Synthesis and activity of some quinoxaline analogues of streptonigrin


Material Information

Synthesis and activity of some quinoxaline analogues of streptonigrin
Physical Description:
x, 156 leaves : ill. ; 29 cm.
Rock, Charles Paul, 1955-
Publication Date:


Subjects / Keywords:
Streptonigrin -- analogs & derivatives   ( mesh )
Quinoxalines -- chemical synthesis   ( mesh )
Quinoxalines -- analogs & derivatives   ( mesh )
Quinoxalines -- toxicity   ( mesh )
Drug Design   ( mesh )
Structure-Activity Relationship   ( mesh )
Antibiotics, Antineoplastic   ( mesh )
Department of Medicinal Chemistry thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Pharmacy -- Department of Medicinal Chemistry -- UF   ( mesh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1992.
Includes bibliographical references (leaves 151-155).
Statement of Responsibility:
by Charles Paul Rock.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002294154
notis - ALP7362
oclc - 49675775
System ID:

Full Text








I acknowledge my great appreciation to Dr. K.V. Rao for his guidance

and supervision throughout the course of this work and to the other members

of my committee, Dr. K.B. Sloan, Dr. Raymond J. Bergeron, and Dr. James A.

Deyrup. I would also like to thank Dr. Stephen G. Schulman for his

participation and interest.

I greatly appreciate all of the encouragement and support that I received

from all of my family, especially my parents and my wife.



ACKNOW LEDGEMENTS ............................................................... ii

LIST OF TABLES ........................................................................ v

LIST OF FIGURES ....................................................................... vi

ABSTRACT ............................................................................... ix


I. INTRODUCTION ............................................................. 1

Mechanism of Action ................................................ 3
Synthesis of Streptonigrin .......................................... 6
Structure-Activity Relationships .................................. 13

QUINOXALINE-5,8-DIONE ............................................. 18

Introduction ............................................................. 18
Quinoxaline Background ............................................ 24
Present Synthetic Approach ....................................... 29
Alternative Scheme for the 2-Phenylquinoxaline ............ 35
Experimental ............................................................ 38

CHLOROQUINOXALINE ................................................ 57

7-Amino-3-Chloro-6-Methoxyquinoxaline-5,8-Dione ........ 57
7-Amino-3,6-Dimethoxy-2-Phenylquinoxaline-5,8-Dione ... 61
Experimental ............................................................ 64

DERIVATIVES .......................................................... 79

Dione ..................................................................... 79
5,8-Dione ................................................................ 83
Experimental ............................................................ 86

SYSTEM ................................................................... 96

6-Amino-7-Methoxy-3-Phenyl-2(1 H)-Quinoxalinone-5,8-
Dione .................................................................. 96
6-Amino-7-Methoxy-1 -Methyl-3-Phenyl-2( 1 H)-
Quinoxalinone-5,8-Dione ....................................... 97
Experimental ............................................................ 102

QUINONES ........... .................................................. 113

6-Piperidinoquinoxaline-5,8-Diones .............................. 113
6-Amino-7-Bromoquinoxaline-5,8-Diones ..................... 114
Experimental ............................................................ 116

VII. STUDIES OF BIOLOGICAL ACTIVITY............................... 126

Introduction ............................................................. 126
Materials and Methods .............................................. 128
Results and Discussion .............................................. 132
Conclusions ............................................................. 145

REFERENCES .................................................................... 151

BIOGRAPHICAL SKETCH .................................................... 156



A Comparison of the Chemical Shifts Between Some
N-methylquinoxalin-2(1 H)-ones and 2-
Methoxyquinoxalines ............................................100

L1210 Cytotoxicity Assay ...........................................135

Antibiotic Assay ...........................................................139

Root Growth Inhibition Assay .........................................144

Overview of Activity ...............................................146-147

Table I.

Table II.

Table III.

Table IV.

Table V.


Figure Page

1.1 Streptonigrin and Derivatives ............................................... 2

1.2 SN Partial Syntheses: AB, CD, and ABC Ring Syntheses ......... 7-8

1.3 W einreb Synthesis of SN .................................................... 10

1.4 Kende Synthesis (SCHEME V) and Boger & Panek Synthesis
(SCHEM E V I) ..................................................................... 12

1.5 Simpler Synthetic Analogues of SN ....................................... 15

2.1 6-Methoxy-7-Aminoquinoxalinequinone Analogues of SN Prepared
in Present Study ......................................................... 20

2.2 Quinoxaline Ring System ..................................................... 25

2.3 Need for Symmetric Starting Materials in Quinoxaline Ring
Synthesis Using the Condensation Method ............................ 25

2.4 Quinoxalinequinones Previously Synthesized .......................... 27

2.5 Regio-Controlled Synthesis of 6-Methoxy-7-
Aminoquinoxalinequinones Via the Quinoxalinone ................... 30

2.6 Alkylation Pathway in to Quinoxaline Synthesis ...................... 36

2.7 NMR Spectrum of Compound 2.1 ......................................... 40

2.8 NMR Spectrum of Compound 2.6 ......................................... 44

2.9 NMR Spectrum of Compound 2.9 ......................................... 47

2.10 NMR Spectrum of Compound 2.11 ....................................... 49

2.11 NMR Spectrum of Compound 2.12 ....................................... 51

Figure Page

2.12 NMR Spectrum of Compound 2.13 ...................................... 52

2.13 NMR Spectrum of Compound 2.14 ...................................... 54

3.1 Synthetic Pathway to Chloroquinoxalineaminoquinone and
Tetrazolo-(1,5)-quinoxalineaminoquinone .............................. 60

3.2 Synthetic Route to Methoxyquinoxalineaminoquinone ........... 63

3.3 NMR Spectrum of Compound 3.2 ....................................... 65

3.4 NMR Spectrum of Compound 3.3 ....................................... 67

3.5 NMR Spectrum of Compound 3.7 ....................................... 70

3.6 NMR Spectrum of Compound 3.12 ...................................... 74

3.7 NMR Spectrum of Compound 3.15 ...................................... 76

3.8 NMR Spectrum of Compound 3.16 ...................................... 78

4.1 Synthetic Pathway to Cyanoquinoxalineaminoquinone ........... 82

4.2 Synthetic Scheme for the Synthesis of Carbomethoxyquinoxaline
am inoquinone .................................................................. 85

4.3 NMR Spectrum of Compound 4.9 ....................................... 91

4.4 NMR Spectrum of Compound 4.15 ...................................... 95

5.1 Reactions Leading to the Quinoxalinoneaminoquinones .......... 98

5.2 NMR Spectrum of Compound 5.3 ....................................... 104

5.3 NMR Spectrum of Compound 5.9 ....................................... 108

5.4 NMR Spectrum of Compound 5.10 ...................................... 111

5.5 NMR Spectrum of Compound 5.11 ...................................... 112


Figure Page

6.1 Various Reactions of Nucleophiles on Quinoxalinequinones ........ 115

6.2 NMR Spectrum of Compound 6.2 ........................................ 118

6.3 NMR Spectrum of Compound 6.3 ........................................ 119

6.4 NMR Spectrum of Compound 6.4 ........................................ 121

6.5 NMR Spectrum of Compound 6.5 ........................................ 122

6.6 NMR Spectrum of Compound 6.6 ........................................ 125

7.1 Compounds Tested in Present Work ..................................... 127

7.2 Comparisons of Compounds 2.12, 3.7, 3.12, 3.16, 4.9, 4.15,
6.7, and 6.6 ..................................................................... 133

7.3 Comparisons of Compounds 6.2, 6.3, 6.4, 6.5, 5.9, 5.10, 5.11,
and IsoPyQ ....................................................................... 134

7.4 Comparisons of Compounds 2.12, 3.7, 3.12, 3.16, 4.9, 4.15,
6.7, and 6.6 ..................................................................... 137

7.5 Comparisons of Compounds 6.2, 6.3, 6.4, 6.5, 5.9, 5.10, 5.11,
and IsoPyQ ....................................................................... 138

7.6 Comparisons of Compounds 2.12, 3.7, 3.12, 3.16, 4.9, 4.15,
6.7, and 6.6 ..................................................................... 142

7.7 Comparisons of Compounds 6.2, 6.3, 6.4, 6.5, 5.9, 5.10, 5.11,
and IsoPyQ ....................................................................... 143

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



Charles Paul Rock

December 1992

Chairman: Dr. Koppaka V. Rao
Major Department: Medicinal Chemistry

Streptonigrin (SN) is a potent anti-tumor antibiotic first isolated from

Streptomvces flocculus. It has displayed a wide range of activity in vitro on

several genera of bacteria, rodent tumor cell lines, and viral strains including

HIV I. Extensive activity has been seen in vivo including that reported in Phase

III clinical trials where significant results were obtained in the treatment of

lymphomas both as a single entity drug or in combination therapy. The toxicity

reported in some of the trials has limited its use.

Structurally, SN is a complex phenyl-pyridylquinoline derivative

containing an aminoquinone moiety and several sites for metal complexation.

In the presence of NADH, the complex is reduced, and the resulting

dihydroquinone is capable of autooxidation. From this, a radical oxygen species

is produced which can attack DNA. This rather unique mechanism of action


has been the focus of many studies in order to determine the structure-activity

relationships involved hoping to find analogues with greater activity/ toxicity


In working toward this goal, the present study was carried out. Several

quinoxaline analogues of a simpler nature were prepared and tested in three in

vitro assays. These analogues contained the same quinone ring substituents

as the parent compound. Also, the substituent at the 3-position was varied

from electron-donating to electron-withdrawing in order to determine the effect

on activity created by this change. The levels of activity were compared to

those of SN, other simpler quinoline and isoquinoline aminoquinone analogues

previously reported, and quinoxalinequinone compounds which did not contain

the aminoquinone moiety.

The cytotoxicity profile seemed to concur with previous findings that the

presence of electron-withdrawing substituents leads to increased activity.

Cytotoxicity can be seen with compounds without the aminoquinone, but this

system is a requisite for antibiotic activity in these compounds.

The synthetic approach to these quinoxaline analogues differed greatly

from all the reported methods for quinoxalinequinones. In order to maintain

positional control of the substituents, the compounds were made via the 2(1 H)-

quinoxalinone. The nitration of the various intermediates gave products which

provided several avenues to the desired 7-amino-6-methoxyquinone system.


Streptonigrin (SN) (Fig. 1.1-A) is an antitumor antibiotic isolated from the

cultures of StreDtomyces flocculus by Rao and Cullen in 1960.1 It was found

to inhibit the growth of both Gram-positive and Gram-negative bacteria at very

low concentrations.2 When tested against twenty-nine transplanted rodent

tumors, it showed a broad spectrum of activity.3 It was also effective against

transplanted human tumors in animals and in cell cultures.4 Inhibition of viral

replication was also seen against Rauscher leukemia and murine sarcoma

viruses,6 and SN was shown to be one of the most powerful inhibitors of avian

myeloblastosis reverse transcriptase and HIV I.r'e

Early testing in human patients confirmed the antitumor activity of SN.

Objective and subjective improvement was seen in a fair percentage of the

terminal patients used in the studies, but toxicity to the bone marrow and

gastrointestinal tract was noted.7'8'9 Later, double-blind studies showed SN to

be comparable in effectiveness to chlorambucil in the treatment of malignant

lymphomas and chronic lymphoblastic leukemia. Toxicity was only slightly

higher in the SN-treated patients, which could be accounted for by factors such

as average age and broader range of initial performance status of that

group.10'11 In the mid-seventies, several clinical trials were conducted where

SN was used in combination therapies with favorable results.12.13'14 One such



A) R = OH (SN)

B) R = OCH3

C) R = NHR'

D) R = OR'

Figure 1.1. Streptonigrin and Derivatives

trial was conducted on patients having either lymphosarcoma or reticulum cell

sarcoma, none of whom had previous chemotherapy. SN was given with

vincristine and prednisone, and this regimen was compared with that of

vincristine-prednisone-cyclophosphamide and vincristine-prednisone alone

(standard treatment). The SN combination was found to be far superior in the

treatment of lymphosarcoma, with nearly half the patients achieving complete

remission. Interestingly, there was less hematologic and gastrointestinal

toxicity with the SN combination than with the cyclophosphamide-containing

combination.12 Further clinical studies have not been conducted.

Mechanism of Action

Since early reports noted that the effects of SN in bacterial cells and

human leucocytes involved degradation of DNA, much work has been done to

understand the mechanism involved.15 Streptonigrin was shown to cause

single strand breaks in DNA as determined by nicking of covalently closed

circular DNA."1 Temperature melting curves (Tm) of SN-treated DNA showed

that cross-linking of DNA was not involved.17 The absence of change in the

viscosity of SN-treated DNA solutions, as well as the lack of a hypochromic

shift in the absorption spectra, eliminated intercalation as an important factor.'1

Fragmentation of the DNA was also not observed, thus indicating that no

double strand scission occurred.18

Certain conditions are said to be required for SN to express its activity

in vitro and, presumably, in vivo. These are 1) availability of 02, 2) a


one-elctron source, e.g. NADH, and 3) a divalent transition metal cation, e.g.

Cu"2 or Fe+2. Under these conditions, the degradation of DNA occurred, which

was attributed to the generation of a reactive radical species in close proximity

to the DNA.19 Early studies demonstrated that SN disrupted mitochondrial

respiration by uncoupling the oxidative phosphorylation process and by

generating H202.20 Electron spin resonance studies showed that SN, in the

presence of NADH, was reduced to a semiquinone radical,21 and spin-trapping

studies showed that SN, in the presence of oxygen, produced a hydroxyl

radical.22 Since many compounds can generate H202 without causing damage

to DNA, it was suspected that either the semiquinone or the hydroxyl radical

was responsible for the observed activity.23

There has been much speculation regarding the role of certain transition

metal cations in the activity of SN. The fact that the metal ion is essential was

proved by the absence of SN-induced cleavage of DNA in the presence of a

chelating agent such as EDTA.20

The precise nature of the interaction of SN with DNA has been the focus

of many studies. Streptonigrin readily complexes with certain divalent

transition metal cations, and this complex binds to DNA to an extent dependant

on the concentration of the SN-metal complex.19'24'25 Inexplicably, difference

spectra indicated that a molar ratio of 7:1 of Zn+2 to SN was needed to obtain

the E6x of this complex. Titration curves of the SN-Zn+2 complex with DNA

further indicated that saturation of this complex was reached with 20-25 moles

of DNA (nucleotides).24 Although there are several possible sites on the SN

molecule which might bind metal ions, it is difficult to rationalize the

requirement of 7 equivalents of the metal ion for saturation.

