RESEARCHES ON THE 
 STRUCTURE OF ANTHRACENE 
 
 By 
 
 ERNEST ATKINS WILDMAN 
 
 B.S. Earlham College, 1912 
 M.S. University of Illinois, 1914 
 
 THESIS 
 
 SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS 
 FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN CHEMISTRY 
 IN THE GRADUATE SCHOOL OF THE UNIVERSITY 
 OF ILLINOIS, 1922 
 
 URBANA, ILLINOIS 
 
\'~3 
 
 UNIVERSITY OF ILLINOIS 
 
 THE GRADUATE SCHOOL 
 
 May 12 
 
 192 - 
 
 I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY 
 SUPERVISION BY ERHB8T ATKINS W ILDMAK 
 
 ENTITLED RESEARCH ES OH T HE S TRUCTURE Qg ANTHRACENE 
 
 BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR 
 
 THE DEGREE OF Doctor of Phil osop hy 
 
 UC A 
 
 / In Cl 
 
 n Charge of Thesis 
 
 uC A AT* 
 
 H^ad ot Depa 
 
 partment 
 
 Recommendation concurred in 
 
 in* 
 
 Committee 
 
 on 
 
 Final Examination^ 
 
 A C'OAQO 
 

 
 
 
CONTENTS. 
 
 I. THEORETICAL DISCUSSION page 1. 
 
 II. DEJECT' OF THIS RESEARCH page 7. 
 
 III. DISCUSSION OF THE EXPERIMENTAL PART page S. 
 
 IV. EXPERIMENTAL PART page 14. 
 
 V. BIBLIOGRAPHY' page 26. 
 
Digitized by the Internet Archive 
 in 2016 
 
 https://archive.org/details/researchesonstruOOwild 
 
X 
 
 I. THEORETICAL DISCUSSION. 
 
 It is the general opinion of chemists that the 
 structure of anthracene is represented either by Formula I. 
 or by Formula II. 
 
 I . II . 
 
 These have been called the meso-linkage or the bridge formula, 
 suggested by Graebe and Liebermann x , and the aromatic or 
 ortho-quinoid formula, suggested by Armstrong", respectively. 
 
 3 
 
 Kurt Meyer considers that it is altogether improbable that there 
 can exist a rearrangement or isomerism between these two forms 
 and therefore that they are not different phases of the same 
 formula. 
 
 Of the several known syntheses of anthracene none of 
 those that involve the intermediate formation of dihydro- 
 anthracene or of anthraquinone are of any value in deciding 
 between the above two formulas. There is no evidence of the 
 disposition made of the valencies of the meso carbon atoms during j 
 either oxidation of dihydroanthracene to anthracene or reduction 
 of anthraquinone to anthracene. The only synthesis that is free 
 from this objection is that of Anschutz and Elsbacher by means 
 of the Friedel and Crafts reaction using benzene and tetrabromo- 
 
 ethane ^ 
 
 /3a -c- <J lx. 
 
 -f I -+- 
 
 rw c-ox, 
 
 i 
 
 H 
 
2 
 
 While this synthesis seems to be an indication in favor of the 
 meso- linkage formula it is nevertheless discredited by some 
 authorities, as Meyer and Jacobson's Lehrbuch^ on the basis that 
 aluminium chloride is a reagent that may cause a rupture of 
 linkage between carbon atoms as, for example, in the reversal of 
 the Friedel and Crafts reaction. 
 
 Assuming the meso- linkage formula and the Kekule 
 formula for the benzene ring there are three phases depending on 
 the positions of the double bonds: 
 
 In other words the two outside rings in such a formula are to be 
 considered as typical benzene rings while the meso carbon atoms 
 are distinctly alifatic in nature. A more proper formula to 
 emphasize this fact is 
 
 true we should expect these two benzene rings to have 
 properties of benzene. While they do have such to a 
 extent there is at least one very striking exception. 
 I , 4-anthraquinone 
 
 is a well known substance and 1 ,2-anthraquinone 
 
 If this is 
 the usual 
 large 
 Whereas 
 
3 
 
 is easily prepared by gentle oxidation of I ,2-dihydroxyanthracene , 
 
 r* 
 
 Lagodzinski 0 found that 2 , 3-anthraquinone , the existence of which 
 is predicted by the meso-linkage formula, is not produced by the 
 oxidation of 2 , 3-dihydroxyanthracene . The fact is interpreted as 
 meaning that there is only a single bond between positions 2 and 3 
 in anthracene and that this single bond is fixed in that position. 
 There is, consequently, no rearrangement or shifting from one 
 phase to another of the double and single unions. This fact cannot 
 be reconciled with the mesoT linkage formula but it is entirely 
 compatible with the ortho-quinoid formula in which there is such a 
 fixation of double and single bonds. It is true a 2,3-quinone may 
 be formulated as follows 
 
 but if this is formed from 2 , 3-dihydroxyanthracene there must be 
 a very extensive rearrangement of valence. Lagodzinski found that 
 mild oxidizing agents did not affect 2 , 3-dihydroxyanthracene and 
 that more energetic agents completely disrupted the molecule. 
 
