76 HOOKER AXD CARNELL : THE COXDENSATIOK XII.-Contwviou of Ortho- iiiti, Pnru-, c w c l of Purci- deusa t i o n of A ldeh y dcs 10 it IL p- Hyt7 I 'oxy- a- imph t Ii CI - pumone. ilzto Ortho-quinollc' Dc~*it~~iti~cs." I. The? COSZ- By SAMUEL C. HOOKEI; and WILLIAM C. CARXELL. Is the course of researches in the lapachol group, already partially communicated by one of us to the Society, nunierous compounds of the general formula 0 0 have been isolated, from which dehFdratiiig agents, preEerably dihtte or concentrated mineral acids, rery readilj- remore wntey or hydrogen * Compare Tram., 1893, 63, foot note p. 431; also Proc., 1893, 13.OF ALDEHPDES WITH B-HYDROXT-a-NAPHTHAQUIXONE. 7 7 chloride or bromide, quantitativeIy converting them, without exception, into internal anhydrides.It was at first supposed tbat the change was a simple one, involving nothing further than the elimination of water; but it was subse- quently found that several of the lapachols, namely, chlorhydrolsp- achol, hydroxyhydrolapachol, and bromhydroxyhydrolapachol, each ga-re two isomeric anhydrides, which proved to be related as follows. 0 Corresponding or a-aniijclride. Pseudo- or 8-anlqdride. In spite of the very varied conditions under which they have been exposed in contact with mineral acids, the remaining lapachols have given in each case only one anhydride, and these have all proved to be P-nzpht,haquinone derivatires. It can, therefore, be generally stated, at least so far as the com- pounds at present known in the lapachol group are concerned, that substances of the general formula I are more readily converted by the action of mineral acids into the p- than into the a-anhydrides, a change which involves the conversion of the para- into the ortho- quinone group.In order to ascertain how far this formation of the &anhydrides would prove general, we have investigated the beha viour of compounds prepared synthetically from P-hydroxy-2-napb tha- quinone. Zincke and Thelen (Ber., 21,2203) first obaeryed that benzaldehydc and /?-h y drox y -a-nap ht haq uinone interact readily, and correctly attributed to the resulting compound the formula- 0 I CbH5 Schoch", one of Zincke's pupils, snbsequen tIy extended the study of this reaction to other aldehydes, and prepared a number of com- pounds having the general formula * " Ueber die Einwirkung Ton Aldehyden und Ketonen auf Oxynaphtochinon," We are indebted to Professor Zincke for Inaugural Disscrtat'm, Marburg, 1858.a copy of this pamphlet.7s HOOKER AXD CARNELL : THE COXDENSATION 0 0 I x In preparing some of these compounds, small quantities of an- hFdridca *are obtained as secondarj- products, t o which the general formula 0 0 0 A/\/\\/V\ I I I I I I v \o/\L&\d'\/ I s was assigned. These anhydrides xwe also in some cases directly ob- tained from the corresponding hycli*oxyl compounds by the action of dehydrating agents ; b u t in every instance Schoch records that the yield as veq- small. The behaviour of the lapachols in forming, as a rule, /%anhydrides in preference to a-anhydrides, rendered questionable the structure attributed by Zincke and his pupils to the anhFdrides obtained by them ; and our experiiiients have conclusively prox-ed that these compounds, or at least all that we haTe examined, are, in realitF, as we anticipated, deriT-atiTes of /%naphthaquinone, to which one or other of the following formula: must be attributed. I * s s c&Aiihydridc.~ ~ - A n h ~ - d r i c l e . The anh-plyides may be prepared quantitatively, or almost so, by the methods successfullF used in t h e lapachol group. In no instance, however, in spite of many experiments. have we been successful in obtaining more than one anhydride from each compound. All the anhjdrides studied condense readily with orthotolu_vlene- diamine, giving characteristic azines, an action which, in this case, can only be explained by the presence of the orthoquinoiie group.Up t o the present time, the azines obtained have resulted from the action of one molecule or the diamine on one of the anhydride, so that, of the abo3-e €ormula;., the first, which is that of an ~/5-anhydi=ideOF ALDEHYDES WITH ~-HYDROXY-C~-NAPHTEAQVINOXE. 