Among the many cations found to complex with SN, Zn+2 and Cu+2 were

the most often studied. Antimicrobial activity of the SN-Zn+2 complex was

found to be greater than that of the SN-Cu+2 complex.'9 Interestingly, these

two complexes differ significantly in their UV spectra and reduction

potentials.26 The addition of Zn"2 to SN was found to lower the reduction

potential of SN while Cu+2 raised it. This fact, coupled with the difference in

UV spectra, would indicate a different site of binding on SN for these two


The proposed events involved in the activity shown by SN is said to start

with the complexation with a divalent transition metal, interaction of this

complex with DNA, reduction of SN to the semiquinone, and autooxidation

back to the quinone with concomitant generation of a radical species capable

of attacking the DNA. The role of the metal may involve either assistance in

the reduction step, stabilization of the semiquinone radical, or assistance in the

autooxidation step.

In a recent study, SN was shown to inhibit the action of topoisomerase

II in vitro. The usual cleaving and rejoining of DNA by this enzyme is

interrupted by SN before rejoining can occur. Streptonigrin was thought to

stabilize the topoisomerase II-DNA complex which has been shown to induce

cleavage of DNA.27

The antiviral activity seen with SN is often difficult to evaluate due to the

simultaneous cytotoxicity seen at concentrations tested in vitro.' Activity is

said to arise from the electron-acceptor ability of SN, where mitochondrial

oxygen consumption is uncoupled from ATP production, leading to generation

of H202.28 Studies showed that inhibition of reverse transcriptase (RT) by SN

in vitro could only be overcome by addition of more enzyme which indicated a

1:1 interaction of SN with RT.29 Several simpler bicyclic quinones also showed

equivalent RT inhibition but without the cytotoxicity seen with SN. Since

several of the quinones, which were inactive in the assay, contained bulky side

chains as compared to the active compounds, the existence of a "quinone

pocket" on the enzyme, accessible only to planar molecules, was proposed as

the site of inactivation of the virus.30

Synthesis of Streotonigrin

The structure of SN was determined through degradative and

spectroscopic studies in 196331 and was later confirmed by X-ray

crystallography.32 Due to the complex and highly functionalized nature of the

molecule, twenty years elapsed before the first total synthesis was reported.33

During that time, several groups made contributions through partial

syntheses.34-43 Two other total syntheses were also subsequently


Liao and Cheng37 synthesized an AB ring precursor using a Skraup

condensation to form (1) (Fig.1.2-SCHEME I). This was converted to



(1 )

1) HNO3/H2SO4
2) H2/PtO2
3) No2Cr207

1) Br2/AcOH
2) NaN3/EtOH
3) H2/PtO2



1) PhPOCl2
2) 5XPd/C
4) NoOH/H20
5) Br2/NoOMe
6) TsOH
7) A920

Figure 1.2. Streptonigrin Partial Syntheses: AB, CD, and ABC Ring Syntheses



0 R R 0
Smooe00 0 mo "e 00M

NOH Br02 Br 02 NO H3
R = H, OMe

(7) (8) (9)

R 0
) 1) No2S204 0
2) Nob04 MO

o,02 N0 H3 3) NN3 NH2 COOMe
4) No2S204 0 NH3

(9) (10)

Figure 1.2. (continued)

6-methoxyquinolinequinone (2) in 3 steps. After bromination, reaction with

ammonia led to displacement of the 6-methoxyl by the amine instead of giving

the expected 6-methoxy-7-aminoquinone. This problem was circumvented by

conversion of the 7-bromoquinone to the 7-azidoquinone, followed by reduction

to the 7-aminoquinone (3).

A CD ring precursor was also prepared by this group which required 18

steps.3" Condensation of (4) (Fig. 1.2-SCHEME II) with ethyl cyanoacetate gave

the pyridone (5). This was converted to a pyridine derivative in two steps, and

after several more steps including hydrolysis of the cyano group to the amide

and Hoffman degradation to the amine, the desired CD ring unit (6) was

obtained. Neither of these bicyclic analogues were used directly in the total

synthesis of SN.

Rao synthesized ABC ring analogues of SN using a modified Friedlander

condensation.39 An eight step sequence provided (7) (Fig.1.2-SCHEME III)

which when reacted with (8), gave the nitrochalcone (9). Reductive cyclization

followed by oxidation with periodic acid gave the bromoquinone which was

converted to the aminoquinone (10) in two steps.4'41

The first total synthesis of SN was reported by the Weinreb group.33 It

required 34 steps and produced a 0.011 % overall yield. As seen in Fig.1.3-

SCHEME IV, the key steps in the synthesis included 1) a Diels-Alder reaction

with an imine to give the CD ring precursor, 2) a ten step sequence used to

introduce the 5'-amino group of the C-ring, which involved a [2,3]- sigmatropic

1) heot
2) BoOH

1) mrCPBA
2) Ac20
3) K2C03/Me ,
4) SOCI2
5) C4HeN-CH2CN
6) K+tBuO-/DMSO

1) Oxalic acid
3) KMnO4
4) Modified


CH3 N 00Me

NHf- CHz

1) NoN SN
3) deprotect

Figure 1.3. Weinreb Synthesis of SN


CH -_C13 +

rearrangement, and 3) introduction of the 7-amino group in the quinone ring via

the iodide.42

The Kende group followed with the second total synthesis, which was

completed in 27 steps with an overall yield of 0.069% (Fig. 1.4-SCHEME V).44

They modified the pyridone synthesis of Liao and Cheng by condensing the

same aminobutenone with methyl acetoacetate instead of ethyl cyanoacetate,

which gave (11), the 3-acetylpyridone derivative. Reduction of the keto group

followed by halogenation/ dehydration gave the 2-chloro-3-vinyl pyridine. This

was further converted to the 2-acetyl derivative (12) the by introduction of a

cyano group and reaction with methyl magnesium bromide. After condensing

this molecule with the appropriate aminobenzaldehyde imine, the tetracyclic

product was subjected to transformations in ring A which provided a route for

conversion to a quinone. Before the quinone was made, however, a six step

procedure was employed to convert the 3'-vinyl group to the desired amino

group, and the 6'-methyl group to a carbomethoxy function (13). After

formation of the quinone, six more steps were needed to obtain the methyl

ester of SN.

Boger and Panek synthesized SN in 17 steps with a 0.5% overall yield.45

Fig.1.4-SCHEME VI shows the two [4+2] cycloaddition reactions which

provided the basic framework of the molecule. The remaining steps included

conversion of the 3'-carbomethoxyl to the amine, and development of the

substitution pattern for the A-ring.





R H 2

1) TFA/ 0oC
2) HN03/ CH3NO2
3) Me2S04



1) 004 MeO
2) Nal04
3) Se2O

4) NaCO03
5) MeOH/ AcCI
6) (PhO)2PON3/ Et3N

R' = -CH2C6H4-p-OMe
X = -CH=NC6H4-p-Me



MeO o



Me02C-/ \)-CO2Me



C02M -



Figure 1.4. Kende Synthesis (SCHEME V) and Boger & Panek Synthesis

1) NaBH4
2) POC13
3) CuCN
4) CH3MgBr

10 steps



R H3

Structure-Activity Relationshios

Numerous analogues have been made and tested over the years in order

to understand the structural features necessary for optimal activity and to find

an analogue with a more favorable activity/ toxicity ratio than that of SN. A

structure-activity profile for SN has been developed based on data from

antimicrobial, antitumor, antiviral and even clinical testing. The scope ranged

from manipulations on SN itself, to the synthesis of simpler analogues

containing quinoline, isoquinoline, quinoxaline and quinazoline units.

The methyl ester of SN (Fig.1.1-B) has been evaluated in clinical trials,

and it showed good activity in the treatment of Hodgkins disease,

lymphosarcoma, and several carcinomas. It was reported to have a greater

therapeutic index and to be better tolerated by patients than SN.46 It showed

no antimicrobial or antiviral activity in vitro, but it retained equivalent activity

on animal tumors at higher concentrations. Another analogue, isopropylidine

azaSN, also underwent clinical trials on the basis of its significant antitumor

activity in animals at doses 102 times of SN.47 However it failed to show any

objective responses.

Based on the results from animal studies, a number of conclusions were

drawn regarding structural elements of SN needed for activity. The amino

group in the quinone ring (ring A) was found to be necessary because activity

was lost when it was replaced with a hydroxyl or a methoxyl. Reductive

acetylation or methylation in ring A to the triacetyl- or tetramethyl-

dihydroquinone also caused loss of activity, and this indicated that the quinone

was necessary. Converting ring A to a 5,6-dimethoxy-o-quinone derivative

caused a loss of activity which may indicate that the para-quinone is important

for activity. The only analogue with variation in ring B in the study was the

tetrahydroSN, and it was found to retain some activity. In ring C, acetylation

of the amino group led to an inactive compound, and esterification of the

carboxyl group gave active products. In ring D, acetylation or methylation of

the hydroxyl was compatible with activity, whereas demethylation to the

trihydroxy compound caused a loss of activity.48

A series of amides and esters were prepared from the carboxyl group of

SN.49,50 The amides [Fig.1.1-C: R'= (CH2)nN(CH3)21 exhibited high activity

against RT with no cytotoxicity or antimicrobial activity. Only a few analogues

showed the latter activities, such as the hydrazide (Fig.1.1-C: R' =NH2) and the

dimethylaminoalkyl esters [Fig.1.1-D: R'=(CH2),N(CH3)2], but only at doses

higher than that of SN.

Rao found that destrioxyphenyl-SN (Fig.1.2-10), in which ring D of SN

was replaced by a hydrogen, was twice as active as the parent compound

against Bacillus subtilis in the disc-plate assay. When ring D of SN was

replaced by a bromine, activity of the analogue dropped to 60% of that of SN.

When the analogue contained an unsubstituted phenyl or pyridyl group as ring

C (Fig.1.5-17, 18), only 25% of the activity was retained. Interchanging the

amino and methoxyl substituents in the quinone ring gave tricyclic analogues

which retained this same percentage of activity."1

(14) (15)

R = NO2, NH2

R = NO2, NH2

(16) o: R = H
b:R = CH3

R = NO2, NH2

R = NO2, NH2


R = NO2, NH2

R = NO2, NH2

R = NO2, NH2

R = NO2, NH2

Figure 1.5. Simpler Synthetic Analogues of SN


Boger et al. compared the antimicrobial and cytotoxic activities of the

bicyclic and tricyclic analogues (Fig.1.5-14, 15, 16a, 16b) with SN and its

methyl ester to assess the role of the free carboxyl group of ring C. Since

esterification of SN greatly reduced cytotoxic potency, and amide formation

eliminated cytotoxicity, analogue 16a was expected to show good activity.

However, this analogue was less active than 14, 15, and 16b. Actually,

analogues 14 and 15 were comparable to SN in cytotoxic activity in the cell

lines tested.52

Lown and Sim prepared several quinoline-5,8-diones with different

substituents in ring A and determined the rate at which each cleaved DNA in

vitro.53 In this assay, 6,7-dibromoquinolinequinone, 6,7-dichloro-

quinolinequinone and 6,7-diaminoquinolinequinone showed the fastest rates.

The 6,7-unsubstituted, 6-methoxy and 6-methoxy-7-halo analogues all cleaved

the DNA faster than SN. The authors correlated the time at which 50% of the

DNA was cleaved with the percent inhibition of W256 tumor in vivo for some

of these compounds as recorded from NCI data.54 Though a correlation was

claimed by using only selected quinolinequinones, a closer analysis shows

otherwise. For example, SN was not correlated, nor was the dibromo or

dichloro compounds, and if the data on these and others were included, they

would clearly show no correlation between the in vivo inhibition of tumor and

in vitro rate of cleavage of DNA. This is unfortunate because subsequent

authors used this in vitro assay to test analogues as though some in vivo

significance could be attached to the results.

A later report by Lown and Sim55 described the synthesis of tricyclic SN

analogues in which ring C consisted of an o-nitrophenyl or an o-aminophenyl,

while the 7-amino-6-methoxy substitution of ring A was changed to

7-amino-6-bromo, 7-bromo-6-methoxy, and 6,7-dibromo analogues (Fig.1.5).

The rate of cleavage of DNA was again measured, and in general the

o-nitrophenyl compounds were faster acting than those of the o-aminophenyl

group, especially when coupled with 6,7-dibromo substitution in ring A.

Other bicyclic and tricyclic analogues containing the isoquinoline-,

quinoxaline-, and quinazolinequinones have also been synthesized and tested.

Some bicyclic isoquinolinequinones showed good inhibition of RT.30 In another

study, several isoquinolinequinones (Fig.1.5) were prepared which displayed

good activity in antimicrobial and plant growth inhibition assays.5" Activity was

generally higher in those compounds which contained an electron-withdrawing

group (e.g. a nitro group) in ring C. The o-nitro and m-nitro derivatives were

twice as potent as SN in the plant growth inhibition assay, but only one-fourth

to one-tenth as potent as SN in the antibiotic assay.



Although many analogues of SN have been synthesized and tested, many

questions are yet unanswered. The basic question remains: what are the

essential structural elements needed for activity? Once this is answered,

identifying any structural changes diminishing its toxicity while maintaining its

activity would elevate this highly active compound to a more useful status in


Information on SN analogues containing structural modifications on ring

B is scarce. The tetrahydroSN derivative was one such analogue, and this

showed good activity on the Walker-256 tumor.54 The only other report

involved the preparation and testing of tricyclic 1-phenylisoquinoline-5,8-dione

analogues, which showed some interesting results.5" For example, moving the

heteroatom from 1 to 2 did not seem to cause any significant loss of activity

in the 1-phenyl and 1-(2'-pyridyl) analogues as compared to their quinoline

equivalents. This definitely confuses the issue concerning the site for metal

complexation. Even more disquieting is the fact that 1-benzylisoquinoline

derivatives showed higher antimicrobial activity than the 1-phenyl and 1-(2'-


pyridyl) isoquinolinequinones, and this puts the need for planarity of the

molecule for metal complexation in question.

In order to pursue the effects of structural changes in the B-ring further,

the current work was undertaken which involves the synthesis and testing of

quinoxalinequinone analogues of SN. The target compounds described herein

are shown in Figure 2.1. They are tricyclic, with ring A being the same as that

in SN, and ring B containing the two heteroatoms, a varied substitution at the

3-position, and a phenyl group (ring C). These compounds are tested in a

group of in vitro assays, and comparisons are made with SN on one hand and

quinoline/ isoquinoline analogues of an equivalent nature on the other.

