摘要:

A GLUCOSIDE FROM TEPHROSIA PURPUREA. 1833CXCVL-A Glucosicle f q e o r n Tephrosia purpurea.By GEORGE CLARKE, jun., and SHRISH CHANDRA BANERJEE.Tephrosia purpurea (Pers.)-nat. ord. Zeguminosae--a, small, woodyannual, is found all over India from the Himalayas to Ceylon, andin Assam, ascending to altitudes of 4000 feet.Native works on Materia Medica describe the plant asdeobstruent, diuretic, and useful in certain febrile attacks commonin the East. When collected for medicinal purposes, the wholeplant is gathered just before flowering, dried, and tied in bundlesfor sale (Watt, Dictionary of the Economic Products of India,Vol. VI, 1893, 15). The Pharmacographia Indica states that itcontains resin, wax, and a yellow principle allied to quercetin orquercitrin. The latter has not been separated or examined.Owingto the similarity of Tephrosia prpurea and the cultivated varietiesof the Indigoferae, it is known in the vernacular as " jungli nil "-wild blue or wild indigo-although it does not contain any indicanor other substance yielding indigo.Ten lbs. of sun-dried leaves, collected in the Cawnpore districtat the end of August, were extracted with cold 95 per cent. alcoholfor seven days. The alcoholic extract was drained off as com-pletely as possible, and evaporated to a small bulk under atmosphericpressure. The green extract was poured into water, and washedwith light petroleum until the green colour disappeared. The darkbrown mother liquor deposited a copious crop of yellow crystals.The yield was about 24- per cent.of the weight of dry leaves.The crude substance was best purified by crystallisation fromdilute acetic acid (1: 3)) and finally from dilute alcohol (1: 1) orwater. It crystallised in yellow needles, soluble in ethyl or methylalcohols or acetic acid, very sparingly so in cold water, and insolublein benzene, petroleum, or ether. It dissolved readily in dilutealkalis, giving a yellow solution, from which it could be precipitatedby careful neutralisation with mineral acids. It began to sinter at182O, and finally melted and decomposed at 184--186O (u-ncorr.).The pure compound contained water of crystallisation, to whichit clung very persistently under atmospheric pressure, and underthese conditions required heating to 160° before the last traces ofmoisture could be expelled.It is easily obtained in an anhydrousstate by heating in a vacuum at the boiling point of ethylenedibromide ( 1 3 1 O ) .The air-dried substance gave on analysis the following results 3 834 CLARKE AND BANERJEE :0.698 lost 0.058 E120 in a vacuum at 1 3 1 O . H20=8.30.0-222 ,, 0-0182 H,O 131'. H,O = 8-19.0.9552 ,, 0.0777 H,O a t 160O. H,O =8*13.1'6222 ,, 0-1300 H,O ,, 160'. H20=8-Ol.The anhydrous substance is very hygroscopic, and regains water0.2650 gained 0.023.Separate fractions of the air-dried substance crystallised from0.2138 gave 0.3780 CO, and 0*1100 H,O. C = 48.22 ; H =5*71.0.1810 ,, 0.3210 CO, ,, 0.0911 H@. Cz48.36; H=5*59.C2,H30016,3H20 requires C = 48.79 ; H = 5-42 per cent.The anhydrous substance, dried in a vacuum at 131°, gave the0.2040 gave 0.396 C = 52.94 ; H = 5.20.0'2136 ,, 0.4180 CO, ,, 0.0974 H,O.C=53*37; H=5.06.C2,H,Ol, requires C = 53.1 1 ; H = 4.91 per cent.As the substance appeared to be a glucoside, it was examined inthe following way. A solution of 20 grams in 1 litre of 2 per cent.sulphuric acid was digested at the boiling point for six hours. ASthe reaction proceeded, a yellow, crystalline precipitate separatedout, and more was deposited on cooling. After being kept over-night, this substance was separated and examined. It was solublein alkalis, giving a deep yellow solution, and yielded orange-colouredacid derivatives by treating in boiling acetic acid with concentratedhydrochloric, hydrobromic, hydriodic, or sulphuric acids.The sulphate prepared in this way, washed with acetic acid anddried at looo until constant in weight, was decomposed by water,and the sulphuric acid estimated:0.343 gave 0.2004 BaSO,.S = 8-02.1.135 ,, 0.659 BaSO,. S=7*97.C1,H,,O,,H,SO, requires S = 8.00 per cent.The acetyl derivative, prepared from acetic anhydride in theusual manner, and crystallised from ethyl alcohol, melted at190-191O (uncorr.). After being dried at looo, it gave the follow-ing results on analysis :C2,H,0,,,3H,0 requires H,O = 8-13 per cent.of crystallisation on exposure to air:H20 = 8.67.C27H30016 requires H,O = 8.85 per cent.water or dilute alcohol gave on analysis the following results:following results :CO, and 0.0956 H,O.0.2070 gave 0.4424 CO, and 0.0760 H,O.The properties of this substance and the analysis of its derivativesC=58-28; H=4*07.C25H20912 requires C = 58-59 ; H = 3-90 per cent.left no doubt that it was quercetinA GLUCOSIDE FROM TEPHROSIA PURPTJREA. 1835The filtrate from which the quercetin had been separated wasneutralised with barium carbonate, filtered, and evaporated tosmall bulk on the water-bath undcr atmospheric pressure.Thesolution examined in the polariscope was dextrorotatory. It WWkept for some days in a partial vacuum over sulphuric acid, but gaveno signs of crystallisation.One gram of the dried syrup in 5 C.C. of water was mixed with2 grams of phenylhydrazine hydrochloride and 4 grams of sodiumacetate in 15 C.C.of water, and heated on the water-bath for threeto four hours. Bright yellow crystals separated, which were washedwith water and dried in the air. The melting point of the phenyl-osazone thus produced was 190-1 92O (uncorr.). When crystallisedfrom dilute ethyl alcohol (1 : l), it melted indefinitely a t 190-195O,but when crystallised from pure ethyl alcohol the melting point was206-207O, and this was not changed when the substance was mixedwith d-phenylglucosazone.Ten grams of the phenylosazone were prepared in the mannerindicated above with specially purified phenylhydrazine hydro-chloride. When fractionally crystallised from ethyl alcohol, itseparated into two osazones. The sparingly soluble one melted at206-207O (uncorr.), and the readily soluble one at 178-180°(uncorr.).The two substances were recrystallised three times eachfrom ethyl alcohol, Their melting points remained unchanged.They differed in appearance, the fraction of higher melting pointconsisting of long, silky needles, characteristic of d-phenyl-glucosazone ; the more readily soluble fraction crystallised in rosette-shaped clusters of small needles.Pure phenylrhamnosazone prepared from pure rhamnose meltedat 180-181O (uncorr.), and the melting point of a mixture of equalquantities of this substance and the readily soluble osazone fromthe Tephrosia glucoside sugars melted at 178-180°.There is thus a, very sharp separation into the two phenyl-osazones of rhamnose and dextrose resembling that described byW.Will in characterising the sugars of hesperidin and naringin(Ber., 1887, 20, 1186).Somewhat prolonged heating is necessary in preparing the mixedosazones, as phenylrhamnosazone appears to be formed more slowlythan the corresponding dextrose derivative.d-Phenylglucosazone (the sparingly soluble osazone), m. p.206-207O, when dried a t looo gave the following results onanalysis :0.151 gave 0.3328 CO, and 0.0861 H,O. C=60*10; H=6-33.0.2694 ,, 0.593 CO, ,, 0.1515 HZO. C=60*03; H=6*24.VOL. XCVlI. G EC,,H,,O,N, requires C = 60.33 ; H = 6-14 per cent1836 A GLUCOSIDE FROM TEPHROSIA PURF’UREA.Phenylrhamnosnzone (the readily soluble osazone), m. p.178-180°, when dried at looo gave on analysis the followingresults :0.1971 gave 0.4537 CO, and 0.1182 H,O.C18H,,0,N, requires C = 63-16 ; H= 6.43 per cent.A solution of the sugars from the hydrolysis of 20 grams of thepure glucoside from Tephrosia puryurea, from which quercetin andsulphuric acid had been separated in the manner described in thepreceding experiments, was evaporated to 500 c.c., and cleared witha little washed animal charcoal.The colourless solution thus pro-duced was treated with a few grams of ordinary brewer’s yeast forfour days a t 35-40°. Active fermentation ensued, and a strongodour of alcohol was produced. After separating the yeast, thesolution was evaporated to small bulk on the water-bath underatmospheric pressure, and finally dried in a vacuum over sulphuricacid. A few crystals were formed, but not enough to separate.Thesyrupy residue was warmed with 300-400 C.C. of ethyl alcohol, and awhite, flocculent substance, which separated out on the addition ofalcohol, was removed, The clear alcoholic solution was evaporatedto about 50 c.c., and set aside to evaporate at room-temperaturef o r a few days. About 2 grams of large, rhombic crystals, charac-teristic of rhamnose, were obtained, and a second crop of smallercrystals separated from the mother liquor.The substance thus produced melted when carefully heated at93-94O (uncorr.). The melting point of pure rhamnose under thesame conditions was 93-94*, and a mixture of equal quantities ofpure rhamnose and the rhamnose from the Tephosia glucoside alsomelted at that temperature.The rhainnose from the Tephrosia glucoside, dried in a vacuumover sulphuric acid, gave on analysis the following results:0.1866 gave 0.2710 CO, and 0.1310 H20.The decomposition of the glucoside into quercetin, rhamnose, andC=62*77; €1=6*66.C=39*60; H=7*80.C6H1406 requires C = 39.56 ; H = 7.69 per cent.glucose takes place in accordance with the following equation :C27H300]6 $- 3H20 =C,,H,,o7 + c6H1206 + C(3H1205’H,0*The decomposition was carried out quantitatively, the quercetin0.544 air-dried glucoside gave 0-246 C15H,,0, = 45.22.C27H3001,,3H20 requires C,,H,,O, = 45-48 per cent.0.639 anhydrous glucoside gave 0.315 Cl5Hl0O7 = 49.29.CZ7H,,Ol, requires C1,H1,07 = 49-50 per cent.Two substances have been described which closely resemble theTephrosia glucoside, namely, osyritrin (Osyris compressa) and rutin,collected in a Gooch crucible, and dried a t 16O0PICKERING : C U PRl CITR ATES.1837which is present in rue (Ru.ta graveoleus) and other plants.Whereas the former, C,,H,,0,,,3Hz0, gives quercetin and dextrose(Perkin, Trans., 1897, 71, 1134), the latter, originally consideredby Schunck (Trans., 1888, 53, 264) to yield only quercetin andrhamnose, has been found by Schmidt (Arch. Pharm., 1908, 246,214) to have the formula C,,H,@,,,3Hz0, and to give, whenhydrolysed, quercetin, rhamnose, and dextrose. Through the kind-ness of Mr. A. G. Perkin, a small sample of rutin was available forcomparison, and as a result of experiment it was found that theTephrosia glucoside is identical with this substance. On the otherhand, it is interesting to note that Perkin (private communication)has found that osyritrin is identical with rutin [this vol., p. 17761.The authors desire to express their thanks to the Managers ofthe Royal Institution of Great Britain for kindly placing the equip-ment of the Davy-Faraday Research Laboratory at their disposalfor completing this investigation, which was begun in India; and toMr. H. Martin-Leake, Economic Botanist to the United ProvincesGovernment, for kindly identifying and growing the material used.DEPARTMENT OF AGRICUL~ URR,DAVY-FAKADAY LABORATORY, RESBARCII LABORATORIES,ROYAL INSTITUTION.UNITED PROVIKCES,1 N 111 A