79 still containing one pal-aquiiione group, is perhaps the more probable, although we do not regard the second as definitely excluded. As a further proof of the presence of the orthoquinone group, it may be urged that there is a marked difference in the colour of the hydroxyl compounds, and of their corresponding anhydrides. The former, like a-naphthaquinone, are yellow ; the latter, like &naphtha- quinone, are orange or ormzge-red.A very considerable number of compounds of the lapachol group hare been studied, and this change in colonr from yellow to orange or red has proved an unfailing indication of a simultaneous change from the para- to the ortho- c yuinone. It has been shown in the lapachol group, that both the a- and &anhydrides take up water when boiled with dilute alkalis, yielding the same hydroxylapachol--a change, which, in the case of the /%anhydrides, involves the reconversion of the ort ho- into the para- qui none. 0 bv SaOH 0 0. by SaOH Precisely the same change occurs in the case of the synthesised @-anhydrides described in this paper only; it is more diEcult to demonstrate, as the hydroxSl compounds examined very rapidly undergo change in contact even with very dilute caustic soda, a reaction which appears full of interest and will be studied in detail by one of us.0 0 /\/\OH J J o / \ A B e n z y l i d e n e d i h ? ! d r o ~ 1 . y l t a p ~ ~ ~ ~ a ~ ~ ~ i ~ i ~ ~ z e 7 1 I I 1 I v y \L.--'\o/v C,H, This compound was first obtained by Zincke and Thelen (Ber., 21 2293), and subsequentiy more fully examined by Schoch (Diss., Marburg, 1888, p. 11). I n preparing the cornpound for our experi- ments, the directions of Zincke and Theleii were only partially followed. 8 grams of benzaldehyde, 8 grams of hydroxynaphtha- quinone, and 24 C.C. of alcchol were heated in a corked flask on a -.wter bath for about 45 minutes." The heating was then discon- tinued, and the solution diluted with 160 C.C. of alcohol ; the compound * Zincke and Thelen, also Schoch, recommend heating for sereral hours; this is unnecessary, as the action is complete in 45 minutes, and if the heating is extended o-i-ei' a longer period, a portion of the compound is conrerted into its anhydride.80 HOOKER AND CARNELL : THE COXDENSATION shortly aiterwards commenced to separate in minute, yellow tuEts of iriicroscopic needles, vliich, after some hours, mere collected and washed with alcohol.A microscopic examination revealed t h e presence, i n small qaantity, oE orange-red needles,* wliich usually nccom,pany to a somewhat greater extent the additionul substance obtained on concentrating the mnther liquor. The red needles are left for the most pai=t undissolved when the crude substance is re- cr~stallised from alcohol, but it is necessary to repeat the operation, sometimes more than once, in order that t'he separation may be rendered complete. We found the melting poiut of benzylidenedihydroxynaphthaquin- one to be about 230" when rapidly heated, but as the substance decomposes when heated for a short t,irne at temperstures slightly below this, its point of fusiou is influenced by variations in the rate of heating.Benzylidenedihydroxynaphthaquinone gives two distinct sodium derivatives, according as the hydrogen of one or both hydroxyl groups is displaced. The monosodium derivatire cryetallises in orange needles, and may be readily obtained by cautiously adding rz 1 per cent. aqueous solution of sodium hydroxide to a boiling and nearly saturated alcoholic solution of the compound.The addition of the alkali is stopped as soon as the colour of the solution, which is a t first orange, becomes permanently slightly carmine, and rz precipitate of tlie carmine or normal salt begins t o form. Small quantities of benzg 1 idenedih y drox ynap h t ha qui none are then gradually added t o the boiling solution until the carmine precipitate redissolves. 011 cooling, the solution first becomes turbid, arid shortly aftern-nrds orange needles of the sodium compound form throughout its entiye volume. The normal salt, previously obtained by Zincke ant1 Thelen, is n dark carmine-red, crystalline powder. When seen under the microscope, each grain appears as an almost perFectlj- formed octahedron. The very slight solubility of the salt, together with its very characteristic appearance and crystalline form, render it possible to identify minute quantities of benzylideDedihydroxy- naphthaquinone with ease and certainty.It is simply necessary to dissolve the compound in a few drops of hot dcobol, arid add a drop or so of dilute aqueous solution of sodium hydroxide ; if the alcoholic solution be a concentrated one, an abundant precipitate will be almost * This substance, which appeals to Esve escaped the observation of Zinckc, Thelen, and Schoch, differs essentially from the anhydride to be presently described : it dissolves in concentrat,ed sulphuric acid, forming an intenkely bluish-green solu- tion, whereas the anhydride under the same conditions gives a reddish-brow1 solution. We made many unsuccemful attempts to obtain the compound in larger quantities.OF ALDEHYDES WITH ~-HPDROXY-C~-NAPHTHAQUIXONE.81 instantly formed, which should then be examined with a high power. in order that its crystalline form may be recognised. ap-Anhydride of BenzyliLEenedihydl.o~~nap1~ tliaquinone .