One major objective in this study is to maintain the same substitution

pattern in ring A as is present in SN. This particular feature is desired because

the antitumor data compiled by the NCI indicated that compounds containing

an amino group on the quinone ring showed the best in vivo antitumor activity.

They have also exhibited strong cytotoxicity, anti-microbial activity, and plant

growth-inhibitory activity in vitro.51,52,56 Those analogues which contained

halogens on the quinone ring, although having the ability to cleave DNA rapidly

in vitro, possessed no in vivo activity.54 These observations clearly show that

the aminoquinone, and perhaps even the methoxy-aminoquinone system, as

present in SN, must be maintained in order to optimize activity. In preparing

analogues of this nature, a proper basis for comparison with the parent

compound and its simpler analogues containing an aminoquinone will be




R = H. OH, Cl. OMe. CN, COOMe

Figure 2.1. 6-Methoxy-7-Aminoquinoxalinequinone Analogues of SN

Prepared in Present Study

The overall effect of substituting a quinoxaline ring for the quinoline

should provide some slight yet subtle changes in the activity. This expectation

is based on a comparison of the properties of quinoxaline versus quinoline

which shows some similarities as well as differences. The pK, for protonation

of the quinoline nitrogen is 4.96,5s whereas the pK. for protonation of the first

heteroatom of quinoxaline is 0.6,58 a vast difference. The electron densities of

many of the ring positions of the two systems are comparable. The two

positions alpha to the heteroatoms of quinoxaline compare well to the 2 and 4

positions of quinoline, all having a low electron density. This property allows

many of the same chemical reactions to occur in both systems, such as

reductions, Reissert-type reactions, and the Chichibabin reaction. The

substitution patterns resulting from electrophilic reactions are also alike, with

the quinoline being a little more complex than the symmetrical quinoxaline.57'58

One significant difference in these two ring systems is the reduction potentials

of their respective 5,8-diones. Shaikh et al. measured the reduction potentials

of the 6,7-dichloroquinones of several bicyclic ring systems.59 These potentials

varied in a descending order with phthalazinequinone > quinazolinequinone >

quinoxalinequinone > isoquinolinequinone > quinolinequinone >

naphthoquinone. In this study, a good correlation was seen between the

reduction potential of these compounds and their rates of cleavage of DNA.

By using assays which will, more likely, relate to in vivo activity, the present

study hopes to determine if the reduction potential has a strong influence on

the activity. Although no potentials will be measured directly in this work, the

effect of varying the substituent at position 3, from electron-withdrawing to

electron-donating, will be considered in comparing the activities (since electron-

withdrawing groups are said to increase the reduction potential and electron-

donating groups to lower it).59

The use of an unsubstituted phenyl group for ring C was decided upon

because this substituent maintained a reasonable degree of activity when

present in other simpler analogues of SN. To begin with, there seemed to be

no difference in activity in the simpler analogues where ring C was a phenyl or

2-pyridyl.51'56 In published reports where changes were made in ring C, the

results were not very definitive as to which substituents on ring C are essential.

For example, the replacement of the trioxyphenyl ring attached to the 4'-

position of ring C (as it is present in SN) by hydrogen did not result in the loss

of antimicrobial activity.51 However, when ring C was 6'-carboxy-2-pyridyl or

a 2',4', or 6'-aminophenyl, the analogues generally showed poor activity.

Collectively, these quinoline and isoquinoline analogues showed less

cytotoxicity, antimicrobial activity, plant growth-inhibitory activity, and slower

rates in cleaving DNA in their respective assays.51,52,56 Therefore, since

previous data showed that analogues with ring C as a phenyl did show a

certain degree of activity, it is felt that its use as ring C in this study will enable

one to draw relevant conclusions regarding the effect of changes made in the

rest of the quinoxaline molecule (i.e. the 2, 6, and 7 positions) on the activity.

There are only a few reports in the literature where analogues of drugs

have been prepared having a quinoxaline ring in the place of the parent ring.

In one, a potent anti-folate compound was obtained when the pteridine ring

was replaced by a quinoxaline ring.' In another, some quinoxaline-1,4-di-N-

oxide derivatives possessing antibiotic and antiviral activity were also tested for

vitamin K activity.61 Finally, some 8-aminoquinoxaline analogues of the active

8-aminoquinoline-type antimalarials were made in a search for compounds with

better antimalarial activity.62

In the cases where the quinoxaline analogues did not show the desired

activity, further investigations led to the preparation of quinoxaline compounds

which contained a 5,8-dione system. Levy and Joullie synthesized 6-methyl-

quinoxalinequinone for comparison with 2-methyl-1,4-naphthoquinone, a

member of the vitamin K family.63 This too was found to be inactive. Fisher

et al. synthesized a variety of 6-dimethylaminoalkylamino-5,8-dimethoxy-

quinoxalines, again looking for antimalarial activity.64 Since the parent quinoline

drug was thought to be converted to a quinone in vivo, the rationale was to

develop a compound which could be more easily converted to the quinone than

the 8-aminoquinoxalines. Here again, no activity was found in the quinoxaline


However, in spite of these unpromising results in other areas, it is still

possible that some of the proposed quinoxaline analogues of SN will be active

since past work has shown that the molecule of SN can accept significant

structural changes and still retain activity.

Quinoxaline Background

The classical method for synthesis of the quinoxaline ring system is the

condensation of an o-phenylenediamine with an a,Sf-dicarbonyl compound. An

alternative to the diamine is the use of an o-nitroaniline under reductive

conditions, while an a-halogenoketone or an a,fi-dihalide can be substituted for

the dicarbonyl (although the product will be a di- or tetrahydroquinoxaline).58

The numbering of the quinoxaline ring is shown in Figure 2.2.

While this condensation method is relatively simple, it is useful only

when there is symmetry in either of the starting materials; otherwise, a mixture

of isomers is obtained. This is illustrated in Figure 2.3 where (A) containing a

symmetrical dicarbonyl and (B) containing a symmetrical o-phenylenediamine

provide a single product. In contrast, (C) shows a mixture of products formed

from asymmetric reactants. In meeting the objectives described above, the

problem of isomer formation must be controlled because the desired analogues

must have the specified substituents at positions 2,3,6 and 7.

A search of the literature on the quinoxalinequinones described no such

method for preparing compounds with unambiguously situated substituents at

positions 2,3,6 and 7. This entire group of compounds was made using the

condensation method. The use of symmetrical dicarbonyls such as glyoxal,

2,3-butanedione, and 1,2-bis(2-pyridyl)ethanedioneorsymmetrical diamines like

4,5-dimethylphenylenediamine and 2,3-diamino-1,4-dimethoxybenzene

prevented the formation of isomers.65'66

5 4

6 3

7 2

8 1

Figure 2.2. Quinoxaline Ring System



-, 2

H 2


+ : -H3

0 CH3 ^


+ A

R 0 3 =) C .

Figure 2.3. Need for Symmetric Starting Materials in Quinoxaline Ring

Synthesis Using the Condensation Method


The procedures for converting these condensation products to their

respective quinones are shown in Figure 2.4. As can be seen, the use of 2,3-

diamino-1,4-dimethoxybenzene not only provided symmetry, but a facile route

to the p-quinone after the condensation. For this, the 5,8-

dimethoxyquinoxaline was subjected to demethylation using aluminum chloride,

followed by oxidation with silver oxide.65 The other diamine used was 4,5-

dimethyl-o-phenylenediamine. After condensation of this diamine with 1,2-

bis(2-pyridyl)ethanedione, nitration produced the 5,8-dinitro compound, which

then yielded the quinone after reduction to the diamine followed by oxidation

with ferric chloride.8e

Also shown in Figure 2.4 is a reported procedure describing the

conversion of 6-methoxyquinoxaline to 6-methoxyquinoxaline-5,8-dione which

was quite similar to that described in the Kende synthesis of SN. Chernyak et

al. found that nitration of this compound produced the 5-nitro-6-methoxy

derivative. Catalytic hydrogenation then gave the amine which was oxidized

to the quinone with Fremy's salt."7 Later, Renault et al. attempted to use this

approach for the preparation of the SN analogue 6-amino-7-methoxy

quinoxaline-5,8-dione from 6-amino-7-methoxyquinoxaline." After acetylation

of the amine, nitration gave the 5-nitro-6-amino derivative due to hydrolysis of

the acetylamino function. Reduction of this gave the 5,6-diamino-7-

methoxyquinoxaline, but only the 5,6-dione was obtained after oxidation with

Fremy's salt. In the same report, the 6-methoxyquinoxalinequinone was

brominated, but attempts to convert this compound to the amine via the azide


0 3
+ 01:3 1

1) H/Pd/C
2) Fremy's salt

Figure 2.4. Quinoxalinequinones Previously Synthesized


1 2) AgCO
2) Ag7,o

CH3O -,



failed.68 This was the only attempt at making a methoxy-aminoquinone in the

quinoxaline series.

The well-known reactivity of the 6- and 7- positions of the quinoxaline-

quinones has also been documented to some extent. In those compounds

where these positions were unsubstituted, it was found that 1,4-addition

reactions occurred with amines and hydrogen halides.69 Thus, reaction of the

quinone with primary and secondary amines gave the 6-aminoquinoxaline-

quinone derivatives directly, with higher yields seen with the secondary amines.

The reaction of the unsubstituted quinone with hydrogen chloride gas produced

the 6-chlorohydroquinone which was subsequently oxidized to the quinone.

The preparation of 6,7-dichloroquinoxaline-5,8-dione obtained by the oxidation

of 5,8-dimethoxyquinoxaline with sodium chlorate in hydrochloric acid was an

unusual method to the dihaloquinone; however, the yield was only 1.1%.59

As another example of reactions of quinoxalinequinones, when 6-chloro-

2,3-dimethylquinoxalinequinone was treated with aziridine or ammonia, it gave

the corresponding 7-amino-6-chloroquinone derivative in a 50% yield. The

reaction was said to involve a 1,4-addition followed by oxidation of the

resulting aminohydroquinone by the unreacted quinone.89 The 6-piperidino-

quinoxalinequinone added hydrogen chloride with similar results, while the 6-

methoxyquinone was unreactive towards hydrohalogenation.70 In contrast,

Renault et al. obtained 6,7-diaziridinyl-2,3-bis(2-pyridyl) quinoxalinequinone

from the corresponding 6-methoxyquinone through displacement of the

methoxyl as well as 1,4-addition.e6 The 1,4-addition which occurred with 6-

amino- and 6-haloquinoxalinequinones did not take place with the 6-

halonaphthoquinones where the halogen was simply displaced by the amine.69

Present Synthetic ADDroach

The synthesis of the lead compound (Fig. 2.1, R = H) was carried out as

shown in Figure 2.5. Thus, coupling of 3,5-dimethoxyaniline and benzoylformic

acid in the presence of dicyclohexylcarbodiimide (DCC) proceeded smoothly to

give N-(3,5-dimethoxyphenyl)benzoylformamide in 85% yield. Treatment of

this a-ketoamide with one molar equivalent of nitric acid in acetic anhydride

gave a mixture of the 2'-nitro and 4'-nitro derivatives in a ratio of 85:15

respectively, as determined by UV absorbance values at 255nm. Although the

desired 2'-nitro isomer could be crystallized from ether readily, complete

separation required chromatography. Since separation at the next stage was

easier, the mixture was subjected to reductive cyclization by sodium dithionite

producing the desired 2(1H)-quinoxalinone in a 66% overall yield (based on the

a-ketoamide), together with N-(4-amino-3,5-dimethoxyphenyl)benzoyl-

formamide. Separation of the quinoxalinone from the highly ether-soluble 4-

amino-a-ketoamide was not a problem. This three step procedure to the

quinoxaline ring system provided the needed positional control of the

substituents. This achievement was due to formation of only one bond in the

cyclization step instead of two, as is seen with the condensation method.

The conversion of 2(1 H)-quinoxalinones to 2-haloquinoxalines has been

described previously.58 Similar reactions were used in the preparation of the 2-



"I- [ 0OOH

1) HNO3/Ac20
2) N02S204/MeOH/H20

DCC 10

1) POCI3 or PBra
2) Zn/AcOH


1) Fe/ AcOH
2) Ce(SO4)2


Figure 2.5. Regio-controlled Synthesis of 6-Methoxy-7-
Aminoquinoxalinequinones via the Quinoxalinone


chloro- and 2-bromoquinoxalines by treating the 2(1H)-quinoxalinone with

phosphorus oxychloride and phosphorus tribromide, respectively, resulting in

yields of 60% or better. Both of these compounds could be reductively

dehalogenated using either magnesium or zinc in acetic acid to produce 6,8-

dimethoxy-2-phenylquinoxaline, the target compound.

To generate the 5,8-quinone system, the scheme commonly described

in the literature- conversion to the 5-nitro derivative, reduction to the amine and

oxidation was considered.87 The introduction of the 7-amino group would then

proceed along the well known path through bromination, azidation, reduction,

and air-oxidation.37 Thus, the nitration of the 2-phenyl-6,8-dimethoxy-

quinoxaline was studied using nitric acid in acetic anhydride. With addition of

2 molar equivalents, the major product isolated was the mononitro derivative.

This compound was assigned the structure 6,8-dimethoxy-5-nitro-2-phenyl-

quinoxaline based on the following reasons: nitration is known to occur at the

5-position in quinoxaline itself as well as 6-methoxyquinoxaline, due to the

higher electron density at this position;58e.7 and the presence of the second

methoxyl at the 8-position should again favor nitration at the 5-position since

the 7-position is more sterically hindered.

A minor product isolated from the above reaction was a more polar

compound which was soluble in aqueous sodium bicarbonate. Its nmr

spectrum showed no protons on ring A, a broad peak at 4.0-4.5 ppm which

exchanged in D20, and one methoxyl group. From this it was concluded that


it was a dinitrophenol. This was further confirmed by the fact that it formed

a mono-acetate.

Initial attempts at reducing the mono-nitro compound with sodium

dithionite were not successful, perhaps due to the low solubility of the

compound in the reaction medium. This difficulty made the using of the

dinitrophenol route more appealing. A simple reduction step to the

diaminophenol, followed by autooxidation, would be expected to give the

desired 6-methoxy-7-aminoquinone. Indeed, reaction with the dithionite

occurred readily, and a quinone product was isolated. However, it was still to

be established whether this product was the ortho- or para-quinone, or one of

the corresponding quinonimines. Elemental analysis has since shown that it

was a quinone and not a quinonimine.