ISSN:0368-1645    

DOI:10.1039/CT9109701833

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

PICKERING : C U PRl CITR ATES. 1837By SPENCER UMFREVILLE PICRERING, M.A., F.R.S.AN examination of the double carbonates of copper and sodium(Trans., 1909, 95, 1409) led to the suggestion that in those ofthem wherein the copper is electronegative, it is present in thequadrivalent condition as :Cu:O, united directly with one 1- of thecarbon atoms. The change of the copper atom into such a positionfrom that which it usually occupies would be brought about bythe conversion of a carboxyl group into two hydroxyls through theaction of an alkaline hydroxide (#3-compounds) or carbonate (a-corn-* Some time has elapsed since this paper was communicated to the Society (seeProc., 1910, 26, 17), and portions of it have been so much elaborated by furtherwork, that it has been considered advisable to reserve these for independenttreatment.f It was originally represented as being united to both the carbon atoms; butthis was unnecessary, and involved the somewhat unacceptable assumption of thegroup :C/ I being present in the molecule.0\O6 3 1838 PICKERING : CUPRICITRATES.pounds), the component atoms of which become added to those ofthe double salt, as indicated by the formulz (I), (2) and (5) :....._O'H ONa NaO.C OiNa Na0.C O/H NaO*COH 0; HNaO*C00 I : I : I ONaNaO*CO@u cu:io cu:;o 7"OHNaO*C00I ..NaO*C00I -..-...NnO*C00(1) Sodiocupric (4) y-Cupri-carbonate. (2) 13-Cupricarbonate. (3) 8-Cupricarbonate. carbonate.I ............__..._.NaO* COO( 5 ) a-Cnpricarbonate.OOC-ONn(6) a-Cupricarbonate.The y-cupricarbonate was not mentioned in the paper itself,but some evidence was given that the action of an alkali could becarried a step further than the formation of the j3-compound(2), the copper, as well as the carbon, becoming hydroxylised, andbeing then in the " cuprite " condition; and it appears now, aswas suggested in an addendum t o the paper, that it was the y-, andnot the j3-compound, which was present in the deep violet solutionobtained, and described as containing the latter, since there wasa large excess of sodium hydroxide present with it.This colour, aswell as the property of combining with cellulose, appears to becharacteristic of copper in the cuprite condition. Some fach havealready been mentioned (Zoc.c i t . , p. 1427) to show that a simplesodium cuprite may be obtained by the action of sodium hydroxideon copper oxide, and, also, that calcium cuprites exist.+The double carbonate, or sodiocupric carbonate, is obtainable asa crystallised salt, but none of the cupricarbonates were isolated,and the evidence in favour of their existence has now been greatlystrengthened by the isolation of numerous similar compounds inthe case of citric acid. The two cases seem to be very strictlysimilar ; double salts of analogous composition, exhibiting very* This group was originally represented as :C<E>O, implying the direct unionof oxygen atoms, which is now obviated.It would be preferable, i n the writer's opinion, t o adhero to the designationof cuprates, fornierly applied, for salts derived from cupric oxide, rather thanto reserve that term fdr the hypothetical derivatives of CuO,.That the latterexist, and are red in colour, can scarcely be accepted, as maintained by Ramsay(Proc., 1910, 26, 18), on the strength of Brauner and Kuzma having obtained a redcompound which was " supposed to contain a tellnrocuyric acid " (Ber., 1907, 40,3362)PICKERING : CUPRICITRATES. 1839similar properties, form the starting point in both series, and thefinal products of the action of alkalis are in both cases violetcompounds, which oxidise dextrose in the cold, and combine withcellulose; whilst, in the case of citric acid, the existence of com-pounds containing potassium in the condition, both of hydroxideand of carbonate, has been established, by these being isolated inthe crystalline condition, and several other similar salts, containingcopper in the place of the alkali metal, have been prepared.I n two particulars a modification of the views expressed as tothe constitution of the cupricarbonates is suggested by the resultsobtained with the citrates.Firstly, that the a-compound isprobably not formed by the combination of a molecule of thealkaline carbonate with a molecule of the double carbonate, asshown in (5), but by the combination of a molecule of sodiocupriccarbonate with two molecules of the double carbonate, as shownin (6). Its empirical formula would, thus, be 3[Na&u(CO,),];and that it does consist of sodium and copper carbonates combinedin equal molecular proportions is in accordance with the fact(Zoc.cit., pp. 1418, 1425) that it is formed in maximum quantitywhen the reagents are present in such proportions. It may alsobe mentioned that what appears to be an analogous compoundcontaining potassium has been isolated as a crystalline salt.Secondly, that the substance present in the solution from whichthe double salt is obtained is really a cupricarbonate, with its copperin the electronegative condition, which renders it possible toattribute the formula for sodiocupric carbonate (1) to the crys-tallised salt, instead of having to have recourse to the view that itis merely a molecular compound.The cupricitrates described below are, with one exception, in whichthe substance is of a constitution different from that in othercases, characterised by certain general features. Unless the pro-portion of copper present in them is very large, they are allextremely soluble, dissolving in about a third of their weight ofwater at, 8O.The solutions are of an intense blue colour, thoughof a dirtier tint than that of copper sulphate; they can be con-centrated by heat, without decomposition, until a scum begins toform on the very viscid liquid; but crystallisation often does notbegin for many days, and proceeds very gradually throughoutseveral days more, until the whole becomes solidified, unless this isprevented by the addition of a little more water. Concentration ofthe liquid can be effected by adding alcohol.The crystals aremicroscopic in size, of a dead, somewhat light blue colour, ands o f t ; on filtration from the liquid they form a putty-like mass, andcan be purified by recrystallisation only with great difficulty. The1840 PICKERINO : CUPRICITRATES.gradually dry a t looo to an extremely hard mass, but most ofthem retain some water at this temperature, which they give upa t 140-150°. The anhydrous compounds are blue and hygro-scopic. On heating to a higher temperature they decomposequietly, this affording a convenient means of analysing them, thecopper being determined gravimetrically as oxide, and the potassiumvolumetrically as carbonate.Elect ronegative Copper.The presence of a deep blue electronegative ion containing copperwas proved in the case of the cupricarbonate by Wood and Jones(Proc.Camb. Phil. SOC., 1907, 14, 174), and in the case of thecupritartrate by Masson and Steele (Trans., 1899, 75, 725): theapparatus devised by the latter-consisting of two tubes connectedby a side-tube filled with agar agar in brine-has been used bythe present writer for examining the cupricitrates, and it has beenfound that they, also, contain a dark blue, slow-moving, electro-negative ion, although, during the electrolysis, a certain amount ofa light blue, more rapidly-moving, electropositive ion makes itsappearance, and results in a deposition of some copper on thenegative plate. Copper in the electronegative condition does not,apparently, attack metallic iron, or react with ferrocyanide; butwith the latter reagent, the electronegative copper appearsgradually to be converted into electropositive, the red colourappearing in a space of time varying between a few seconds andmany hours. None of the cupricitrates, except the one containingpotassium in the alkaline condition, fail to react with ferrocyanideon long keeping; but that is not the case with certain other cupri-compounds examined ; potassium cuprisaccharate, for instance,although quite neutral, gives no trace of red with ferrocyanide,unless the solution has been previously boiled, but the boiled solu-tion loses its power of reacting after the lapse of two days.When no red ferrocyanide is formed-which is also the case ifalkali hydroxide is present, either free or as an integral part of thesalt-the liquid, if dilute, turns yellow, then green, and, after a longtime, an orange precipitate, or ring at the surface of the liquid,forms, which is insoluble in acid, but dissolves in strong alkalis,being first converted by them into a blue substance. Excess offerrocyanide is required for the formation of this compound.The presence of free or combined carbonate does not interferewith the formation of the red copper ferrocyanidePICKEliIh’G : CUPRICITRBTES.1541Compounds 0 b tained.(1) Potassiocupric Citrate, (C6H,07),CuK,.-when 100 grams ofpotassium citrate (C6H,07K,,H20) are dissolved in a small quantityof water on a water-bath with 8 grams of citric acid and 40 gramsof copper citrate-which is a basic salt, (C,H~Oi),Cu3,Cu0,2H,0-the liquid becomes converted into a mass of minute, irregular,hexagonal crystals, which, unlike the cupricitrates, are hard andgritty, and often adhere strongly to the dish.This substancecontinually makes its appearance when the reagents are taken inproportions other than those mentioned above, and also when nocitric acid figures amongst them. It is of a very light bluecolour, and although it dissolves to a large extent in water tQ forma deep blue solution--100 C.C. of a, solution at 8O, containing 44grams of the salt-it does so only very slowly, the result of whichis that, when a stream of water is directed on to it on a filter,the moist blue substance becomes white wherever the water falls,and then gradually resumes its blue colour; a behaviour whichrenders it easily recognisable.After drying over sulphuric acid or at looo, various preparationsof it gave as a mean:Fonncl : Cu, 10.56 ; I<, 25.95 ; Ratio, 1 : 4.00.Calculated: Cu, 10.63 ; K, 26.15 ; Ratio, 1 : 4.This salt, consisting of two molecules of potassium citrate withtwo of the potassium atoms displaced by a copper atom, is analogoust o the double carbonate of potassium and copper, and, as in thecase of the citrate, the substance can hardly be represented as amolecular compound-the constitution of which would have to be4C,H,07K, + 3[(C6H,O7),Cu:]-the presumption is that the carbonatealso is not a molecular compound. The nature of the two sub-stances, and the similarity of their behaviour with water, givesfurther support to this view; as they are both light blue, crystallinesubstances, which in solution form very dark blue liquids.That these liquids cannot be mere dissolutions of the crystallisedsubstances is demonstrated in the case of the carbonate by the factthat the crystals will not redissolve in the mother liquor, and aredecomposed by water.The liquids, it, is suggested, contain a cupri-compound, formed by the addition of the elements of water to thedouble salt in a manner similar to that in which the alkalineP-cupri-compound is formed by the addition of the elements ofEHO; the citrate would thus be(2) A P-cupricitrate, (C,,H,07)2K413,C~~0, with a constitutionsimilar to that represented by the structural formula (3), p.18381842 PIC I( EK I NG : C U PR 1 C ITK ATES.The objection to such a view is that the copper in thedissolved substance appears to be in the electropositive con-dition, reacting completely with ferrocyanide at once, or verynearly so; this, however, has been proved to be due to theelectronegative copper being in a very labile condition, for, onelectrolysis, it is found that the copper, or the greater part of it,is really present in the electronegative ion, just as in the case ofthe other cupricitrates. It is also found that it has the samecolour intensity and peculiarities as the copper in these othercupricitrates, this intensity being eighteen times that of thecopper in copper sulphate (p. 1850). I n the case of the cupri-carbonate, tfie presence of electronegative copper was similarlydemonstrated, even in those solutions which contained no excess ofsodium carbonate, and the reaction of which with ferrocyanide hadpreviously led to the conclusion that the copper in them mustbe electropositive (Trans., 1909, 95, 1419).*(3) Potassium P-mpricitrate, (C,H,O,),K,HKCuO, has not beenisolated with certainty.On one occasion a small crop of micro-scopic, but comparatively large, crystals was obtained, which werestrongly alkaline, and contained no carbonate; these were probablythe compound in question, but they were not obtained in quantitysufficient for analysis. On mixing potassiocupric citrate with therequisite amount of potassium hydroxide, and concentrating byexposure over sodium hydroxidean alkaline solution cannot beconcentrated by heat, as copper oxide is thereby separated-a clear,deep blue, and almost solid syrup was obtained, which was stillstrongly alkaline.After some months, during which a little waterhad been added several times to it, and the liquid re-evaporated,it gradually became converted into a crystalline magma, which wasquite neutral, aiid consisted probably of a mixture of some of thecupricitrates described below.(4) Potassium P-Cuyricitrate, (C,H,07),K,Cu,6H,0.-Luff ob-tained this salt by mixing potassium hydroxide, copper sulphate,potassium citrate and citric acid, and concentrating by shaking upwith alcohol (Zeitsch. Ges. Brauwesen, 1898, 21, 319). Jeffers, who,however, has not published his results, improved the method bysubstituting copper acetate for the copper sulphate.A still simplerprocedure is to dissolve 40 grams of copper citrate in a hot, nearlysaturated solution of 100 grams of potassium citrate, and to addto it, when the liquid has cooled to 40--50°, 20 grams of potassium* It appcars probable that some of the siniple copper salts of organic acids mayhave, in solntioii, a coiistitution similar to that suggested for the double &ate insolution. Salts of ironexhibit like peculiarities ; thus, ferric citrate, when quite iieutrd, giyes no reactionwith ferrocyanidc.It is remarkable that copper acetate does not act on ironPICICERING : CUPHIClTlt ATES. 1843hydroxide in very strong solution. A copious crystallisation occurs,either at once, or on cooling, the crystals consisting of hexagonaltablets, generally very large, clear and regular, and of a fine violet-blue colour.They may be washed with a little water or alcohol,dried between blotting paper, and then by heating at looo for ashort time. The substance cannot be recrystallised, as in strongsolution it soon decomposes, although in a dilute solution (con-taining about 0.1 per cent. of copper) it is quite stable, even onboiling. It is insoluble in a strong solution of potassium citrate.It loses a little water very slowly at 100°, and 6H,O at 120-160°,forming a lavender-coloured mass, which redissolves in water, pro-ducing a solution identical in every respect with that obtained bydissolving the hydrated crystals.A t 170° it becomes green, andon dissolution forms what is apparently a green solution, but thiscolour is due to the presence of minutely divided cuprous oxide,which eventually subsides, leaving a blue liquid. A similar decom-position occurs with other similar cupri-compounds, and has beenmisinterpreted in several cases as indicating the formation of somesubstance which is really green.It has no action on metallic iron, and gives no red colorationwith Terrocyanide, but the orange substance, mentioned above, isgradually formed if the ferrocyanide is in excess. It does notoxidise dextrose in the cold, but when boiled with i t for a fewminutes, cuprous oxide is precipitated, which redissolves slowly aftercooling.Many preparations of the salt were analysed, and all gave con-cordant results, the mean of which was:Found: Cu, 8-09 ; K, 29’74 ; H,O, 14.10 ; Ratio, 1 : 5.97 : 6-15.Calculated: C ~ I , 8’13 ; K, 29.99 ; H,O, 13-82 ; Ratio, 1 : 6This salt has been accepted as being potassium citrate with coppersubstituted for the hydrogen of the alcoholic hydroxyl; but sucha constitution is quite inadmissible, for it is found that thepotassium present in it is in two different conditions, one of thesix atoms being present as alkaline hydroxide.The alkalinitywas determined with sulphuric acid, using phenolphthalein asindicator, and evaporating the liquid repeatedly until the residueshowed no further red colour on the addition of water: themean result of many determinations gave 0-94K as being in thealkaline condition. The only formula that appears capable ofexpressing this fact is the following, in which the copper functionsas in the case of the other cupricitrates, but also acts partly bydisplacing the hydrogen atom in one of the alcoholic hydroxylgroups, the formation of such a compound being explained by theaddition of 2KOH t o the molecule of potassio-cupric citrate, instead: 61844 PICKEKING : CUPRICITRATES.of 1XOH only, as in formula (2), p.1838, which, resulting as itwould in the formation of the unstable group :Cu< I , gives riseto rearrangement, with the liberation of the elements of water, andthe formation of00An objection to this formula is that it represents the presenceof only five molecules of water of crystallisation, so that, of thesix lost on drying, one must be derived from the elements composingthe body of the molecule.This is not very improbable, for ondehydration at a temperature a little above looo, the water lostwas found to be considerably less than 6B,O (5.3 t o 5.7H,O invarious cases), indicating that all the water present is not on thesame footing. I n any case, however, no argument based on theapparent water contents can counterbalance that depending onthe alkalinity of part of the potassium in the salt.I n solution, this salt shows a colour intensity similar to €hat ofthe other cupricitrates, but its colour in the solid condition, as wellas the size, hardness, and general appearance of the crystals,differentiates it entirely from the latter.Bullnheimer and Seitz (Ber., 1900, S3, 817) obtained a saltwhich they represent as consisting of two molecules of Luff's saltcombined with one molecule of a similar salt with K, in place ofCu; but no details respecting it were published, and the writer hasnot a t present succeeded in preparing it.(5) Potassium a-Cztpm'citrat e ; empirical focmula,[(cGH,0,)2CUK~l,,=2~u(co~)~*-When a concentrated solution of potassium carbonate is added tocopper citrate dissolved in potassium citrate, some carbon dioxide isevolved, and, on evaporation, a mass of microscopic, acicular crystalsis eventually obtained.They are alkaline in reaction, due to thepresence.of carbonate, not of hydroxide. The copper present iselectronegative, as proved by electrolysis, and gives no reactionwith ferrocyanide until after about one minute.Four preparations, made with various proportions of reagents,gave the following values for the anhydrous substancePICKEKING : CUPRICITRATES.1845Found. Calculated.p-. zzr5ZCu .................. 13 '82 3.00 13-08 8K (total) ............ 28.34 10.00 26.82 10K (alkaline) ...... 5.51 1'94 5-36 2CO, ................. 4'92 1.55 6 04 2The water present in the specimens, after drying at looo, was2H,O in one case, and about 6H20 in the others.The proportions of copper, total potassium, and alkaline potassiumagree well with a formula similar to No. 6, p. 1838, representing asubstance derived from two molecules of potassiocupric citrateunited by one molecule of potassiocupric carbonate.The propor-tion of carbon dioxide is low, 1-55 instead of 2, due apparently tothe fact that some of the compound to be next mentioned is alwaysformed with the a-cupricitrate, and cannot be effectually separatedfrom it by recrystallisation. In a preparation which evidently con-sisted of a mixture of these two substances, the proportion ofalkaline potassium to CO, was found to be 1: 1.07, which agreeswell with the ratio 1: 1 required on the view that K,Cu(CO,), ispresent in the molecule.In accordance with this view, it was found that potassiocupriccarbonate dissolved easily and completely in a solution of potassio-cupric citrate, even when dilute, although the double carbonate isentirely decomposed by water.The product thus obtained in thecase of the citrate was not examined, but that obtained in thecorresponding case of the tartrate was found to be a substancecontaining as part of its composition the elements of potassio-cupric carbonate, as here indicated. In the case of the citrate, thedissolution of the double carbonate is always attended by theevolution of some carbon dioxide, so that some substance otherthan the a-cupricitrate must be formed at the same time, thuscreating the difficulty mentioned above in obtaining the a-cupri-citrate pure. With the tartrate there is no such evolution of gas,and no formation of other substances.The addition of alkali hydroxide to the a-cupricitrate abstractscarbon dioxide from it, and precipitates a basic citrate (No.10,p. 1848), which dissolves in excess of alkali, forming a deep violetsolution. This action is precisely similar to that occurring in' thecase of the a-cupricarbonate.The a-cupricitrate often makes its appearance in alkaline liquidswhich have been exposed to the air. From the liquids, also, whichcontained much potassium carbonate, were obtained on severaloccasions large, acicular crystals of K2C0,,3H20, which are remark-able in being, unlike the dihydrate and anhydrous salt, non-hygroscopic (Morel, BuU. SOC. franq. Min., 15, 7)1846 PlCKERTNG : CUPRICITRATES.(6) Tetrapotassio-cupric P-Cupricitrate, (C,H,O,),K,Cu,CuO.-Onthe solidification of solutions of copper citrate in potassium citrate,there was repeatedly obtained a substance closely resembling thea-cupricitrate in general appearance, but very different in crystallineform : the crystals resembled straight hairs with truncated ends,the length of which was often fifty times greater than the breadth.They gradually grew together into rounded masses, projecting con-siderably above the liquid.Four samples on analysis after re-crystallisation gave :cu. R. H,O at 100". Ratio.Found : 16-63 22-85 5.22 2 : 4.36 : 2-22Calculated : 17 -82 21 '92 5.96 2 : 4 : 2These values are not very satisfactory, m there always seems tobe some of the a-cupricitrate present, which cannot be eliminatedby recrystallisation. A mixture of these two substances was re-crystallised five times, whereby the relative proportion of thea-cupricitrate was greatly reduced, although there was still a con-siderable quantity of it left.Ascertaining the amount of this fromdeterminations of the alkaline potassium and carbon dioxidepresent, and deducting the corresponding quantities of copper andpotassium from the totals, the composition of the residue wasgiven as:Found: Cu, 19-48 ; K, 23.98 ; Ratio, 2 : 4-01,Calculated: Cu, 18.76 ; K, 23-08 ; Ratio, 2 : 4.The ratio here is satisfactory, and the discrepancy between thefound and calculated percentages would be accounted for by acomparatively small error in the water determination.The constitutioll of this compound may be represented as that ofa /3-cupricitrate similar to No. 3, p. 1838, with the hydrogen atomsof the two hydroxyls displaced by an atom of copper.It is probablethat this copper is quadrivalent, and that two such molecules arejoined together by means of it:-0 -*>cu :.<;I,for the whole of the copper in the compound appears to be electrenegative, not acting on iron, and giving no colour with ferrocyanideuntil one or two minutes after it is added.This substance may also be obtained by the decomposition of thefollowing compound.(7) and (€3) Dipotassio-dicupric 8-Gupricitrate,(C,H,O,),K,Cu,CuO.-When copper citrate is heated with a strong solution of potassio-cupric citrate, it dissolves, but in a few minutes the whole graduallysolidifies to a light blue, crystalline mass. After a preliminarPICKERINCX : CUPRICI'I'RATES.1847washing to remove soluble impurities, this is found to consist of(a) a small quantity of a dense, dark blue, scaly deposit, oftenadhering firmly to the dish, and only sufficiently soluble to givea faint reaction with ferrocyanide; and ( b ) , a light blue solid,which, on washing with water for many days, yields continuouslya solution containing 0-08 to 0.09 per cent. of copper. Analysesof both these solids, and of the solution obtained from ( b ) , all gavethe same ratio for the copper and potassium present:Ratio. m: 2.06 : 0-47 1; : 2.04 : - .................. Dense solid ( a )Light solid ( b ) .................. r% 1 2-08 : -: 2.13 : 0.92 ..... I * 1.99 : 0.82 Solution of ( b ) evaporatedThe mean results for (u) and ( b ) gave, for the anhydroussubstance :Found : Cu, 27.29 ; K, 1153.Calculated: Cu, 28'77 ; K, 11.80.The proportion of water found in (a) indicates that the formulashould be doubled, as suggested for the compound last described,(6), from which the present substance differs only in having anatom of copper substituted for two atoms of potassium.Theselatter may be different atoms in the case of the two compounds( a ) and ( b ) , thus accounting for the existence of two substanceswith the same formula. It acts, although very slowly_, on iron, andgives some red colour a t once with ferrocyanide; the greater partof the copper in it appears, however, to be electronegative, the fulldepth of colour being developed only on keeping. The colour-intensity of the copper in the solution is about eleven times that ofcopper in copper sulphate.In all these rqpects the characteristicsof the substance tally with the formula suggested, which representsthe presence of both electronegative and electropositive copper (seep. 1850).The solution of ( b ) on evaporation leaves a residue which isdecomposed by water, the cupricitrate, No. 6, (C,H,O,),K,Cu,CuO,passing into solution,* whilst ordinary basic copper citrate remainsundissolved. This reaction indicates that the latter substance isreally a(9) Copper cupricitrute, (C,H,0,),Cu3,Cu0, forming with the twolast-described compounds a series of cupricitrates in which successivepairs of potassium atoms are displaced by copper atoms, with* This woald probably bc the best method for obtaining this compound in a stateof pu L' i ty 1848 PICKERINO : CUPKICITRATES.progressive decrease in solubility. The peculiar lavender-bluecolour of copper citrate indicates that it is probably not an ordinarybasic salt.(10) The basic copper citrat e, (C,H,07),Cu,,4Cu0, precipitatedby the action of alkali hydroxide on the a-cupricitrate (p.1845),appears, on the other hand, to be a true basic salt. It is of a pureblue colour, and, unlike the cupricitrates, it loses all its water a.tlooo, and begins to decompose at 160O.(11) A Potassio-cupric P-Cupricitmte, (C6H507)3K5Cu2,2Cu0.-During the preparation of one of the cupricitrates, another sub-stance was obtained in considerable quantity, consisting of veryregular lenticular crystals, which were so small that they could notbe resolved except under a high power of the microscope.Unlikeany of the other compounds obtained, this substance is decomposedslowly by water, which accounts for the deficiency of potassiumfound in it on analysis, the sample having been slightly washed:Found: Cu, 22.38 ; I<, 15'55 ; Ratio, 4 : 4-64.Calculated: Cu, 23'25 ; I<, 18'64; Ratio, 4 : 5.It contained about 3H,O at looo, which it lost at 160°, but thiscould not be determined sat.isfactorily, as it began to decomposeat, or slightly above, this temperature. The substance is probably,as indicated, similar to No. 6, but ilerived from three, insteadof two, citric nuclei; and the existence of such suggests thepossibility of there being many more cupricitrates of complexcharacter.By the action of water on it, a nearly insoluble residuewas obtained, which was found to be the cupricitrate No. 8, whilsta blue solution was obtained containing copper and potassium inthe ratio of 1 : 3, but whether in the form of a definite compound,or not, was not determined.(12) Potassium y-Cup=icitrate.-Attempts to isolate the deepviolet substance present -in solution, when excess of alkali is addedto any of the cupricitrates, failed. The liquid decomposes withthe liberation of cupric oxide when concentrated, either by heat,or by exposure over sulphuric acid. Alcohol abstracts the excess ofalkali from it, and Luff's salt is the only product obtained. Thecolour intensity of the copper in it is about eighty times that ofcopper in copper sulphate, but increases with the strength of thesolution, and with the proportion of alkali added. As regards itscolour, the precipitation of cuprous oxide from it by dextrose inthe cold, and its combination with cellulose, it is closely similar towhat has been described as the y-cupricarbonate, as well as t o othercompounds (cupritartrates, etc.) obtained under like conditions.(13) Potassio-cqric Hydrogen Citrate,(C,H,O,),K,Cu,C,H~~,KH,,H,OPICKERING : CUPRI CITR ATES. 1849-On several occasions the -residual liquid, after the crystallisationof some of the cupri-compounds, was of a light, greenish-blue colour,especially when citric acid had been taken as one of the reagents,and, on further evaporation, yielded a crop of fairly large, lightgreen, hard crystals, with an acid reaction.They were stable inair, and began to decompose at 1 5 0 O . On analysis, after drying atlooo, they gave values agreeing with the formula given above,which represents a molecular compound of potassio-cupric citratewith monopotassium citrate :Fonnd: Cu, 7.34 ; I<, 23.29 ; Ratio, 1 : 5 04.Calculated: Cu, 7.52 : I<, 23-10 ; Ratio, 1 : 5.There is no reason for regarding this salt as a cupri-compound.The Constitution a9td Colour of Cupri-compounds.The isolation of so many cupricitrates agreeing in their pro-perties, and, apparently, in their constitution, with the non-isolatedsodium cupricarbonates, must lend considerable support to the viewsput forward as to the nature of the latter, especially when the saltsisolated are found to include some in which the potassium is presentin the condition of hydroxide and carbonate.A considerableamount of additional evidence has already been obtained, all tendingin the same direction. Thus, several potassium cupricarbonateshave been isolated in the crystalline condition, and these agree withthe formulae suggested for the sodium salts, and are, moreover,closely similar in their general characteristics to the cupricitrateshere described. According to the views hitherto accepted, thecupricitrates would be represented, either as ordinary copper com-pounds with the metal displacing the carboxylic hydrogen, or ascompounds with the metal displacing the hydrogen in the alcoholicgroup; according to the view now suggested, the copper is presentin the molecule as :Cu:O, and the molecule contains, for eachatom of copper present, the elements HHO over and above thosepresent according to the ordinary views.It should be possible,therefore, to settle between the rival theories by determining themolecular weight of the compounds. The smallness of the dif-ferences to be measured, and the difficulty in obtaining the cupri-citrates in a condition of sufficient purity, has rendered suchdeterminations unsatisfactory in their case ; but other similar com-pounds have now been obtained where such difficulties do not exist,and in every one of these instances, now numbering six, themolecular weight of the substance agrees closely with that requiredaccording to the present views1850 PICKERINQ : CUPRICITRATES.I n cases where a copper atom displaces the hydrogen of the twohydroxyl groups, as in No.6, the properties of the substance seemto be conclusive against its being represented according to thehitherto accepted views; for according to these it would either bepotassio-cupric citrate with Cu.0 added on to it-that is, an ordinarybasic s a l t - o r potassio-cupric citrate, with copper displacing thehydrogen atoms of the alcoholic hydroxyl groups. Numerous com-pounds, however, have been obtained with copper in the " alcoholic "portion of the molecule, and these are all essentially different intheir nature from this, or any of the other cupricitrates; whilst theextreme solubility of this compound must effectually negative theview that it can be a basic salt.The accepted theory as to the constitution of cupri-salts affordsno explanation of the formation of further products by the actionof excess of alkali on them, such as must exist in the deep violetsolutions thus obtained (y-cupri-salts), or of the existence of com-pounds containing as part of their constitution the elements ofalkali carbonates (a-cupri-salts).Some interesting further evidence on the subject has been obt.ainedfrom a study of the colour-intensity of these and other copper salts,to which a brief reference only can be made here.The molecularcolour-intensity of copper in salts formed from strong acids hasapproximately the same value in all cases, and is independent ofthe concentration of the solution throughout the wide range overwhich comparison is possible. Taking this intensity as unity, thatof copper salts derived from weak acids is much greater, reachingthe value of 10 in some cases, and in every instance it diminishesas dilution increases, falling to 2 or 3 in most cases, and sometimeseven to 1. With the cupri-compounds, however (Nos. 2, 4 and 6, forinstance), the colour phenomena are very indifferent, indicatingthat the copper present must be in some peculiar condition, forthe intensity is as high as 18, and remains constant down toextreme dilution. With the compound No. 7, p. 1846, the colourintensity, although still very high, is less than 18, and diminishesslightly on dilution, this being quite in accordance with the viewthat it is a cupri-compound, but one containing also some copperin the electropositive condition.The view that copper may act as a tetrad is not new, althoughits behaviour as such receives very uncertain support from theexistence of the oxide CuO,. The extraordinary facility with whichit enters into combination with carbon compounds would receivesome explanation if its quadrivalent character is admitted, and itsposition in the periodic system, which is in any case anomalous, caPlCKERlNG : THE CONSTiTUTlON OF BASIC SALTS. 1851afford no argument against such a view. That compounds in whichcopper is directly united to carbon should be explosive, becausecopper acetylide, the constitutdon of which is very uncertain, is such(see Ramsay, Proc., 1910, 26, 19), is an argument which canscarcely carry much weight.HAILPENDEN