 7 
 
 Hinsberg has found that the stability of azines with 
 increasing numbers of linearly conjugated rings grows less with 
 such rapidity that the five ring azine is not capable of existence. 
 The synthesis of phenazine from ortho-phenylene diamine and ortho- 
 benzoquinone indicates that without doubt the nitrogen atoms are 
 linked in the ortho-quinoid manner. This is further indicated by 
 the intense color of such compounds as naphtho-phenazine which is 
 probably caused by the accumulation of quinoid linkages as shown 
 

 
 
 
 
 
 
 
 
 
 
 
 
i 
 
 On the other hand the stability of the dihydrogenated derivatives 
 of the azines increases with the number of rings. Both these facts 
 are to be expected from Baeyer's strain theory. The addition of 
 a ring in the azines is in fact the attachment of a four atom 
 conjugate system while it is a true benzene ring in the dihydro- 
 genated compounds. Their stability may be compared to that of 
 dihydrobenzene and benzene. Hinsberg believes that an analogous 
 reasoning also holds for anthracene and related compounds. 
 
 K. H. Meyer°f avors the ortho- quinoid formula since it 
 seems more in accord with the keto-enol tautomerism which he has 
 demonstrated exists between anthranol and anthrone . Furthermore 
 anthranol is distinctly phenolic in character, a fact which is 
 represented by the ortho-quinoid formula whereas the meso-iinkage 
 formula indicates the hydroxyl group as attached with the typical 
 union of the. tertiary alifatic alcohol. 
 
 Scholl^ has obtained several reduction products of 
 flavanthrene in which either one or both anthraquinone nuclei are 
 reduced. He found that the ortho-quinoid formula for anthracene 
 must be assumed in order to formulate the reduction products in 
 accord with their properties. Dihydrof lavanthrene (VI) forms an 
 addition product with water to produce the mono-hydrate (VII) 
 which is acidic and forms a sodium salt. If the meso-iinkage 
 
 VI 
 
 VII 
 
5 
 
 formula is used to represent the mono-hydrate the compound 
 formulated should not have acidic properties. The evidence in 
 this case is of the same nature as that referred to above in 
 connection with the acidic properties of anthranol. 
 
 Schlencfcr and his associates have produced quite 
 strong evidence of the ortho-quinoid formula in showing that 
 anthracene adds metallic sodium very readily in a manner analogous 
 to the ethylenic hydrocarbons. Schlenck takes the position that 
 the meso-linkage formula expresses a condition of valence 
 saturation for the meso carbon atoms --presumably analogous to the 
 manner of linkage in cyclobutane- -and that therefore the fact of 
 the addition of sodium proves the existence of the conjugated 
 system of two double unions. 
 
 Recently von Weinberg has argued from the basis of 
 heat of combustion determinations that anthracene is completely 
 conjugated as is rather generally accepted to be the case in 
 benzene. Such a conclusion is in harmony with the ortho-quinoid 
 formula rather than the meso-linkage formula. The results of 
 another physical method, that is, refractive index determination, 
 point in the same direction. Auwers "*" 6 has shown that meso- 
 isoamylanthracene has refractive and dispersive powers much 
 
 exalted over those for dihydr.o-meso- isoamylanthracene . 
 
 13 
 
 Finally Colver with Noyes have synthesized anthra- 
 cene froin naphthalene by means of a series of reactions in which 
 a third ring was built up on a dihydronaphthalene nucleus. The 
 work is in harmony with the ortho-quinoid formula although it is 
 not claimed to throw any positive light on the problem. 
 
 The ortho-quinoid formula was early recognized by 
 

 
G 
 
 Armstrong as unsymmetrical . While there is sufficient evidence 
 from several syntheses that dihydroanthracene and anthraquinone 
 are symmetrical there is no proof that anthracene is. It is true 
 that anthraquinone mono-sulfonic acids and mono-amines when reduced 
 to the corresponding anthracene derivatives might he expected to 
 yield two isomers if an unsymmetrical structure is generated, yet 
 such have never been observed. This is net a weighty objection, 
 however, for the chemistry of the substitution products of 
 anthracene is still to a large extent undeveloped due to the much 
 greater technical importance of anthraquinone derivatives. 
 