In order to establish that the yielded compound was indeed the desired

para-quinone, it was felt that the compound must be prepared by an

unambiguous route. Reduction of the 5-nitroquinoxaline derivative was finally

accomplished by heating it in acetic acid with reduced iron. The resulting

amino compound was extracted into chloroform and then directly oxidized to

the quinone by shaking this extract with an acidic solution of ceric sulfate. The

presence of the quinone function in the product was detectable by its behavior

on tic (characteristic darkening of the spot after exposure to ammonia vapors),

and by its ir spectrum, which revealed the strong carbonyl bands expected for

a quinone. Additionally, elemental analysis agreed with its formulation as the


The next step was to convert this quinone into the desired 7-

aminoquinone, and the possibility of introducing the 7-amino group in one step

was explored. An example of such a one step reaction reportedly occurred in

the conversion of 7-methoxy-2-phenylquinoline-5,8-dione into the

corresponding 6-amino-7-methoxy-5,8-dione by reaction with sodium azide in

dimethylformamide in the presence of catalytic amounts of acetic acid.39 In

similar examples the mechanism postulated for this reaction involves

protonation of the heteroatom which can then hydrogen-bond with the carbonyl

function in the peri position. Through this bonding, an electrophilic center

develops at the unsubstituted 6-position of the quinoline. The azide can then

attack this position and undergo an acid-assisted decomposition to the amine.71

Although this reaction was successful with the 7-methoxy-2-phenyl-quinoline-

5,8-dione, the corresponding 6-methoxyquinolinequinone did not yield the 7-

aminoquinone under the same conditions. One might explain this lack of

reactivity in terms of resonance, because a resonance structure cannot be

drawn showing an electrophilic center at the 7-position.

Since there are two heteroatoms adjacent to the quinone ring with the

quinoxalinequinones, one might expect that the unsubstituted 7-position should

be susceptible to nucleophilic attack if the appropriate heteroatom is

protonated. However, of major concern is the low basicity of these

heteroatoms which may affect their ability to protonate/ hydrogen-bond under

the influence of such a weak acid as acetic acid.

In a paper which described some reactions of nucleophiles on

quinoxalinequinones in acetic acid,70 it was stated that strong nucleophiles such

as piperidine would displace a methoxyl group on the quinone rather than

attack the unsubstituted position. The opposite was seen when a chlorine was

present on the quinone ring instead of a methoxyl. Thus, the reaction of 6-

methoxyquinoxaline-5,8-dione with piperidine in acetic acid gave 6-

piperidinoquinoxaline-5,8-dione, and the reaction of 6-chloroquinoxaline-5,8-

dione with piperidine gave 6-chloro-7-piperidinoquinoxaline-5,8-dione. It was

thought that protonation of N-1 was occurring preferentially in the

methoxyquinone because of the electron-donating ability of the methoxyl group

and its ability to stabilize the resultant positive charge. With the 6-

chloroquinone, the chlorine is expected to have a destabilizing effect when N-1

was protonated; therefore, N-4 is the expected site for protonation. This

hypothesis was partly confirmed by substitution of the piperidine at the C-7

position. Since the azide was not discussed as one of the nucleophiles, it was

not known whether the desired 7-substitution would occur, the methoxyl group

be displaced, or any reaction take place at all.

Against this background, the reaction of 6-methoxy-2-phenylquinoxaline-

5,8-dione with sodium azide in acetic acid was studied. After 1 hour at 55 C,

the desired aminoquinone was produced, albeit in a low yield. When the

reaction was followed by tic, it appeared to proceed through the azidoquinone

(faster-moving, brown spot) which slowly disappeared with the appearance of

the slower-moving (and darker brown) aminoquinone. Several minor spots were

also seen which indicated the occurrence of side reactions. This reaction was

carried out on many of the methoxyquinones in this study. The major side

product isolated in each case was the hydroquinone of the starting material.

This compound is produced when the initial addition product, azidohydro-

quinone, is converted to the azidoquinone by the starting material. The starting

quinone was readily recovered by oxidation with ceric sulfate. Infrared and nmr

spectra showed this product to be the same as that obtained by reduction of

the dinitrophenol, the desired aminoquinone.

Alternative Scheme for the 2-PhenylQuinoxaline

One of the aminoquinones originally targeted for synthesis was a

dihydroquinoxalinederivative. Saturation in ring B could show some interesting

activity, since good in vivo activity was noted with the tetrahydro-SN analogue

on some transplanted rodent tumors.54 The dihydro- and tetrahydro-

quinoxalines are known to readily oxidize to the quinoxaline,58 therefore the

objective was to make an N-acetyldihydroquinoxaline in order to block such

conversion to the quinoxaline.

The alternative synthetic scheme (Fig. 2.6) began with the alkylation of

3,5-dimethoxyaniline by 2-bromoacetophenone. This was carried out in

dimethylformamide with sodium bicarbonate at 50C for two hours. A slight

excess of the bromoacetophenone was used to assure maximum use of the

dimethoxyaniline. The major product obtained was the desired secondary

amine, along with small amounts of the dialkylated material. The product was





Figure 2.6. Alkylation Pathway in to Quinoxaline Synthesis

1) Ac2O
2) HN03

3) No2S204


then acetylated using acetic anhydride at 800C. At the completion of the

reaction, the mixture was cooled to 0-50C and treated with nitric acid.

Nitration occurred in a manner similar to that seen with the a-ketoamide

described earlier. Two mono-nitro compounds were obtained, with the desired

ortho-nitro derivative being the major product. Reductive cyclization of this 2-

nitro derivative was accomplished with dithionite to yield the dihydroquinoxaline

acetate. Further nitration of this compound, a first step in the generation of the

quinone system, led to nitration as well as deacetylation/ oxidation to give the

mononitroquinoxaline, identical to the 5-nitro-6,8-dimethoxy-2-phenyl-

quinoxaline described earlier.

Although this procedure did not afford the desired dihydroquinoxaline

aminoquinone, it did provide a second, more efficient method to the 3-H

quinoxaline compounds. The success seen here prompted attempts to use

other a-substituted 2-bromoacetophenones to generate quinoxalines with the

desired 3-substituents already present. Unfortunately, except in the case of 2-

bromopropiophenone, such reactions showed limited success, and the scheme

was abandoned.

In summary, starting with the amide prepared from 3,5-dimethoxyaniline

and benzoylformic acid, nitration followed by reductive cyclization gave the 3-

phenyl-6,8-dimethoxyquinoxalin-2(1H)-one, which was converted to the

quinoxaline via halogenation at the 2-position and reduction. Nitration of the

quinoxaline gave the 5-nitro derivative and the 5,7-dinitro-8-hydroxy compound

depending on the conditions. The former was converted to the 5,8-dione by

reduction and oxidation, and the dione was aminated at the 7-position by

reaction with sodium azide in acetic acid. Alternatively, the dinitrophenol was

reduced and oxidized to yield the same 7-amino-6-methoxy-2-

phenylquinoxaline-5,8-dione. The 2-phenylquinoxaline system was also

generated by the alkylation of 3,5-dimethoxyaniline with 2-bromoacetophenone

followed by nitration and reductive cyclization.


Nuclear magnetic resonance (nmr) spectra were recorded on a Varian EM

390 with chemical shifts reported in parts per million (ppm) relative to

tetramethylsilane (0.0 ppm), the internal standard. Some spectra were

obtained from samples which were not of analytical grade, and peaks marked

with (X) indicate impurity or solvent. Infrared (ir) spectra were recorded on a

Beckman Acculab III as KBr pellets and reported as cm-1. Melting points were

determined on a Fisher-Johns hot stage melting point apparatus and are

uncorrected. Elemental analyses were carried out by Atlantic Microlabs,

Atlanta, Ga. Mass spectral analyses (M.S.) were performed on a MS80RFA

(Kratos Analytical). FAB Spectra used a FAB11NF (Ion Tech Ltd.) source using

xenon at 6 keV and 1 mA, and El Spectra were obtained at 70 eV and 0.3mA.

N-(3.5-Dimethoxvyphenvl)benzolvformamide (2.1)

Benzoylformic acid sodium salt (95% pure) was purchased from Aldrich.

An aqueous solution of the salt was acidified, partially saturated with sodium

chloride, and extracted three times with 35% acetone/ benzene. The combined

extract was dried (sodium sulfate) and concentrated to dryness. The extent of

extraction was followed by uv absorbance of solvent-free extracts at 253nm.

After 3 extractions, nearly 96% of the uv absorbance was found in the extract.

To the extract from 18 g of benzoylformic acid sodium salt (containing

app. 14.5 g free acid in 200 ml 1:1 acetone/ benzene) was added 3,5-

dimethoxyaniline (14.5 g) with stirring. The solution was cooled to O0C in a

methanol/ ice bath, and dicyclohexylcarbodiimide (DCC, 21 g), dissolved in

benzene (100 ml), was added in portions over three minutes (min). The

mixture was stirred for 15 min, and for another 15 min, at room temperature.

After the addition of cold water (150 ml), the contents of the flask were stirred

for 1 hour (h) and then filtered. The filtrate was separated, and the aqueous

layer was extracted with benzene. The precipitate was washed with

chloroform until most of the yellow color was in the filtrate, leaving mostly a

white solid (dicyclohexylurea) behind. The chloroform wash and benzene

extract were combined and concentrated. Crystallization from 9:1 ether/

methanol gave the amide as yellow needles. Yield: 22.9 g (85%); m.p.: 126-

1280C; nmr (CDCI3): 8.80 (1H, br=broad,NH), 8.27-8.38 (2H,Ar.m.=

aromatic multiple, C-2,6), 7.45 (3H,Ar.m., C-3,4,5), 6.88 (2H,d doublett, C-

2',6',J = 2Hz), 6.27 (1 H,dd doublett of doublets, C-4', J = 2Hz), 3.73 (6H,s =

singlet,OCH3); ir (cm-): 3345, 2935, 1675, 1650, 1585, 1515, 1430, 1405,

1230, 1175, 1155, 1135, 1045, 800, 660. Anal. Calc. for Cl,eH5N04: C,

67.35; H, 5.30; N, 4.91. Found: C, 67.28; H, 5.34; N, 4.95.



-4 E

Lr 2




N-(3.5-Dimethoxv-2-Nitrophenvl)benzovlformamide (2.2)

To a suspension of 2.1 (9 g) in acetic anhydride (40 ml) at 0-50C, was

added 70% nitric acid (7 ml) dropwise with stirring over 5 min. The stirring

was continued for an additional 30 min. The reaction mixture was diluted with

ice/water and stirred for 1 h. The precipitate was filtered, dissolved in CHCI3,

and washed with aqueous bicarbonate. After drying, the chloroform was

concentrated to dryness. The solid which crystallized from ether was filtered

and washed with 1:1 ether/ ligroin to give 2.2. Yield: 7.4 g (71%); m.p.: 163-

165C; nmr (CDCI3): 10.73 (1H,br,NH), 8.28-8.37 (2H,Ar.m.), 7.77

(1H,d,J=3Hz, C-6'), 7.40-7.63 (3H,Ar.m.), 6.40 (1H,d,J=3Hz, C-4'), 3.90

(6H,s,OMe); ir (cm1): 3315, 3115, 2940, 2920, 1710, 1680, 1605, 1590,

1540, 1530, 1495, 1430, 1355, 1325, 1295, 1260, 1230, 1160, 1100,870,

685. Anal. Calc. for C1,H14N2Oe: C, 58.18; H, 4.27; N, 8.48. Found: C,

58.08; H, 4.24; N, 8.48.

N-(3.5-Dimethoxy-4-Nitroohenvl)benzovlformamide (2.3)

This is the minor product from the above reaction, purified by column

chromatography (silica/benzene) of the mother liquor of 2.2. After

concentration of the appropriate fractions, the product separated as orange

needles from 2:1 ether/ligroin. m.p.: 158-160OC; nmr (CDCI3): 9.03 (1H, br,

NH), 8.30-8.37 (2H,Ar.m.), 7.33-7.77 (3H,Ar.m.), 7.05 (2H,s,C-2',6'), 3.87

(6H,s,OMe); ir (cm-): 3345, 3075, 2980, 2940, 1700, 1675, 1615, 1600,

1525, 1460, 1450, 1410, 1375, 1280, 1240, 1200, 1140, 845, 740, 685.

Anal. Calc. for C1,H14N20: C, 58.18; H, 4.27; N, 8.48. Found: C, 58.10; H,

4.24; N, 8.50.

N-(3.5-Dimethoxy-2.6-Dinitrophenyl)benzovylformamide (2.4)

A suspension of 2.1 (0.55 g) in acetic anhydride (5 ml) was stirred in an

ice bath. Nitric acid (70%, 1.0 ml) was added dropwise, and the mixture was

allowed to react for 20 min, brought to room temperature, and stirred for 15

min. Ice was added with stirring, and after 30 min the solid was filtered and

washed with water. The precipitate was dissolved in chloroform and washed

with aqueous bicarbonate. The chloroform was dried and concentrated. The

solid was crystallized from ether and filtered to give 2.4. Yield: 0.6 g (82%),

m.p.: 221-225C; nmr (CDCI3): 11.00 (1H,br,NH), 7.95-8.07 (2H,Ar.m.),

7.40-7.70 (3H,Ar.m.), 6.82 (1H,Ar.s.= aromatic singlet), 4.00 (6H,s,OMe); ir

(cm-1): 3365, 3140, 3050, 2970, 1730, 1675, 1600, 1550-1500, 1350-

1330, 1270, 1225, 1190, 1175, 1105, 685. Anal. Calc. for C1eH13N3O,: C,

51.20; H, 3.49; N, 11.20. Found: C, 51.14; H, 3.51; N, 11.14.

N-(3.5-Dimethooxy-2.4-Dinitrophenyl)benzoylformamide (2.5)

To a suspension of 2.3 (0.1 g) in acetic anhydride (3 ml) was added

nitric acid (0.2 ml) with stirring at room temperature. After 15 min, ice was

added, and after 1 h the precipitate was filtered, dissolved in chloroform, and

washed with aqueous sodium bicarbonate. The chloroform was dried and

concentrated to dryness. Crystallization from ether gave 2.5 as yellow

crystals. Yield: 0.085 g (75%); m.p.: 147-148oC; nmr (CDCI3): 10.40

(1H,br.,NH), 8.30-8.40 (2H,Ar.m.), 8.27 (1H,Ar.s.), 7.40-7.80 (3H,Ar.m.),


4.03 (3H,s,OMe), 3.97 (3H,s,OMe); ir (cm-1): 3325, 3120, 2955, 1715,

1680, 1600, 1545, 1540, 1500, 1470, 1450, 1415, 1350, 1320, 1280,

1225, 1170, 1125, 920, 910, 870, 835, 830, 750, 700, 690, 605.