ISSN:0368-1645    

DOI:10.1039/CT9109701837

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

PlCKERlNG : THE CONSTiTUTlON OF BASIC SALTS. 1851CXC V I I I. - Tlz e Constitution of Basic Salts.By SPENCER UMFREVILLE PICICERING, M.A., F.R.S.No class of compounds has proved less attractive to chemists duringthe last half-century than basic salts; they have been accepted asnecessary evils, cropping up where least required, and undeservingof any hypothesis to explain their constitution. They are, perhaps,regarded in a hazy way as being analogous to hydrated salts,although such a view can hardly be maintained, for the chiefcharacteristics of hydrated salts are their solubility and powers ofcrystallising, whilst basic salts are mostly insoluble and amorphouscompounds; in hydrated salts, moreover, the bond of union betweenthe water and the salt is, in all probability, the oxygen atoms,whereas in basic salts it would appear to be the metallic atom.It is significant that no basic salts of univalent metals are known.The want of interest evinced in these compounds is largely dueto the fact that there seems to be no guiding principle governing intheir formation, axd that, in the large majority of cases, even theircomposition is a matter of considerable doubt.This may be saidwithout any disparagement of the accuracy of the work on whichwe have to depend for our knowledge respecting them, for a t thetime when this work was done, ideas as to what constituted validevidence of the individuality of a compound were very differentfrom what they are now. Certainly our present neglect of basicsalts is not justified by any inferiority in the part which they playin chemical changes; for in many manufactures, ia the operationsof agriculture, and in geological processes, they evidently play, orhave played, a very prominent part, and we can hardly claim asatisfactory knowledge of simple salts until we know something ofthe numerous progeny of basic salts to which they give rise.Some light appears to be thrown on the matter by a study ofthe basic sulphates of copper, in connexion with the facts establishedVOL.XCVII. 6 regarding the cupri-salts, for it is found that these basic sulphatesgradually pass by successive steps from substances akin to ordinarycopper sulphate to others wherein the copper is electronegative,ending with compounds which contain no sulphur a t all, and whichare, apparently, simple cuprites.The series of substances formed by adding different proportionsof lime-water to copper sulphate have already been described(Trans., 1907, 91, 1989), and additional evidence respecting them0*0021Of~I150*000300008ubstan.ces Precipitated f r o m Copper Sulphate b y Lime ?Yater.Down t o ( 6 ) the liquids all contain 0.05 per cent.of copper.niln i lnilnilnilniinil-Mols. CaCtakcii to1 inol. cuso1.I.(2). 0.750.76O'ib'(3). 0.80%0%1 *o1 '11-15(5). 1'2(4). 0.91 3251.301-352 .O5a). 3.03 '54 '04 '5(6). 25(7). g1.50Snbstancepresent.I I.4 c n o , 8 0 3Mixture .. (O'OGCaSOj]( 5C;b,SO,"i loc'h0,sO;' 10c'~0,s0,'I (025!hSO~]Mixture ..1 1.3Ca8O4Mixture ..,5 , a.0.3CaO,CaW,9 1 1 1, l 1111 $ 19 7 I 1 Mixture . ,8 , .*.9 , ...ccu'6,zcaojCuO,3CaOY1 11I 1 J J* Not determined.-Val. 01:ipitatt111.PercantRye ofcopper insolution.In I Inwater. dextroseIV. I v.Clinnge induxi rosein cold.VI.0.0023O*oOl%0900500000000000000000000 -24 houir34 1, % 111 hour1 11 4 11 --Change inclexti oscheated.VII.nilnilnilnil} ;:::,y1Dirtygreeu andC I I ~ O de- / posited2CII~O dc-posited 1i - ----Coloiircliangeboiliiig.VIII.0 I1~.nil 1nil 1nil 1nil 1ci% l anilnilnilnilnil* 42e .-m 2 Elnil -t That is, all the copper dissolved.-Grn1n.sof ironlissolvedIX.1 If freshly precipitated, these become greener and denser.1 Increasing in amount with increase of basicity.2 Tlie amount of green decreasing, and that of deposi+.ed oxide incrensing, with increase3 Tlie blackening increasing to a maximum a t No.4 and then decreasing.4 Tlie least and most basic precipitate exhibiting a paitial clrnnge only.of basicity.was given in the Eleventh Report of the Wob'urn ExperimentalFruit Farm. These results have been further elaborated recently,but a summary of them will be sufficient here.The members are numbered (2) to (7) in the above abbreviatedtable, and to these must be added two other basic sulphatesobtained by other means. The first four contain some looselPIcKERING : THE CONSTITUTION OF BASIC SALTS, 1853combined calcium sulphate, but the amount is so inconsiderable inthe case of the first two members, that its presence may be accidental.The proportions of Cu0:CaO in Nos.6 and 7 are uncertain;Cu0,3CaO seems to be constant in composition throughout a widerange of proportions of lime-water added, but the existence ofCuO,2CaO is doubtful; all that is certain is that some compound ofthe two oxides with less calcium than Cu0,3CaO must exist, as theprecipitate becomes destitute of sulphate before it attains the com-position of Cu0,SCaO. All the liquids after passing No. 4 arealkaline.Columns I11 to IX contain details as to the behaviour of thevarious precipitates obtained by adding increasing amounts of lime-water to the same weight of copper sulphate, the total volume ofthe mixture being the same in all cases.When first formed, whatever be the proportions taken, the pre-cipitate is the same in appearance, being bulky, and of a full bluecolour; but this soon changes to a light blue in the case of thethe two lowest sulphates (2 and 3), which, in a few days, becomegreenish-blue and more dense,The next two members permanently retain their bulky bluecharacter, and, even after being kept for many months, occupy fowand a-half times the space occupied by the same amount of copperin the form of the less basic sulphates.If milk of lime, instead oflime-water, be used for preparing the more highly basic precipitates,they are invariably violet, instead of blue, this colour being, as wehave reason to believe, distinctive of a cuprite.The various members of the series, after precipitation, may beconverted one into the other by the addition of more lime, or ofmore copper sulphate, although the conversion occupies some littletime.In column I11 is given the relative volume occupied by the variousprecipitates under similar conditions.This increases rapidly up tolOCuO,SO,, remains nearly constant to 10Cu0,S03,3Ca0, and afterdecreasing gradually for some distance, begins to fall rapidly.*The volume occupied by 5Cu0,SO3 shows that it cannot be a meremixture of the neighbouring compounds, for such a mixture wouldoccupy 5.1 volumes, instead of 6.5; another series, quoted in theWoburn Report, illustrated this point more forcibly.From column IV it will be seen that the least basic precipitatesare sufficiently soluble in water for the amount of copper in solutionto be determined by ferrocyanide ; the solubility, however, extends* The series quoted in the Woburxi Report were less complete, and the rapiddecrease was erroneously taken as starting at 10Cu0, S0,,3CaO.6 ~ 1854 PICKERING : THE CONSTITUTION OF BASIC SALTS.considerably beyond this point, rn is shown by the fact that themore basic liquids act on iron.Column IX contains the results ofstrictly comparable experiments wherein iron was left in the mix-tures for two days. I n this action, copper is first deposited onthe metal, and then the solution becomes electrolysed, hydrogenbeing evolved and oxide of iron formed.This action goes on wellbeyond the point (10Cu0,S03) at which the liquids become alkaline.I n more strongly alkaline liquids-beyond 10Cu0,S03,3CaO-theiron gradually becomes dulled by the formation of a white depositon it, which is probably a ferrite.When the liquids containing these precipitates are boiled *(column VIII), no change takes place with the lower members,beyond that of the solids becoming more compact and green, asthey do, also, on long keeping. As the basicity increases, a pointis reached where boiling causes blackening of a portion of theprecipitates, owing to the liberation of cupric oxide. With furtherincrease in basicity, more of the precipitate blackens, the whole ofit< doing so at 10Cu0,S03; then the blackeni’ng diminishes, andceases altogether at 10Cu0,S03,3Cs0.At a further degree ofbasicity another change begins, this consisting of the precipitatebecoming violet and dense, and eventually a point is reached whereboiling again produces no change.Up to a, certain point dextrose has no action, and does not affectthe solubility of the basic sulphate in the liquid (columns IVand V) ; but when a certain basicity is reached, we get, apparently,a direct action (column VI), the precipitates being reduced by thedextrose in the cold. This action is evidenced by the precipitatesbecoming in part of a dirty green colour, due to the mixture ofcuprous oxide with the blue precipitate.When a further degree of basicity is attained, an action of atotally different character occurs ; the copper dissolves in thedextrose solution (column V) ; this becomes violet, and, after beingkept for a certain time, turns yellow, owing to the spontaneousseparation of cuprous oxide (column VI).The copper, while dis-solved, is electronegative, and is absorbed by cellulose; it is present,no doubt, as cupridextrose.When the basic sulphates are heated to boiling with dextrose solu-tion, the action begins one stage earlier, column V I I ; the turninggreen of the undissolved precipitate occurs as soon its 5Cu0,S03 ispassed, and the next stage consists of this same change, togetherwith the dissolution of some of the precipitate, and the gradual* They slionld have been prepared some time before boiling, or they may behaveLime itself has no such action.anomalouslyPICKEKING : THE CONSTITUTION OF BASIC SALTS, 1855deposition of an increasing amount of cuprous oxide from thesolution: in the last stage the cuprous oxide is deposited entirelyfrom the liquid.The evidence as to the individuality of the various compoundsmay be summarised as follows:(1) 3CuO,SO,.-A greenish-blue precipitate, obtained by boilingcopper sulphate solution (Pickering, Chem.News, 1883, 47, 181).Also obtained by Shenstone (Trans., 1885, 47, 375) in a crystallinecondition by heating copper sulphate with a little water in sealedtubes at 200O; and by Friedel. Cesaro and Buttgenbach found itas a, mineral. A series of experiments was made to ascertainwhether this, or any less basic sulphate, was formed during thepartial precipitation of copper sulphate by lime, but no suchindications were obtained.(2) 4CuO,SO3,(0-06CaS0,).-A light blue precipitate, obtainedby adding to copper sulphate alkalis in any quantity up to thatsufficient to throw down all the metal; also as a crystalline pre-cipitate by decomposing copper sulphate with an a.cetate (Pickering,Zoc.cit.); and as a crystalline mineral, bronchontite. It turnsgreen on keeping or heating.(3) 5Cu0,S03,(0*25CaS0,).-Similar. in character to the former.Its appearance coincides with the point (1) at which the copperin solution ceases to be recognisable by the ferrocyanide test(column IV in the table), (2) at which cupric oxide begins to beliberated on boiling (column VIII), and (3) at which cuprousoxide begins to separate when heated with dextrose (column VII).The volumes, also, show that it cannot be a mixture of the higherand lower su1phat.e.(4) 10Cu0,S0,,1~3CaS04.-A full blue, bulky precipitatg, thecomposition of which is established by its coinciding with the pointat which the liquid becomes alkaline when alkali is added to coppersulphate; also by the action of dextrose in t.he cold, which, as soonas this point is past, begins to act on the precipitate, liberatingcuprous oxide from it (column VI), or, when boiled with it, begins todissolve some of the copper (column VII).The volumes (column111) also indicate its existence.( 5 ) 10CuO,S0,,3CaO,CaSO,.-Composition proved by analyscsin former communication, and also as being the point (1) up towhich some copper is still in solution, as shown by the action ofiron (column IX) ; (2) at which copper begins to dissolve in dextrosein the cold to form a violet solution, depositing cuprous oxide(column VI), and (3) at which boiling ceases t o liberate cupric oxide(column VIII)1856 PICKERING : THE CONS'l'ITUTION OF BASIC SALTS.According to the analyses previously quoted, the composition ofthis precipitate remains constant with increasing amounts of lime,till the excess of the latter reaches a certain limit (about the point5a in the table).The change in volumes (column 111) and theresults on boiling (column VIII) also indicate that some other, asyet unidentified, compound makes its appearance at this point.(6) Cu0,2CaO.-Composition uncertain, see above.(7) Cu0,3CaO.-Approximate composition of the precipitateobtained on adding from 50 to 500 CaO to each molecule of coppersulphate.(8) 15CuO,SO,.- A precipitate obtained by decomposingcuprammonium carbonate and sulphate with water (Pickering,Trans., 1909, 95, 1417).Developing a scheme for the representation of these various sub-stances by the introduction of successive CuO groups into themolecule of copper sulphate, the introduction of one and two suchgroups gives the members A2 and A3, of which the latter is thelowest basic sulphate known, and is really the orthosulphate, theformer being unknown, but analogous to the so-called monohydrateof the sulphate, in which, there can be little doubt, the water is notordinary water of crystallisation.The introduction of further CuOgroups would give the three members of the B series, and of thesethe first two have been isolated. The members of both these seriesshould all exhibit similar properties, inasmuch as the copper inthem is in the same condition. This is the case, for they all reactwith iron or ferrocyanide, and are not decomposed by dextrose :Not fully examined.*Al. 8 2 . A3.Not knownCUSO,. (analogous to OaS04,H,0). 3CuO,SO,.* It was suggested (Trans., 1907, 91, 1995) that 10CnO,S0,,10CaO,SO, mightexist; this was based on the view, which is now unacceptable, that the highermembers of the series were molecular compounds of two basic sulphates of copperand calcium, the compound No.5 (with the calcium sulphatc added to it) being10Cu0, SO,, 4Ca0, SO,PICKERISG : THE CONSTITU'TION OF BASIC SALTS. 1857B1. 132. B3.ocu>os &ll >cu\-oc2.>o>o,. ocu*ocu--ocu*ocu//ocu*ocus --\.X : O C u 'ocu>oF2.p > o/ocu>o";\:>cus q o c u5cu0,s0,. Not known.or S1OCU0, so,.15Cu0, SOS1858 PICKERING : THE CONS'I'I'I'UTION OF BASIC SALTS.The introduction of further CuO groups may be effected in eitherof the ways indicated in C2, in the first of which the copper is dyad,and in the second, tetrad, as in the case of the cupri-salts. Thesecond member of this series, containing two pairs of such CuOgroups, corresponds with the next basic sulphate isolated. Theready separation of cupric oxide by heat from this compound(column V I I I in the table) would be explained equally well byeither of the formulze, but its action on dextrose (column VII),analogous to that of the cupricitrates, seems to be best expressed bythe second formula.Its small solubility, and feeble action on iron(column IX) is also more in consonance with this formula, whichrepresents the presence of only -OCU>()-0cu one whereas,according to the first formula, there are three such groups present,and such a compound would be expected to act at least asenergetically as the members of the B series.By the further action of an alkali, the external CuO groupswould become converted into " cuprite " groups, as indicated inthe alternative formulze F2, which represent the next member ofthe series isolated.Such a substance would be analogous to they-cupri-compounds, and, like these, it dissolves in dextrose solution,forming a blue liquid, from which cuprous oxide is deposited (columnVII), and which combines with cellulose. The basic sulphateobtained by the decomposition of cuprammonium sulphate fits inas the third member of this series, F3.Although the members of the F series are less stable in thepresence of dextrose than are those of the preceding series, inconsequence of their containing the cuprite groups, they are morestable under the action of heat, the external CuO groups beinghedged in, as it were, by the outlying electropositive element.Theremarkable differences in behaviour of the various basic sulphateson being heated is thus explained.The series F is derived from the series C by the introduction of3M"O, and there might be two intermediate series with MI10 and2M"O, respectively. That some other compounds exist which havenot been isolated is almost certain, as has already been pointed out.The final products of the action of lime on copper sulphate seemto be simple cuprites, and the oxidising action of these on dextroseis still more energetic than that of highest basic sulphate.Two members of the sulphate series, C2 and F2, contain a con-siderable proportion of combined calcium sulphate ; the presenceof this can be explained if the alternative formulz areadopted, but not otherwise, the calcium sulphate becoming aPICRERING : THE CONSTITUTION OF BASIC SALTS.1859orthosu1phat.e by connexion with t.wo :CuO groups of the quadri-valent copper : -cn-' \ ...................................E>b<u,>ca....... / .......................The maximum amount of sulphate thus capable of being introducedwould be two molecules, and this is the amount actually found whenthe basic sulphate F2 is precipitated in the presence of excess ofsodium sulphate, its composition then being10Cu0,S03,3Ca0,2 (Na,Ca) SO,Ferrit es.How far the explanation which seems to be satisfactory with thecopper compounds will apply to those of other metals remains tobe seen; but, in the case of iron, Fjome evidence has been obtainedthat an analogy exists.I n a paper on emulsions (Trans., 1907, 91,2001) it was mentionedthat when ferrous sulphate is precipitated by excess of lime, andchurned up with paraffin, the greenish-black precipitate of basicsulphate often becomes quite white, or does so when the emulsion iskept for some time.This was attributed to the chemical actionof some impurity in the oil; but it has since been found that thesame change occurs, although much more slowly, when the basicsulphate is left with milk of lime only, or even with excess of clearlime-water, decolorisation in that case requiring many days tobecome complete. The white substance retains the flocculentcliaracter of the basic sulphate, and remains quite unchanged ina closed vessel, but when exposed to the air it very graduallybecomes orange, through the formation of ferroso-ferric oxide.Ondigestion with a solution of dextrose, a small portion of it dissolves,and in this the iron is in the electronegative condition, as it givesno blue colour with ferrocyanide until acid is added.To obtain some information on the subject, quantities of onelitre of lime-water, diluted with 20 per cent. of water, were mixedwith different amounts of ferrous sulphate, and left in closedbottles for six weeks, after which the lime still in solution wasdetermined. I n this way the molecular proportions of lime used up(loc. cit.)1860 PICKERING : THE CONSTITUTION OF BASIC SALTS.in the precipitation of the iron were determined, and these are givenin t'he accompanying table.The proportions of lime to iron takenare also given, and, as will be seen, the lime in all cases was in excess.It was only in the last three experiments that the greenish-blackbasic sulphate became white; it very nearly did so in No. 4, whilstin No. 3 there was a smaller, but st.ill considerable, lightening incolour. Taking the last three experiments, where the change wascomplete, the proportion of lime used up is evidently in excess ofthe one molecule which would suffice for the complete decom-position of the ferrous sulphate, so that some of the lime must havegone into combination with the ferrous oxide. Omitting the lastexperiment, as being uncertain owing to the very small proportionsin which the iron had t o be taken, the excess of lime used is0'27Ca0, which would indicate the formula of the compound as4FeO,CaO, if no sulphur is' left in it; this, however, was not thecase, for, in No. 5, there was found to be O*153SO3 present to eachatom of iron, and in No. 6, 0'0550,; this may be present simplyas calcium sulphate, but, if it is combined with the iron in theform of a basic sulphate, the proportions of FeO to CaO in theprecipitate would be less tlian the 4 : 1 mentioned above.Precipitation of Ferrous Sulphate by Lime.Proportions taken.Fe : CaO.1. 1 : 1.0332. 1 : 2.073. 1 : 5.154. 1 : 10.35. 1 : 25236. 1 : 51.57. 1 : 103.0Proportions used up.Fe : CaO.1 : 0'3341 : 0'5431 : 0.6411 : 1'0301 : 1.2321 : 1.3111 : 1'116Further work would be required, of course, before the natureof the substances here present could be established, but it is evidentthat ferrous compounds analogous t o the alkaline basic salts ofcopper or to the cuprites do exist, and it is noticeable that theirformation seems to necessitate approximately the same large excessof alkali, for with the iron compound between 10 and 26 equivalentsof lime to each equivalent of iron are required, whilst in the case ofcopper, about 25 equivalents were necessary (loc. cit.).The solubility of basic metallic salts of this description in organicsubstances probably plays an important part in their assimilation byplants.MARPENDEN

ISSN:0368-1645    

DOI:10.1039/CT9109701851

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

INTERNATIONAL COMMITTEE ON ATOMIC WEIGHTS, 1911, 1861The Council has ordered the following letter and report to beprinted in the Journal and Proceedings of the Society :IMPERIAL COLLEGE OF SCIENCE AND TECHNOLOGY,SOUTH KENSINGTON,LONDON, S.W.August 22nd, 1910.GENTLEMEN,I beg to forward the Annual Report of the International Com-mittee on Atomic Weights for 1911, to which I have appended, bytheir desire, the signatures of Professors Ostwald and Urbain.The Committee, it will be seen, have acceded to the wish of theCouncil of the Chemical Society, and have prepared the Report insuch time that it can be published prior to the commencement ofthe ensuing academic year.Slight changes have been made in the atomic weights of argon,helium, krypton, lithium, neon, phosphorus, platinum, strontium,vanadium, and xenon, but otherwise the new table remains verymuch as in 1910.I have the honour to be, Gentlemen,Your obedient servant,T.E. THORPE.To the Hon. Secretaries,The Chemical Society,Burlington House,London, W.Report of the International Committee on Atomic Weights, 191 1.I n the autumn of 1909 the Council of the Chemical Society ofLondon voted unanimously in favour of issuing the annual reportof the International Committee on Atomic Weights in September orOctober instead of in January as heretofore. In that propositionthe Chemical Society of France has concurred, and Americansentiment has also been favourable to the suggested change. There-fore the change is now made.The reasons offered for the new policy are very simple.First, theschool year, at least in most educational inst.itutions, begins in theautumn. It is desirable that teachers should then have the lates1862 REPORl' OF THE INTERNATIONALtable of atomic weights at their command, in order to avoid changesafter school work has begun. Secondly, publishers of text-booksare accustomed to issue their new works in the autumn, and oftenrequest early information as to changes which are likely to bemade. The proposed chaiige in the time of issuing the table istherefore an aid to teachers, students, and publishers, and no dis-advantage to anyone else. The immediate usefulness of the tableis increased, and to attain that end should be the main purposeof the Committee.Since the preparation of the report €or 1910, a, number ofimportant memoirs upon atomic weights have appeared.Theresults obtained are, in brief, as follows:ChZorine.-The density, composition by volume, and compressi-bility of hydrochloric acid have been measured by Gray and Burt(Trans., 1909, 95, 1633) with great care. From the density andvolumetric composition, when H = 1.00762, C1= 35'459. From thedensity and compressibility, C1= 35.461. The mean, 35.460, is thevalue given in the annual table of atomic weights for the past twoor three years.The density of hydrochloric acid has also been determined byScheuer (Zeitsch. physikal. Chem., 1909, 68, 575), who givesmeasurements made under varying conditions. His final con-clusi'on, based upon his own work after comparison with that of Grayand Burt, is that C1=35.466.Lithium.-Richards and Willard ( J .Amer. Chem. SOL, 1910, 32,4), in their important research upon the atomic weight of lithium,measured three distinct ratios, namely, silver to lithium chloride,silver chloride to lithium chloride, and lithium perchlorate tolithium chloride. From these ratios, without the intervention ofany others, the following independent values for three atomicweights are obtained :Li = 6.939.GI = 35'454.Ag=l07*571.The value for silver varies from the accepted value, 107.88, byabout one part in 12,000, which is probably less than the actualuncertainty. That for chlorine diverges more widely, namely, byabout oiie part in 6000. The new figures are undoubtedly entitledto great weight, but in view of the excellent work done by othersit would be unwise to make any hasty change in the table.Forlithium, however, the value 6-94 may be taken, replacing the old7-00.Strontium.-Thorpe and Francis (Proc. Roy. SOC., 1910, 83, A COMMITTEE ON ATOMIC WEIGHTS, 1911. 1863277), in their determinations of the atomic weight of strontium,measured six ratios, and obtained the following results :Ratio 2Ag to SrBr, ............ Sr=87*645,, 2AgCl to SrCI, ......... ,, = 87 ‘645,, 2AgEr t o SrBr, ......... ,, =87’653,, 2Ag t o SrCl, ............ ,, =87-642,, SrBr, to SrSO, ........... ,, =87*629,, SrC1, to SrSO, ............ ,, =S7.661Sr = 87*646--Mean of all .........The value adopted by the authors is 87-65.Richards’s figure is87.62.Phosphorus.-Atomic weight redetermined by Baxter and Jones(J. Amer. Chem. SOC., 1910, 32, 298). From the ratio betweensilver and silver triphosphate, the authors find P = 31.043, whenAg=107.88.Vanadium.-From the ratio between silver chloride and vanadyltrichloride, Prandtl and Bleyer (Zeitsch. anorg. Chem., 1910, 65,152) find V=50.963 and 51.133 in two series of experiments. I na later paper, Prandtl and Bleyer (Zeitsch. anorg. Chem., 1910,67, 257), also from analyses of vanadyl trichloride, find V = 51.061.From reductions of V,O, to V203, they found V~51.374. Thelatter method, however, they regard as uncertain. The valueV = 51.06 may be provisionally adopted.Tellurium.-Marckwald and Foizik (Ber., 1910, 43, 1710 ; seealso Browning and Flint, Amer.J . Sci., 1909, [iv], 28, 347, whoadduce evidence to show that tellurium is possibly complex), by asomewhat complex volumetric process, based on the oxidation ofTeO, by KMnO,, conclude that Te=127*61. This agrees withmany of the other recent determinations of the constant, but is notsufficiently exact to supplant the value given in the table.Rhodium.-Two inaugural dissertations upon the atomic weightof rhodium have been issued from Gutbier’s laboratory at Erlangen.Renz reduced rhodium pentamine bromide in hydrogen and foundRh = 102-92. H. Dit.tmar (reproduced in Sitaungsber. phys. med.Soz. Erlangen, 40, 184), by similar reductions of the correspond-ing chloride, found Rh = 102.93.Platinum.-The very elaborate investigation of Archibald (Proc.Roy.SOC. Edin., 1909,29, 721) upon the atomic weight of platinumwas based upon analyses of the chloroplatinates and bromoplatinatesof potassium and ammonium. I n all, 28 ratios were measured,giving values for Pt ranging between 195.19 and 195.25. Theirarithmetical mean gives Pt = 195.22. Archibald, however, in hisAn intermediate value, 87.63, is adopted in the new table.The rounded-off figure 31.04 is to be adopted1864 REPORT OF TEE INTERNATIONALfinal discussion, uses only 12 ratios, giving, in mean, Pt=195-23.The figure 195.2 is given in the fable.The Inert Gases.-The densities and molecular weights of heliumand neon have been redetermined by Watson (Trans., 1910, 97,810).For the atomic weights he finds He = 3.994 and Ne = 20.200.I n another paper (ibid., 97, 833) he applies the critical constantsof krypton and xenon to their densities as determined by Moore,and finds Kr=82.92 and Xe=130-22. There are also new deter-minations of the density of argon by Fischer and Hehnel (Ber.,1910, 43, 1435). Their mean value, referred to 0=16, is 19.945,a figure rather higher than that given by Ramsay and Travers. Itcorresponds to an atomic weight of A = 39-89.It is also to be noted that a third, revised edition of Clarke’s‘ I Recalculation of the Atomic Weights ” has recently been publishedby the Smithsonian Institution.The annual table of atomic weights for 1911 follows, with butfew changes from that of the preceding year.F.W. CLARKE.W. OSTWALD.T. E. THORPE.G. URBAINCOMMITTEE ON ATOMIC WEIGHTS. 19 1 1 .1911 .In t erncct ionat? Atomic Weights .Alumiiiiuin ................. A1Antimony ..................... SbArgon ...................... AArsenic ..................... AsBarium ........................ BaBismuth ..................... BiBoron ........................ P;Bromine .................... BrCadmium ..................... CclCasiuiu ....................... CsCalcium ........................ CaCarbon ........................ CCerium ........................ CeChlorine ..................... C1Clironiiiiin .................. CrCobalt ....................... CoColumbium ................. CbCopper ........................ CuErbium ........................ErEuropium ..................... ELIFluorine ..................... FGadolinium .................. GdGallium .................... GaGermanium .................. GeGluciiiuin ..................... GIGold ........................... AuHelium ........................ HeHydrogen ..................... HIndium ....................... InIodine ........................ IIridium ....................... I rIron ........................... FeKrypton ..................... IirLanthanum .................. LaLead ........................... PbLithium ..................... LiLutecium .................... LuMagnesium Md,oManganese .................. MnMercury Hg................. Dysprosium DY.......................................0 = 16 .27.1120.239.8871.96137.37208.011.079'92112.40132.8140.0912.00140.2535.4658.058.9793.563.57162.5167'4152.019 .0157.369.972 59.119'1'23-991 '008114.8126-92193-155.8582.9139-0207.106 '94174'024-3254-93200 . 01865Molybdenum ............... MoNeodymium ................. NdNeon ........................... NeNickel ........................ NiNitrogen ..................... NOsmium ..................... 0 sOxygen ........................ 0Palladium ..................... PJPhosphorus .................. PPlatinum ..................... PtPotassium ..................... I<Praseodymium ............... PrRadium ........................ RaRhodium .....................XhRubiclinm ..................... RbRutlieniuin .................. RuSaniariuni ..................... SaScandium ..................... ScSelenium ..................... SeSilicon ........................ SiSilver A,aSodium ........................ NaStrontium .................. SrSulphur ..................... STantalum .................... TaTellurium ..................... TeTerbium ..................... TbThallium ..................... T1Thorium ..................... ThThulium : .................... TmTin ........................... SnTitanium ..................... TiTungsten ..................... WUranium ..................... UVanadium ..................... VXenon ........................ XeYtterbium (Neoytterbium) YbYttrium ..................... YZinc ........................... ZnZirconium ..................... Zr........................0 = 16 .9 6 0144-320-2586814-0116-0031.0439'10190.9106.7195'2140-6226-4102-9101.7150'444-179-22 8 3107-8823.0087'6332.0785.45181'0127.5159'2204'0232.0168.5119.048.1184 *O238-5130'2172089.065-3790.651.06YOL . XCVII . 6