 
 
 
 
 
 . 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
7 
 
 II. OBJECT OF THIS RESEARCH 
 
 This research is "based upon the supposition that if 
 the ortho-quinoid structure is correct it should be possible to 
 oxidize a proper anthracene derivative and obtain as the result 
 a derivative of naphthalene in much the same way as Bamberger and 
 
 i A 
 
 Praetorius have oxidized anthragallol and obtained hydroxy- 
 alpha-naphthoquinone acetic acid. It is the opinion of the/ writer 
 that if a naphthalene derivative were obtained in this way from 
 an anthracene derivative the result would be quite conclusive 
 evidence of the correctness of the ortho-quinoid formula. 
 
 In view of the relative ease with which the meso 
 positions of anthracene are attacked it is necessary to render one 
 of the outside rings more susceptible to oxidation. This it has 
 been hoped to accomplish by preparing I , 2 , 3-trihydroxyanthracene , 
 in which the ring containing the three hydroxyl groups should be 
 readily oxidized with the probable formation of haphthalene- 
 2 , 3-d.icarboxylic acid. 
 
 There is also included a description of three attempts 
 made several years ago to synthesize anthracene, one of which 
 aimed at the meso-lihkage structure and the other two at the orthc- 
 quinoid structure. Since none of these were successful the main 
 part of the work has been that outlined above. 
 
6 
 
 III. DISCUSSION OF THE EXPERIMENTAL PART. 
 
 Attempt to synthesize the meso-Iinkage structure. 
 
 Whereas Anschutz and Elsbacher* used a method which has been 
 discredited in the opinion of some it seemed that this objection 
 would be removed if instead of benzene , ortho-dibromobenzene were 
 condensed with tetrabromoethane by means of Fittig's reaction. 
 
 The reaction was expected to take place at least in part as 
 indicated in the following equation: 
 
 The ortho-dibromobenzene used was prepared according to the 
 
 following series of transformations: 
 
 from products of reaction. 
 
 Attempt to synthesize the ortho-quinoid structure 
 
 from alpha-naphthol and succinic anhydride. 
 
 7« T itt and Braun have prepared l-hydroxy-2-acetonaphthalene 
 by the action of acetic anhydride on alpha-naphthol in the presence 
 of zinc chloride. If the same sort of condensation is of general 
 application as one might suppose since it has also been used by 
 

 
 
 
 
 
 
 
 
 
 
 
 
9 
 
 Hantzsch' 1 ' for the preparation of i-hydroxy-2-propionaphthalene , 
 it is reasonable to expect that succinic anhydride might also 
 be used and that it would form a product as shown in the 
 equations that would condense further to i , 4 ,9- trihydroxy - 
 anthracene. This result was not realized for no pure substance 
 could be isolated from the tar that was formed. 
 
 C-H do 
 
 l y. 
 
 
 Attempt to build up the third ring with ethane 
 tetracarboxylic ester. 
 
 The procedure which was outlined for this synthesis is 
 indicated in the following series of transformations and equations 
 
 roV3E3SKZZKE2Ur3 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 o 
 
 rV> _ h 
 
 10 
 
 1 
 
 o 
 
 H 
 
 ^1 
 
 
 o H 
 
 u 
 
 
 II 
 
 /\/ 
 
 
 — > 
 
 , H 
 H 
 
 kj 
 
 
 
 
 ft 
 
 0 
 
 The scheme was 
 
 carried 
 
 o 
 
 out in pra 
 
 ct ice 
 
 to the 
 
 o\\ 
 
 point 
 
 of attempting to condense the alpha-methoxy-beta-naphthoyl 
 chloride with the mono-sodium salt of ethane te tracer boxy lie 
 ester in ether solution. There //as no reaction. The work v/as 
 then discontinued and when several years later the original 
 samples were examined with greater care it was found that they 
 were quite impure on account of the fact that the hydroxy- 
 naphthoic acid had been incompletely methylated. The methoxy- 
 acid was present in the mixture, however, and was isolated in 
 pure form by careful fractional crystallization. The synthesis 
 ha§ not been carried farther in view of the decision to attempt 
 the degradation of an anthracene compound to a naphthalene 
 compound . 
 
 Degradation of anthracene to naphthalene. 
 
 An examination of the literature indicated that the best 
 method of preparing I ,2 ,3-trihydroxyanthracene is probably by 
 demethylation of its tri-methyl ether which should be produced 
 by reduction of anthragallol trimethyl ether. There are two 
 methods on record for the synthetic production of the latter 
 substance. 0n§ , described in D. R. F. 15627b consists in heating 
 1 , 3-dinitro-2-methoxyanthra^uinone with sodium methylate in 
 methyl alcohol solution for several hours. Some preliminary 
 
11 
 
 trials that were made with this method showed it to he unsatis- 
 factory. It seems that the 1 , 3-dinitro-derivative , if formed 
 at all on nitrating 2 -methoxyanthraquinone , is at "best only a 
 
 part of several products. A more feasible method seemed to be 
 
 17 
 
 that of Bock . The latter heated anthragallol with sodium 
 carbonate in nitrobenzene until the sodium salt of the anthra- 
 gallol was formed. Methyl sulfate was then added and the heating 
 continued. On cooling the filtered solution anthragallol di- 
 methyl ether separated in crystalline form. This was converted 
 into its sodium salt and the latter heated with methyl sulfate 
 at 180° in order to complete the methylation of the hydroxyl 
 group in position 1. This method was tried by the writer but 
 given up on account of the difficulty of handling the hot 
 nitrobenzene and methyl sulfate solutions and also because Bock 
 was not able to obtain a pure product in this way. 
 
