) C. K. Johnson,

ORTEP II: "A Fortran Thermal-Ellipsoid Plot Program

for Crystal Structure Illustrations".

 

 

23.

24.

25.

26.

27.

28.

120

E. Goldish, K. Hedberg, and V. Schomaker, J. Amer.
Chem. Soc., 78, 2714 (1956).

P. Friedlander and J. Robertson, J. Chem. Soc., 3083
(1956).

"Tables of Interatomic Distances and Configuration in
Molecules and Ions", The Chemical Society, London
(1968).

I. Haller and R. Srinivasan, J. Chem. Phys., 41, 2745
(1964).

T. Goodwin, M. Przybylska, and J. Robertson, Acta
Cryst., 3, 279 (1950).

R. Stewart et al., J. Chem. Phys., 42, 3175 (1965).