--T his corn- pound is formed in small quantity in preparing benzylidenedi hydroxg-- naphthaquinone if the time of digestion recommeuded above be much prolonged; i t was in tliis way that Zincke and Thelen obtained it (Ber., 21, 2203); Schoch found that it was produced wheu an alco- holic solution of the hydroxyl compound was heated with a little sulphuric acid ; but even in this case Schoch states that only small quantities were formed (Diss., Marburg. 1888, p. 14). We have found that the compound can be prepared practically quantitative1~- in the following manner :-4 grams of benzylidenedi- hydrosynaphtliaquinone are dissohed in 50 C.C. of boiling acetic acid, to which 2.5 C.C.of concentrated hjdrochloric acid are then verF gradually added ; on continuing the boiling for a few minutes, a heavy precipitate of small, orange-red crystals rapidly forms, and may be collected when the solution has cooled. Wheu slightly washed with acetic acid, i t is obtained in a condition pure enough for all ordinary purposes. A microscopic examination of t h e crystals revealed de- tached, well-formed, acicular prisms. The compound was purified foi- analysis by recrjsfallisation from acetic acid. 0.1837 gave 0.5203 CO, and0.583 H,O. C = 77.24; H = 3 . 3 . C,,H,,O, requires C = 77-52 ; H = 3-35 per cent.When heated in a capillary tube, it commences to darken a t about 245". It dissolves with great difficulty in the ordinary solvents, being almost insoluble in ether, and only very slightly soluble in alcohol, rather more so in chloroform, beiizene, and acetic acid. The anhydride is reconverted into beiizylideiiedihydroxpaphtha- quinone by boiling alkalis. small quantity was moistened with aqueous 1 per cent. sodium hydroxide, thoroughly ground into a, paste ; more of the alkaline solution was then added, and the whole boiled for a few minutes ; after filtration from the unattacked sub- stance, the carmine colcured solution was acidified with acetic acid, and the precipitate collected, and identified, by its characteristic sodium salt, in the manner above described.The anhydride and OrthotoIuylenedianri~~e interact very readily, a thoroughly typical azine being formed : 2 grams of the anhydride, 1 gram of orthotoluylenediamine hydrochloride, and 3 grams of crys- tallised sodium acetate are heated in 70 C.C. of acetic acid. Soon after the acid commences to boil, and before the anhjdride has entirely passed into solution, the azine cgmmences to separate in small clusters of microscopic needles. The conversion is complete i n a few minutes, but, before the boiling is finally discontinued, it is well to82 HOOKER AND CAMELL: THE CONDENSATION ascertain, with the help of the microscope, that no unchanged crystals of the anhydride are present. The substance is then collected, washed with acetic acid or alcohol, and finally with water.For analysis, the azine was twice crystallised from benzene, in which, as in other ordinary organic solvents, it dissolves with considerable difficulty. It was thus obtained in orange-yellow, microscopic needles which can be heated to 245O without undergoing change. 0.1701 gave 8.5 C.C. moist nitrogen at 30° and 764 mm. 0.2189 ,, 0.6516 CO, and 0.0815 H,O. C = 81.17 ; H = 4.13. C,H,,N,O, requires C = 80.95 ; H = 3-96; N = 5.55 per cent. The azine dissolres in concentrated sulphuric acid yielding a, brownish-violet coloured solution, from which an orange coloured salt is precipitated on slight dilution. When zoistened with concentrated hydrochloric acid, it. becomes more deeply orange ; and d e n heated with zinc-dust, and alcohol acidified with hydrochloric acid, a violet coloured solution is obtained, which readily absorbs oxygen, again becoming orange.The azine mas heated with orthotoluylenediamine, under varied conditions, but as no int’eraction occnrred, we were unable to demon- strate the presence of a second orthoquinone group. N = 5-43. 