 In casting about for a satisfactory method an attempt 
 was next made to condense 2,3,4- or 3,4,5-trimethoxybenzoyl- 
 benzoic acid^, obtained by Friedel and Crafts reaction from 
 phthalic anhydride and pyrogallcl trimethyl ether, directly to 
 anthragallol trimethyl ether. The above acid was prepared in 
 rather poor yield and on this account ortho-anisoylbenzoic acid 
 was prepared for use in determining the best conditions for 
 carrying on the further condensation. 
 
 The best method previously reported for the prepara- 
 
 19 
 
 tion of ortho-anisoylbenzoic acid is by Meyer and Turnan , who 
 obtained a 35,T yield( calculated on the aniscl used) from 0.675 
 mols of phthalic anhydride, 0.79 mols of aniscl and 0.71 of a 
 
 formula weight of Aid 3 res.cting in 15o cc . of nitrobenzene. 
 
' 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
The latter, they found, is a much better solvent for phthalic 
 anhydride than carbon disulfide and consequently assists the 
 reaction considerably. However Meyer and Turnan used approximately 
 only one mol of aluminium chloride whereas Rubridge and qua~^ 
 have shown that in reactions involving the use of phthalic 
 anhydride two formula weights of A1C1._- should be used for each mol 
 of the former, that is, one for each carbonyl group. When the 
 reaction is performed in this manner the writer has obtained a 
 66/J yield of ortho-anisoylbenzoic acid. 
 
 p. 
 
 Lagodzinski was able to condense 3 ,4-dimethoxy , 
 benzoylbenzoic acid to hystazarin dimethyl ether in nearly 
 quantitative yield by short heating in solution in concentrated 
 sulfuric acid on the water bath. When this method was used on 
 ortho-anisoylbenzoic acid almost complete saponification of the 
 methoxy group occurred, no doubt due to the fact that the con- 
 densing ring hydrogen atom is not in a reactive x osition, being 
 rr.eta to the methoxy group. 
 
 The same result was obtained when 2,3,4- or 3,4,5- 
 trimethoxybenzoylbenzoic acid was heated on the water bath in- 
 concentrated sulfuric acid solution. The material was very 
 readily condensed to an anthraquinone compound but even more 
 easily saponified so that the product was a partially methylated 
 anthragallcl . 
 
 Finally there was devised a method of methylating 
 anthragallcl with methyl sulfate, which, while extravagant in the 
 amount of methyl sulfate required was convenient in manipulation 
 and gave a product of good quality. 
 
 Reduction of anthragallol trimethyl ether to 1,2,3- 
 

 
 
 
 
 . 
 
 ■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 j 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
13 
 
 trimethoxyanthracene was first tried by the method used by 
 Lagodzinski^ in reducing hystazarin dimethyl ether to 2,3- 
 dimethoxy anthracene , but with unsuccessful results, his method 
 involved preparing the material in a very fine state of division 
 and then treatment with zinc dust and ammonia with continual 
 shaking and heating on the water bath for several hours. No 
 satisfactory reduction could be obtained in this way even when 
 the process was continued six or eight hours and the material 
 kept in continual agitation with a mechanical stirring device. 
 After many trials it was found that the anthragallol trimethyl 
 ether was suitable for use in the form in which it crystallized 
 from solution and that the reaction proceeded satisfactorily 
 in about thirty or forty minutes time only. 
 

 
 
 
 
 
 
 
 
 
 
 
 
14 
 
 IV. EXPERIMENTAL PART. 
 
 para-Nitrobromobenzene . 
 
 A mixture of 51 cc. of fuming nitric acid (1.5) and 
 108 cc. of concentrated sulfuric acid was added through a dropping 
 funnel slowly to j-oo grams of bromobenzene kept at 50° on the 
 water bath. After the nitrating mixture was all added the temper- 
 ature was gradually raised to 120° with frequent shaking. The 
 material was then poured into water and filtered, washed free from 
 acid and then dissolved in 600 cc. of boiling 95 $ alcohol and 
 allowed to cool. The crystalline para-nitrobromobenzene was 
 filtered off and a second crop obtained on concentrating the 
 solution. Total yield, 03 grams. M. P. 125°. The alcohol 
 solution on evaporation left 70 grams of impure ortho-nitrobrcmo- 
 benzene . 
 