ISSN:0368-1645    

DOI:10.1039/CT9109701861

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

1866 MILLS AND BAIN: OPTICALLY ACTIVE SALTS OFCXC1X.- Optically Active Salts of 4-Oxim~nocyclo-hexanecarboxylic Acid ar2d the ConJiyuration ofthe Oximino-Group.By WILLIAM HOBSON MILLS and ALICE MARY BAIN.THE theory which Hantzsch and Werner put forward (Ber., 1890,23, 11) to account for the isomerism of the oximes provides SOsatisfactory and consistent an explanation of the facts that it hasreceived practically universal acceptance. The fundamental assump-tion, however, on which this theory is based-namely, that in theoximino-group the three valencies of the nitrogen atom are not inone plane-has rested up to the present on indirect evidence only.It seemed to us that it might be possible to put this hypothesisto direct experimental test by investigating an oximino-compoundso constituted that its molecule would possess a plane of symmetryor not according as the oximino-group has a plane or a, trihedralconfiguration.This condition is fulfilled by compounds of the type:“>c: C:NOH, bwhich might be regarded as oximes of the unsymmetrical ketens :;>c:co.A molecule of the configuration (I) is clearly superposable on itsmirror-image, whilst one of configuration (11) is not,n b\,, ..**’CA\/(I.) c/ ”.\..., \ ,.aiNa bc \,,/“/\\/(11.) c / ‘.,\ /.,?N\013IO HI n these diagrams the dotted lines are intended t o indicatevalencies lying behind the plane of the paper, while valencies infront of it are represented by thick lines4-OXIMINOCYCLOHEXANECARBOXY LIC ACID, ETC.1867A substance of configuration (11) stands in close relationship taallene derivatives of the type :">c: b c:c<i,the molecular asymmetry of which was first pointed out by van't Hoffin 1875 (La Chim/ie duns EJE7space, p. 29), for, provided that theoximino-group has the configuration assumed by Hantzsch andWerner, it can be substituted for one of tphe carbon atoms of suchan allene compound, as shown in the diagram, without destroyingthe asymmetry of the molecule:D C HO NThe preparation and manipulation of a compound of this typewould doubtless offer even greater difficulties than those whichhave hitherto prevented the experimental realisation of van't Hoff '8prediction in its original form (see Dimroth and Feuchter, Ber.,1903, 36, 2238; Lapworth and Wechsler, Trans., 1910, 97, 38).These difficulties can, however, be avoided by means of the samedevice as has been successfully employed by Perkin and Pope(Trans., 1908, 93, 1075; Perkin, Pope, and Wallach, Trans., 1909,95, 1789; compare Marckwald and Meth, Ber., 1906, 39, 1171)in order to obtain a compound (methylcyclohexylidene-4-acetic acid)which reproduces the essential spatial characteristics of the allenederivatives without having their instability-that is, by the expan-sion of the two-membered ethylene ring into the hexamethylenering.In place of keten oximes, one then has the oximes of sub-stituted cyclohexanones of the types :If the three valencies of the nitrogen atom in the oximino-group6 a 1868 MILLS AND BAIN: OPTICALLY ACTIVE SALTS OFare not in one plane, then the oxime of such a ketone must consistof an equimolecular mixture of two enantiomorphous forms :R R R R'CH,/' CH2 CH,/)CH,UH,/ ICH, CH2\,CH2,C\/C/ Hu\OHIf, on the other hand, they are in one plane, these oximes, likethe ketones from which they are derived, will have a plane ofsymmetry, and will therefore be incapable of exhibiting opticalactivity.Hence, if a compound of such constitution can be obtainedin an optically active condition, it will follow as a necessary con-sequence tliat the three valencies of the nitrogen atom in theoximino-group do not lie in one plane.I n order to facilitate resolution, it is desirable that the substanceshould contain the carboxyl group.The simplest compound of thenecessary type is therefore :C0,II H > ~ < C H z * C H 2 > ~ : CH,*CH, NOH.This is the oxime of the important cyclohexanone-4-carboxylicacid which was syntliesised by W. H. Perkin, jun. (Trans., 1904,85, 416), and served as the basis of his well-known syntheses ofterpin, terpineol, and dipentene (Trans., 1904, 85, 654). It hasalso been obtained by Lumsden (Trans., 1905, 87, 87) by thereduction of anisic acid. It is shown in the present communicationthat this oximino-acid (or rather its salts) is, in fact, capable ofexisting in two enantiomorphously related forms, the separation ofwhich has been effected with the aid of morphine and quinine.I f a molecular proportion of morphine is added to the ethyl-alcoholic solution of the acid, a morphine salt separafes whichcontains a dextrorotatory form of the acid, for on decomposing itwith excess of ammonia and suitably removing the morphine, theammonium salt which remains in the solution is strongly dextro-rotatory.The ammonium salt prepared in this manner has amolecular rotation [MI, in aqueous solution varying from about50° to 60° in different preparations. By means of quinine thecorresponding lzevorotatory ammonium salt has similarly beenobtained, the salt which the I-acid forms with the alkaloid being i4-OXIMINOCYCLOH EXANECAHBOXY LIC AC1 D, ETC. 1869t'his case the more sparingly soluble of the two diastereoisomerides.This lzvorotatory ammonium salt has a molecular rotation of[MI, -70° to -SOo in aqueous solution; the separation effectedby quinine is thus more complete than that attained with the aidof morphine.The activity of these salts is evanescent, and its rateof disappearance falls off in accordance with the unimolecularformula, the velocity constants being very different for differentsalts of the acid. The rate of racemisation is greater, in the saltsinvestigated, the weaker the base from which the salt is derived,and the presence of an excess of alkali, which in so many casesaccelerates racemisation, here greatly retards it, while acidificationcauses the rapid disappearance of the rotatory power. Thus themorphine salt of the d-acid when dissolved in water shows muta-rotation, obviously on account of the rapid diminution of thedextrorotation due to the acid, the period of half-change being aboutone minute.For an approximately 0*2N-solution of the ammoniumsalt in water, the time of half racemisation is thirteen minutes, andfor the sodium salt under similar conditions twenty-four minutes.The presence of ammonia in N / 10-concentration increases the per-sistence of the activity of the ammonium salt approximately forty-fold, raising the period of half-change to 8.5 hours, and a corre-sponding addition of sodium hydroxide to the solution of the sodiumsalt has a still greater relative cffect on the rate of racemisation,lengthening the period of half-change from twenty-four minutes totwenty-two hours.The rapid racemisation of the salts of the acid with weak basesrenders intelligible the results observed in preparing the morphineand quinine salts from the inactive acid.In the case of the lattersalt, for example, the crystals which separate yield a lzvorotatoryammonium salt after decomposition with ammonia and removalof the quinine, but there is no corresponding amount of quinined-acid salt in the mother liquor, the solution of ammonium salt,obtained after removal of the quinine from the latter, being quiteinactive. However, on concent.ration of the mother-liquor, anothercrop of quinine Z-acid salt is deposited of similar activity to the first,and so on. The behaviour of the morphine salt is analogous-notonly the first crystallisation, but also the subsequent crops obtainedon concentration of the alcoholic mother liquor all yield solutdons ofdextrorotatory ammonium salt after removal of the morphine.The process therefore appears to be one of activation rather thanof simple resolution, and recalls the behaviour of the methylethyl-propyl tin d-camphorsulphonate and bromocamphorsulphonatedescribed by Pope and Peachey (Proc., 1900, 16, 42, 116).It seemed inevitable from the foregoing that any attempts t 1870 MILJS AND BAIN: OPTICALLY ACTlVE SALTS OFisolate the optically active forms of the acid itself would be unsuc-cessful.It was therefore necesmry, in order to make these experi-ments as conclusive as possible, to obtain some definite proof thatthe substances, to which these optically active solutions owe theirrotatory power, are indeed the normal ammonium salts of dextro-and lzevo-rotatory forms of oximinocyclohexane-4-carboxylic acid.This was done by treating the active solutions with an aqueoussolution of silver nitrate. The suhstances thus precipitated, fromthe dextro- as well as from the lzvo-rotatory solutions, were foundto be normal silver salts of this a-cid in analytically pure condition.Further, they were optically active, for on treating them with anaqueaus solution of sodium chloride, the solutions of sodium saltobtained were respectively dextro- and lzevo-rotatory: the rotationsobservea corresponding with molecular rotations [MID 74’5O and - 79’9O respectively.These results show conclusively that the salts of this oxime-acidare capable of existing in two enantiomorphously related forms.Moreover, the evanescent character of the optical activity makes itcertain that their molecular asymmetry is determined in somemanner by the oximino-group.It remains to be consideredwhether any method of accounting for the asymmetry is possibleother than that indicated a t the beginning of this paper.Apparently the only other hypothesis by which it might, at firstsight perhaps, appear explicable is that the compound can exist inthe two tautomeric forms (111) and (IV), and that the opticallyactive salts are derived from the latter, which contains an ordinaryasymmetric carbon atom :(111.) (IV.)This explanation, however, cannot be maintained. If the activecompound has the formula (IV), it is a @substituted hydroxyl-amine, and must possess the power of reducing Fehling’s solution,which is characteristic of compounds of this class.A comparison ofthe behaviour of an active solution in this respect with that ofacetoxime on the one hand, and of P-phenylhydroxylamine on theother, showed that although its reducing power was slightly greaterthan that of acetoxime, it was of an altogether different order fromthat of phenylhydroxylamine, and it is quite certain that noappreciable quantity of a, P-substituted hydroxylamine .could havebeen present in the solution.The conclusion therefore seems unavoidable that the molecularasymmetry, to which these salts owe their optical activity, is deter-mined by the peculiar configuration of the doubly linked tervalen4-OXIMIWOCYCLOHEXANECARBOXYL~C ACID, ETC.1871nitrogen atom. They accordingly constitute a new type of opticallyactive substances ; and through their optical activity a direct experi-mental proof is afforded that tho three valencies of tervalentnitrogen, however they may be disposed in a, compound of the typeN-b, /" are directed in the oximino-group, as was postulated by\cHantzsch and Werner, along the three edges of a trihedral angle.These substances may also possess a certain interest in that theyprovide another example of compounds of which,_like d- andZ-inositol,* and d- and Z-l-methylcyclohexylidene-4-acetic acids(Perkin, Pope, and Wallach, Zoc. cit.), the molecular asymmetry ismore fittingly defined with reference to the configuration of themolecule as a whole than expressed in terms of the presence of somo" asymmetric atom."EXPERIMENTAL.The cyclohexanone-4-carboxylic acid employed for the preparationof the oxime the properties of which form the subject of this com-munication, was synthesised according to the method described byW-.H. Perkin, jun. (Zoc. cit.), and Kay and Perkin (Trans., 1906, 89,1640). The acid was converted into its oxime by mixing its solutionin ten parts of absolute alcohol with one molecular proportion ofhydroxylamine hydrochloride dissolved in about an equal volumeof the same solvent, and adding one molecular proportion ofanhydrous sodium acetate. After three days an equal volume ofdry ether was added, the precipitated sodium chloride was removed,and the solution then evaporated, finally in an exhausted desiccator.The oximineacid was extracted from the solid residue in a Soxhletapparatus, and purified by recrystallising from dry ether in thesame manner.It separates from ether in crystalline-crusts, meltingat 148--148*5O, and otherwise agreeing with Perkin's description(loc. ckt.).Quinine Salts of Oxim~nocyclo?~exane-4-ca~~bozylic Acid.A quinine salt containing a preponderance of the lzvorotatoryform of the acid may be obtained by heating the solution of theinactive acid in 30 parts of water with 2.1 parts (slightly morethan one molecular proportion) of anhydrous quinine, the smallexcess of quinine being removed by filtration of the hot l i p i d .The salt separates on cooling, usually in hemispherical clusters of* The configuration of the active inositols was, apparently, first given in Stereo-chcmie (van't Hoff-Meyerlioffer, 1892), p.91 ; see also Bouveaalt, Bull. SOC. him.,1894, [iii], 11, 1451872 MILLS AND BAIN: OPTICALLY ACTIVE SALTS OFfine needles.water of crystallisation :After being air-dried, it contains 24 molecules of1.2756 air-dried salt lost 0.1070 H20. H20 = 8.39.1.0693 ,, ,, ,, 0.0894 H,O. H20=8.36.C27H3,0,N,,2~Et20 requires H20 = 8.56 per cent.Under the above conditions, approximately 80 per cent. of thetotal quantity of salt crystallises out. Recrystallisation can beconveniently effected from 10 parts of hot water.The investigation of the optical activity of the acid component ofthe salt was carried out by placing a weighed quantity of the salt(0.3 to 0.5 gram) in a small separating funnel, and adding 10 C.C.ofan aqueous solution of sodium or ammonium hydroxide of suchconcentration that the amount of alkali left after decompositionof the salt would give an approximately N/lO-solution. The quininewas then completely removed by extracting three times with chlorGform, 12-15 C.C. of chloroform being used for each extraction. Thealkaline solution of the sodium or ammonium salt was then filtered,and after washing the separating funnel and filter with successivesmall quantities of N/lO-sodium or ammonium hydroxide, as thecase might be, the mixed filtrate and washings were examinedpolarimetrically. For example, 0.66 gram of air-dried quinine saltwas treated with an aqueous solution (10 c.c.) of 0.12 gram ofsodium hydroxide. After proceeding in the above manner, thefollowing polarimetric observation was made, the final volume ofthe aqueous solution being 14.00 C.C.:I =2, c * = 1.606, U, - 1-47', [MID - 81.9'.On examination of the various preparations of quinine salt inthis manner, it was found that the sodium or ammonium saltsobtained from them possessed molecular rotations lying for the mostpart between [MI, -70° and [MIn -80'.It was naturally useless to attempt to obtain the optically purequinine I-acid salt by repeated recrystallisation in the usual way,since it is clear from the observations described below (pp. 1880,1881) that even with the most rapid working it must be impossibleto avoid racemising the salt extensively with regard to its acidcomponent every time that it is brought into solution.The con-ditions were rather to be sought under which the pure salt wouldcrystallise out directly free from its diastereoisomeride. Theseshould presumably be such as would secure that the rate ofracemisation should be a maximum compared with that of crys-tallisation, since in this manner the concentration of quinine I-acid* The concriitratioiis given tlirongliout this paper refer to the sodium or am-Tliey nie calciilated from the monium salt present in tho solution examined.amount of quinine or other salt taken4-OXIMINOCYCLOHEXBNECARBOXYLIC ACID, ETC. 1873salt would be kept as high, and that of quinine d-acid salt as low aspossible during the crystallisation.Accordingly, it was endeavouredto diminish the rate of crystallisation by slow cooling and acceleratethat of racemisation by acidification, and, in fact, although theexpekiments in this direction are not complete, the salt of highestactivity (with regard to the acid component) hitherto observedwas obtained by crystallisation from a solution somewhat stronglyacidified with acetic acid, the sodium salt prepared from it havinggiven the following numbers :Z=2, ~=0'727, a, -0*74O, [MI, -91.0'.I n preparing the quinine Z-acid salt from inactive acid, it is foundthat the excess of the dextrorotatory form of the acid, which shouldbe present in the mother liquor in order to correspond with theexcess of the laevorotatory form removed from the solution in com-bination with quinine, has disappeared, for the mother liquor (atany rate if a short time is a.llowed to elapse before its separationand examination) contains the quinine salt of the inactive acidonly, since after the addition oi a small excess of sodium hydroxidesolution and removal of the quinine by chloroform, the solutionof sodium salt obtained is inactive.On concentrating themother liquor, the second crop of quinine salt obtained containsagain an excess of the I-acid salt. For example, after converting3.2 grams of inactive acid into quinine salt in the manner describedabove, the crop, which crystallisea directly, weighed 8.5 grams(about 80 per cent.of the total quantity of salt produced) and gaverise to sodium salt of molecular rotation [MIL, - 71*5O (c = 0.731).The second crop, obtained by concentrating the mother liquor,weighed 0.96 gram (another 10 per cent. of the total quantity ofsalt), and gave rise to sodium salt of molecular rotation [MI, - Slago(C = 1.607).These results are clearly due to the racemisation phenomenawhich have already been mentioned. Of the two diastereoisornericquinine salts, that of the I-acid must be the less soluble in water,and thus crystallises first from an aqueous solution containing equalquantities of the two salts. The excess of quinine &acid salt therebyleft in solution, however, racemises very rapidly, so that in spiteof the removal of the quinine I-acid salt, an approximate equalityis maintained between the quantities of the two salts in the solution.An excess of the quinine Z-acid salt is accordingly depositedthroughout the process of crystallisation.I n our earliest experiments (Proc., 1909, 25, 177) the quinineI-acid salt was obtained by combining the inactive acid with quininein ethyl acetate solution and allowing crystallisation t o take placefrom that solvent.This procedure was abandoned, since it wa1874 MILLS AND BAIN: OPTICALLY ACTIVE SALTS OFfound that this salt could be obtained more rapidly and certainly,and, moreover, in a state of higher optical activity, by using wateras solvent. These experiments were, however, of Borne interest,since by means of them not only the quinine I-acid salt, but alsoin some cases a, salt containing a preponderance of the diastereo-isomeric quinine d-acid salt was obtained.Thus, in one experiment,5.4 grams of quinine (anhydrous) were dissolved in 300 C.C. of dryethyl acetate, and to the boiling solution 2-61 grams (one molecularproportion) of the finely powdered acid were added. The quininesalt formed separated very slowly. After five days it was collected,and a portion dried in a vacuum was polarimetrically examined insolution in ethyl acetate, racemisation taking place in this solventfar more slowly than in water or in alcohol:I = 2 ; C = 1.6512 ; a: - 3-24'; [a]: - 98.1".Since the specific rotation in ethyl acetate of the quinine salt ofthe inactive acid (prepared by dissolving 0.2732 gram of anhydrousquinine and 0,1324 gram of the acid in 25 C.C. of ethyl acetate)was found to be as follows:Z=2; ~ = 1 * 6 2 2 4 ; a: -4.28'; [ c c ] ~ -131.9'the acid component of the salt was clearly markedly dextro-rotatory.On recrystallisation of this preparation from ethylacetate, some of the dextrorotatory power was, however, lost, thespecific rotation having risen to [a] - 106.5O. The dextrorotatoryammonium salt obtained from this quinine salt was examined withthe following result :Z=2, ~=0*947, c ~ D 0.44'9 [MID 40'4'.These results recall the interesting observations recently madeby McKenzie and Clough (Trans., 1909, 95, 783) on the crys-tallisation of the morphine salt of phenylchloroacetic acid, a givenethyl-alcoholic solution of morphine (1 mol.) and r-acid (2 mols.)depositing either morphine d-acid or morphine I-acid salt, accordingto the exact conditions under which crystallisation takes place.Silver 1 : 4-Oximinocyclohexanecarboxylate.-To the solution ofammonium Z-acid salt, obtained by adding a solution of 0,348 gramof ammonia in 14 C.C.of water to 6-61 grams of quinine 2-acid salt,and removing the quinine by extracting with chloroform, an aqueoussolution of 2-75 grams of silver nitrate wa8 added. The denseprecipitate of silver salt thus produced was collected, thoroughlywashed with water, rapidly passed out on a porous tile, and thendried for five days over sulphuric acid in an exhausted desiccator,the whore of the operations being carried out in non-actinic light.2-20 Grams of silver salt were thus obtained in the form of a whitepowder, which apparently can be preserved indefinitely if protecte4-OXIMINOCYCLOHEXANECARBOXYLIC ACID, ETC.1875from light. On being gently heated, it decomposes almostexplosively, but if mixed with a large proportion of powdered cupricoxide, its analysis presents no special difficulty :0.3860 gave 0.4542 CO, and 0.1368 H;O.0.27830.2978 ,, 0.1616 AgC1. Ag=40*85.(.?= 32.09 ; H = 3.94.N=5.37. ,, 13.35 C.C. N, at 24O and 760 mm.C,H,,O,NAg requires C = 31.82 ; H = 3.79 ; N= 5.30 ;Ag = 40.90 per cent.This silver salt is too sparingly soluble to allow of its rotatorypower being measured directly. It was, however, possible todemonstrate its optical activity by converting it into sodium saltand determining the rotatory power of the latter.Dry silver salt(0.8901 gram) was digested with 7 C.C. of an aqueous solutionof sodium chloride containing 4.4 grams of sodium chlorideand 0.4 gram of sodium hydroxide in 100 c.c., the sodiumhydroxide being added in order to check the racemisationof the dissolved sodium salt, which proceeds rapidly in neutralsclution (see p. 1877). The resulting solution was decanted fromthe solid residue through a filter, the residue repeatedly extractedin the same way with successive small quantities of the alkalinesolution of sodium chloride, and the total filtrate polarimetricallyexamined :I = 2, c = 0.4065, a, - 3*82’6, [MI;- 79.9O.equi-molecular proportions of morphine and oximinocyclohexane-4-carboxylic acid are mixed in hot ethyl-alcoholic solution, thesparingly soluble morphine salt which separates contains a pre-ponderance of the dextrorotatory form of the acid.5.8 Grams ofthe acid, dissolved in 50 C.C. of absolute alcohol, were added to ahot solution of 11.19 grams of morphine in 200 C.C. of the samesolvent. The salt (14.04 grams) was gradually deposited, on cooling,in rosettes of small prisms. It can be recrystallised from about30 parts of boiling absolute alcohol. It can also be recrystallisedfrom methyl alcohol, but it is very sparingly dissolved by ethylacetate, chloroform, or other common organic solvents, althougheasily soluble in cold water. The optical activity of the acidcomponent of the salt was examined in a manner similar to thatemployed in the case of the quinine salt.The very finely powderedsalt was treated with an aqueous solution of so much ammonia asto leave an approximately decinormal solution of the latter afterthe decomposition of the salt. The separated morphine was thenremoved by filtration, and washed with successive small quantitiesof a N/lO-aqueous solution of ammonia. The filtrate and washingswere freed from dissolved morphine by repeated extraction withM orp?hin e d-0 ximinoc ycloh e xane-4-car71 ox yla t e. - Whe18'76 MILLS AND BAIN: OPTICALLY ACTIVE SALTS OFtheir own volume of chloroform, and then polarimetrically examined.2.6 Grams of morphine salt digested with a solution (8.4 c.c.) con-taining 0.114 gram of ammonia and treated in this ma.nner gave asolution which was examined in the polarimeter with the followingresult :Z=2, ~ ~ 7 .3 1 , UD 5-43'> [MI, 64.6'.The salt racemises as regards its acid component very rapidly inalcoholic solution. It is therefore not possible to obtain the moresoluble morphine Z-acid salt from the mother liquor after thecrystallisation of the less soluble d-acid salt. On concentration ofthe mother liquor, the crystals which separate still contain anexcess of the latter salt. Also, the degree of separation of thotwo diastereoisomeric salts attained depends greatly on the exactconditions under which crystallisation takes place. Different cropsof crystals examined in the manner described above gave ammoniumsalts of the following molecular rotations: [&t]D 43O, 48O, 64.6O,50°, 68O, 52O, 53'5O.powderedmorphine salt (6 grams), which had been twice crystallised fromalcohol, was treated with an aqueous solution (8.8 c.c.) containing0,253 gram of ammonia.The separated morphine was removed inthe manner already described, and the resulting solution, afterhaving been polarimetrically examined :I = 2, c = 16.634, a, 9*57O, [MI, 50'1°,was treated with an aqueous solution of 2.55 grams of silver nitrate.The silver salt was collected, washed, and dried exactly as in the caseof the silver salt of the I-&id. It weighed 2-38 grams, and wasanalysed with. the following results :Silver d-0 x~m~izocycloh exa ue-4-curb ox ylu t e.- Finely0.3961 gave 0.4623 CO, and 0.1387 H,O.0.49680.4876' ,, 0-2652 AgC1. Ag=40*94.C = 31.83 ; H= 3.89.N=5.14. ,, 21.75 C.C. N, (moist) a t 14O and 759 mm.C7HIoO3NAg requires C = 31-82 ; H = 3.79 ; N = 5.30 ;Ag = 40.90 per cent.A portion of the same preparation (1-0019 grams), on treatment>with an alkaline solution of sodium chloride as described in the caseof the silver salt of the I-acid, gave a solution of dextrorotatorysodium salt, which was polarimetrically examined, with the followingresult :Z= 2 ; c = 4 8523 ; a:' 2.68"; [MILG 49.4".Dextrorotatory silver salt of higher activity was subsequentlyobtained from a solution of ammonium salt of molecular rotation[MI, 64.6O (c=7*31) made from a preparation of morphine salt(2.6 grams), which had been four times recrystallised from alcohol4-OXIMINOCYCLOHEXANECARBOXYLIC ACID, ETC.1877On treatment with silver nitrate, this gave 0.917 gram of silver salt,0.6151 gram of which was converted into sodium salt in the mannerdescribed. The rotatory power of the resulting solution was it9follows :z-2, ~ ~ 2 . 9 7 9 , a, 2*48O, [MI, 74.5'.Reducing Power of the Optically Active Sodium Salt.-For reasonsgiven in the introduction (p. 1870), it appeared desirable to comparethe reducing power of an active solution of the sodium salt of theoximino-acid towards Fehling's solution with that of P-phenyl-hydroxylamine on the one hand and that of acetoxime on theother.1.62 Grams of quinine I-acid salt were treated with a solution(15 c.c.) of 0.2 gram of sodium hydroxide, and the separated quininewas removed by extraction with chloroform.For comparison withthe solution of the sodium salt of the I-oximino-acid in N/lO-sodiumhydroxide thus obtained, solutions of equimolecular quantities ofphenylhydroxylamine (0-37 gram) and of acetoxime (0.246 gram)in 15 C.C. of NjlO-sodium hydroxide were prepared. The threesolutions were respectively added a t 20° t o three quantities ofFehling's solution, each containing 1 gram of crystallised coppersulphate (1.2 mols.), 5 grams of Rochelle salt, 1.5 grams of sodiumhydroxide, and 60 grams of water. The solution to which thephenylhydroxylamine had been added was, of course, conipletelyreduced in a few seconds. The other two were kept well stopperedfor twenty-four hours in a thermostat at 20°, and the smallquantities of cuprous oxide which had been precipitated in eachwere collected and their respective amounts determined :Precipitated by acetoxime : 0.013 gram Cu,O ; 0.065 atom Cii.precipitated by sodium salt of oximino-acid : 0.0316 gram Cu,O ;0.132 atom Cu.The reducing power of the sodium salt of the oximino-acid inalkaline solution is therefore, as was to be expected, of the sameorder of magnitude as that of acetoxime, and of an altogetherdifferent order from that of phenylhy3roxylamine.There is there-fore no reason to suppose that the hydroxylamine residue containedin these optically active salts possesses a constitution different fromthat usually assigned to it in the kectoximes.T7t e €2 ac e.misation Ph enona ena.The Sodium Salt.-A solution of 1aevor.otatory sodium saIt freefrom excess of alkali was prepared by digesting active silver saltwith a slight excess of an aqueous solution of sodium chloride.2-40 Grams of quinine Z-acid salt (giving sodium salt of molecula1878 MILLS AND BAIN: OPTICALLY ACTIVE SALTS OProtattion [q, - 7 6 .3 O , c = 0.633) were converted into ammoniumsalt in the manner already described, and the weakly ammoniacalsolution of the ammonium salt thus obtained was treated with anaqueous solution of silver nitrate (0.935 gram). The precipitatedsilver salt, after having been carefully washed free from ammonia,was digested while still moist with an aqueous solution (20 c.c.) of0.292 gram of sodium chloride.After thirty seconds, the solutionof sodium salt thus obtained (which would be approximately0-2-norma1, and would also contain the slight excess of sodiumchloride used) was filtered as rapidly a,s possible from silver chlorideand examined in a, 2-dcm. tube. The rotation was seen to diminish,the diminution taking place, as is shown in the accompanying table,in accordance with the unimolecular formula. The temperature was20.4O :t (niins.). Rotation. l/tloga/(a - x).0 .o - 1'44" -2.75 1 *34 0'0115 -85 1 *22 0'01210.40 1 *08 0.01215.0 0.94 0'01220 -3 0.80 0-01325 -7 0.67 0.01335 *9 0.51 0 -01 347 *3 0.36 0.013Taking 0.0125 as the value of the constand, the period of half-change is twenty-four minutes.Sodium Salt in Presence of Sodium Hydroxide.-The solu-tion on which the observations were made was prepared in thefollowing manner.2-50 Grams of quinine E-acid salt were treatedwith an aqueous solution (10 c.c.) of 0.246 gram of sodium hydroxide.The liberated quinine was removed with chloroform, and theseparating funnel washed with small quantities of N 10-sodiumhydroxide solution. The resulting solution was placed in a 2-dcm.tube, which was kept between the observations in a thermostat at20°. The rotation was found to be continually decreasing, againin accordance with the unimolecular formulE, but very much moreslowly than in the previous experiment.t (hours.).0.01-563'955.667.759.8821-6224'2426.7529-7546'23Rotation.- 3 *42" * 3-203.002.852-662-491 *711 *571 '441 *330'793/1logn/(n -2).-0'01470'01440.01400'01410*0140O'G1390'01390.01400'01380'0134-OXIMLNOCYCLOHEXANECARBOXYLIC ACID, ETC. 1879Taking the value of the constant t o be 0.0139, the time of half-change is 21.7 hours-fifty-four times as great as that for thesodium salt in neutral solution, The concentration of the sodiumhydroxide was determined after the polarimetric observations werefinished by titrating 10 C.C. of the solution with N/lO-sulphuricacid, using phenolphthalein as indicator. It wits thus found to be0.098-normal. The concentration of the sodium salt, calculatedfrom the quantdty of quinine salt taken, was 0-32-N.Ammonium Salt.-An approximately 0.2-N-solution of laevo-rotatory ammonium salt was prepared by a method exactlyanalogous to that employed in the case of the sodium salt.The samequantity of quinine Z-acid salt (2.4 grams) was converted intoammonium salt (the molecular rotation of which was [MID -72*8O,c = 5.674) and the silver salt obtained from it by precipitation withsilver nitrate was digested with a solution of 0.267 gram ofammonium chloride in 20 C.C. of water. The resulting solution oflzevorotatory ammonium salt was observed as rapidly as possible ina 2-dcm. tube, the temperature being 18.5O. The rate of diminutionof the rotation was approximately twice as great as in the corre-sponding solution of sodium salt :t (mins.).0.01 -873'434.736.477'629.1310.5812.3714-5517.3521.8325.7739.65Rotation.- 1.72"1.561 *431'34I -231-151-080.990.900 T i0.680.550-420'19l/tloga/(n - -0.0230.0230.0230 0230.0230.0220 a0230.0230 '0240.0230.0230 0240.024The constant 0.023 corresponds with a period of half-change of13-1 minutes.Ammonium Salt in presence of Ammonium Hydroxide.-Theseobservations were made on a dextrorotatory solution prepared from3-0 grams of morphine d-acid salt by digestion with 8-18 C.C.of anormal solution of ammonia, the morphine being removed in themanner described by filtration and extraction of the filtrate andwashings with chloroform. The final volume was 14-5 C.C. Theexcess of ammonia present could not readily be determined, but theabove quantities were adjusted so as to give an approximatelyN / 10-solution.The concentration of the ammonium salt calculatedfrom the quantity of morphine salt used was 0.47-N. The solutio1880 MILLS AND BAIN: OPTICALLY ACTIVE SALTS OFw m placed in a 2-dcm. tube, and kept between the observationsin a thermostat at 20°:t (hours.).0.01 '662-473 *574'735'536 *728 -30Rot ation.5-00"4.364.083.733.393'122'892-54l/tlogn/(a - x). -0.03580.03580.03560 03570.03510.03540 0354The mean value of the constant (0.0355) corresponds with aperiod of half-change of 8-48 hours. The rate of racemisation ofthe ammonium salt is therefore thirty-nine times less in the presenceof this excess of ammonium hydroxide than under the conditions ofthe previously described experiment.Mutarotation of the Morphine &Acid Salt in A p e o u s 8oZution.-The morphine salt is readily soluble in cold water.If the aqueoussolution is polarimetrically examined as soon as possible after itspreparation, its lzvorotatory power is found to be quickly increasing.The change is, however, rapidly completed (practically), and afterabout seven minutes no further alteration in the rotation is to bedetected. I f the solution is now decomposed by the addition of aslight excess of ammonia and the morphine removed, the resultingammonium salt is quite inactive. The mutarotation is thereforedue to the racemisation of the acid component of the salt.One gram of finely powdered morphine salt, giving ammoniumsalt of molecular rotation [MI, 51.7O (c=0'844) was added to 20 C.C.of water.As soon as most of the salt had dissolved, the solutionwas filtered, and its rotatory power observed in a 2-dcm. tube, thefirst observation being completed 1-75 minutes after first wettingthe salt. The temperature was 26'5O:t (mins. ).0.01 -031.912-724.776.7012.1300Rot at ion. - 7.19"7 *41:'SOI -587.667 -667'677.67A. A (calc.).0.54" I0'26 0.27"0.17 0.150 09 0 090.01 0 '020 '01 0 010-00 0 .oo - -Under A are given the differences between the rotations observed,and the value finally attained. The numbers in the last columnare calculated by means of the unimolecular formula, k being takenas 0.29, corresponding with a period of half-change of 1-04 minutes.They agree with the observed differences within the limits ofexperimental error4-OXIMINOCYCLOHEXANECARBOXYLIC ACID, ErC.1881Tlie Mutarotation of tJbe Quinine Salts of the d- and l-Acids.-The quinine salts are not sufficiently soluble in water to allow ofsatisfactory observations being made in that solvent, and althoughathey possess a high solubility in alcohol, when brought into contactwith it they form gummy masses, which pass comparatively slowlyinto solution. I n the case of each salt, therefore, 0.8 gram was firstfinely ground up with 6 C.C. of ethyl acetate, and 12-14 C.C. ofabsolute alcohol then added to the mixture, when solution tookplace very rapidly.The solutions thus obtained were filtered andobserved in a 2-dcm. tube, the temperature being 20° in each case.The quinine Z-acid salt used was a sample which gave sodium salt ofmolecular rotation [MI, - 67-3O ( c = 0*731), the quinine d-acid saltgave ammonium salt of molecular rotation [MI, 40'4O ( c = 0.947).Quinine d- Oxinainoc y clohexa ne-4-ccwbox~Znte.t (inins.).0.03 03 *75 .o9.615.320'400Rota tion.- 9-16"9'359-389.479 '569'609.599 60A, A (talc.).0-44" 0.46"0 2 5 0 *240 22 0'200'13 0.150.04 0.050 .oo 0.010.01 G.00 - -Quinine 1- Oxinainocy clohexane-4 carboxykate.t (inins.).0.02 '13.15-26 - 47 99.812.521.700Rotation. A.- 11 *70" 0 *89"11-37 0.5611-23 0.4211.08 0.2710.98 0.1710-96 0.1510.92 0'1110.87 0'0610.82 0.0110'81 -A (cdc.).0-550'440 '270.200'140 -090.050'01--The hvorotation was seen to increase in the case of the quinined-acid salt and to diminish in that of the quinine Z-acid salt, therate of change agreeing (within the probable limits of error ofthe single rapidly made observations) with that calculated fromthe unimolecular formula when the velocity constant is faken as0.1 in each case, corresponding with a period of half-change ofthree minutes.The character of the differences among the velocity constantsobserved in these experiments might suggest that the rate ofracemisation of the salts was connected with their degree ofhydrolysis in the different solutions investigatecl-that the non-hydrolysed portion of the salt racemises very slowly, and that theobserved loss of activity is mainly brought about through therapidly racemising free acid present in larger or smaller proportionaccording to the degree of hydrolysis.It would appear, however,that this simple hypothesis is inadequate, for whereas the rates ofracemisation in the aqueous solutions of the sodium and ammoniumsalts investigated were found to be approximately in the ratio of2 to 1, the relative proportions of free acid present in the twosolutions, calculated from the dissociation constants of water andVOL. XCVLL. (i 11882 nONNAN AND POTTS: KINETICS OF THE REACTIONammonia, were of the order of 100 to 1. Other factors musttherefore probably be taken into account. For example, it is con-ceivable that the interconversion of the d- and I-mordificationsshould ta.ke place through an iseoximino-form :-C- -C- -C-and t,hat the effects which acids and alkalis produce on the ratsof racemisation are brought about through their influence on therelative proportion of the latter form present in the solutions.The authors propose to extend this investiga,tion to other com-pounds containing doubly-linked tervalent nit,rogen. I n conclusion,they desire to express their indebtedness to the Research FundCommittee of the Chemical Society for a grant by which theexpenses of this work have been largely defrayed.NORTHERN POLYTECHNIC INSTITUTE,HOLLOWAY, LONDON, N