5.7-Dimethoxv-3-Phenvlouinoxalin-2(1H)-One (2.6)

To a 1:4 MeOH/ H20 mixture (500 ml) was added 2.2 (17 g) and sodium

dithionite (45 g., 5 eq). The suspension was boiled under reflux for 30 min.

After cooling the mixture by diluting with water, the solid was filtered and

washed, first with water and then with ether, to give 2.6 (8.2 g) as light yellow

crystals. The filtrate was concentrated and extracted 3 times with chloroform;

and the extract was concentrated. The solid was crystallized from 3:1

acetone/ ligroin to give an additional amount of 2.6 (1.73 g). Yield 9.93 g

(72%); m.p.: 256-258oC; uv maxima: 365nm (major), 265nm (minor); nmr

(CDCIa/DMSO-de): 12.12 (1H,br.,NH), 8.19-8.27 (2H,Ar.m.), 7.31-7.38

(3H,Ar.m.), 6.25 (2H,2d=2 doublets,J=2Hz), 3.95 (3H,s,OMe), 3.80

(3H,s,OMe); ir (cm'): 3450, 3100, 2940, 1645, 1620, 1590, 1505, 1460,

1405, 1390, 1290, 1210, 1165, 1150. Anal. Calc. for CeH14N203: C, 68.07;

H, 5.00; N, 9.92. Found: C, 68.02; H, 5.00; N, 9.94.

N-(4-Amino-3.5-Dimethoxyvhenvl)benzovylformamide (2.7)

Chromatographic separation of the mother liquor from 2.6 using silica

and eluting with benzene gave 2.7 as the first compound to be eluted.

Concentration of the appropriate column fractions gave 2.7 as orange crystals

from ether. Yield: 1.79 g (13%); m.p.: 141-143OC; nmr (CDCI3): 8.70-8.87

(1H,br.,NH), 8.30-8.40 (2H,Ar.m.), 7.30-7.63 (3H,Ar.m.), 6.93 (2H,Ar.s.),


3.83 (6H,s,OMe), 3.50-4.00 (2H,br.,NH2); ir (cm-): 3420, 3340, 3000, 2960,

1680,1670,1625,1520,1455,1450, 1220,1190, 1175, 1075, 1015,835,

740, 685. Anal. Calc. for C,,HeN204: C, 63.99; H, 5.37; N, 9.33. Found:

C, 64.07; H, 5.36; N, 9.26.

3-Chloro-6.8-Dimethoxy-2-Phenvlquinoxaline (2.8)

To 2.6 (6 g) in a dry flask was added phosphorus oxychloride (10 ml),

and the mixture was heated at reflux for 30 min. After cooling for 30 min in

an ice bath, ice water was added with vigorous stirring. After another 30 min

the solid was filtered, dissolved in chloroform and washed with aqueous

sodium bicarbonate. The chloroform was dried over sodium sulfate and

concentrated to dryness. The product was crystallized from ether as off-white

crystals. Yield: 6.0 g (94%); m.p.: 188-190C; uv maxima: 365nm (minor),

265nm (major); nmr (CDCI3): 7.70-7.78 (2H,Ar.m), 7.37-7.43 (3H,Ar.m.),

6.88 (1H,d, J=2.5Hz), 6.83 (1H,d,J=2.5Hz), 3.97 (3H,s,OMe), 3.90

(3H,s,OMe); ir (cm-): 3060, 2940, 1615, 1540, 1470, 1410, 1335, 1235,

1210, 1200, 1160, 1145, 1040, 825, 770, 695. Anal. Calc. for

CleH13N202CI: C, 63.90; H, 4.36; N, 9.31. Found: C, 63.14; H, 4.33; N,


6.8-Dimethoxv-2-PhenylQuinoxaline (2.9)

To a solution of 2.8 (0.98 g) in acetic acid (6 ml) was added magnesium

turnings (freshly washed with 0.1 N HCI, 0.9 g); and the mixture was stirred

at 85C for 30 min. Water was added, and the mixture was extracted with

chloroform. The chloroform was washed with aqueous sodium bicarbonate,

dried and concentrated to dryness. The solid was crystallized from 9:1 ether/

acetone to give 2.9 as a light-brown solid. Yield: 0.72 g (85%); m.p.: 138-

140C; uv maxima: 278nm (major); nmr (CDCI3): 9.15 (1H,Ar.s.), 8.07-8.13

(2H,Ar.m.), 7.43-7.50 (3H,Ar.m.), 7.00 (1H,d,J= 2Hz), 6.74 (1H,d,J= 2Hz),

3.87 (3H,s,OMe), 3.77 (3H,s,OMe); ir (cm-1): 3060, 3000, 2940, 2835,

1610, 1580, 1535, 1470, 1450, 1415, 1340, 1325, 1240, 1210, 1150,

1130, 770, 695. Anal. Calc. for C1eH14N202: C, 72.16; H, 5.30; N, 10.52.

Found: C, 71.90; H, 5.27; N, 10.47.

6.8-Dimethoxv-5-Nitro-2-PhenylQuinoxaline (2.10)

To a stirred suspension of 2.9 (0.98 g) in acetic anhydride (5 ml) was

added nitric acid (70%, 0.8 ml) dropwise at room temperature. The initial red

solution gradually turned light brown, at which time tic indicated the completion

of the reaction. Ice was added with stirring, and after 30 min the solid was

filtered, dissolved in chloroform, and washed with aqueous sodium bicarbonate,

dried and concentrated to dryness. The solid was crystallized from acetone/

ligroin to give 2.10 as yellow crystals. Yield: 0.72 g (63%); m.p.: 267-

270C; nmr (CDCI3/ DMSO-d6): 9.33 (1H,Ar.s.), 8.12-8.20 (2H, Ar.m.), 7.08-

7.23 (3H,Ar.m.), 4.20 (3H,s,OMe), 4.14 (3H,s,OMe); ir (cm-1): 3060, 3020,

2980, 2950, 2850, 1620, 1580, 1540, 1530, 1485, 1475, 1350, 1215,

1180, 1110, 935, 820, 810, 765, 685. M.S.-FAB: [M+1]*= 312; Calc. for

CleH13N304: MW = 311.


- v -I -- a

L.. E




__^ 0

\^ J~ 00

5.7-Dinitro-8-Hydroxv-6-Methoxv-2-Phenvlauinoxaline (2.11)

To a suspension of 2.9 (0.2 g) in acetic anhydride (3 ml), was added

nitric acid (70%, 0.4 ml) with stirring at room temperature. The reaction was

monitored by tic to determine the absence of the mononitro derivative and the

presence of the product as a yellow spot at the origin. Ice was added, and the

solid was filtered. The solid was dissolved in chloroform, and washed with

aqueous sodium bicarbonate, whereby, the uv-absorbing material passed into

the aqueous layer. The aqueous layer was acidified, and the pale-yellow solid

which precipitated was filtered. Crystallization from ether gave 2.11. Yield:

0.12 g (75%); m.p.: 218-220C; nmr (CDCI3): 9.40 (1H,Ar.s.), 8.26-8.33

(2H,Ar.m.), 7.45-7.53 (3H,Ar.m.), 4.70 (1 H,br,OH), 4.07 (3H,s,OMe); ir (cm-1):

3360, 2950, 1630, 1580, 1540, 1470, 1415, 1350, 1300, 1250, 1145,

1060, 1000, 815, 795, 770, 685. Anal. Calc. for C15H10N406: C, 52.64; H,

2.95; N, 16.37. Found: C, 52.60; H, 2.98; N, 16.42.

A small portion of 2.11 (0.1 g) was heated with acetic anhydride (2 ml)

at 100C for 15 min. Water was added, and the solid was filtered. After

drying the solid, it was crystallized from ether/ ligroin to give the acetate as

pale-yellow crystals. m.p.: 225-228oC; nmr (CDCI3): 9.27 (1H,Ar.s.), 7.90-

8.10 (2H,Ar.m.), 7.37-7.55 (3H,Ar.m.), 4.05 (3H,s,OMe), 2.43 (3H,s,OAc).

7-Amino-6-Methoxv-2-PhenvlQuinoxaline-5.8-Dione (2.12)

To a solution of 2.11 (0.1 g) in methanol (3 ml) was added 5% aqueous

sodium bicarbonate (3 ml) and sodium dithionite (0.6 g). The mixture was

heated at reflux until the starting material was absent. The reaction mixture





was acidified (pH 4) and extracted with chloroform. The chloroform was

shaken with water containing 5% ceric sulfate/ 2M sulfuric acid (4 ml). The

reddish-brown solvent layer was concentrated to dryness, and the residue was

crystallized from 2:1 acetone/ ligroin to give 2.12 as a dark-brown crystalline

solid. Yield: 0.027 g (35%); m.p.: 260-2700C (dec.); nmr (CDCI3/ DMSO-de):

9.33 (1H,Ar.s.), 8.03-8.30 (2H,Ar.m.), 7.33-7.67 (3H,Ar.m.), 6.70

(2H,br,NH2), 3.93 (3H,s,OMe); ir (cm1): 3485, 3260, 3100, 2960, 1690,

1645, 1590, 1580, 1570, 1555, 1520, 1440, 1340, 1250, 1200, 1095,

1020, 1010, 925, 745, 690. Anal. Calc. for C15H,,N303 Y2 H20: C, 62.13;

H, 4.05; N, 14.09: Found: C, 61.92; H, 4.08; N, 14.22.

6-Methoxv-2-Phenvlouinoxaline-5.8-Dione (2.13)

To acetic acid (10 ml) was added 2.10 (0.3 g) and reduced iron (0.2 g),

and the mixture was heated at 85C for 10 min. The mixture turned dark and

gradually lightened to a grayish color. At the completion of the reaction, water

(100 ml) was added, and the contents were extracted twice with chloroform

(2x70 ml). The chloroform was filtered through cotton into a 2N sulfuric acid

solution containing 5% ceric sulfate. These were shaken together vigorously

for 2-3 min with sufficient ceric sulfate added to complete the oxidation. The

CHCI3 layer was washed with water, dried over sodium sulfate, and

concentrated to dryness. The product was crystallized from acetone/ ether and

filtered. Yield: 0.198 g (77%); m.p.: 180-183oC; nmr (CDCI3): 9.30

(1H,Ar.s.), 8.00-8.20 (2H,Ar.m.), 7.33-7.57 (3H,Ar.m.), 6.40 (1H,s,vinyl),









L. 2









3.93 (3H,s,OMe); ir(cm-1): 3045, 2915, 2840, 1690, 1660, 1600, 1560,

1515, 1315, 1280, 1250, 1205, 1110, 1065, 1040, 850, 775, 685.

7-Amino-6-Methoxy-2-PhenvlQuinoxaline-5.8-Dione (2.12)

In a loosely stoppered tube, 2.13 (0.094 g) was dissolved in acetic acid

(5 ml), and sodium azide (0.1 g) was added. The contents were heated at

550C with stirring for 1 h. The absence of starting material was noted from

tic, and spots corresponding to the aminoquinone and azidoquinone were seen.

The contents were diluted with water and extracted with chloroform.

Separation of the components was accomplished by chromatography using

silica and eluting with 1:1 chloroform/ benzene. The fractions containing the

aminoquinone were concentrated, and the product was crystallized from 2:1

acetone/ ligroin. Yield: 0.015 g (15%); Spectral data showed this to be the

same compound as that obtained by reduction of the dinitrophenol.

N-(3.5-DimethoxvDhenyl)-a-Benzoylmethvlamine (2.14)

To a flask containing 3,5-dimethoxyaniline (1 g) and sodium bicarbonate

(1 g) in dimethylformamide (12 ml) was added a-bromoacetophenone (1.3 g).

The mixture was stirred at 50C in a water bath for 2 h and then diluted with

water (100 ml). The precipitate was filtered and washed with water. The solid

was dissolved in chloroform, dried over sodium sulfate, and concentrated to

dryness. Crystalization from ether gave pale-yellow crystals. Yield: 1.55 g

(88%); m.p.: 112-115C; nmr (CDCI3): 7.77-8.00 (2H,Ar.m.), 7.27-7.60

(3H,Ar.m.), 5.87 (3H,Ar.s.), 4.53 (2H,s,CH2), 3.73 (6H,s,OMe), 3.1-3.5

(1H,br,NH); ir (cm1): 3380, 3060, 3000, 2955, 2835, 1690, 1620, 1580,







CD z





1515, 1480, 1440, 1315, 1250, 1200, 1170, 1155, 1070, 995, 955, 930,

815, 790, 760, 685, 675. Anal. Calc. for C1,H17NO3: C, 70.83; H, 6.32; N,

5.16. Found: C, 70.71; H, 6.30; N, 5.09.

N-(3.5-DimethoxvDhenvl)-N-Benzovlmethvlacetamide (2.15)

To the secondary amine, 2.14 (0.1 g), was added acetic anhydride (2

ml), and the mixture was heated in a water bath at 800C for 10 min. Ice was

added, then water, and a gummy material formed. After 30 min, the water

was decanted and fresh water added to the gum. When no change was seen,

the gum was partitioned between chloroform (2x10 ml) and water twice, and

the chloroform was dried over sodium sulfate and concentrated to an oil. The

product did not crystallize even after chromatograophy. Yield: 0.068 g (59%);

nmr (CDCI): 7.77-7.93 (2H,Ar.m.), 7.23-7.53 (3H,Ar.m.), 6.50 (2H,d,J = 2Hz),

6.37 (1H,dd,Ar.), 5.03 (2H,s), 3.73 (6H,s,OMe), 2.03 (3H,s,OAc).

N-Benzovlmethvl-(3.5-Dimethoxv-2-NitroDhenvl)acetamide (2.16)

After the completion of the above acetylation reaction on 2.14 (3.65 g),

the solution was cooled in an ice bath for 15 min. At 5C, nitric acid (3.0 ml)

was added dropwise with stirring, and the reaction was stopped after 30 min

with the addition of ice and water. After 30 min, the solid was filtered,

dissolved in chloroform, dried, and concentrated. After chromatography, the

major product was crystallized from ether/ ligroin. Yield: 3.47 g (72%); m.p.:

113-115OC; nmr (CDCI3): 7.73-7.93 (2H,Ar.m.), 7.27-7.57 (3H,Ar.m.), 6.90

(1H,d,J=2Hz), 6.53 (1H,d,J=2Hz), 5.63 (1H,d,J=18Hz), 4.27 (1H,d,J=


18Hz), 3.85 (3H,s,OMe), 3.82 (3H,s,OMe), 2.00 (3H,s,OAc). Anal. Calc. for

C1-H1,N20e: C, 60.33; H, 5.06; N, 7.82. Found: C, 60.30; H, 5.05; N, 7.81.