 O *1 
 
 1 ,2-Dibromo-4-nitrobenzene . ‘-" t 
 
 20 grams of para-nitrobromobenzene, 5 grams of anhy- 
 drous ferric chloride and 6.4 cc . of dry bromine were sealed in a 
 tube and heated at 98° for 50 hours. The dark brown liquid was 
 thrown into water and washed free from ferric chloride and acid. 
 
 It was purified by distilling under reduced pressure. At 22 mm.. 
 
 52 grams were collected between 175 c and 185°. This solidified 
 in the receiver. It was crystallized from alcohol. M.P. 57 u . 
 
 o o 
 
 3 ,4-Dibromoaniline . 
 
 50 grams of 1 ,2-dibromo-4-nitrobenzene were mixed ’with 
 
15 
 
 one liter of water and 30 grams of iron dust and kept at 40 u while 
 grams of concentrated sulfuric acid were added a few drops at a. 
 time over a period of three hours. The solution was made alkaline 
 and the base was extracted with ether which on evaporation left a 
 brown oil that was not purified. Yield, 40 grams. 
 
 ortho-Dibromobenzene. 
 
 105 grams of the hydrochloride of the base obtained 
 above were mixed with a solution of 12C cc. of hydrochloric acid r* 
 (1.2) in 95 cc . of water. The mixture was cooled in an ice and 
 water bath and a saturated solution of sodium nitrite added slowly 
 until the solution contained an excess of free nitrous acid. After 
 about forty-five minutes the diazcnium salt suddenly crystallized 
 cut. The mass was poured slowly into 1500 cc. of boiling 95 /j 
 alcohol. A rapid evolution of nitrogen followed. Since ortho- 
 dibromobenzene distils readily with alcohol vapor it had to be 
 separated by dilution with water. The precipitated oil was dis- 
 tilled at 53 mm. and 50 grams collected. E ,P. 130-132°. 
 
 Attempt to condense ortho-dibromobenzene 
 with tetrabromoethane . 
 
 The tetrabromoethane used was dried, over calcium 
 chloride and distilled under reduced pressure several times. It was 
 obtained as a colorless oil 'which boiled at i.C2° at 2 mm. .and at 
 65° at i mm. 
 
 10 grams of ortho-dibromobenzene and 14.7 grams of 
 tetrabromoethane were dissolved in 100 cc . of anhydrous ether and 
 
16 
 
 4 grams of clean sliced sodium added. The mixture was warmed slight- 
 ly and showed a fairly vigorous reaction. A considerable amount of 
 mono-hromoacetylene was evolved. After twenty-four hours the ether 
 was distilled off and left a red oil which still contained enough 
 bromoacetylene so that it ignited in the air. The red liquid was 
 distilled at ordinary pressure. 8 grams of ortho-dibromobenzene 
 were recovered and only a trace of residue left. 
 
 The experiment was repeated with the modification that 
 the tetrabromoethane was dissolved in a small amount of anhydrous 
 ether and dropped slowly into the other materials during 48 hours, 
 meanwhile the mixture was kept gently boiling. After removal of 
 the ether the oil was distilled at 45 mm. pressure. 16 grams of 
 the two oily substances came over between 110 u nand 130 . About 
 0.1 gram of residue was left in the flask. This was dissolved in 
 benzene and filtered with a small amount of charcoal. A color test 
 with picric acid gave a negative result. 
 
 The alifatic compounds in the distillate were destroyed 
 with sodium ethylate and the ortho-dibromobenzene recovered in pure 
 form. 
 
 A similar experiment w as performed with the substitution 
 of "Natur Cupfer 0" for sodium as the condensing agent . No reaction 
 occurred when the materials were heated at 16C° for two hours. 
 
 Reaction of succinic anhydride 
 and alpha-naphthol . 
 
 14 grams of alpha-naphthol, 1C grams of succinic 
 
17 
 
 anhydride and 7 grams of freshly fused zinc chloride were heated 
 together in an oil bath at 140° for fifteen minutes. The material 
 turned into a dark red rosin- like mass from which no single 
 crystalline compound could be obtained. 
 
 alpha-IIydroxy-beta-Naphthoic. acid . 
 
 Sodium naphthyl oxide was made by dissolving one mol 
 of sodium in 200 cc. of absolute methyl alcohol and adding one mol 
 of alpha-naphthol in solution in ICC cc . of absolute methyl alco- 
 hol, and then evaporating to dryness finally under reduced pressure. 
 The solid was quickly powdered and placed in a calorimeter bomb 
 modified for the purpose and sealed with a lead gasket. The bomb 
 was attached to a carbon dioxide tank containing liquid carbon 
 dioxide and the carbon dioxide allowed to enter it slowly until the 
 pressure in the bomb was the same as in the tank. Considerable 
 heat was evolved making it necessary to cool the bomb in cold water 
 during the absorption, otherwise the sodium naphthyl carbonate 
 may rearrange and cut down the yield on account of the following 
 
 reaction : 
 
 o "*■<«- 
 
 
 ,, o 
 
 C'' o >v<x_ 
 
 The temperature of the bomb was not allowed to rise above 65 " . The 
 contents were shaken during the absorption which was complete in a 
 about 15 minutes. 
 