ISSN:0368-1645    

DOI:10.1039/CT9109701866

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

I882 nONNAN AND POTTS: KINETICS OF THE REACTIONCC.-Kinetics of the Reaction between Silver Salts*] _ L 1and Aliphatic Iodides.By FREDERICK GEORGE DONNAN and HAROLD EDWARD POTTS.THE reaction between silver nitrate and alkyl iodides in alcoholicand aqueous-alcoholic solutions has been studied from a kinetic pointof view by Burke and Donnan (Trans., 1904, 85, 555; Zeitsch.physikal. Chem., 1909, 69, 148). The conclusion was drawn thatthe reaction is pseudo-bimolecular in type, and that the course ofthe reaction is affected by an accelerating influence due to someproduct. The following experiments have been carried out withthe purpose of throwing further light on the mechanism of thereaction. They deal with the reaction between ethyl iodide andsilver salts other than the nitrate, the reaction between ethyl iodideand silver nitrate in a non-hydroxylic solvent (acetonitrile), andthe reaction between silver nitrate and aliphatic polyiodides (inalcoholic solution).The diminution in the velocity-coefficient with decreasing initialconcentration in the reaction between ethyl iodide and silver nitratein alcohol led Burke and Donnan to the conclusion that the velocityof the reaction is accelerated by some product (although it wasshown that this product could not be ethyl nitrate, ether, or nitriBETWEEN SILVER SALTS AND ALIPHATIC IODlDEY.1883acid). This fact is clearly shown by Fig. 1, in which theordinates represent concentrations, and the abscissze times.Curve A B represents the timeconcentration curve for the reactionN/40-AgN0,+N/40-C2H,I in dry alcohol, whilst CD is the corre-sponding curve for the reaction N/80-AgN03 + A7/80-C2H,I in thesame solvent at the same temperature (Burke and Donnan, Trans.,Zoc.cit.). I n the figure, the time-zero for the second curve (CD)has been shifted to t.he time-point (abscissa) when the primaryreactants corresponding with the first curve ( A B ) have fallen to aconcentration of N / 80. The non-coincidence of these curves andthe greater steepness of AU show that the velocity of the reactionFIG. 1.Time (minutes).is accelerated by some product. The reaction is therefore from akinetic point of view a “disturbed” one, and the fair degree ofconstancy of the velocity-coefficient of the simple bimolecularequation in any particular experiment can only be explained as dueto some compensating effect.A special experiment made to test this conclusion showed thatin the filtered reaction fluid resulting from a previous reaction, andcontaining no silver salt or alkyl iodide, the reaction proceeded (onaddition of fresh portions of the reactants) more rapidly than inthe pure solvent, the acceleration observed being in the ratiocalculable from the results of Burke and Donnan.6 a 1884 DONNAN AND POTTS: KINETICS OF THE REACTIONI n view of this disturbance, the cpnstancy of the velocity-ooefficientof the ‘( bimolecular ” equation in any given case does not provethat the reaction is really one of the second order.I n the absenceof the accelerating effect referred to, it is clear that the velocity-coefficient calculated from the simple bimolecular equation,-dc/dt=kcz, would fall during the course of the reaction. Thisshows that the exponent of c, that is, the order of the reaction,must be higher than 2.I n fact, although the Noyes-van’t Hoffmethod is not strictly applicable t o a case such as the present wherethe products exercise an accelerating influence, if we apply it toonly a 10 per cent. change in the N / 4 0 - and N/80-reaction curvesshown in Fig. I, the value n = 2.8 is obtained for the exponent of c.It is possible that this result may be explained by considering thevarying electrolytic dissociation of the silver nitrate, as describedon p. 1889 (results with silver lactate in aqueous alcohol).Units and Method of Measurement.The (‘ dry )’ alcohol used in the following experiments was pre-pared by long-continued boiling of the ordinary ( ( absolute ” alcoholwith excess of freshly prepared calcium oxide.The alcohol soobtained was then freed from reducing substances by distillationfrom a, small quantity of silver nitrate..The reaction-velocities were all measured at 25O, and the amountof silver salt in solution estimated by titration with thiocyanatesolution. The numbers given in the tables are expressed in termsof the following units. Unit of volume = 1 C.C. ; unit of mass = frac-tion of a, molecule contained in 1 C.C. of a N/lOO-solution. Thus,the concentration of a N/lO-solution = 10. The velocity-coefficientscalculated with these units are six times greater than thosecalculated with the units employed by Burke and Donnan in theirfirst paper.The unit of time employed is one minute. The valuesof k, the velocity-coefficient, have been calculated from the ordinarybimolecular equation except when otherwise stated.The reaction fluids were previously warmed in separate vesselsDefiniteequal volumes of the mixture were then quickly transferred bymeans of a previously warmed pipette into a series of small flasksplaced in the thermostat.. in the thermostat, aud at a given time-point rapidly mixed.Experiments with Silver Lactate in Aqueous AZcohoZ.It has been shown in the case of silver nitrate that, althoughthe velocity-coefficient of the bimolecular equation remains fairlyconstant in any given reaction, the velocity of the reaction isundoubtedly increased by some reaction product.BETWEEN SILVER SALTS AND ALIPHATfC IODIDES.188.5Experiments were therefore made with another silver salt in orderto see if this peculiar behaviour could be associated with thepresence of the NO,-group. *Owing to the difficulty of getting silversalts sufficiently soluble in dry alcohol, the experiments were madewith silver lactate in a composite solvent, obtained by mixing 40volumes of water with 60 volumes of dry alcohol. It has beenshown by Burke and Donnan that in a solvent of this compositionthe behaviour of silver nitrate and ethyl iodide resembles that indry alcohol.TABLE I.Silver lactate, 3N/200 x 0.95; ethyl iodide, 3Ay/200.t ......0 4 - 5 9.5 14.5 19.5 33 39.5 6 8 5cAg ... 1-43 l.285 1-16 1.07 1'00 0.835 0.785 0.615k ...... - 0'0161 00160 0'0152 0'0145 0.0141 0.0136 0.0126It will be observed that the bimolecular velocity-coefficient fallssteadily during the course of the reaction.Determinations of the amount of free acid produced during thereaction were made, the results of which are shown in the followingtable.TABLE 11.Acid producedt . cacia. CAg. Silver used up0 0.0 1'43 -26 0.345 0 -95 7030 0.365 0.89 6734 0.39 0.85 662 weeks 0.97 < 0.01 70The ratio of free acid produced to silver salt used up in thereaction remains constant. It is noteworthy that this ratio ispractically the same as that observed by Burke and Donnan inthe case of the reaction between silver nitrat; and ethyl iodide indry alcohol, although in this case the acid is different.The follow-ing two tables contain reaction-velocity measurements made withanother sample of silver lactate at two different initial con-centrations. The values of k are here calculated from point topoint, and not from the beginning of the reaction to each time-point as in the previous table.TABLE 111,Silver lactate, 3 N / 100 x 0.96 ; ethyl iodide, 3 N / 100.t ... ...... 0 7 11 16 20 30 40k . . . . . . . . . - 0.0302 0'0209 0.0200 0.0147 0.0157 0'0121CAg ...... 2'89 1.76 1.525 1.31 1-21 1.00 0.81886 DONNAN AND POTTS: KINETICS OF THE REACTIONTABLE IV.Silver lactate, 3N/200 x 0.96; ethyl iodide, 3N/200.t ......0 10 25 45 72 I05 133 160k ....,. - 0.0205 0'0120 0.0100 0'0099 0.0103 0.0095 0'0098CAg ... 1.445 1.118 0.917 0.765 0.627 0.508 0'442 0'39I f these two sets of times and concentrations are plotted, and thetime-zero for the 3 N / ZOO-reaction made to coincide with the time-F I G . 2.,150 50 100Tivze (minutes).point where the primary reactants of the 3 N / 100-reaction havefallen to 3N/200, as shown in Fig. 2, it will be seen that the twosets of points lie on the same continuous curve. This is in strikingcontrast to the case previously discussed (reaction between silvernitrate and ethyl iodide), and shows that in the present case thepreviously observed acceleration due to the products is absent.It seems to be pretty certain therefore that in the case of silvernitrate the apparent constancy of the bimolecular reaction-coefficienBETWEEN SILVER SALTS AND ALIPHATIC IODIDES, 1887is due to a compensating effect caused by the acceleration referredto.This is well shown by the following table, which contains theresults of some further measurements of this reaction in the com-posite solvent employed above.TABLE V.Silver nitrate, 3 N / 100 ; ethyl iodide, 3 N / 100.t . . I . . . . . . 0 6 19 49 90 124 164G A ~ ...... 3‘00 2.48 1‘843 1.125 0.703 0.528 0’41k ..I...... - 0.0117 0*0108 0’0116 0’0131 0’0136 0.0134Here one observes a fair constancy of k, with, however, a tendencyto increase towards the latter part of the reaction. A similartendency is observable in some of the results previously given byBurke and Donnan (Zoc.cit.).The practically constant ratio of lactic acid produced to silverlactate used up (66 t.0 70 per cent.) shows, as in the case of silvernitrate, that there must be a t least two simultaneous reactions, inone of which the free acid is formed (Wegscheider’s criterion).Denoting the negative ion of the silver salt by A, we mayrepresent these two reactions thus :A , . . . , AgA+EtI -+ AgI+EC;A.B , , , AgA+EtI+EtOH + AgI+Et,O+HA.C , . AgA+EtI+H*OH --+ AgI+EtOH+HA.In the case of aqueous solvents, there will also occur the reaction :In the case of alcoholic or aqueous-alcoholic solvents, the reactionB (or B+C) is the main one, as only about 30 per cent. of the“ theoretical ” amount of the ester Et,A is produced.In order to explain this “ abnormal ” production of acid,Wegscheider and Prankel (Sitzungsb er.’CYierz. Akad., 1907, 116,Abt. IIb) have proposed the following hypothesis. They supposethat the essential cause of the reaction is the affinity of the silverand halogen atoms, and that this leads to a series of not sharplydistinguishable intermediate stages in which there occurs a Fpatialapproximation of the silver and halogen atoms. These may beroughly formulated as follows :C2H5-I . . . . Ag-NO,.As a result of this, the ‘‘ affinit,y” between CzH5 and I on theone hand, and Ag and NO, on the other is weakened, so that theabove system may rearrange itself into tho definite end-productsAgI and CzH5*N03, or may react with a molecule of the solventto give (in the case of ethyl alcohol) the end-products AgI,(C,H5),0, and HNO,.It would seem difficult to distinguish between this hypothesis an1888 DONNAN AND POTTS: KINETICS OF THE IlEACTIONwhat is usually understood by the assumption that a transientintermediate compound between the silver salt and the alkyl iodideis the first stage in the reaction. Moreover, Wegscheider’s hypothesisdoes not lend itself to analytical treatment in the present conditionof chemical science, so that it is impossible t o apply t o it.any exactexperimental test. We shall content ourselves therefore withexamining the consequences of more definite hypotheses. I n thefirst place, let us suppose that the velocity of the reaction is con-trolled by a direct interaction between one molecule of undissociatedsilver salt and one molecule of alkyl iodide.Considering onlyequivalent concentrations, and calling a the degree of dissociation ofthe silver salt, this leads to the equation :-clc/dt = k ( l - a ) c 2 .As ct will increase, and therefore 1 - a decrease, with the progressof the reaction, the coefficient k.(l - a ) of the simple bimolecularequation will diminish as the reaction progresses. This is exactlywhat the experimental results show in the case where the reactionis not disturbed by the products. Let us now make the furtherassumption that the simple law of mass-action holds for theequilibrium between the undissociated silver salt and its ions. Thisleads t o the equation X(1- a) =a%, which, combined with theprevious one, gives : - clc Jclt = kfa2c3.Now, as a increases with the dilution, we should expect thecoefficient k’a2 of the ‘‘ termolecular ” equation to increase with theprogress of the reaction, but as, a t the dilutions employed, a willnot differ very far from unity, it is obvioins that the increase ofQa2 will be considerably smaller than the decrease of k ( 1 - a).The following table contains the values of k3 corresponding tothe equation -dc/dt =1c3c3? as calculated from the curve given inFig. 2.The values of k, are calculated from point to point.TABLE VI.t ......... 0 4-2 11.1 30.1 48-5 84.7 119’0c .........2.89 2’0 1.5 1 -0 0 ‘8 0-6 0.5k3 ......... - 0.0298 0*0281 0.0292 0-0306 0.0356 0.0356The coefficient of the simple biniolecular equation falls from0.030 to 0.010 over the same range of concentra.tion.The constancyof k3 is surprising, and, indeed, even greater than could have beenexpected on the above hypothesis. This will be made clear by thefollowing table :a ............. 0.7 0 *8 0.9a2 ............... 0 -49 0.64 0 ‘81l - a ............ 0 ‘30 0 -20 0’1BE’I‘WEEN SILVER SALTS AND ALIPHATIC IODIDES. 1889This table shows that if (1 - a ) falls in the ratio of 3 to 1, a2 willincrease by more than 50 per cent. Now, whilst the fall of thebimolecular velocity-coefficient is in reality in the ratio 3 : 1, thecorresponding increase in the termolecular velocity-coefficient is onlyabout 20 per cent. Too much stress need not be laid on thisdiscrepancy, as the assumption that the simple law of mass-actionholds for the electrolytic dissociation of the silver salt may be onlya very rough approximation.It is necessary also to remark here that if the simple law of mass-action be assumed, the equation -dc/dt=k’aV can be equallywell reconciled with the view that only the ions Ag’ a.nd NO,’ reactwith the alkyl iodide.The results appear to show, however, that thereaction velocities due to undissociated silver nitrate on the onehand, and the ions Ag’ and NO3’ on the other, must be very different,so that the main effect must be due to either one or the other.Experiments carried out in alcohol as solvent have shown (Burkeand Donnan, ZeZtsclL. plqsikal. C ~ E ? ? ? ., 7oc. cit.) that highly dis-sociated nitrates, such as ammonium nitrate, considerably increasethe velocity of the reaction between silver nitrate a.nd ethyl iodide.This seems to indicate that it is the undissociated silver nitratemolecules which are effective (at all events, in alcoholic solution).Concerning the question of the formation of an intermediatecompound, the experimental results are not decisive. I f , as seemspossible from the above, it is the undissociated silver salt whichreacts, then it is a plausible hypothesis to assume that the velocityof the whole reaction is determined by the rate of formation of anintermediate compound between the silver salt and the alkyl iodide,this intermediate compound then rapidly changing to alkyl esterand silver iodide on the one hand, and, on the other, reactingwith a molecule of solvent to form silver iodide, free acid, and anether (or alcohol).There is, however, no experimental evidencewhich would enable one t o test this hypothesis.Experirneiats wit7~ Silver Nitrate in Acetonitrile.The foregoing experiments, together with those of Nef (Annuten,1899, 309, 126), Burke and Donnan (Zoc. cit.), and Wegscheidcrand Frankel (Zoc. cit.), have shown that when a silver salt reactswith an alkyl iodide in a solvent which contains the hydroxyl group,free acid is one of the main products. It was decided therefore tocarry out some experiments in a non-hydroxylic solvent in order t oexamine the kinetics of the reaction when no free acid can beformed.As a few preliminary experiments showed that thereaction between silver nitrate and ethyl iodide proceeds with con1890 DONNAN AND POTTS: KINETICS OF THE REACTIOXvenient speed in acetonitrile, and that the precipita-te producedconsists of silver iodide, a quantity of this solvent was shaken with1 3phosphoric oxide, andthen distilled fromthis reagent into abottle provided withit phosphoric oxideguard-tube and mer-cury-sealed outlet-tube.The velocity of the r eaction was measured inthe apparatus shownin Fig. 3, the reaction-vessel being immersedi n the thermostat.Twenty-five C.C. of thedry acetonitrile, con-taining a weighedquantity of silvernitrate, were intro-duced into the reactionbe vessel, then a weighedquantity of ethyliodide (contained in asmall stoppered tube)dropped in, the mer-cury-sealed stopper ofthe reaction-vessel re-placed, and the con-tents thoroughly mixedby blowing in a-currentof dry air.A t definiteintervals, portions ofthe reaction - liquidwere blown over intot,he burette, rapidlymeasured off, andtitrated.UI n the following tables, li is the velocity-coefficient of the simplebimolecular equation :TABLE VII.Silver nitrate, N / 10 ; ethyl iodide, N / 10.t ...... 0 15 36 66 88 115CAg ... 10.00 7'94 6.27 4.95 4.37 3.88rE ...... - 0'00173 0-00164 0.00155 0.00146 0'00137k (mean) =_0'00155BETREEN SILVER SALTS AND ALIPHATIC IODIDES. 1891TABLE VIII.Silver nitrat,e, N j 20 ; ethyl iodide, AT/ 20.t .........0 15.5 41 84 139 196 244k ......... - 0.00090 0.00107 0*00120 0.00107 0*00105 0.00105k (mean) = 0.00105.CAg ...... 5'00 4.68 4-09 3'32 2.87 2.47 2-20TABLE IX.Silver nitrate, N / 4 0 ; ethyl iodide, N / 4 0 .......... 0 91 248 349k ......... - 0 *00081 0*00081 0 ~00081k (mean) = 0.00081.I n these experiments the reaction-liquid a t the end of the reactionwas found to contain a small quantity of acid, but this was tracedto the solvent, which was found t o contain a small quantity ofphosphoric acid sufficient to explain the acidity of the reaction-fluid.To obviate this, the acetonitrile, which had been previously driedby phosphoric oxide as described, was treated with freshly ignitedlime and redistilled. The following experiment was carried outwith the now neutral distillate :Ag ......2-50 2.11 1'66 1'47TABLE X.Silver nitrate, 1 / 2 0 ; ethyl iodide, N / 2 0 .t . . ....... 0 98 130 157 166CAg ...... 5.00 3.33 2-95 2-73 2-73k ........ - 0-00102 0.00106 0*00106 0 *00100k (mean) = 0*00104.This value of k agrees very well with that previously found. Itwas noticed, however, that at the end of the experiment the reaction-fluid contained about the same (small) quantity of free acid as inthe previous experiments. Probably the acetonitrile, after treatmentwith phosphoric oxide, contained a little ethyl phosphate (due totraces of alcohol), and this on distillation from lime regenerated thealcohol, which afterwards reacted to form free acid.The above experiments prove that whilst the reaction proceedsnormally (that is, without production of free acid) in acetonitrileas solvent, the reaction-kinetics closely resemble those obtained withalcohol as solvent,.The simple bimolecular velocity-equation holdsgood for each experiment (although in the case of the N/lO-reactionthere is a decided drop), but the velocity-coefficient varies with theinitial concentrations of the reactants. It is clear that the peculia1892 DONNAN AND POTTS: KINETICS OF THE REACTIONkinetic behaviour observed in hydroxylic solvents has nothing to dowith the acid formation. As explained previously, this behaviour isprobably to be accounted for by the varying dissociation of thesilver salt, and, in the case of the nitrate, by the formation of someproduct which acts as an accelerator.If time-concentration curves for the experiments with variousinitial concentrations be plotted as described on p.1883, it willbe found that in the case of acetonitrile as solvent the various curvesso obtained cannot be made to superimpose, although the divergenceis not nearly so marked as in the case of alcohol as solvent. Thiswould seem to show that the accelerating effectl referred to aboveis much smaller in the experiments with silver nitrate in acetonitrile.Experiments with Iodoform in A!coholic Solution.Preliminary experiments showed that in dry ethyl alcohol a t 25O,silver nitrate and iodoform (both N/30) react with convenient speed,the reaction occurring in the stoicheiometric ratio 3AgN0, : CHI,,and the precipitate produced being silver iodide.During thereaction free nitric acid is produced, and this slowly liberates iodinefrom iodoform, but if the experiment does not last more than afew hours, this secondary reaction may be neglected. The followingtables contain the results of the velocity-measurements (k = coefficientof simple bimolecular equation) :TABLE XI.t ...... 0 30 45 60 120 - 210k ...... - 0'00136 0 00136 0'00146 0'00135 0'00147Mean = 0*00140.CAg ... 3.765 3'255 3 -05 5 2.825 2 34 1 '74TABLE XII.t ...... 0 52 70 82 92 116 155 173C A ~ ... 3.185 2.595 2.55 2-47 2.445 2.29 2'06 1.965k ...... - 0*00136 0'00111 0'00111 0*00103 0'00106 0'00110 0*00113Mean = 0*00113.TABLE XIII.N/40-AgN03 ; N/40-(cH13/3).t .........0 52 71 94 148 167CAg ..... 2'45 2 265 2.175 2.13 1.945 1 *87k ...... - 0-00066 0-00073 0.00065 0-00072 0*00076Mean = 0*00070BETWEEN SILVER SALTS AND ALIPHATIC IODIDES. 1893TABLE XIV.N / 30-AgNOs ; N/40-(CHI3/ 3).t ......... o 19.25 42 5 a5 98 210 232 303CAg ... 3'33 3.185 3.015 2'725 2.675 2.185 2.035 2.015k ...... - 0.00090 0-00095 0.00107 0.00103 0-00110 0.00122 0 00097Mean = 0'00104.TABLE XV.t ......... 0 28 109 142 180 231 257 ...... 1.93 1.835 1-645 1.54 1-495 1.375 1-345k ......... 0 00062 0'00064 0.00070 0-00063 0*00071 0'00063 ,Mean = 0.00065.An experiment was tried with it solution of iodoform in alcoholThis solutionThe following table summarises the results obtained with iodo-which had been preserved in the dark for eleven days.reacted with silver nitrate with the same speed as a fresh solution.form :TABLE XVI.Initial concentration.Value of k x lo5.N/25 ..................... 140N/30 ..................... 113N/30-AgNO3 ; N/40-(CHJS/3).. 104Xi40 ..................... 70N/40-AgNO, ; N/52-(CH13/3).. 65It will be observed that the behaviour of iodoform closelyresembles that of an alkyl monoiodide. The reaction is pseudo-bimolecular in type, the velocity-coefficient increasing rapidly withincreasing initial concentrations of the reactants, and being mainlydependent on the initial concentration of the silver nitrate. I ftime-concentration curves be plotted, the curves representingreactions starting at different initial concentrations cannot be super-imposed, the reaction which starts at a lower initial concentrationalways proceeding markedly slower than a reaction which hasreached the same concentration of the reactants from it higher initialconcentration.As shown in Fig. 4, this divergence is greater thanin the case of the reaction between silver nitrate and ethyl iodidein alcohol as solvent.The constancy of the bimolecular velocity-coefficient in spite ofsuch a great acceleration due to the products proves that in theabsence of this disturbing effect the order of the reaction would bemuch higher than 2. Indeed, an application of the Noyes-van't Hoffmethod to the initial portions (10 per cent. change) of the variousreaction-curves plotted above yields, for the order of the reaction1894 DONNAN AND POTTS: KINETICS OF THE REACTIONnumbers the mean of which is about 3.8.It is probable thereforethat the real order of the reaction is bebween 3 and 4, and nearer 4than 3. With the data at hand, it is not possible to go furtherthan this. In particular, it is not possible to decide whether thereaction proceeds in a series of successive stages, though t.he some-what analogous case of the alkali hydrolysis of the triglycerideswould render this view probable.As in the case of the alkyl iodides, the reaction proceeds withformation of free acid. For the reaction starting at N / 3 0 -FIG. 4.Time (minutes).equivalent concentrations of silver nitrate and iodoform, the freenitric acid found at the completion of the reaction was equivalentto 85 per cent.of the original silver nitrate. This is a, higheramount than that observed in the case of ethyl iodide.The following experiment is a good example of a “rapid”reaction, in which one of the reactants is a carbon compound. Ifa, small quantity of finely powdered silver nitrate be cautiouslymixed with the equivalent amount of solid iodoform, a scratch orlight blow is sufficient to set up a violent reaction, in which cloudsof iodine vapour and oxides of nitrogen are evolvedBETWEEN SILVEK SALTS AND ALIPHATIC IODIDES. 1895Ezperiments with Metlqlene Iodide and Carbon Tetruiodide inAECO~LOE~C Solution.The experiments recorded in the previous section show thatiodoform reacts with silver nitrate in alcoholic solution a t 2 5 O abouteight times more slowly than methyl iodide.A few experimentscarried out with methylene iodide and silver nitrate in alcoholicsolution showed that the velocity-coefficient is about one hundredtimes smaller than the corresponding coefficient for iodoform. I npoint of reactivity with silver nitrate, niethylene iodide does nottherefore occupy a position intermediate between methyl iodide andiodoform, but is characterised by a very great sluggishness.Carbon tetraiodide was prepared by a modification of Gustavson’smethod, but it was found impossible to carry out any satisfactorymeasurements of its velocity of reaction with silver nitrate, as inalcoholic solution the tetraiodide loses iodine with comparativerapidity. Some special experiments on this point showed that in asolution containing 0.1 gram of tetraiodide in 20 C.C. of alcohol,95 per cent. of the iodine of the tetraiodide is present as free iodineafter sixteen hours.On adding an alcoholic solution of silver nitrate t o an alcoholicsolution of carbon tetraiodide there is an immediate diminution inthe amount of dissolved silver salt, due to interaction with the freeiodine. After this, the concentration of the silver salt diminishesvery slowly, so that the carbon tetraiodide itself does not react a tall rapidly with silver nitrate in alcoholic solution.I n conclusion, the authors desire to express their best thanksto the Research Fund Committee of the Chemical Society for agrant which helped to defray the expenses of this investigation.Note.---Since the completion of the above work, a paper hasappeared by G. Senter (Trans., 1910, 97, 346), in which this authorfinds that in the reaction between silver nitrate and methyl iodidein alcoholic solution, precipitated (or colloidal) silver iodide exertsa slight accelerating effect on the velocity of the reaction. It maybe possible therefore that this is the product of the reaction whichproduces the accelerating effect discussed in the foregoing paper. Itis somewhat difficult, however, to reconcile this view with theapparent absence of the accelerating action in the reaction between -silver lactate and ethyl iodide in aqueous-alcoholic solution. Thematter requires further investigation before this point can bedefinitely set;tled.MUSPRATT LABORATORY OF PHYSICAL AND ELF,CTI:O-CIIEMISTltY,UNIVERSITY OF LIVERPOOL