4-Acetvl-6.8-Dimethoxv-2-Phenyl-3.4-Dihydroquinoxaline (2.17)

In a flask containing 2.16 (3.0 g) dissolved in methanol (15 ml) was

added sodium dithionite (7 g) dissolved in water (40 ml), and the mixture was

heated at reflux for 30 min. After further dilution with water, the contents

were filtered. The solid was dissolved in chloroform, washed with water, dried,

and concentrated. The product was crystallized from ether as white crystals.

Yield: 1.7 g (65%); nmr (CDCI3): 7.77-8.03 (2H,Ar.m.), 7.20-7.53 (3H,Ar.m.),

6.40 (2H,br,Ar.), 4.70 (2H,s), 3.93 (3H,s,OMe), 3.83 (3H,s,OMe), 2.27




The 6,8-dimethoxy-2-phenyl-3-chloroquinoxaline (2.8) obtained as an

intermediate in the synthesis of the quinoxalinequinones proved to be a very

useful compound. Not only was it readily dehalogenated, but it was also

readily displaced by other nucleophiles providing a pathway to more 3-

substituted analogues. The conversion of the quinoxalinone (2.6) to the 3-

chloro derivative proceeded in a much higher yield than to the 3-bromo

compound, so that the former was used in those reactions where it showed

sufficient reactivity.

Under the same conditions used in the nitration of 6,8-dimethoxy-2-

phenylquinoxaline described in chapter 2, the 3-chloro analogue gave only the

mononitro derivative. Adding a molar excess of nitric acid did not produce any

of the corresponding dinitrophenol as was observed with the quinoxaline, 2.9.

The mononitro derivative of the 3-chloro analogue did not appear useful. Again,

attempts at reducing the 5-nitro group with sodium dithionite were

unsuccessful (Fig. 3.1-1), and when the reduction was carried out using iron

in acetic acid, only the dehalogenated amine was produced. However, the

compound did find use in making the methoxyquinone analogue 2.13. Not only


did nitration stop at the mononitro stage with the chloroquinoxaline, but the

reduction of the nitro group and dehalogenation could be combined into one


In an attempt to produce further nitration of the 3-chloroquinoxaline, it

was observed that addition of a very small amount of sulfuric acid to the

reaction mixture, containing essentially the mononitro compound, permitted

further reaction to take place. A mixture of products was obtained from this

which are shown in Fig. 3.1-2. The major product isolated was not the

expected nitration product at all but 3-chloro-2-phenyl-6-methoxyquinoxaline-

5,8-dione. Although this had nearly the same Rf as the mononitro compound,

they were readily distinguishable from each other on the tic slide after exposure

to ammonia vapors. Other products isolated from this reaction were the

expected dinitrophenol and a dimethoxydinitro compound. The

mononitrophenol was also isolated, and it was converted into the 6-methoxy-

5,8-dione in two steps: 1) reduction to the aminophenol with dithionite, and

2) oxidation with ceric sulfate.

Reduction of the dinitrophenol was carried out in the same manner as

described in chapter 2. The spectral and analytical data obtained on the

product of this reaction agreed with those expected for the 7-aminoquinone of

chloroquinoxaline. As can be seen in Fig.3.1, this was the only route which led

to this compound.

For the conversion of the 6-methoxy-5,8-dione to the desired 7-

aminoquinone, the three step procedure consisting of bromination, azidation,


and reduction/ oxidation was attempted. The bromination reaction did not

proceed to completion when carried out in acetic acid. Yields were low due to

the need for chromatographic separation. However, bromination using a freshly

prepared solution of bromopyridinium bromide in pyridine readily proceeded to

completion and gave acceptable yields of the 3-chloro-2-phenyl-7-bromo-6-

methoxyquinoxaline-5,8-dione. When this 7-bromoquinone was combined with

sodium azide in dimethylformamide, a dark-brown product was obtained in

excellent yield. It showed a strong peak at 2120 cm1- in its ir spectrum which

indicated that it was an azido derivative. The product obtained after reduction

of this dark-brown compound with sodium dithionite still showed an azide peak

in its ir spectrum, but the carbonyl peaks corresponding to the quinone were

absent. This indicated that the reduction reaction only proceeded to the

dihyroquinone. Reduction under more vigorous conditions gave yet another

compound in which both the azide peak and the carbonyl bands were absent

in its ir spectrum. Unlike the characteristic auto-oxidative behavior expected

for the aminohydroquinones of this type, the present compound resisted air-

oxidation to yield an aminoquinone. The oxidation of this compound with ceric

sulfate gave a purple compound whose ir and nmr spectra contained the peaks

expected for the 7-aminoquinone. These developments are shown in Fig. 3.1-3


The elemental analysis of this aminoquinone showed an unusually high

nitrogen content. Such a result could be accounted for if the proposed

tetrazole ring was formed. It was found in the literature that 3-

1O HN03


2.8 33


3.2 3.8

NoN3/ AcOH

0 0 1) No2S204/
) '' NaHCO3/H20
4) N 2)C,(S 4)2

o I I


Figure 3.1. Synthetic Pathway to Chloroquinoxalineaminoquinone
and Tetrazolo-(1,5)-quinoxalineaminoquinone


chlorquinoxalines readily form tetrazolo-1,5-quinoxalines when reacted with

sodium azide through displacement of the chloro group and cyclization.57 Also,

the tetrazole ring does not show any absorption in the azide region of the ir.72

From this, it is apparent that both halogens must have been displaced in

the azidation reaction with the chloroquinoxaline bromoquinone. Since an azide

peak was clearly present in the ir, the bromine must have been displaced. The

nmr spectra were the best indication for the displacement of the desired 3-

chloro group. The chemical shifts of the ortho-protons of the phenyl ring had

moved downfield. The poor resolution seen in the spectra of the

dihydroquinones made the interpretation of these spectra difficult. Structure

elucidation remained in doubt until the oxidation reaction to the aminoquinone.

The reaction of sodium azide with the two other chloroquinoxaline

compounds was carried out in order to establish this nmr shift as a

characteristic trend. At higher temperatures, this reaction proceeded readily

with 3-chloro-6,8-dimethoxy-2-phenylquinoxaline (2.7) to produce a tetrazole

derivative. In the reaction of sodium azide in acetic acid on the 6-methoxy-3-

chlorquinoxalinequinone 3.2, where an excess of sodium azide is used, the only

aminoquinone product obtained was the tetrazoloquinoxaline.


As mentioned earlier, the chloro group of the chloroquinoxaline

intermediate was readily substituted by nucleophiles. Aside from the

dehalogenation and azidation, this group was also displaced by hydrazine and

methoxide ion. The preparation of the 3-methoxy derivative and its conversion

to the corresponding aminoquinone are described here.

Formation of 2-phenyl-3,6,8-trimethoxyquinoxaline from the 3-chloro

derivative was accomplished by refluxing 2.8 in methanol containing sodium

hydroxide. Nitration of this compound was difficult to control. The major

product formed after the addition of one equivalent of nitric acid was a red

compound which was too insoluble in all solvents to obtain an nmr spectrum.

Traces of the starting material, the mononitro compound, and even the quinone

were seen, but in very low yields. As shown in Fig. 3.2, this problem was

circumvented by reacting the mononitrochloroquinoxaline (3.1) with methanolic

sodium hydroxide. Excellent yields of the desired mononitromethoxyquinoxaline

derivative were then obtained.

The reduction of the mononitro compound to the amine and its

conversion to the quinone were carried out as previously described for

compound 2.13 (p. 52). The introduction of the amino group onto the quinone

ring was accomplished using sodium azide in acetic acid in 40 percent yield.





Meo Me
NoOMe I ) Fe/ AcOH
MeOH 2) Ce(SO4)2

OMe 3.14


> AcOH

Synthetic Route to Methoxyquinoxalineaminoquinone

Figure 3.2.


3-Chloro-6.8-Dimethoxv-5-Nitro-2-Phenvlouinoxaline (3.1)

To a suspension of 2.8 (1.27 g) in acetic anhydride (6 ml) was added

nitric acid (70%, 1.5 ml) at 5-10C. After 45 min, ice was added to the

reaction contents with stirring and further diluted with water (50 ml). The

reaction work up was the same as that described for compound 2.10 (p. 48).

The product was crystallized from acetone and filtered. Yield: 1.17 g (74%);

m.p.: 212-214C; nmr (CDCI3/DMSO-de): 7.68-7.78 (2H,Ar.m.), 7.45-7.52

(3H,Ar.m.), 6.80 (1H,Ar.s.), 4.10 (6H,s,OMe); ir (cm1): 2970, 1615, 1525,

1470, 1375, 1345, 1220, 1200, 1160, 1115, 815, 690. Anal. calc. for

C1eH12N304CI: C, 55.60; H, 3.50; N, 12.16. Found: C, 55.67; H, 3.48; N,


3-Chloro-6-Methoxy-2-Phenylquinoxaline-5.8-Dione (3.2)

At room temperature, 2.8 (1.0 g) was dissolved in chloroform (6 ml) and

acetic anhydride (8 ml). To this was added nitric acid (70%, 0.8 ml), and the

mixture was allowed to react until no starting material was present on tic. At

this point, a 1:1 mixture of H2SO4/ HNO3 (0.4 ml) was added. After 10 min,

the reaction was stopped by the addition of ice and then water. After 1 h, the

precipitate was filtered and extracted with chloroform. The chloroform was

washed twice with aqueous sodium bicarbonate and then with water and

concentrated to dryness. The residue was stirred with ether until it became

crystalline. Yield: 0.4 g (40%); m.p.: 262-264C; nmr (CDCI3): 7.92-7.83





(2H,Ar.m.), 7.50-7.58 (3H,Ar.m.), 6.40 (1H,s,vinyl), 3.97 (3H,s,OMe); ir (cm

1): 3050, 3020, 2945, 1700, 1666, 1610, 1540, 1520, 1305, 1260, 1230,

1180, 1160, 1065, 950, 860, 720, 695. Anal. Calc. for C15HN203CI: C,

59.91; H, 3.02; N, 9.32. Found: C, 59.94; H, 3.01; N, 9.32.

3-Chloro-5.7-Dinitro-8-Hydroxy-6-Methoxv-2-Phenvlauinoxaline (3.3)

This was obtained from the sodium bicarbonate washings of the above

reaction workup. After two or three such washes, the combined aqueous

layers were acidified, and the solid was filtered. Crystallization from ether/

ligroin yielded 0.11 g (9%); m.p.: 186-187C; nmr (CDCI3): 7.78-7.85

(2H,Ar.m.), 7.53-7.60 (3H,Ar.m.), 4.13 (3H,s,OMe), 3.75 (1H,br,OH); ir (cm-1):

3380, 3065, 3020, 2940, 1640, 1545, 1535, 1480, 1465, 1450, 1410,

1390, 1360, 1330, 1300, 1170, 1135, 1100, 1040, 1030, 975, 900, 820,

745, 695, 650.

A portion of the dinitrophenol (0.1 g) was heated with acetic anhydride

(2 ml) at 1000C for 15 min. The cooled mixture was diluted with water, and

the solid was filtered after 10 min. and crystallized from ether. m.p. 173-

1750C; nmr (CDCI3): 7.75-7.85 (2H,Ar.m.), 7.50-7.57 (3H,Ar.m.), 4.17

(3H,s,OMe), 2.43 (3H,s,OAc); ir (cm-): 2960, 2940, 1800, 1625, 1545,

1355, 1320, 1160, 1150, 1130, 1095, 810, 800, 695. Anal. Calc. for

C17HjN407CI: C, 48.76; H, 2.65; N, 13.38. Found: C, 49.01; H, 2.70; N,









3-Chloro-6.8-Dimethoxv-5.7-Dinitro-2-Phenvlquinoxaline (3.4)

This was obtained from the mother liquor as described under 3.2. The

residue was subjected to column chromatography (silica/benzene). Fractions

containing the first band, on concentration and crystallization from ether, gave

3.4 as a pale-yellow crystalline solid. Yield: 0.07 g (5.4%); m.p.: 178-181 C;

nmr (CDCI3): 7.73-7.88 (2H,Ar.m.), 7.47-7.57 (3H,Ar.m.), 4.47 (3H,s,OMe),

4.10 (3H,s,OMe); ir (cm-1): 2960, 2930, 2875, 1610, 1550, 1530, 1470,

1380, 1350, 1320, 1190, 1145, 1095, 1035, 850, 805, 710, 695, 645.

3-Chloro-8-Hvdroxy-6-Methoxv-5-Nitro-2-Phenvlouinoxaline (3.5)

This was obtained from the chromatographic column (silica/benzene)

described under 3.4 by elution with a 2% acetone in benzene. The fractions

containing this component were concentrated to dryness, and the solid was

crystallized from acetone/ ligroin. m.p.: 214-215C; nmr (CDCI3): 7.70-7.85

(2H,Ar.m.), 7.40-7.57 (3H,Ar.m.), 7.02 (1H,Ar.s.), 4.03 (3H,s,OMe), 2.60-

3.30 (1H,br,OH); ir (cm'): 3410, 2960, 2920, 2860, 1630, 1520, 1480,

1340, 1200, 1175, 1160, 960, 820, 720, 690. Anal. Calc. for C15HIoN304CI:

C, 54.31; H, 3.04; N, 12.67. Found: C, 54.29; H, 3.01; N, 12.58.

3-Chloro-6-Methoxv-2-Phenvlauinoxaline-5.8-Dione (3.2)

The nitrophenol 3.5 (0.4 g) was reduced with sodium dithionite (0.3 g)

in 5% aqueous bicarbonate/ methanol (1:1, 10 ml) by boiling under reflux for

15 min. The cooled reaction mixture was extracted with chloroform, and the

chloroform was dried and concentrated to dryness. The solid was subjected

to chromatography on silica gel using 2% acetone in benzene. Fractions


containing the first band were concentrated to dryness, and this aminophenol

(3.6) was crystallized from ether. Yield: 0.246 g (68%); m.p.: 175-1780C;

nmr (CDCI3): 7.67-7.90 (2H,Ar.m.), 7.37-7.58 (3H,Ar.m.), 7.03 (1H,Ar.s.),

3.97 (3H,s,OMe), 3.0-4.3 (2H,br.,NH2); ir (cm-1): 3395, 3300, 3000, 2985,

2940, 2850, 1615, 1540, 1495, 1470, 1395, 1340, 1305, 1250, 1200,

1175, 1115, 960, 830, 770, 705.