 The bomb was then heated in an oil bath at 135° for 
 
 four hours in order to bring about the rearrangement of sodium 
 

 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I litJ V i’l 
 
 - 
 
 
 
 
 
 
 
 
 
16 
 
 naphthyl carbonate to sodium alpha-hydroxy-beta-naphthoate . The 
 dry salt obtained was nearly white. It was washed with a little 
 ether to remove traces of naphthcl, then dissolved in water and 
 acidified with sulfuric acid. The free acid which was obtained 
 was nearly white and melted at 20C-2C2 0 . Yield, 170 grams, 90 % 
 of the theory. A sample was purified by charcoal treatment in 
 alcohol solution and then recrystallized from the alcohol. It 
 melted at 206-9°, and remained constant at that point after 
 successive crystallizations from alcohol, benzene and ligroine . 
 Various melting points have been reported in the literature, the 
 
 o r z 
 
 highest being by Anschutz’ 00 as 191-2°. 
 
 al pha-Me thoxy -beta- naph tho i c acid. 
 
 60 grams of alpha-hydroxy-beta-naphthoio acid and 13.8 
 grams of sodium were dissolved in 200 cc. of methyl alcohol and 
 boiled under a reflux condenser. 76 grains of methyl sulfate were 
 added in small portions through the condenser. The product was 
 thrown out by adding water and extracted with cold dilute sodium 
 hydroxide. The insoluble material, was then distilled and the 
 fraction boiling from 188° to 195° at 18 mm. was collected and 
 crystallized from dilute alcohol. Later work showed that this 
 sample was not pure, methyl ester of methoxynaphthoic acid. 
 
 It was readily hydrolyzed by boiling with dilute 
 sodium hydroxide solution. The acid which was precipitated when 
 the alkaline solution was acidified was recrystallized several 
 times from benzene and then from alcohol. A sample was obtained 
 which melted at 123- 4° and was shown by analysis to be methcxy- 
 
19 
 
 naphthoic aoid. 
 
 v . 1432 grams of substance gave 0.3756 grams of carbon dioxide 
 and 0.0663 grams cf water. 
 
 Calculated for Cb.-H^O-^, C=71.25$, H=4.98)£. Found, C=7i.53/C, 
 
 H=5 .18$. 
 
 In view of the fact that impure compounds were used 
 an account of the further work along this line is of no value. 
 
 orthc-Anisoylbenzoic acid. 
 
 29 grams (0.2 mol) of phthalic anhydride, 22 grams 
 (0,2 mol) of anisol and 50 cc. of nitrobenzene were mixed and war m3 
 until solution was complete. Then 53 grams (0.4 mol as AlClg) of 
 aluminium chloride were added at one time. The reaction took place 
 rapidly with violent evolution of hydrogen chloride and appeared to 
 be complete in about five minutes. It 'was heated on the steam bath 
 however, for about twenty minutes. Water and hydrochloric acid 
 were then added and the mixture steam distilled until all the 
 nitrobenzene and unchanged anisol were removed. The oily residue 
 was extracted with dilute ammonia in which nearly all dissolved. 
 
 The ammoniacal solution was treated with dilute hydrochloric acid 
 until it was nearly neutralized and a small quantity of black tar 
 which came out was filtered off. The filtrate which was nearly 
 colorless was entirely acidified with hydrochloric acid which 
 threw down an oil that quickly solidified. The material was 
 recrystallized from alcohol and obtained as a white substance, 
 
 M.P. 148°. The yield was 34 grams or 66$ cf the theory. 
 

 
 
 . 
 
 
 
 
 
 
 . 
 
 . 
 
 ' 
 
 
 
 
 
 
 
 
 
 
20 
 
 Attempt to condense ortho-aniscylbenzoic acid to 
 3-methoxyanthraquinone . 
 
 20 grams of ortho-anisoylbenzoylbenzoic acid were 
 dissolved in 100 cc . of concentrated sulfuric acid and heated on 
 the water bath for two hours. The solution was poured onto cracked 
 ice and the precipitated material washed and warmed with dilute 
 ammonia. Al}. but a trace of the substance dissolved. The insol- 
 uble part was crystallized from hot alcohol and gave less than 0.1 
 gram of yellow needles -which were 3-methoxyanthraquinone. The 
 ammoniacal solution was acidified and the precipitated material 
 crystallized from 50 % alcohol. There was obtained in this way 10 
 grams of slightly colored material which melted at 143 c and was the 
 unchanged original substance. The 3 7 hydroxyanthraquinone in the 
 mother liquor was not isolated. 
 