ISSN:0368-1645    

DOI:10.1039/CT9109701882

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

1896 DAWSON: CHANGES IN VOLUME IN THECCI.--Changes in Volume in the Formation of DiluteSolutions. Part IT. Relationship between Changein Volzirne and Constitutio.12.By HARRY MEDFORTH DAWSON.IN Part I of this work (this vol., p. 1041) an account was givenof the volume changes which occur in the formation of dilutesolutions of iodine and naphthalene. From the fact that thesequence of the solvents, when arranged according to the magnitudeof the volume changes, is quite different for the two dissolved sub-stances, the conclusion was drawn that these changes in volumecannot, in general, be interpreted in terms of any property orproperties of the pure solvents. This was shown to be the casewhether it is assumed that the observed volume change is entirelydue to a change in the volume of the dissolved substance, or whetherthe change is debited entirely t o the solvent.There can be no doubt that neither of these limiting views iscorrect, and according to Tammann (Zeitsch.piiysikal. Chem., 1896,21, 529), the volume change Av occurring in the formation of asolution must in general be represented by an equation of the form :represent respectively the alterationsin volume experienced by the solute and the solvent when thepressures acting on these are changed from the values representingthe internal pressures of the pure substances to the common internalpressure of the solution. The term 2A+ represents the furtherchange in volume which is supposed to take place when the solventand solute are then mixed together without any alteration inthe value of the pressure to which the components of the systemare subjected.SA4, Tammann cites the experi-mental measurements of Braun (Ann.Physilz, 1888, [iii], 34, 943),according to which the mixing of gases, even at atmosphericpressure, is accompanied by appreciable volume changes. Theinference drawn by Tammann from these observations does notappear, however, to be very conclusive. I n the first place, it shouldbe noticed that in Braun's experiments, the different gases wereallowed to mix without change in the manometrically measuredpressure. F o r one and the same external pressure, the gases are,however, subjected to different internal pressures or total pressures,and this is more particularly so in the case of the more easily con-densible gases, such as sulphur dioxide and carbon dioxide, which,hv=A+,+A+,+XA+.. . . . . (1).I n this equation A+l andI n justification of the terFORMATION OF DILUTE SOLUTIONS. PART 11. 1897when mixed with equal volunies of hydrogen or nitrogen, were foundto give rise to relatively large changes of volume a t atmosphericpressure. Furthermore, the internal pressure of the mixed gaseswill, in general, be different from the internal pressures of theunmixed gases, although the manometrically measured pressureretains the same value. Some idea of the magnitude of the volumechanges, which may result from the operation of these internalpressure differences, is obtained if we consider the variation of theproduct pu a t low pressures.According to measurements of Fuchs(Castell-Evans, Pliysico-chernical Tables, p. 450), the value of p vchanges in the ratio 1: 1.0251 for sulphur dioxide, 1: 1.0063 forcarbon dioxide, and 1 : 0.9988 for air when the pressure is reducedfrom 1000 to 250 mm. of mercury.From the above considerations, it would appear that thedeviations from the additive law of gas volumes observed by Braunare not directly applicable to the process of the mixing of liquids(or gases) under the conditions stipulated by Tammann.We may now consider the admixture of two different liquids A,and A, a little more closely from the point of view of internalpressure, or rather from that of the molecular forces of attractionwhich give rise to this pressure.If the specific molecular attractionof the pure liquids is denoted by a, and a?, and the correspondingattraction between the molecules of A, and A, by and if themolecular volumes of the unmixed liquids are vul and v2 and thevolume of the mixture is v,.~, then the internal pressures of A,, A,,and (A, + A2) are given by a1/v12, ~ ~ / v ~ ~ , andrespectively. I n the ideal process of mixing, from the considerationof which equation (1) is derived, the two internal pressures a1/v12and a2/u2, are supposed to be reduced to the common value(a, + a, + Za,.,) / v,.~, before the actual mixing is carried out. Ifthese conditions are fulfilled, it is difficult to see on what groundsa change in volume corresponding with 2A+ is t o be anticipated.I n the author’s opinion the conditions of the process are such as topreclude the possibility of such a volume change, and equation (1)reduces therefore to the simpler form :(a1/u1.,2 + nz/v,.,2+ 2ap,/q.,2)AV=A+,+A+z .. . . . . (2) -I n reference t b the validity of this equation as an expression ofthe volume change which occurs in the mixing of two liquids, itshould be noted that it is tacitly assumed that the mixing processis not accompanied by any change in the molecular camplexity ofeither of the two liquids, that is to say, by association or dissociationphenomena, nor yet by any change in the nature of chemicalcombination between the two sets of molecules. Pairs of substancesYOL. XCVII. 6 2808 DAWSON: CHAKGES IN VOLUME IN THEwhich satisfy these conditions may for convenience be referred t o as" normal " substances, and the solutions obtained by mixing themas " normal " solutions.fromequation (l), the expression for the change in volume which occursin the formation of a solution is greatly simplified.I f the specificvolumes of the two components (solute and solvent) in the pureliquid condition are denoted by q511 and $2, and the true volumesof these components in the solution containing x parts of A, and(1 - x) parts of A, are represented by 4 , s and +,s, the change ofvolume in the formation of unit quantity of solution may bewritten :As a consequence of the elimination of the term S A +Azl=x(+l~-+l~)+ ( l - x ) ( $ 2 q ! q ) . . . . (3)-If P, and P, are the internal pressures of the pure liquids, P, thatof the solution, and if b1 is the mean compressibility of A, (referredto unit quantity) between the pressure limits P, and P3, and Pz isthe corresponding value of the compressibility of A, between P,and P,, the above equation may be written in the form:~\'"=~;P~(P,-P~)+(I-x)P~(P~-P~) .. (4).According to this equation, the change in volume is dependenton the proportions of the two substances in the solution and ontheir compressibilities between certain limits of pressure. Of theselimiting pressures, P, is common to the two substances, and is afunction of x, the value of which determines the composition ofthe solution. According to results obtained by von Biron ( J . Rztss.Phys. C7bem.SOC., 1910, 42, 188) for pairs of liquids which appearto conform to the above definition of " normal " pairs, P, is alinear function of x, and its connexion with the internal pressuresP, and P2 of the pure liquids is given by P3=xP1 + (1 - x)P2. I fthis result is accepted as a first approximadion, substitution for P,in equation (4) leads t o :Av=2(1-X)(Pl-P2)(Pl-Pe) . . . . . (5).I n the case of a dilute solution in which one of the components(the solvent) is present in large excess, x may be neglected in com-parison with unity, and equation (5) may then be written in theform :in which A v / x is the change in volume referred to unit quantity ofthe solute.According to this, the volume change which accompanies theformation of dilute solutions of one and the same substance(P, =constant) in different solvents is dependent on the internalpressure P, of the solvent and on the compressibilities and P , ofA?l/z=(P,-P,)(P1-PJ .. . . . (6)PORMATTON OF DILUTE SOLUTIONS. PART 11. 1899the solute and solvent respectively. The compressibilities are, ofcourse, dependent, on the pressure limits which are involved in aparticular case, and in the absence of a sufficiency of data connectingpressure and compressibility it is not possible to subject equation(6) to a rigorous experimental examination. We may, however,inquire to what extent the relative magnitudes of the changes ofvolume observed in the formation of the dilute solutions investigatedin the first part of this paper are in agreemeilt with the require-ments of this equation, according to which the change in volumeis influenced by two independent factors, one of which may bereferred to as the internal pressure factor, and the other as thecompressibility factor.I f now we compare solvents of approximatelyequal internal pressures, but considerably different compressibilities,then acoording t o equation (6) we may expect to find a connexionbetween the relative volume changes and the compressibilities ofthe solvents. On the other hand, if solvents are compared whichhave approximately equal compressibilities and different internalpressures, it is to be anticipated that the volume changes will berelated to the internal pressure values.Such relationships may not, however, be apparent unless theconditions stipulated in connexion with equation (2) are fulfilled.The solutions must be “ normal ” in the sense already defined;that is to say, the formation of the solutions must not be accom-panied by changes in the degree of complexity of the solute orsolvent, and the solute must not enter into chemical combinationwith the solvent.I n the case of dilute solukioas of iodine which will be speciallydiscussed in this paper, molecular-weight determinations have shownthat the first condition in regard to the solute is fulfilled quitegenerally.Whstover the nature of the solvent, dissolved iodineappears to be present in tlie form of diatomic molecuIes. Experi-ment furnishes no evidence in respect of the constancy of the degreeof complexity of the solvent, but for dilute solutions it is safe toassume that this condition is generally satisfied by the differentsolvents. It is, however, an entirely different matter in regard tothe absence of chemical combination.Recent work relating to the cause of the differences in colourof iodine solutions appears to point to the conclusion that thesedifferences are due to variations in the extent to which iodine iscombined with the solvent.From these investigations, in referenceto which attention may be called to a paper by Waentig (Zeitsch.physikal. Chrm., 1909, 68, 513), it is probable that violet-colouredsolutions of iodine are those which approximate more closely towhat have been termed “normal” solutions than do the red and6 1 1900 DAWSON: CHANGES IN VOLUME IN THEbrown solutions.As the ability to transmit tho more refrangiblevisible rays diminishes, the deviation from this “ normal ” typeincreases, and an increasing proportion of the dissolved iodine ischemicaIIy combined with the solvent.The extent to which the formation of solvates may modify thevolume changes which occur in the formation of solutions cannotbe estimated in any particular case. It is very probable, however,that th.ese complex molecules are formed with a diminution involume the magnitude of which will depend on the specific natureof the solvent. This view is supported by the results of experimentsin which iodine was dissolved in solvents containing soluble poly-iodides. At the same time, some idea is obtained of the extent towhich the chemical combination of the dissolved iodine mayinfluence the observed volume changes.Experiments were made with solutions of potassium iodide andiodine in nitrobenzene and in ethyl acetate.I n these solutions theadded iodine combines to a large extent with the dissolved sub-st,ance or substances already present to form a higher polyiodideor polyiodides, and the difference between the “ solution volumes ”represents therefore the ‘‘ molecular solution volume ” of combinediodine as distinguished from that of free iodine, the correspondingvalues for which are appended for comparison in the table:Nitro- EthylBenzene. acetate.(‘ Solution volume ” of KI + 31 ......... 198.3 C.C.KI+41 ............ 147’3 C.C.“ Molar sol. vol.” of combined iodine 65.2 ) ) combined iodine. 63.2 ),............ Y Y ,, ,, K I + 5 I ...... 133’1 ), I<I + 61 210.5 ,,), . ,, ,, free iodine ...... 67.2 ,, free iodine ...... 64 ‘5 ,,Since the polyiodide solutions in the case of both solvents are goodconductors, and therefore ionised to a considerable extent, and sincechanges in the degree of ionisation are accompanied by appreciablechanges in volume (compare Walclen, Zeitsch. physikal. Chem.,1907, 60, 87), conductivity measurements were made before andafter the addition of iodine to ensure that no changes in volumedue to this factor were involved. From the values of the “ molecularsolution voIume ” of the free and combined iodine, it is evidentthat the combination of the iodine is accompanied by a contractionof 2 C.C.per gram-molecule in nitrobenzene solution, and of 1.3 C.C.in the solution in ethyl acetate. Volume changes of this order ofmagnitude may therefore be expected in the formation of so-calledsolvates.If the change in volume resulting from the chemical combinationis relatively small, the effect of the solvate formation may not besufficient to modify appreciably the influence of those factors whichotherwise determine the connexion between the total volume changFOIlMATION OF DILUTE SOLUTIONS. PART 11. 1901and the nature of the solvent. On the other hand, if the changein volume due to solvate formation is relatively large, the influenceof the internal pressure and the compressibility of the solvent onthe observed change in volume may be completely overshadowed.From equation (6) and the above considerations relating tosolvate formation, it follows that, in general, the volume changeswhich occur when the same quantity of iodine is dissolved indifferent solvents to form dilute solutions are dependent on (1) theinternal pressure of the solvent, (2) the compressibility of thesolvent, and (3) the contraction which accompanies the chemicalcombination of a smaller or larger proportion of the iodine with thesolvent .Whatever the nature of the dissolved substance, these are thodetermining factors, and the only reason for the selection of iodinefor consideration is the fact t.hat more evidence in regard to itsmolecular condition in solution in different types of solvents hasbeen obtained than in the case of any other single substance. Theadvantages attaching to the examination of it coloured substance,such as iodine, which is characterised by specific activity towardsvisible light rays, are not inconsiderable from the point of view ofthe general theory of solutions.‘Other specific properties, such asoptical rotatory power, are of still greater value in connexion withquantitative measurements, and the importance of optical activityon the part of one of the components of a solution in connexionwith the investigation of the nature of solutions has been rightlyemphasised by Winther (Zeitsch. ph~~siJi.~Z. Chent., 1907, 60, 590,641, 685).From the experimental data for solutions of iodine and naph-thalene recorded in Part I.(Zoc. cit.), the conclusion was drawnthat the changes in volume associated with different solvents cannotin general be accounted for by reference to the internal pressuresor other properties of the solvents. This conclusion is consistentwith the result which is now obtained as a consequence of a closerexamination of the factors which are operative in determining thesevolume changes. I n view of the fact that the volume change is, ingeneral, determined by three independent factors, it is not possibleto subject the experimental results to any very detailed analyses.Certain relationships having an obvious connexion with the viewdeveloped in this paper may, however, be referred to.When the fifteen solvents investigated are arranged according tothe magnitude of the change in volume which occurs when iodineis dissolved in them, we obtain a series in which nitrobenzene isthe first member, and ethyl ether the last.Per gram-molecule ofiodine, the formation of a solution in nitrobenzene is accompanie1902 DAWSON: CHANGES IN VOLUME IN THEby an expansion of 8.7 c.c., and in ethyl ether by a contraction of8.0 c.c.* Of all the solvents exa.mined, nitrobenzene is the leastcompressible, and etlhyl ether is by far the most compressible, andsince the internal pressures of the two liquids are not very different,the observed volume changes are obviously in accordance with whatwould be anticipated froin equation (6).With naphthalene asdissolved substance, the relative volume changes in the case of thesetwo solvents are quite similar, in that nitrobenzene, as ths secondmember of the series of fifteen solvents, shows the next largestexpansion, whilst ethyl ether gives rise to by far the largestcontract ion.Carbon tetrachloride and carbon disulphido are two other solventswhich may bo compared. Both yield violet solutions, and thedissolved iodine is therefore probably present, for the most part,in the free condition; their conipressibilities also appear to beapproximately equal. I n these circumstances a connexion betweenthe volume changes and the internal pressures of the solvents maybe expected. According to Walden and Traube: the interfialpressure of carbon disulphide is about 50 per cent.greater thanthat of carbon tetrachloride, and in agreement with this it is foundthat the solution of one gram-molecule of iodine in carbon tetra-chloride is accompanied by an expansion of 7.3 c.c., whereas thecorresponding expansion for the solution in carbon disulphide isonly 2.6 C.C. When the volume changes associated with theformation of solutions of naphthalene in these two solvents arecompared, the relative positions are, however, reversed. For agram-molecule of naphthalene, there is a contraction of 0.75 C.C.in carbon tetrachloride, and an expansion of 1.35 C.C. in carbondisulphide. I n the case of naphthalene there is no evidence ofthe molecular condition of the dissolved substance apart from itsdegree of complexity, and the relative volume changes lead to thesupposition that the naphtha.lene enters into combination withcarbon tetrachloride t o form addition compounds to a greater extentthan it does in the case of carbon disulphide.In reference to the view that brown solutions of iodine containa considerable proportion of the iodine in combination with thesolvent, the fact may be mentioned that all solutions which containcombined iodine in the form of polyiodides of the alkali and alkalineearth metals and of hydrogen appear to have a brown colour.Incertain cases, as, for example, the brown solution in pyridine, directevidence of combination of iodine with the solvent has been obtainedby the isolation of a crystalline compound (Waentig, Zoc.c i t . ) .In others, the brown colour is probably due in it large measure to* These changes in volnnia refer t o the liquid soluteFORMATION OF DTLUTE SOLUTIONS. PART 11. 1903iodine compounds which are forEed as a result of double decom-position. Acetone, which has been frequently used as a solventfor iodine in investigations relating to the cause of the differencesin colour exhibited by iodine solutions, belongs to this class.It has been shown (Dawson and Leslie, Trans., 1909, 95, 1860)that iodine reacts rapidly with acetone, forming hydrogen iodideand acetone, and that a condition of equilibrium is established asa consequence of the reversibility of the reaction. In t.he case ofa solution, prepared by dissolving one gram-molecule of iodine in100 gram-molecules of acetone, about 45 per cent. of the iodineis converted into hydrogen iodide. I n the resulting solution, theiodine, which has not been acted on by the acetone, is largelycombined with the hydrogen iodide in the form of a polyiodide,and, in all probability, the colour and the absorption spectrum ofthis solution correspond with that of iodine, which is thus combinedwith one of the products resulting from the interaction of thesolute and the solvent. Other ketones react similarly with iodine,and it is not unlikely that the colour of solutions of iodine in othersolvents may be in part due to the same cause.These facts add weight to the conclusion that solutions of iodinedo not, in general, belong to the type of ‘(normal” solutions asdefined in this paper, and on this account the absence of any generalconnexion between the ineasured changes in volume and theformation of these solutions and the internal pressures and com-pressibilities of the solvents is not surprising. It may be thatsolutions of naphthalene approximate more closely to this type,but, in any case, the observed lack of parallelism between the volumechanges for these two substances in the same series of solvents isconsistent with the view that the dissolved substances combine toan appreciable extent with the various solvents.PI~YSICAL CHEMISTRY LABOI~ATOILY,THE UNIVERSITY,LEEDS