A solution of the aminophenol (0.2 g) in chloroform (10 ml) was shaken

with aqueous ceric sulfate/ 2M sulfuric acid (10 ml). The chloroform was

concentrated to dryness, and the product was crystallized from ether. The

product was found to be identical with 3.2.

7-Amino-3-Chloro-6-Methoxv-2-Phenvlouinoxaline-5.8-Dione (3.7)

A 5% sodium bicarbonate solution (8 ml) containing sodium dithionite

(0.6 g) was added to a methanolic solution (2 ml) containing the dinitrophenol,

3.3 (0.1 g), and the mixture was refluxed for 10 min. The contents were

diluted with water and extracted with chloroform. The chloroform was dried

over sodium sulfate and concentrated. The solid was purified by

chromatography (silica, 1:1 CHCI3/ benzene), and after concentration of the

appropriate fractions, the product was crystallized from acetone and filtered.

Yield: 0.029 g (35%); m.p. 260-2750C (dec.); nmr (CDCI3): 7.67-7.83

(2H,Ar.m.), 7.33-7.53 (3H,Ar.m.), 6.00-6.25 (2H,br,NH2), 4.00 (3H,s,OMe);

ir (cm-): 3470, 3340, 2950, 1685, 1640, 1580, 1555, 1510, 1440, 1350,

1280, 1195, 1140, 1035, 880, 745, 695. Anal. Calc. for C15HioN303CI: C,

57.06; H, 3.20; N, 13.31. Found: C, 56.79; H, 3.36; N, 13.12.





7-Bromo-3-Chloro-6-Methoxv-2-Phenvlauinoxaline-5.8-Dione (3.8)

The brominating reagent was prepared by adding bromine (0.2 ml),

dropwise with stirring, to well-cooled pyridine (3 ml) at O0C. This reagent was

added to a suspension of 3.2 (0.84 g) in pyridine (5 ml) at 0-50C with stirring.

The mixture was stirred for 45 min at this temperature until a clear red solution

resulted. The contents were poured into 2N HCI (50 ml) and stirred for 5 min.

This mixture was diluted further with water and extracted with chloroform

(2x50 ml). The combined extract was washed with sodium bisulfite which

changed the color of the extract from red to pale yellow. After washing with

aqueous sodium bicarbonate, the chloroform was concentrated to dryness.

Crystallization of the residue with ether/ ligroin gave 3.8. Yield: 0.58 g (55%);

m.p.: 240-2500C (dec.); nmr (CDCI3): 7.88-7.97 (2H,Ar.m.), 7.47-7.55

(3H,Ar.m.), 4.37 (3H,s,OMe).

7-Azido-8-Methoxv-4-Phenvl-1.5-Tetrazoloquinoxaline-6.9-Dione (3.9)

The bromoquinone 3.8 (0.44 g) was dissolved in DMF (10 ml), and

sodium azide (0.1 g) was added to the solution. An immediate darkening of the

solution was seen, and within 5 min the reaction was complete. The mixture

was diluted with water (50 ml), and the precipitate was filtered and washed

with water. The solid was dissolved in chloroform, dried, and concentrated.

The product was crystallized from ether to yield 0.37 g (92%) of 3.9 as dark-

brown crystals. nmr (CDCI3): 8.60-8.80 (2H,Ar.m.) 7.43-7.67 (3H,Ar.m.),

4.20 (3H,s,OMe); ir (cm-): 3400, 2940, 2860, 2120, 1680, 1675, 1610,

1580, 1500, 1300, 1230, 1180, 1030, 685.

7-Azido-6.9-Dihydroxv-8-Methoxv-4-Phenyl-1.5-Tetrazoloauinoxaline (3.10)

The azidoquinone, 3.9 (0.2 g), was suspended in 4:1 water/ methanol

(10 ml) and sodium dithionite (0.6 g) was added. The mixture was heated at

800C for 10 min. The tic revealed a conversion to a slower-moving, yellow

spot which darkened on standing. The reaction contents were extracted with

chloroform, and the chloroform was dried and concentrated. The product was

crystallized from acetone/ligroin. Yield: 0.17g (85%). m.p.: 163-164C; nmr

(CDCIz): 8.57-8.83 (2H,Ar.m.), 7.70-8.50 (2H,br), 7.37-7.67 (3H,Ar.m.), 4.17

(3H,s,OMe); ir (cm-): 3400, 3060, 2920, 2840, 2110, 1600, 1530, 1500,

1480, 1350, 1325, 1090, 1030, 800, 690.

7-Amino-6.9-Dihvdroxv-8-Methoxv-4-Phenyl-1.5-Tetrazologuinoxaline (3.11)

To 3.10 (0.1 g) was added 10 ml of a 4:1 water/ methanol solution

containing sodium dithionite (0.3 g) and sodium bicarbonate (0.1 g). The

mixture was heated at 80C for 10 min. The tic showed the absence of the

starting material and the appearance of a slower brown spot. After several

minutes of air exposure, it was noticed that this spot had turned purple. The

reaction contents were diluted with water and extracted with chloroform. The

chloroform layer was filtered through cotton, dried over sodium sulfate, and

concentrated. The product, 3.11, was obtained as a yellow-brown solid (0.07

g, 76%) after crystallization from ether. nmr (CDCI3/ DMSO-de): 8.63-8.83

(2H,Ar.m.), 7.90-8.63 (2H,br), 7.27-7.53 (3H,Ar.m.), 4.50 (2H,br,NH2), 3.97

(3H,s,OMe); ir (cm1): 3480, 3460, 3410, 3360, 2940, 1600, 1515, 1480,

1345, 1325, 1140, 1010, 690, 625.

7-Amino-8-Methoxv-4-Phenvl-1.5-Tetrazoloauinoxaline-6.9-Dione (3.12)

Since the ir spectrum of 3.11 showed strong absorption in the hydroxyl

region and no absorption in the carbonyl (quinone) region, 3.11 (0.07 g) was

dissolved in chloroform (10 ml) and shaken with 0.5M sulfuric acid solution

(1Oml) containing ceric sulfate (0.1g). The chloroform layer rapidly turned to

a deep purple color, and tic showed the absence of the starting material. The

chloroform layer was filtered, dried over sodium sulfate, and concentrated to

dryness. The product, 3.12, was crystalized from acetone/ligroin (3:1). Yield:

0.062 g (90%); m.p.: 190-192C (dec.); nmr (CDCI3): 8.53-8.80 (2H,Ar.m.),

7.40-7.67 (3H,Ar.m.), 6.75-7.03 (2H,br,NH2), 3.83 (3H,s,OMe); ir (cm-1):

3455, 3340, 2940, 1710, 1645, 1620, 1585, 1560, 1500, 1440, 1390,

1335, 1220, 1100, 1090, 995, 820, 755, 690, 630. Anal. Calc. for

C15HioNe03: C, 55.90; H, 3.13; N, 26.08. Found: C, 56.30; H, 3.28; N,


2-Phenvl-3.6.8-Trimethoxvauinoxaline (3.13)

To methanol (30 ml) was added 2.8 (0.87 g) and sodium hydroxide

pellets (0.4 g), and the mixture was refluxed for 30 min. Water (50 ml) was

added, and the precipitate was filtered. After washing with water, the solid

was dissolved in chloroform, dried, and concentrated. Crystallization of the

residue with ether afforded 3.13. Yield: 0.71 g (83%); m.p.: 148-150C;

nmr (CDCI3): 7.80-8.07 (2H,Ar.m.), 7.25-7.50 (3H,Ar.m.), 6.73 (1 H,d,J = 2Hz),

6.55 (1H,d,J=2Hz), 4.06 (3H,s,OMe), 3.97 (3H,s,OMe), 3.90 (3H,s,OMe);

ir (cm'): 3050, 3000, 2965, 2940, 2850, 1610, 1575, 1550, 1460, 1440,






1370, 1340, 1300, 1280, 1265, 1220, 1150, 1130, 1045, 980, 825, 760,

690. M.S.(FAB): M+ = 296. Calc. Mol. Wt. = 296.

5-Nitro-2-Phenvl-3.6.8-Trimethoxyvuinoxaline (3.14)

Sodium methoxide (0.3 g) and 3.1 (1.5 g) were added to methanol (30

ml), and the mixture was refluxed for 1 h. The reaction work up was the same

as that described above for compound 3.13. Crystallization from acetone gave

3.14. Yield: 1.3 g (88%); m.p.: 197-200OC; nmr (CDCI3/ DMSO-de): 7.83-

8.10 (2H,Ar.m.), 7.27-7.53 (3H,Ar.m.), 6.65 (1H,Ar.s.), 4.06 (6H,s,OMe),

4.03 (3H,s,OMe); ir (cm'): 3060, 3010, 2980, 2950, 2850, 1620, 1580,

1560, 1520, 1475, 1465, 1450, 1435, 1380, 1350, 1260, 1230, 1220,

1145, 1120, 1005, 980, 810, 740, 690. Anal. Calc. for C17H15N3Os: C,

59.82; H, 4.43; N, 12.31. Found: C, 59.91; H, 4.41; N, 12.32.

2,6-Dimethoxv-2-PhenvlQuinoxaline-5.8-Dione (3.15)

Reduced iron (0.8 g) was added to acetic acid (20 ml) containing 3.14

(1.7 g), and the mixture was heated at 800C for 15 min until the contents

turned a light gray color. The reaction procedure that was followed was the

same as described for compound 2.13 (p. 52). The quinone thus obtained,

3.15, was crystallized from acetone. Yield: 1.03 g (70%); m.p.: 223-2260C;

nmr (CDCI3): 8.03-8.23 (2H,Ar.m.), 7.33-7.53 (3H,Ar.m.), 6.27 (1H,s,vinyl),

4.21 (3H,s,OMe), 3.94 (3H,s,OMe); ir (cm1): 3050, 2950, 2840, 1700,

1655, 1605, 1530, 1445, 1375, 1350, 1325, 1255, 1230, 1180, 1120,

1060, 1000, 795, 760, 685, 670. Anal. Calc. for CleH13N304: C, 64.86; H,

4.08; N, 9.45. Found: C, 64.91; H, 4.07; N, 9.47.


CY) ---





-LC 2


7-Amino-3.6-Dimethoxv-2-Phenvlauinoxaline-5.8-Dione (3.16)

Sodium azide (0.2 g) was added to acetic acid (10 ml) containing 3.15

(0.26 g), and the mixture was heated at 550C for 1 h. The reaction contents

were diluted with water (100 ml) and extracted with chloroform (2x50 ml).

This extract was washed with water, dried, and concentrated to dryness. The

solid was chromatographed on silica eluting with 1:1 chloroform/ benzene, and

the fractions containing the band corresponding to the aminoquinone were

collected and concentrated to dryness. After crystallization from acetone, the

filtered crystals were washed with ether and dried. Yield: 0.08 g (29%); m.p.:

177-179oC; nmr (CDCI3/ DMSO-de): 8.00-8.17 (2H,Ar.m.), 7.30-7.47

(3H,Ar.m.), 5.03-5.30 (2H,br,NH2), 4.25 (3H,s,OMe), 4.06 (3H,s,OMe); ir

(cm-): 3460, 3320, 2940, 2840, 1680, 1640, 1600, 1570, 1520, 1440,

1365, 1345, 1250, 1230, 1060, 995, 970, 790, 745, 685. Anal. Calc. for

CleH13N304: C, 61.72; H, 4.22; N, 13.50. Found: C, 61.82; H, 4.23; N,




J -1" 0


L_ E



I c



The method used most often for generating a 2-(or 3-) cyanoquinoxaline

system has been to start with the appropriate precursor already containing the

cyano function, as for example, by condensation of a diamine with an a-halo-B-

keto nitrile. Only two examples were found in the literature where the cyano

group was introduced by nucleophilic displacement. In these, the reaction of

sodium cyanide with quinoxalines containing a methylsulfonyl or

trialkylammonium salt yielded the cyanoquinoxaline derivative. Here, since the

quinoxalinone, 2.6, was readily converted into the 3-chloro analogue by

reaction with phosphorus oxychloride, the possibility of displacement to form

the 3-cyano analogue was explored.

Unfortunately, reaction of the 3-chloroquinoxaline derivative with an

alkali cyanide under a variety of conditions produced little or no reaction. Since

cuprous cyanide has also been used as an alternative for the displacement of

an active halogen by cyanide, this approach was also tried under a variety of

conditions. Even when vigorous conditions were used (e.g. refluxing in

nitrobenzene for 16 hours), only about 10% of the desired product formed.

The reaction was also attempted with the nitration product 3-chloro-6,8-


dimethoxy-5-nitro-2-phenylquinoxaline, 3.1, with cuprous cyanide in refluxing

dimethylformamide for several hours, but here again there was little or no


This lack of reactivity of the 3-chloro derivative made it necessary to

prepare the 3-bromo derivative and test its reactivity. Based on experience

with a similar reaction in the pyridine series,41 it was felt that this derivative

might possess enough reactivity for the conversion. Preliminary experiments

indicated that this was the case, and this prompted a study of the optimal

conditions for the production of the 3-bromo derivative.

Reaction of quinoxalinone 2.6 with phosphorus tribromide in

nitrobenzene, chlorobenzene, and others led to various degrees of

decomposition. The best results were obtained when the quinoxalinone was

heated in phosphorus tribromide, neat, at 130-1350C for 1-2 hours. A side

product isolated from the reaction was identified as the 3-bromo-8-hydroxy-6-

methoxy-2-phenylquinoxaline formed by demethylation at the 8-position. This

compound showed a base-induced shift in its uv spectrum, and it was oxidized

to the same quinonoid compound that was obtained from the nitration of 3-

bromo-6,8-dimethoxy-2-phenylquinoxaline using the method described for

obtaining compound 3.2 (p. 65).

Conversion of the 3-bromoquinoxaline to the 3-cyanoquinoxaline was

accomplished by starting with 1.2 equivalents of cuprous cyanide in refluxing

DMF for 2 hours. There was some difficulty in monitoring this reaction because

of the closeness of the Rf values of the starting material and the product.

However, the reaction proceeded to completion when 1.5 equivalents of

cuprous cyanide were used.

The nitration of the 3-cyanoquinoxaline compound followed the same

course as that seen with the 3-chloroquinoxaline. Nitration stopped at the

mononitro compound when the reaction was carried out in acetic anhydride and

nitric acid. Since the reduction of this mononitro derivative with sodium

dithionite also failed, the iron/ acetic acid method was used. The reduction

took place, and after oxidation of the resulting amine a quinone was obtained.

Surprisingly, the nmr and ir spectra of this product were identical to those of

compound 2.13, signifying the loss of the cyano group during the reduction.