 2,3,4- or 3 , 4 , 5- trimethoxybenzoy lbenzoic acid. 
 
 The pyrogallol trimethyl ether used in this experiment 
 was prepared by treating pyrogallol in sodium hydroxide solution 
 with methyl sulfate according to the method of Bargellini and 
 
 24 pc 
 
 Leda^ and of UHmann" 0 . From 126 grams of pyrogallol there was 
 
 o 
 
 obtained 00 grams of the trimethyl ether boiling between 235 and 
 237°. 
 
 15 grams (0.1 mol) of phthalic anhydride, 17 grams 
 ( 0 . 1 mol) of pyrogallol trimethyl ether and 35 cc. of nitrobenzene 
 were mixed and warmed until the anhydride was dissolved. Then 27 
 grams (0.2 mol as A1C1-) of aluminium chloride were added at one 
 time. The reaction proceeded rapidly for about five minutes end 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
21 
 
 was then completed by twenty minutes warming on the water bath. 
 
 After adding water and hydrochloric acid and steaming out the 
 nitrobenzene the oily residue was extracted with warm dilute 
 ammonia. The dark colored ammonia solution was neutralized with 
 
 I 
 
 dilute hydrochloric acid, a few drops in excess added and the 
 solution boiled with charcoal and filtered. The material was then 
 completely precipitated with hydrochloric acid and a solid obtained. 
 This was collected and crystallized from dilute alcohol. Neaply 
 white crystals. Yield, 3 grams or 9.5 %. M.P. 169°. 
 
 Attempt to condense 2,3, 4- or 3 ,4 , 5-trimethoxybenzoyl- 
 benzoic acid to anthragallol trimethyl ether. 
 
 5 grams of 2,3,4- or 3 ,4^5-trimethoxybenzoylbenzoic 
 acid were dissolved in 25 cc. of concentrated sulfuric acid and 
 heated on the water bath for ten minutes with constant shaking and 
 then poured onto cracked ice. The precipitated material was en- 
 tirely soluble in alkali. On fractional crystallization from 
 glacial acetic acid it yielded two indefinitely melting fractions, 
 one at about 156° and the other above 3i0°. 
 
 Preparation of anthragallol. 
 
 4CC grams of gallic acid, 800 grams of benzoic acid and 
 4350 cc. of concentrated sulfuric acid were placed in a ten liter 
 flask and gradually heated to 140° during three and one half hours 
 time and then maintained at that temperature two hours longer. The 
 solution was then cooled, poured into ice water, filtered and 
 washed with water. The precipitated material was composed of 
 
22 
 
 anthragallol, ruffigallic acid, a little black oxidized substance 
 and a considerable amount of benzoic acid. It was made into a paste 
 with water and treated with concentrated ammonia sufficient to 
 convert the benzoic acid into its salt. The formation of the blue- 
 black salt of anthragallol serves as an indicator for the end-point 
 of the reaction. The substance was then filtered and well washed 
 with boiling water. Portions of the dark brown filter cake were then 
 extracted with boiling alcohol, which dissolves the anthragallol and 
 not the ruffigallic acid, and the alcohol solution allowed to cool. 
 Red-brown to chocolate-brown crystals were obtained according to 
 the rate at which the solution cooled. Occasionally the color was 
 a light yellow-red. The yield was about 40 % of theory. M.P .320°oorr. 
 
 The inethylation of anthragallol. 
 
 10 grams of anthragallol were covered with ICO cc. of 
 methyl alcohol in a flask connected with a reflux condenser and also 
 a dropping funnel. The methyl alcohol was kept continually boiling 
 while 50/b potassium hydroxide solution in water was added in small 
 portions through the condenser alternating with small portions of 
 methyl sulfate through the dropping funnel. 130 cc. of methyl 
 sulfate were added, a methyl sulfate portion being added last so 
 that the material was left in the form of a yellow partially 
 methylated anthragallol which mostly crystallized out of the hot 
 aolution. Two volumes of water were added and the mixture cooled 
 and filtered. The yellow crystalline material 'was covered as before 
 with IOC cc. of methyl alcohol and given the same treatment using, 
 however, only 50 cc. of additional methyl sulfate. The product from 
 

23 
 
 this treatment was isolated as above and dissolved in 200 co. of 
 boiling benzene. The benzene solution was poured into 200 cc. of 
 aqueous solution of sodium hydroxide which had been warmed to 
 60° and the two shaken together. The partially methylated 
 material formed an insoluble red crystalline sodium salt. The 
 mixed solvents were then filtered from, the sodium salt, the latter 
 washed with hot benzene, and the benzene solution separated from 
 the aqueous layer. The benzene was then distilled from the solution 
 of the trimethyl ether and the latter transferred to a small beaker 
 by dissolving in boiling glacial acetic acid (10-15 cc.). 100 cc. 
 
 of alcohol were then added to this hot solution and the mixture 
 stirred. Within a moment the solution became thick and pasty with 
 fine needle-like yellow crystals of anthragallol trimethyl ether, 
 which were filtered off and dried. Yield, 5.6 grams, 50^ of the 
 theory. M.P. 173°. The dimethyl ether obtained as sodium salt was 
 then further treated as above or as follows. 
 