ISSN:0368-1645    

DOI:10.1039/CT9109701896

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

1904 BAMFOBD AND SIMONSEN : THE CONSTITUTIONC C I I. - The Corist it u tion of the Be nxe IZ c. t e t r a car box y 1 icAcids.By HANNAH BAMFORD and JOHN LIONEL SIMONSEN.DURING the course of an investigation which is described in thesucceeding paper (p. 1910), it became necessary definitely to deter-mine the constitution of two of the benzenetetracarboxylic acids,namely, prehnitic and mellophanic acids. The three isomericbenzenetetracarboxylic acids were first obtained by Baeyer(AnnuZen, 1873, 166, 325) during the course of his classicalresearches on mellitic acid. It is unnecessary here to enter into thereasoning by which he deduced the formulze to be ascribed to theseacids, and it will suffice to point out that whilst he proved con-clusively (Zoc. cit., p.344) pyromellitic acid to be benzene-1 : 2 : 4 : 5-tetracarboxylic acid, he was unable to decide whetherprehnitic acid was benzene-1 : 2 : 3 : 4- or -1 : 2 : 3 : 5-tetracarboxylicacid. He inclined to the latter view, however (Zoc. cit., p. 344),because he found that this acid only yields a monoanhydride,whilst i f it had possessed four adjacent carboxyl groups it shouldgive a dianhydride.The question remained undecided until 0. Jacobsen (Rer., 1884,17, 2517) by the oxidation of isodurene obtained an acid whichwas shown by direct comparison to be identical with Baeyer’s mello-phanic acid. Now, since isodurene was obtained from bromo-mesitylene, it followed that mellophanic acid must be benzene-1 : 2 : 3 : 5-tetracarboxylic acid, and hence prehnitic acid would bethe 1 : 2 : 3 : 4-isomeride.Purther evidence in support of this viewwas obtained by Tohl (Bey., 1888, 21, 907), who oxidised prehniteneand isolated an acid which showed all the properties of prehnitic . - acid.As the present authors had obtained results which seemed tothrow some doubt on these experiments, it was decided to endeavourto devise methods for preparing prehnitic and mellophanic acidswhich would leave no doubt as to their constitution.Me Me C0,H C0,Me/\Bc /\CO,H /\CO,T€ /\CO,TIMe!,,!Me Me\lMe CO,H~,)CO,H CO,MJ \/ ICO,FI(1.) (11.) (111.) UV.1C0,Et/\CO,HCO,Et/ IU0,Et(V. 1\OF THE BENZENEl'ETRhCBRBOXYLIC ACIDS. 1905A ready means of carrying this out appeared to be to preparemesitylenecarboxylic acid and to oxidise this to the correspondingtetrabasic acid.Mesitylenecarboxylic acid (11) was found to be most convenientlyobtained by the action of magnesium and carbon dioxide on bromo-mesitylene (I), and on oxidation with nitric acid under the con-ditions described in the experimental part of this paper (p.1907),it was found to be converted quantitatively into benzene-1 : 2 : 3 : 5-tetracarboxylic acid (111).The substance obtained in this way showed all the propertiesascribed by Baeyer to the acid called by him prehnitic acid, andthere can therefore be little doubt that this acid is really benzene-1: 2 : 3 : 5-tetracarboxylic acid. This view also receives supportfrom the following stereochemical considerations : (a) The acid yieldsonly a monoanhydride (Baeyer, Zoc.c i t . ) ; ( b ) when esterified in thecold with methyl alcohol and hydrochloric acid it gives it dimethylester (compare Meyer and Sudborough, Rer., 1894, 27, 1591), whichprobably has the constitution (IV); (c) when, however, the acid isesterified by means of alcohol and sulphuric acid by heating onthe water-bath in the usual manner, it yields a triethyl ester whichundoubtedly has the constitution (V).Having shown conclusively that prehnitic acid is benzene-1 : 2 : 3 : 5-teti-acarboxylic acid, it seemed that mellophanic acidmust be benzene-1 : 2 : 3 : 4-tetracarboxylic acid, but in order to becertain it was decided to synthesise this acid.For this purpose a large amount of 1 : 4-dirne%hylnaphthalene wasprepared, and this, on oxidation with nitric acid, gave a mixture ofacids, from which a tetrabasic acid was separated; this possessedall the properties of Baeyer's mellophanic acid.When the silver salt of the acid was heated with methyl iodide,a neutral tetramethyl mellophanate (m.p. 1 3 3 O ) was obtained.Since mellophanic acid is 1 : 2 : 3 : 4-benzenetetracarboxylic acid,it should, in accordance with the usual views on steric hindrance,yield only a diacid ester by the direct process of esterification.This is probably the case, since no trace of neutral tetraethyl mello-phanate was produced on heating the acid with alcohol and sulphuricacid. A barium salt was isolated by treatment of the reactionproduct with baryta and subsequent evaporation, and this gave ontitration results which indicated it to be a barium salt of thediacid ester with two or three molecules of water of crystallisation.The direct determination of the water of crystallisation gave entirelyanomalous results, a.nd this important point, will therefore be re-investigated.It is clear, however, that the acid cannot be com-pletely esterified by the usual process1906 BAMFORD AND SIMONSEN : THE CONSTITUTIONAs the result of these experiments there can therefore be littledoubt that prehnitic acid and mellophanic acid are respectivelybenzene-1 : 2 : 3 : 5- and -1 : 2 : 3 : 4-tetracarboxylic acids.It is somewhat difficult to see how Jacobsen (Zoc. c i t . ) obtainedfrom isodurene the acid which he showed t o be identical withBaeyer's mellophanic acid, since he prepared his hydrocarbon bythe action of methyl iodide and sodium on bromomesitylene.It ishardly likely that any rearrangement of groups would take placeduring this reaction, as such rearrangement only appears to takeplace in the presence of aluminium chloride. We are thereforedriven to the conclusion that the hydrocarbon used by Jacobsen inhis experiments cannot have been mesitylene, and must have con-sisted mainly of q-cuinene.It has already been mentioned that Tohl (Zoc. c i t . ) obtainedprehnitic acid by the oxidation of prehnitene, which hydrocarbon isprepared by the degradation of pentamethylbenzene by means ofsulphuric acid. He, however, offers no proof of the constitution ofhis hydrocarbon apart from the fact that it was identical withthat previously obtained by Jacobsen (Ber., 1886, 19, 1213; 1887,20, 901).Jacobsen isolated this hydrocarbon by the action ofhydrochloric acid on a, sulphonamide prepared from durene and bythe degradation of pentamethylbenzene, two reactions which offerno proof of constitution, and the only grounds for considering thatprehnitene is 1 : 2 : 3 : 4-tetramethylbenzene is because it is isomericwith the hydrocarbon he obtained by the action of methyl iodideand sodium on bromomesitylene, and that it does not give pyro-mellitic acid on oxidation. I f , then, the hydrocarbon obtained byJacobsen from bromomesitylene is 1 : 2 : 3 : 4-tetramethylbenzene,then prehnitene must be 1 : 2 : 3 : 5-tetramethylbenzene.Further experiments on this interesting subject are in progress.Unfortunately, the wrong conception of the nature of prehniticacid and of prehnitene, which has so long'prevailed, has led tomuch confusion in assigning constitutional formulz on the basis ofthese substances being 1: 2 : 3 : 4-substitution derivatives ofbenzene.EXPERIMENTAL.Prehnitic Acid (Benzene-1 : 2 : 3 ; 5-tetracarboxylic Acid).Mesitylenecarb oxylic A cid.-Broniomesitylene (20 grams) wasadded to magnesium (2-4 grams) suspended in dry ether, themagnesium having first been treated with it small amount of methyliodide in order to render it reactive.A vigorous reaction ensued,and when this had ceased, the mixture was heated on the water-bath until all the magnesium had passed into solution.After beingkept cold overnight, a slow stream of dry carbon dioxide waOF THE BENZENETETRACARBOXYLIC ACIDS. '1 907passed into the ethereal solution of the magnesium compound, whena heavy oil slowly separated. Ice and dilute hydrochloric acid werethen added, and the ethereal layer was separated and washed withsodium csrbonate solution.On acidifying the latter solution, a colourless, crystalline pre-cipitate (7.5 grams) separated, which was collected and dried. Themesitylenecarboxylic acid obtained in this way was purified bycrystallisation from light petroleum, from which it separated inglistening prisms, possessing the correct melting point, 151-152'.(Found, C = 72.9 ; H = 7.4.Oxidation of Mesitylenerarboxylic Acid t o I'rehnitic Acid.-Mesitylenecarboxylic acid (1 gram) was mixed with 10 C.C.of dilutenitric acid (D 1-15>, and hea.ted in a sealed tube f o r eight hoursat 170-180°. After removing the nitric acid by evaporation onthe water-bath, the prehnitic acid was purified by repeatedcrystallisation from hydrochloric acid, from which it separated insmall prisms. After drying at looo to remove water of crys-tallisation, i t was analysed. (Found, C = 46.8 ; H = 2.3. Calc.,C =47.2 ; 11 = 2.3 per cent.)Prehnitic acid melts at 252O, previously softening at about 240°,and is converted into the anhydride, melting at 238O, whereasBaeyer (Annalen, 1873, 166, 328) states that it begins to melt at237O, becoming completely liquid a t 250°, whilst the anhydridemelts at 238O.That it was a tetrabasic acid was shown by titrationwith standard sodium hydroxide, when it was found that 0.096neutralised 0*058NaOH, whereas a tetrabasic acid, C,,H,O,, shouldneutralise 0'06008NaOH. When barium chloride was added to anaqueous solution of prehnitic acid and the solution warmed, animmediate precipitate of the barium salt of the acid was obtained.After drying at the temperature of the laboratory for two days, itwas analysed, when it was found to have the composition givento the barium salt of prehnitic acid by Baeyer (Annalen, 1873,166, 332):Calc., C = 73.4 ; H = 7.3 per cent.)0.2354 lost 0.0122 H,O at looo.0.1514 gave 0.0502 BaSO,. Ba=19*4.(CloH508)2Ba,3H20 requires Ba= 19.6 ; 2H20 =5.4 per cent.The barium salt, when dried at looo, gave the following result:0.2197 gave 0.07'15 BaSO,.(CloH,08)2Ba,H20 requires Ba = 20.8 per cent.Tetramethyl Prehnitate, C,H2(CO2Me),.-Tetramethyl prehnitatehas already been isolated by Baeyer ( A n d e m , 1873, 166, 333) bythe action of methyl iodide on the silver salt of the acid.Arepetition of this process yielded the ester in beautiful, colourlessH20=5.2.Ba= 20.701908 BAMFORD AND SIMONSEN : THE CONSTITUTIONneedles, which, after crystallisation from methyl alcohol, melted at108-109° (Baeyer gives 104-109°). (Found, C = 54.1 ; H = 4-5.Calc., C=54*2; H=4.5 per cent.)Trie t hyl Hydrogen Prelinitat e, C,H,( CO,E t),*CO,H.This ester, which does not seem t.0 have been previously obtained,is readily formed when prehnitic acid is treated with alcohol andsulphuric acid in the usual manner.It was purified by repeatedcrystallisation from a mixture of benzene and light petroleum,when it separated in colourless clusters of needles, which melt a tabout 108-llOo :0.123 gave 0.2625 CO, and 0.0601 H20.Triethyl hydrogen prehnitate is readily soluble in most organicIts probable con-C=58*2; H=5-4.C,,H,,O, requires C = 58.3 ; H = 5.1 per cent.solvents with the exception of light petroleum.stitution has already been discussed in the introduction (p. 1905).DintetAyl Di?Aydrogen Pre?Lnitate, CGH2(CO2Me),(CO,H),.*It has already been pointed out by Meyer and Sudborough (Ber.,1894, 27, 159) that when a methyl-alcoholic solution of prehniticacid is trea-ted with hydrogen chloride in the cold, only a dimethylester is formed.This result has been confirmed, but the meltingpoint of the ester when crystallised from a mixture of acetone andlight petroleum was found to be about 19l0, and not 176-177O,as stated by them. The silver salt was analysed:0.0519 gave 0.0225 Ag. Ag=43*3.CI2H,O8Ag2 requires Ag = 43.5 per cent.MeZlop?ianic Acid (Benzene-1 : 2 : 3 : 4-tetracarboxylic Acid).Oxidation of 1 : 4-Dimethyln.aphthaZene. - 1 : 4-Dimethylnaph-thalene (2 c.c.), prepared as described by Giovannozzi (Gazzetta,1882, 1'2, 147), was mixed with 40 per cent. nitric acid (20 c.c.),and heated in a sealed tube at 170-180° for eight hours. Theclear solution was evaporated to dryness on the water-bath, whena semi-solid, crystalline mass was obtained, which evidently con-sisted of a mixture of acids.I n order to separate the mellophanic acid, a somewhat tediousprocess of purification had to be adopted.The crude acid wasdissolved in as little water as possible, and the solution treated withexcess of stannous chloride and concentrated hydrochloric acid fortwo hours on the water-bath. The strongly acid mixture was then* By an oversight, Stewart states (.S'tcreochcmistry, p. 341) that prehriitic acidforms a tetramethyl ester when estcrified in the usual manlierOF THE BENZENETETRACARBOXYLIC ACIDS. 1909evaporated to dryness, the residue dissolved in water, and the tinprecipitated with hydrogen sulphide. After separating the tinsulphide, the solution was again evaporated to dryness, and themixture of phthalic and mellophanic acids extracted with dry etherin a Soxhlet apparatus, when it was found that the arhino-acids,which were present ;ts hydrochlorides, were not dissolved.As pre-liminary experiments showed i t to be a matter of some difficultyto separate the mellophanic and phthalic acids by fractionalcrystallisation, they were converted into their silver salts in theusual manner, and the dry silver salts were digested with methyliodide in benzene solution for some hours on the water-bath. Afterseparating from the silver iodide, the benzene was removed, whena viscid oil was obtained, which on trituration with ether partlysolidified. The solid ester, which was found to consist of methylniellophanate, was separated and purified by repeated crystallisationfrom methyl alcohol :0.124 gave 0.2455 CO, and 0.0552 H,O.C,,H,,O, requires C = 54.2 ; H = 4.5 per cent.Me.thyZ melloph.anate, which does not appear to have been pre-viously described, separates from methyl alcohol in long, glisteningneedles, melting at 133-135O.It is readily soluble in benzene,ethyl acetate, or acetone, but only sparingly so in ether, lightpetroleum, or cold methyl alcohol,I n order to obtain the free acid, the methyl ester was hydrolysedin the usual manner with alcoholic potassium hydroxide, and, afterremoving the alcohol on the water-bath, the solution was acidified,evaporated to dryness, and the acid extracted with ether in aSoxhlet apparatus. The crude acid was purified by crystallisationfrom hydrochloric acid, when it was found to soften slightly at225O, and melt and decompose at 238O with formation of theanhydride (compare Baeyer, Zoc. cit.). (Found, C =47*02 ; H = 2.3.Calc., C=47*2; H=2.3 per cent.) The basicity of the acid wasdetermined by titration with normal sodium hydroxide, when it wasfound that 0.0648 neutralised 0*0404NaOH, whereas a tetrabasicacid, C,,H,08, should neutralise 0*04081NaOH.Mellophanic acid is readily soluble in water or acetone, but onlysparingly so in most organic solvents.C = 54.0; H = 4.9.THE UNIVERSITY,MAXGHESTER.THE PRESIDERCY COLLEGE,MADRAS