This result made it necessary to use the stronger nitrating conditions, i.e.

the method used in obtaining compound 3.2 (p. 65). The addition of sulfuric

acid to the mononitro-cyanoquinoxaline again produced the dione, but it was

not the major product. Instead, a mixture of nitration products was formed

(Fig. 4.1-2) consisting of the 5-nitro-6-8-dimethoxy, the 5-nitro-8-hydroxy-6-

methoxy, the 5,7-dinitro-6,8-dimethoxy, and 5,7-dinitro-8-hydroxy-6-methoxy

derivatives, and the nitrophenol compounds predominated. The nitrophenol of

cyanoquinoxaline was readily converted to the cyanoquinoxalinequinone in the

same manner as the corresponding chloroquinoxaline mononitrophenol (p. 69).

Thus, reduction of this compound with dithionite did not suffer decyanation,

and the resulting aminophenol was oxidized to the desired quinone.

The reaction of the 6-methoxyquinone with sodium azide in acetic acid

gave the desired 7-amino-6-methoxyquinoxaline-5,8-dione, another one of the



OMg 4.4


OH 4.7 N

+ mononitrophenol

o) NaN3/ AcOH

b) Br2/ pyr
NaNa/DMF Mo0o L
Na2S204 1

c) N02S204

Figure 4.1. Synthetic Pathway to Cyanoquinoxalineaminoquinone

target compounds (Fig. 4.1-3). In order to confirm the structure of the 7-

aminoquinone, its preparation by the alternative scheme consisting of

bromination of the 6-methoxyquinone, azidation, and reduction/ autooxidation

was studied. Bromination using bromopyridinium bromide in pyridine gave the

7-bromo-6-methoxyquinone in good yield. Reaction of this compound with

sodium azide in warmed acetic acid produced the same 7-amino-6-

methoxyquinone as above. The reduction of the dinitrophenol followed by air-

oxidation also formed the same aminoquinone.

A comparison of the reactivity patterns of the 3-chloro and the 3-cyano-

quinoxalaline derivatives will be instructive at this point. During the nitration,

the 3-chloro compound gave the 6-methoxyquinone as the major product, while

in the case of 3-cyanoquinoxaline, the nitration products constituted the

majority. Reduction of the mononitro derivatives of each using iron in acetic

acid led to dehalogenation and decyanation products. The 3-chloro-7-

bromoquinone showed reactiviy at the 7-position and the 3-position with

respect to sodium azide, while the 3-cyano-7-bromoquinone gave only

displacement of the bromine under the same conditions.


One of the main reasons for preparing the 3-cyanoquinoxaline derivative

was that it can lead to others such as the 3-carbomethoxy, 3-carboxy and the

3-carboxamido analogues. Based on the previously described procedure in

which a 2-cyanopyridine was converted directly to a 2-carbomethoxy derivative

by the action of methanolic sulfuric acid,39 several attempts were made to

adapt this reaction to the present 3-cyanoquinoxaline derivative. Little success

was seen here because, under the conditions of the reaction, demethylation of

one or both of the methoxyls became a competing reaction.

The compound showed resistance to hydrolysis when refluxed in

concentrated hydrochloric acid for 6 hours. However, under base conditions

(2N NaOH) hydrolysis was more successful, yielding the 3-carboxamide after 2

hours of reflux. Further hydrolysis of the amide to the carboxylic acid was

slow in base and rapid in acid, and conversion of the amide to the

carbomethoxy derivative directly using methanolic sulfuric acid proceeded in

good yield.

Thus, a two-step sequence was employed for the conversion of the 3-

cyano to 3-carbomethoxy derivative: 1) basic hydrolysis (2 h reflux with 2N

sodium hydroxide) to the 3-carboxamide, and 2) reaction of the carboxamide

with methanolic sulfuric acid (1 hour, 900C) to yield the 3-carbomethoxy-

quinoxaline (Fig. 4.2).

Nitration of the 3-carbomethoxy-6,8-dimethoxy quinoxaline gave the 5-

nitro derivative in good yield. Attempts at further nitration using the stronger

reaction conditions did give some quinone in trace amounts, but it also

produced several other compounds. The 6-methoxyquinone was, however,

obtained from the mononitro derivative in two steps by reduction to the 5-

amino compound using iron in acetic acid followed by oxidation with ceric

sulfate without isolation of the amine.

. MeOH




0 4.14



1) Fe/ AcOH

2) Ce(SO4)2



Figure 4.2. Synthetic Scheme for the Synthesis of

This 3-carbomethoxy-6-methoxyquinone was then converted directly to

the corresponding 7-amino-6-methoxyquinone using the sodium azide/acetic

acid reaction as described previously in satisfactory yield.

A similar sequence of reactions was attempted with the 3-

carboxamidoquinoxaline, but nitration of this compound led to extensive

breakdown even under mild conditions, and only low yields of the desired

mononitro product were obtained.


3-Bromo-6.8-Dimethoxv-2-Phenvlguinoxaline (4.1)

Phosphorus tribromide (10 ml) was added to 2.6 (0.98 g), and the

mixture was stirred at 1350C in an oil bath for 1.25 h. At completion of the

reaction, the solution was chilled, diluted with ice water, and stirred at 0-5C

for 0.5 h. The precipitate was filtered, dissolved in chloroform, and washed

with aqueous sodium bicarbonate. The chloroform layer was dried over sodium

sulfate and concentrated to dryness. The product was crystallized from ether

and filtered. Yield: 0.62 g (50%); m.p.: 193-194C; nmr (CDCI3): 7.63-7.83

(2H,Ar.m.), 7.33-7.57 (3H,Ar.m.), 6.97 (1H,d,J= 2Hz), 6.73 (1H,d,J= 2Hz),

4.02 (3H,s,OMe), 3.95 (3H,s,OMe); ir (cm'): 3050, 3000, 2960, 2915,

2820, 1615, 1525, 1465, 1445, 1400, 1325, 1230, 1200, 1150, 1140,

1040, 825, 685. Anal. Calc. for C1eH13N202Br: C, 55.67; H, 3.80; N, 8.11.

Found: C, 56.19; H, 3.86; N, 8.20.

3-Bromo-8-Hydroxy-6-Methoxy-2-Phenylauinoxaline (4.2)

This compound was obtained as a by-product from the above reaction.

The mother liquor from 4.1 was subjected to column chromatography (silica

gel/ benzene). The first fractions contained this faster moving phenol. Yield:

0.05 g, (5%); m.p.: 194-195OC; nmr (CDCI3): 7.60-7.80 (2H,Ar.m.), 7.57

(1H,s,OH), 7.33-7.55 (3H,Ar.m.), 6.80-6.97 (2H,2d,Ar), 3.90 (6H,s,OMe); ir

(cm-'): 3465, 3060, 3020, 2990, 2940, 1740, 1640, 1570, 1525, 1480,

1455,1410,1390,1300,1280,1230, 1200, 1175, 1155, 1130, 1030, 950,

860, 820, 790, 760, 690, 660. Anal. Calc. for C15H11N202Br: C, 54.40; H,

3.35; N, 8.46. Found: C, 54.63; H, 3.41; N, 8.37.

3-Cvano-6,8-Dimethoxy-2-Phenylquinoxaline (4.3)

A mixture of 4.1 (2.57 g) and cuprous cyanide (0.85 g) (1.27 eq) in DMF

(15 ml) was boiled under reflux for 1.25 h. The contents were diluted with

water, and the precipitate was filtered. The gummy precipitate was dissolved

in chloroform and refiltered to remove the cuprous bromide/ cyanide. The

filtrate was dried over sodium sulfate and concentrated to dryness.

Crystallization from ether gave 4.3. Yield: 1.95 g (90%); m.p.: 227-2280C;

nmr (CDCI3): 7.80-8.07 (2H,Ar.m.), 7.33-7.60 (3H,Ar.m.), 7.0 (1H,d,J = 2Hz),

6.82 (1H,d,J=2Hz), 4.03 (3H,s,OMe), 3.97 (3H,s,OMe); ir (cm-1): 3070,

2975, 2945, 2220, 1615, 1565, 1480, 1410, 1335, 1245, 1215, 1190,

1160. 1120, 1045, 825, 770, 695. Anal. Calc. for C17H13N302 Y2 H): C,

68.05; H, 4.79, N, 14.01. Found: C, 68.73; H, 4.39; N, 13.83.

3-Cyano-6.8-Dimethoxy-5-Nitro-2-Phenvlauinoxaline (4.4)

To a suspension of 4.3 (0.33 g) in acetic anhydride (4 ml) at 0-5C was

added nitric acid (70%, 0.4 ml) dropwise, and the reaction allowed to proceed

for 30 min. The reaction work up was the same as that described for

compound 2.10 (p. 48). The product was crystallized from acetone. Yield 0.30

g (79%); m.p.: 262-264OC; nmr (CDCI3/ DMSO-de): 7.90-8.04 (2H, Ar.m.),

7.47-7.67 (3H,Ar.m.), 7.31 (1H,Ar.s.), 4.20 (6H,s,OMe); ir (cm-'): 3110,

3080, 3060, 2950, 2850, 2240, 1620, 1530, 1525, 1345, 1230, 1205,

1115, 975, 815, 690. Anal. Calc. for C17H12N404: C, 60.71; N, 16.66.

Found: C, 60.94; N, 15.78.

3-Cyano-6-Methoxy-2-Phenylquinoxaline-5.8-Dione (4.5)

To a solution of 4.3 (0.48 g) in chloroform (5 ml) and acetic anhydride

(4 ml) at 250C was added nitric acid (0.65 ml), and the mixture was stirred for

20 min. When the starting material was consumed, a mixture of 1:1 H2SO4/

HNO3 (0.2 ml) was added. After 10 min, the reaction was stopped by addition

of ice, followed by water. The precipitate was filtered, dissolved in chloroform,

and washed with aqueous sodium bicarbonate. The chloroform was dried and

concentrated to dryness, and the product was crystallized from 9:1 acetone/

ligroin to give 4.5 as light brown crystals. Yield: 0.12 g (25%); m.p.: 250-

253C; nmr (CDCI3/DMSO-de): 7.97-8.20 (2H,Ar.m.), 7.47-7.70 (3H,Ar.m.),

6.60 (1H,Ar.s.), 4.00 (3H,s,OMe); ir (cm1): 3060, 2950, 2920, 2850, 2240,

1710, 1655, 1600, 1535, 1320, 1270, 1230, 1205, 1110, 1070, 860, 695.


Anal. Calc. for CleHN303 H20: C, 62.13; H, 3.59; N, 13.59. Found: C,

61.59; H, 3.33; N, 14.65.

3-Cvano-8-Hvdroxv-6-Methoxy-5-Nitro-2-Phenvlauinoxaline (4.6)

This compound was the major product remaining in the mother liquors

from the above reaction. It was extracted into an aqueous ammonium

hydroxide solution from benzene. After acidification, it was re-extracted into

chloroform, dried, and concentrated. Crystallization from acetone/ ligroin gave

orange crystals. Yield: 0.17 g (32%); m.p.: 220-223C; nmr (CDCI3/ DMSO-

de): 7.80-8.04 (2H,Ar.m.), 7.35-7.60 (3H,Ar.m.), 7.13 (1H,Ar.s.), 4.03

(3H,s,OMe); ir (cm1): 3380, 3100, 2950, 2240, 1635, 1530, 1485, 1450,

1340, 1200, 1170, 1115, 980, 815, 775, 730, 690. Anal. Calc. for

CeHioN404: C, 59.63; H, 3.13; N, 17.39. Found: C, 59.16; H, 3.21; N,


3-Cyano-5.7-Dinitro-8-Hvdroxv-6-Methoxv-2-Phenvlauinoxaline (4.7)

This compound was obtained from the bicarbonate washings in the

above reaction work up by acidification and filtration. The precipitate was

dissolved in chloroform, dried, and concentrated. The solid was crystallized

from ether to give 4.7 as pale yellow crystals. Yield: 0.144 g (24%); m.p.:

141-143oC; nmr (CDCI3): 7.97-8.16 (2H,Ar.m.), 7.43-7.60 (3H,Ar.m.), 5.40-

6.0 (1H,br,OH), 4.10 (3H,s,OMe); ir (cm1): 3440, 2960, 1630, 1540, 1480,

1390, 1340, 1240, 1125, 1060, 970, 810, 770, 750, 685.

Acetate- Anal. Calc. for C18H12N507: C, 52.82; H, 2.71; N, 17.11. Found: C,

52.89; H, 2.74; N, 17.03.

3-Cvano-6.8-Dimethoxv-5.7-Dinitro-2-PhenvlQuinoxaline (4.8)

This compound was also obtained as a by-product in the preparation of

4.5 and of 4.6. The chloroform extract after the ammonium hydroxide

extraction still contained a fast-moving spot and was therefore subjected to

column chromatography (silica/ benzene). Fractions containing this band were

concentrated, and the product was crystallized from ether/ ligroin to give 4.8

as pale yellow crystals. Yield: 0.016 g (3%); m.p.: 122-124oC; nmr (CDCI3):

7.83-8.06 (2H,Ar.m.), 7.43-7.67 (3H,Ar.m.), 4.50 (3H,s,OMe), 4.16

(3H,s,OMe); ir (cm-1): 2960, 2920, 2850, 1615, 1545, 1540, 1360, 1320,

1200, 1135, 1050, 1030, 805, 720.

7-Amino-3-Cvano-6-Methoxv-2-Phenvlauinoxaline-5.8-Dione (4.9)

Sodium azide (0.1 g) was added to a solution of 4.5 (0.14 g) in acetic

acid (4 ml). The mixture was heated in a water bath at 60C for 60 min. Ice

water was added, and the contents were extracted with chloroform. The

extract was washed with phosphate buffer (pH 7), dried, and concentrated to

dryness. The material was chromatographed on silica, with the major product

eluting with 1:1 chloroform/ benzene. Crystallization from acetone gave 4.9.

Yield: 0.06 g (41%); m.p.: 264-266oC; nmr (CDCI3/DMSO-de): 7.70-8.0

(2H,Ar.m.), 7.33-7.53 (3H,Ar.m.), 6.63 (2H,br,NH2), 3.93 (3H,s,OMe); ir

(cm1): 3480, 3315, 2950, 2850, 2230, 1680, 1630, 1570, 1550, 1515,

1440, 1340, 1280, 1210, 1195, 1040, 840, 790, 750, 700, 685. Anal. Calc.

for C16H1oN403: C, 62.74; H, 3.30; N, 18.30. Found: C, 62.72; H, 3.33; N,