0 0 Schoch has already studied the action of acetaldehyde on hydroxy- naphthaquinone (Diss., Marburg, 1888, p. 24), and obtained a product fusing at 154-157”, which refused to crTstallise, and on analjEis gave figures agreeing with the above formula.. In spite of the analytical figures, the compound must have been in a very impure condition, as ethjlidenedihydroxynaph thaquinone crystallises readily, and fuses at about 190” when rapidly heated ; its fusing point is not? however, quite constant.After a number of preliminary experiments, we have siicceeded in satisfactorily preparing the compound in the following manner :- 5 grams of pulverised hydroxynaphthaq~zinone were heated under pres- sure in a steam bath (a5 about 100°) for five hours with 50 C.C. of alco- hol and 50 C.C. of Schuchardt’s “coucentrated aldehyde” (not absolute). After the heating had been in pi-ogress a short time, the bottle was slightly agitated, so as to ensure the complete dissolution of the hjdroxynaphthaquinone. The new compound separated, as the liquid cooled, in a comparatirely pure condition, in small, heavy crystals ;OF ALDEHYDES WITH ,@-HFDROXT-~-SAYHTHAQUINONE. $3 3.3 grams were obtained. The mother liquor, on evaporation, yielded an additional quantity of the substance, but the condition of this was less satisfactory, and, in order to purify it, it was dissolved as com- pletely as possible in n cold, aqueous 1 per cent.solution of sodium hydroxide. After filt~ation from the insoluble residue,* the sub- stance was reprecipitated without unnecessary delajf by adding ncetio acid, and was tben collected, dried, and crystallised %rom alcohol. E thylidenedihydroxynaphthaquinone can be readily distinguished from hydroxynaphthaquinone by the carmine-red colour of its solu- tion in dilute sodium hydroxide, which presents a sirong contrast to the orange-red solution of the latter. The new compound was recrystallised from alcohol for analysis. It was obtained in irreiular, heavy, golden-yellow crystals which melt when rapidly heated, but not perfeclly sharply, at about 190".The melting point is influenced to some extent by the manner of heating, aiid the fused compound gradually darkens and decomposes even as the temperature falls. 9.1920 gave 0.4950 C02 and 0.0655 H,O. C = 70.31; H = 3.79. C,,Hl,Oe requires C = 70.58; H = 3.74 per cent. up-Anhydride of EthyliderLedihyclroxynaphthnqziinoIae. E thylidenedihydroxynaphthaquinone is readily and almost quanti- tatively converted into its pseudo-anhydride by the action of con- centrated sulphuric acid. 8-5 grams of the compound were dissolved in 15 C.C. of concentrated sulphuric acid, and, after standing about 10 minutes, the solution was poured into a relatively large quantity of water.The orange precipitate was collected on a filter, and as soon as the acid had been mostly displaced by water, it was washed with a 1 per cent. solution of sodium hydroxide to remove any un- changed ethylidenedihydroxynayhthaquinone, and then again with water. When dry, the compound was crystallised from acetic acid; it separated in orange needles, which lost some of their brilliancy of colour when exposed to diff uscd daylight. Analysis gnve the following figures. 0.1881 gave 0,5089 C02 and 0.0606 H20. C = 73-78 ; H = 3.57. C2HI2O,$ requires C = 74.15 ; H = 3.37 per cent. * This consists of a dark red compound which dissolves in concentrated sulphuric ixcid, affording a green solution, and probably corresponcls with the red substance (compare footnote, p.80) observed in the prepnration of bcnzjlidenediliy droxy- naphthquinone, which also g i x s a green solution when dissolved in sulphuric acid. t Ethjliclenedihjdroxynaphthaquinone gradually undergoes change eveu in cold dilute alkaline solution. j: By heating paraldehyde with hydroxynaplithaquinone, Schoch obtained an on analysis approximate to' those required by thc formula C22H1sO; (Diss., Mar-a4 HOOKER AND CARNELL : THE CONDENSATION When boiled with a diInte aqneoas solution o€ sodium hydroxide, the anhydride gradually passes into solution, being reconverted into e t h ylidenedih ydrox y nap hthaqui none, which then rapidly undergoes change. The anhydride interacts rery readily with orthotoluylenediamine ; the resulting compound was obtained in essentially the same manner as the corresponding azine from the /3-anhydride of benzylidencdi- hydroxynaphthaquinone.Two preparations mere made for am- lysis : the one ( a ) was crystallised from benzol ; the other ( b ) from chloroform. ( a ) 0.1276 gave 6.9 C.C. moist nitrogen a t 27" and 762 mm. N = 5-95!. (a) 0.1705 ,, 0.4'317 GO, and 0.0728 H,O. C = 78.65; H = 4-74. (a) 0.1636 ,, 0.4744 ,, ,, 0.0669 ,, C = 79.08; H = 4.54. ( b ) 0.1652 ,, 0.4i43 ,, ,, 0.0675 ,, C = 78 90; H = 4.51. C2,Hl,X20, requires C = 78.73 ; H = 4.07 ; N = 6-33 per cent. The compound crjstallises in minute, orange, woolly needles which