 Silver oxide and methyl iodide method. 
 
 5.5 grams of the dimethyl ether, 8 grams of silver oxide 
 and 20 cc. of methyl iodide were boiled together for twenty hours. 
 
 On separating the di- and tri-ethers in the same way as above there 
 was obtained 2.6 grams of the trimethyl ether. 
 
 Anthragallol itself cannot be satisfactorily methylated 
 from the beginning with silver oxide and methyl iodide since there 
 is too much oxidation of the anthragallol. 
 
■mxxrrxr rm.TJ'ixrrzz 
 
 24 
 
 Reduction of anthragallol trimethyl ether to 
 1 ,2 , 3- trimethoxy anthracene . 
 
 5 grams of anthragallol trimethyl ether were suspended 
 in a mixture of ICO cc. of concentrated ammonia water and 200 cc . 
 of water. 25 grams of zinc dust were added and the mixture heated 
 on the water hath with continual shaking and rotating for forty 
 minutes. Within five minutes a deep red color developed in the 
 solution. This gradually disappeared until at the end the clear 
 solution was straw color. The mixture was then cooled and 
 filtered and the zinc residue extracted with about 100 cc. of 
 boiling alcohol. The alcohol solution was evaporated to a small 
 volume, about 25 cc. of glacia}. acetic acid added, heated to 
 boiling, and water added drop by drop while scratching the side 
 of the beaker and allowing it to cool. A crop of nearly colorless 
 prisms amounting to 1.6 grams was obtained. On recrystallizing 
 from acetic acid and a little water 1.4 grams were obtained which 
 consisted of colorless crystals having a blue-green fluorescent 
 cast. The melting point remained constant at 01-2° through 
 several crystallizations. In alcohol solution the material was 
 strongly fluorescent. 
 
 O.j. 306 grams gave 0.3650 grams of CO.? and 0.C734 grams of H?0. 
 0.1291 grams gave C.36C1 grams of CO? and 0.0757 grams of H 2 O. 
 Calculated for C i7 H i6 0 3 , C=76.G9£, H=6.02^. 
 
 Found, C=76.iG/£ and 76.07^. H=6.26^o and 6.56/S. 
 
 Since all the intermediate reduction products of 
 
 anthraquinone will react in one way or another with the Grignard 
 

 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
25 
 
 reagent the material was tested in this way. On adding the 
 substance to a solution of methyl magnesium iodide there was no 
 evidence of a reaction. The mixture was boiled a time and 
 decomposed as usual . Much methane was evolved and the x jr> ecipitated 
 substance was found to be the original material. 
 
 1,2 , 3 7 trihydroxy anthracene . 
 
 0.87 gram of I 3- trimethoxy anthracene was warmed 
 gently with ±.0 cc. of 57 % hydr iodic acid for thirty minutes during 
 the time that methyl iodide was distilled out. The solution was de- 
 colorized with sodium bisulfite and filtered. A solid black 
 material was collected. It was soluble in alkali with a dark brown 
 color. A portion was crystallized from benzene but oxidation in 
 solution and in drying turned the substance dark brown. 
 
 The investigation of this substance will be continued 
 in the future. 
 

 
 
 * 
 
 . 
 
 
 
 . 
 
 . . 
 
 
 
27 
 
 VITA. 
 
 The writer was born in Chattanooga, Tennessee, July 
 2, 1889. His early education was received in the public schools 
 of Clarke and Green Counties in Ohio. In 1908 he entered Earlham 
 College, Richmond, Indiana, and was graduated from that institution 
 in 1912. The same year he entered the Graduate School of the 
 University of Illinois and continued there for three years as a 
 student and assistant in the Chemistry Department. He received 
 the Master of Science degree in 1914. From 1915 to 1917 he was 
 an assistant and a fellow on the staff of the Rockefeller 
 Institute for Medical Research. From 1917 to 1919 he was research 
 chemist and director of chemical research work for Eli Lilly and 
 Company of Indianapolis, Indiana. From 1919 until the present 
 time he has been director of the Chemistry Department at 
 Earlham College. 
 
 ACKNOWLEDGEMENT . 
 
 The writer wishes to express .to Professor W. A. Noyes 
 
 his sincere appreciation for the suggestions and encouragement 
 which have inspired him in the carrying on of this research.