dissolve in concentrated sulphuric acid yielding a brownish-violet. coloured solution, and in this, as in other respects, precisely re- sembles the azine already described from the ,%anhydride of benzyl- idenedihydroxynaph thaqiiinone. ( b ) 0.2216 ,) 11.7 ,, $ 7 24.5 ,, 770 ,, N = 5.9s.a 8- Anhydride of Amy lidened ihy droxy nnpht ha q uinone. A mixture of 3 grams of hydroxpaphthaquinone, 3 C.C. of KaliI- baum's valeraldehyde, and 9 C.C. of alcohol mas boiled for three hours. a reflux condenser being used. A drop of the solution was then diluted with alcohol, and mixed with a weak aqueous solutioil of sodium hydroxide ; the intense crimson colour developed showed that the hydroxynnphthaquinone had undergone change, and, although the cornpouud was not. isolated, the solution undoubtedly contained amylid- enedihydroxynaphthaquinone.* To convert this into its ;I./3-anhydr- ide, the solution was evaporated to dryness i n a dish, on a water orange-coloured substance which lie xas unable to crjst alike.The figures obtained burg, 1883, p. 26). * In the hope of obtaining lapauhol s-ynthetically, I hare studied the behariour of liydroxynaphthaquinone in contact with valerzldehjde under Taried conditions, aiid hare found that if these compounds are heated in acetic acid solution, in the presence of B sufficiently large quantity of hydrochloric acid, an interaction occiirs entirely different from the abore. The resulting compound has been analysed by Mr. C. C. Burger, who has also prepared and analysed its acetjl derivatire. The figures obtained proTe that under these conditions, interaction occurs in the sense of the equation CloH603 + C,HS.COH = C,HI~O, + H,O. needles, melting a t 119-120". and with the alkalis forms violet salts mliich crys- talLise very readily. f l 1 he compound is isomeric with lapacliol ; it crpstallises in brilliant orange-red The acetyl deriratiye fuses at 74"'OF ALDEHYDES WITH b-HYDROXY -#-NAPHTHAQUINOSE.85 bath, to drive off the excess of valeraldehyde, and the residue was dis- solved in 20 C.C. of warm acetic acid, and transferred to a flask which was then immersed in ice-cold water. A mixture of 60 C.C. of Concentrated sulphnric acid and 20 C.C. of acetic acid, also previously cooled, was then added, the flask being kept, during the whole opera- tion, in water cooled by ice ; after a few minutes, it was poured into a large volume of water, and the precipitate which formed was col- lected on a filter and thoroughly washed. When dly, it was twice crystallised from acetic acid and analysed.0,1729 gave 0.4'759 CO, and 0.071 1 H,O. C = 75.06 ; H = 4-56. Cz,H,,05 requires C = 7.5-37 ; H = 4.52 per cent. The anhydride was obtained i n small, orange needles (sometimes also in small, reddish-brown plates), which commence to darken at a temperature near to 200", and fuse and decompose at a somewhat higher temperature. The presence of an orthoquinone group was demonstrated by the behaviour of the compound with orthotolnplenedianiine, the azine being prepared essentially in the manner previously described (p. 81) in the case of the corresponding benzylidene compound. For analysis, it was purified by crystallisation from benzene, in which it is much more soluble than either the corresponding benzylidene or et hylidene compounds. It was obtained in orange-coloured needles which gave precisely the same colour tests with sulphuric acid as the azines already described. 0.1618 gave 0.4705 C02 and 0.0753 H,O. 01653 ,, 8.20c.c. moist nitrogen at 26" and 767 mm. N = 5.55. C32H2rN203 requires C = 79.34 ; H = 4.95 ; N = 5.23 per cent. Cumiuic, Xalicyl ic, aud Ciwiamic Aldehydes. C = 79.30; H = 5-17. Tbese nldehjdes and hydroxpnaphthaquinone interact very rcadily (compare Schoch). By following the methods already given, we have succeeded in converting the resulting compounds, apparentIy quantitatively, into the ap-anhydrides. In all three cases, the anhydrides proved to be bright red, crystalline compounds, which were readily acted on by orthotoluylenediamine, giving characteris tic azines. The compounds were not analjsed, and for that reason are not described mere fully here. I shall defer for a short time the discussion of the constitution o€ the compound and its bearing on that of lapachol; but I may state that I hare succeeded in converting it by a number of interesting changes into compounds containing the full number of carbon atoms preeent in lapachol, and which I have also prepared from layschol itself. The behaviour of other aldehydes under s'milar conditions %-ill also be studied.-% c'. H.