ISSN:0368-1645    

DOI:10.1039/CT9109701904

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

1910 SIMONSEN : ETHYL 6-METHYL-2-PYROXE-3 : 5-DICARBOXYLATEC C I I I. - E t h y I 6 - $Ie t h y I - 2 -p y r o ri e - 3 : 5 - die a "r" b o xg la t eand its Convcmion into Jfetlzyltrir/zesic Acid.By JOHN LIONEL SIMONSEN.IN a recent communication (Trans., 1908, 98, 1022) the authordescribed the preparation of ethyl 6-methyl-2-pyrone-3 : 5-di-carboxylate, and pointed out that on hydrolysis with alkalinereagents it gave rise to an acid to which the empirical formulaC7H604 was given, and i t was suggested that this might possiblybe 1-methyl-A1 :%cyclobutadiene-Z : 4-dicarboxylic acid, internal con-densation having taken piace in the following manner :Since an acid possessing this constitution would be of theoreticalinterest from several points of view, further experiments wereinstituted with the object of subjecting it to a more searchinginvestigation.It seemed, in the first place, desirable t o devise a mea.ns ofdirectly synthesising the ethyl ester of a-acetylglutsconic acid (I),since, if the above reasoning were correct, this ester should, on treat-ment with suitable reagerib, readily be converted into 1-methyl-A1 :3-cycZobutadiene-2 : 4-dicarboxylic acid (11).This synthesis waseasily carried out by the condensation of ethyl propiolate and ethylsodioacetoacetate, when the reaction proceeds quite normally withthe formation of ethyl a-acetylglutaconate,CO,Et.CiCH + CH,Ac*CO,Et -+CO,Et*CH: CH-CHAc.CO,Et,(1.1the reaction proceeding on somewhat similar lines to the con-densation of ethyl phenylpropiolate and ethyl acetoacetate soexhaustively investigated by Ruhemann (Trans., 1899, 75, 251).A t the same time, however, a further reaction takes place, anda small quantity of a crystalline solid was isolated, which meltedat 48O, and was found t o be identical with the ethyl ester of theacid previously obtained by the hydrolysis of ethyl 6-methyl-2-pyroae-3 : 5-dicstrboxylate (loc.cit., p. 1028). The isolation of thisester was at first thought to be additional evidence in favour oAND ITS CONVERSION 1XTO ME'I'HYLTRIMESIC ACID. 1911the original view of the constitution of the acid C7H604, whichwas, moreover, supported by the fact that ethyl a-acetylglutaconate,on hydrolysis with barium hydroxide, yielded, besides glutaconicacid, a small quantity of the same acid (see p.1914). As the resultof further experiments this view had to be abandoned, and theconclusion was drawn that this acid, supposed t o be C,H604, wasin all probability a methylbenzenetricarboxylic acid, C,,H,O,.During the course of his classical investigation of coumalinic acid,von Pechmann (Annalen, 1891, 264, 293) showed that when themethyl ester of coumalinic acid was hydrolysed with aqueoussodium hydroxide it was converted into the monomethyl ester oftrimesic acid, the reaction taking place on the following lines:Taking these facts into consideration, it appeared probable thatthe hydrolysis of ethyl 6-methyl-2-pyrone-3 : 5-dicarboxylate hadproceeded in an analogous manner wit,h the formation of a methyl-trimesic acid :-+ CH*CO,Et OH*EI'Ic=%( CO,E L) : c Me.OH @H*CO,Et(1. )C . < ~ ' i ~ " , l , " , i ~ ~ ~ ~ ~ * C O , H c H<C( CK-CO cO,&): Et CMe. (,H - +CH(OH):CH*CO,Et + C€I,Ac*CO,Et.The 2-pyrone (111) is Converted, in the first place, into thediethyl ester of a'-carboxy-a-acetylglutaconic acid (IV), which, withloss of carbon dioxide, passes into ethyl a-acetylglutaconate (I).Half of the ethyl a-acetylglutaconate then undergoes decompositioninto ethyl formplacetate (V) and ethyl acetoacetate, and the ethylformylacetate thus generaked combines with the unchanged ethyla-acetylglutaconate, with loss of two molecules of water andformation of methyltrimesic acid (VI).I f this constitution be accepted for the acid, the formation of itsethyl ester by the interaction of ethyl propiolate and ethyl aceto-P.1912 SIMONSEN : ETHYL 6-METHYL-2-PYRONE-3 : 5-DICARBOXTLATEacetate is readily explained. It may be assumed that two moleculesof ethyl propiolste condense with one of ethyl acetoacetate inaccordance with the following scheme :Me Me I0 c!/\\/FCO,Et*CH, C*CO,Et CO, Et*F 8- CO,E t.--3 OH CH Ct1 CHN. c; C0,Et C0,EtOn examining the literature, it was found that only one othermethylbenzenetricarboxylic acid is known. This was obtained byDoebner (Annulen, 1900, 31 1, 136) by the condensation of pyruvicacid and glyoxylic acid and fusion of the resulting phthnlidetri-carboxylic acid with potassium hydroxide :0,HOH*? *OHy H 3 H YH3C02H/\I 1 CO,H*CO CO*CO,H ---+ \,CO,H --+OHp 2C0,HC0,HC0,HC0,H(VIJI.)/\CO,EICH\,hO,H -+I I 0-0(XI.)\/ C0,H(VII.)C0,H C0,H-+ /'\Me -+!,!CO,HC02H UO,H(IX.) (X.1C0,H CO,H/\ /\\/(XII.) (XI11 )MA ~ O , H -+ CO,H//CO,FICO,H C02HDoebner considered that benzene-1 : 2 : 4-tricarboxylic acid (VII)was first formed, which then condensed with a further molecule ofglyoxylic acid to give phthalidetricarboxylic acid (VIII), and this,on subsequent fusion with potassium hydroxide, gave l-methyl-benzene-2 : 3 : 6-tricarboxylic acid (IX). He based this view of thaconstitution on the fact that on oxidation with potassium per-manganat,e, phthalidetricarboxylic acid gave prehnitic acid, whicAND ITS CONVERSION INTO METHPLTRIMESIC ACID.1913he regarded as benzene-1 : 2 : 3 : 4-tetracarboxylic acid (X). Now,since it has been shown in tke preceding communication (p. 1904)that prehnitic acid is benzene-1 : 2 : 3 : 5-tetracarboxylic acid(XIII), it follows that the acid obtained by Doebner cannot bel-methylbenzene-2 : 3 : 6-tricarboxylic acid.Probably the condensation described by Doebner proceeds on thefollowing lines. Benzene-1 : 2 : 4-tricarboxylic acid (VII) is firstformed, and this then condenses with glyoxylic acid to give theisomeric phthalidetricarboxylic acid (XI), which, on fusion withpotassium hydroxide, would give l-methylbenzene-2 : 3 : 5-tri-carboxylic acid (XII), and on oxidation with potassium per-manganate, prehnitic acid (XIII). This view of the condensationis also more probable on stereochemical grounds.I n order to leave no doubt as t o the constitution of the methyl-trimesic acid obtained in this investigation, it was oxidised withnitric acid or potassium permanganate, when the tetracarboxylicacid was formed, and was shown to be in every way identical withprehnitic acid (benzene-1 : 2 : 3 : 5-tetracarboxylic acid).EXPERIMENTAL.Condensation of Ethyl A cetoacetate and Ethyl Propiolate.Foi.mation of E t Tzyl Met hylt rimesat e and Ethyl a-A cet ylglut aconat e.Sodium (1.3 grams) was dissolved in alcohol, and, after thea.ddition of ethyl acetoacetate (7.8 grams), ethyl pr*opioIate (5.9grams) was added to the well-cooled solution.Much heat wasgenerated, and the reaction mixture rapidly became deep red.After being kept overnight, water was added, and the oil whichseparated was dissolved in ether, and the ethereal solution washed,dried, and evaporated.The residual oil was carefully fractionatedunder diminished pressure (14 mm.), when, after a small amountof unchanged ethyl acetoacetate had passed over, it was found todistil between 210° and 220O. The distillate (3 grams) was cooledin a freezing mixture, when it rapidly solidified, and after drainingon porous porcelain it was crystallised from light petroleum, whenit melted at 48O:0.1315 gave 0.3013 CO, and 0.0796 H,O.C16Hl,0, requires C = 62.7 ; H = 5'9 per cent.The ethyl methyltrimesate obtained in this manner was found tobe in every respect identical with that previously described (Zoc.cit.,p. 1028), and a mixture of the two melted sharply a t 48O.In order to remove any doubt as to the identity of these twosubstances, a small quantity of the pure ester (1 gram) was hydro-lysed with dilute hydrochloric acid, and after removing the excessC=62.5; H=6*7.VOL. XCVII. 6 1914 SIMOKSEN: ETHYL 6-METHYL-2-PYRONE-3: 5-DICARBOXYLATEof hydrochloric acid by evaporation, the methyltrimesic acid waspurified by repeated crystallisation from hot water. After dryingat l l O o , it was analysed :0.1097 gave 0.2152 CO, and 0.0395 H,O.C,,H,O, requires C=53*6; H=3.6 per cent.On melting, it behaved in the manner characteristic of this acid,commencing to sinter at 268O, and decomposing completely at 300O.The alkaline mother liquor, from which the ethyl methyltrimesatehad been separated, was rendered just acid with dilute hydrochloricacid, when an oil separated. This was dissolved in ether, the etherealsolutdon washed with dilute sodium carbonate," dried, andevaporated, and the residual yellow oil fractionated underdiminished pressure (12 mm.), when it was found to distil veryconstantly at 158-160°.Yield 5 grams:0.1535 gave 0.3244 00, and 0.0997 H,O.CllH,,0, requires C = 57.9 ; H = 7.0 per cent,Ethyl a-acetylglutacomfe, which does not appear to have beenpreviously described, is a colourless oil, possessing a pleasant etherealodour. Its alcoholic solution gives with ferric chloride an intensepurple-violet coloration.Hydrolysis of Ethyl a-Acetylg1utaconate.-For the hydrolysis ofthis ester, barium hydroxide was found to yield the most satisfactoryresults.Ethyl acetylglutaconate (5 grams) was mixed with a concen-trated solution of barium hydroxide [lo grams crystallisedBa(OH),], and afhr boiling for two hours on the sand-bath, thecanary-yellow solution was cooled and extracted ten times withpure ether.On removing the ether an oil was obtained whichrapidly solidified, and after draining on porous porcelain wascrystallised from dry ether, when it was found t o separate in twodistinct forms, consisting of fine, colourless needles and small, hard,yellow nodules. These were separated, as far as possible, mechani-cally, and the needles, which weighed less than 0.1 gram, were foundto consist of methyltrirn esic acid, since they showed the characteristicmelting point of this acid, and when esterified with methyl alcoholand sulphuric acid yielded an ester melting at 107O.The main portion of the acid, which separated in hard, yellownodules, was purified by repeated crystallisation from ether withthe aid of animal charcoal, when it was obtained in colourlessprisms, melting at 138O, and evidently consisted of glutaconic acid,which melts at this temperature (Buchner, Ber., 1890, 23, 703).(Found, C =45.9 ; H = 4.7.* The sodlum carbonate washings, on acidification, yielded a small amount of aviscid oil which was not further invcstigated.C=53*5; H=4*0.C = 57.6 ; H = 7.2.Calc., C = 4 6 2 ; H =4.6 per cent.AND ITS CONVERSION INTO METHYLTRIMESIC ACID. 1915The properties of methyltrimesic acid have already been described(Zoc. cit., p.1027), but in view of the somewhat unsatisfactoryanalytical data obtained, further experiments were instituted withthe view of obtaining this acid in a purer state. As it waa notreadily purified by repeated crystallisation from hot water, the acidwas dissolved in a slight excess of sodium carbonate and oxidisedwith a dilute solution of potassium permanganate in the cold untila permanent pink colour was obtained. After removing the excessof potassium permanganate with sulphur dioxide, the manganesedioxide was removed, and the alkaline solution concentrated andacidified. The precipitated acid was collected and repeatedlycrystallised from hot water, from which it separates, when pure, inlong, prismatic needles, melting at the temperature previouslygiven :0.1606 gave 0.3146 (20, and 0.0532 H,O.C = 53.5 ; H =3*7.0.1392 ,, 0.2747 CO, and 0.0459 H20. C=53*8; H=3.6.ClOH8O6 requires C=53*6; E=3*6 per cent.The silver salt separates as a caseous, white precipitate whensilver nitrate is added to a faintly alkaline solution of theammonium salt. After drying at looo, it wils analysed:0.1915 gave 0.114 Ag. Ag=59*5.Methyt Methyltrimesate.-This ester, which has been previouslydescribed, when pure melts at 107O, and not a t 106O (Zoc. cit.,p. 1028). A pure specimen was prepared for the determinationof the refractive index; on analysis it gave the following figures :C,,H,0,Ag3 requires Ag = 59.4 per cent.0.1349 gave 0.2888 CO, and 0.0626 H,O.C13H,,06 requires C = 58.6 ; H = 5.2 per cent.Methyl methyltrimesate is also obtained in a yield of 96.5 percent.when methyltrimesic acid is esterified by the method usedby Meyer and Sudborough (Ber., 1894, 27, 1591). The acid (0'5gram) was dissolved in methyl alcohol (10 c.c.), and the solutionsaturated in the cold with hydrogen chloride. The acid immediatelyseparated out, but on keeping in the cold slowly redissolved, andafter twelve hours the liquid was filled with a mass of feltedneedles. These were dissolved in ether, when 0.56 gram of themethyl ester was obtained. This result is of some interest, as, fromstereochemical considerations, it seemed unlikely that the trimethylester would be formed, especially since under similar conditionsprehnitic acid only yields a dimethyl ester (see p.1908).Mr. R. T. Hardman kindly carried out a molecular-weight deter-C=58*4; H=5.1.6 ~ 1916 ETIlYJ; G-RIETHYL-2-PYRONE-3 : 5-DICARBOXYLATE.mination of this ester by the cryoscopic method, benzene being usedas the solvent:0.267, in 17.537 benzene, gave A t = - 0 . 3 O . M.W. = 253.7.0.4052 ,, 17.57 77 ,, A t = - 0.456’. M.W. ~ 2 5 3 . 0 .C,,H,,06 requires M.W. = 266.I am much indebted t o Dr. I d a Smedley for very kindly deter-The determination was mining the refractive p-ower of this ester.carried out in chloroform solution :Percentagestrength of solution. Ma. nq3. My .My-=.6.5986 65.47 67-77 68-79 3 32Calculated value C,,H,,0’’30’, 311 = 63.23.It seems likely that the high value is due to the conjugation ofthe carbonyl groups with the benzene ring.Distillatwn of the Barium Salt of Methyltrimesic Acid-Thebarium salt (30 grams) was mixed with barium oxide (120 grams),and carefully distilled from a retort, when an oil slowly passedover.This was purified by repeated distillation over sodium, whenit was found to boil constantly a t l l O o , and evidently consisted ofnearly pure toluene. Calc., C = 91.3 ;H = 8.7 per cent.)(Found, C = 90.7 ; H = 8.7.Oxidation of Methgltrimesic Acid t o Prehnitic Acid.I. With Nitric Acid.-Methyltrimesic acid (2 grams) was mixedwith 40 per cent. nitric acid (20 c.c.), and heated in a sealed tubefor six hours at 170-180°. The clear solution was evaporated todryness, and the solid acid thus obtained was purified by repeatedcryst allisation from hydrochloric acid, when it separated in micro-scopic prisms.After drying at looo, it was analysed. (Found,C = 47.2 ; H = 2.5, Calc., C = 47.2 ; H = 2.3 per cent.) Prehnitic acidobtained in this manner melted a t 250-251°, softening slightly at2 4 7 O , with formation of the anhydride, which melted at 238O. Itwas found to be identical in every way with prehnitic acid, obtainedby the oxidation of mesitylenecarboxylic acid (see p. 1907). Thebasicity of the acid was determined by titration with standardsodium hydroxide, when 0.1095 of the acid neutralised 0*068NaOH,whereas this amount of a, tetrabasic acid, C,,H,08, should neutralise0.0 685 NaO H.11.With Potassium Permanganate.--In carrying out thisoxidation, methyltrimesic acid (5 grams) was dissolved in a slightexcess of dilute sodium carbonate solution, and after the additionof potassium permanganate (10 grams) dissolved in water (500 c.c.),the mixture was boiled in a, reflux apparatus for twelve hours,when the oxidation was complete. After removing the manganesTHE VOLATILE CONSTITUENTS OF COAL. 1917dioxide, the alkaline solution was concentrated on the water-bath,acidified, and extracted several times with ether, the etherealsolution being dried and evaporated. The crystdlline acid thusobtained was purified by crystallisation from hydrochloric acid, whenit was found to melt at 250-251°, with formation of the anhydride,which melted at 238O. This acid was in every way identical withthat obtained in the oxidation with nitric acid. The silver saltseparates as a caseous, white precipitate on the addition of silvernitrate t o a slightly alkaline solution of the ammonium salt. (Found,C=17*2; H=0*7; Ag=62*9. Calc., C=17*6; H=0*3; Ag=63*4per cent.)In order to leave no doubt as to the identity of this acid withprehnitic acid, the characteristic barium salt was prepared by theaddition of barium chloride to a warm aqueous solution of theacid. After drying for two days in the air it was analysed. (Found,Ba = 19.6 ; H,O =5*0. (C,,H,0,)2Ba,3H,0 requires Ba= 19.6 ;2H,O=5'4 per cent.) It was also analysed after being dried atlooo. (Found, Ba= 20.6. (C,,H,O&Ba,H,O requires Ba= 20.8 percent.)Tetramethyl prehnitate was prepared in the usual manner by theaction of methyl iodide on the silver salt of the acid, and aftercrystallisation from methyl alcohol melted at 108-109°. (Found,C=53.7; H=4.6. Calc., CTx54.2; H=4.5 per cent.)The author wishes to thank the Research Fund Committee of theChemical Society for a grant which has defrayed much of theexpense of this and the preceding communication.THE UNIVERSITY,MANCHESTE

ISSN:0368-1645    

DOI:10.1039/CT9109701910

出版商:RSC    

年代:1910

数据来源: RSC

摘要:

THE VOLATILE CONSTITUENTS OF COAL. 1917CC1V.-The Volatile Constituents of Coal.By MAURICE JOHN BURGESS and RICHARD VERNON WHEELER.THIS investigation into the nature of the '' volatile constituents "of coal has been undertaken in connexion with experiments nowbeing conducted by the Mining Association of Great Britain.These experiments--" the British Coal Dust Experiments "-havefor their object the study of the phenomena occurring during theexplosive combustion of mixtures of fine coal dust and air, with aview to discover a means of preventing such explosions.An investigation into the nature of the volatile constituents ofcoal is of particular value in connexion with such a problem,because of the obvious relationship thak exists between the behaviou1918 BURGESS AND WHEELER :on heating of the volatile matter contained in any sample of coaland the degree of inflammability of the coal dust.By determining such factors as (a) the temperature at which gasis most readily evolved, ( b ) the stage in the heating at which themost inflammable mixture of gases makes its appearance, and ( c )the shortest time of heating that will allow any gas at all to bedistilled, it should be possible to draw a distinction between dustsof different degrees of inflammability or liability to propagateexplosion.The interest that attaches to the study of the action of heat oncoal, however, is not confined to any particular problem, and,although in carrying out the present investigation we have hadthe problem of coal dust explosions alone in mind, we believe thatthe experimental methods adopted render our results of more wide-spread application.We hope ultimately to obtain definite information regarding thecomposition and chemical constitution of coal.Destructive Distillation of Cod at Different Temperatures.This part of our investigations, the ultimate object of which isthe elucidation of the nature of “coal,” yields only a, littleinformation in that direction when considered by itself.We findit necessary, therefore, t o reserve our conclusions regarding thecomposition of coal for a future communication, and merely toindicate the bearing that this part of our work has on thesubject.The principal facts brought to light by the distillation of differentsamples of coal at different temperatures are as foIlows :1.With all coals, whether bituminous, semi-bituminovts, oranthracite, there is a well-defined decomposition point, at a tem-perature lying between 700° and 800°, which corresponds with amarked increase in the quantity of hydrogen evolved. Withbituminous coals, the increase in the quantity of hydrogen evolvedfalls off at temperatures above 900°, but with anthracitic coals itis maintained up to 1 1 0 0 O .2. Evolution of hydrocarbons of the paraffin series ceases prac-tically entirely at temperatures above 700O.3. Ethane, propane, and butane, and, probably, higher membersof the paraffin series form a large percentage of the gases evolvedat temperatures below 450O.From these facts it may be concluded that coal of whatevergeological age contains a compound which undergoes decompositionat temperatures above 700° (under atmospheric pressure), and yieldshydrogen as its principal gaseous product.It seems probable, dsoTHE VOLATILE CONSTITUENTS OF COAL. 1919that this same compound is responsible for the hydrocarbons ofthe paraffin series that make their appearance at low temperatures.I nthe manufacture of lighting gas there is, as yet, no decided opinionas to the best carbonising temperature to use, nor has the influencethat temperature has on the nature and quantity of the productsof distillation been thoroughly investigated.The efficiency of a, boiler is largely dependent on the quantityand the nature of those constituents of coal that can be gasified ata, comparatively low temperature.The chief loss of efficiency whengenerating steam by the conibustion of coal arises from the coolingeffect of the boiler surface on the gases evolved from the coal,whereby their temperature is reduced below the point necessaryfor complete combustion. According to the nature of the coal, thetemperature at which gases can be distilled, and the rate at whichthe distillation takes place, varies; and the effect of cooling by theboiler surface will also vary with the nature of the gases evolved.It is possible that on the last factor the smoke-producing tendenciesof some varieties of coal largely depend.Previous work of this nature appears to be very scanty, butsince this investigation was begun an account has appeared in thetechnical journals of some work that lilts been carried out by13.C. Porter and F. K. Ovitz, of the Technological Branch of theUnited States Geological Survey during 1907-1908 (1. Gasliglztirzg,1908, 107, 343), which is very similar in character to that describedin this part of our work, though done with a different object inview.Details are given of the distillation at different temperatures oftwo samples of coal, as follows:The technical application of our results is readily apparent.Ziegler Coal, ZZlinois.Moisture ..................... 7.67 per cent.Volatile matter.. ............. 30 -38 , ,Fixed carbon.. ................ 54-32 , ,Ash .............................. 7-63 ,,Retort temperature (") .........500 600 700 800 900Maximum tenipcrature incoal (") ........................ 390 480 585 685 811Gases evolved from 10 grains(c.c.) ........................... 130 535 978 1550 2335Analysis of Gas.CO, ........................... 21.0 6.7 5.7 3'6 2'2" Illuminants" ......... 5 - 6 4.3 3.6 3.0 3.0CO .......................... 13'3 13'9 19'1 15.7 14-5H, ........................... - 14'2 24'1 38'5 48'7CH, ........................ 16'4 34'1 30-0 27-1 20.7C,H, ........................ 14.8 13.6 7.7 5.0 6 - 0N, ........................... 28'9 13.2 9.8 7.1 4.91000 1100920 10262700 31202.5 1.73'6 3-714.4 13.952.6 54.618.1 18'93'9 4.24'9 6.1920 BURGESS AND WHEELER :Connellsville Coal, Pocahontas.Moisture .....................1.10 per cent.Volatile matter.. ............. 30 -67 ),Fixed carbon .................. 60'35 ),Ash .......................... 7'88 ),Retort temperature(") ......... 500 600 700 800 900 1000 1100Maximum temgcratnre inGas evolved from 10 gramsAnalysis of gas.coal (") ........................ 390 474 559 705 812 922 1010(c.c.) ........................... 150 625 1220 1723 2080 2900 3530COa .......................... 11.2 3'4 2'7 2'0 1'1 1.1 1'0CO ........................... 5.2 5.3 5.3 7.0 7.2 5.9 7'0H, ........................... - 10.9 27-0 38.1 52.0 56.0 58'2CH, ........................ 4.5 41'2 36'2 36'0 25'2 24.4 21.5C,H, ....................... 32.5 18'9 12.9 9'5 7.1 3.0 3.5N, ...........................41.3 14'6 12'1 3'1 2.7 5.3 3.6illuminants " are taken to be the benzene and the ethylenehydrocarbons, whilst the figure for ethane includes all higherparaffin hydrocarbons calculated as C2H6.The distillations were conducted in each case on 10 grams ofair-dried coal in an atmosphere of nitrogen, which was passedthrough the retort before heating until the exit gases containedless than 1 per cent. of oxygen.The most important conclusion drawn by Porter and Ovitz fromtheir work is that the nature of t-he volatile products distilled fromdifferent samples of coal at low temperatures in the early stages ofheating varies in accordance with the smoke-producing tendenciesof the coal.They inclxde among the smoke-producing constituents tar,benzene, ethylene, and the higher homologues of methane.Fromthe figures reproduced here it will be seen that the Connellsvilleyields a larger quantity of these gases than the Ziegler coal at lowtemperatures, while it is found in practice that there is a greaterdifficulty in burning coals of the Connellsville type without smoke.We are inclined to believe that the smoke-producing constituentsconsist practically entirely of the higher hydrocarbons of theparaffin series; for we have isolated small quantities of propaneand butane from the gases evolved from coal at low temperature,and have obtained evidence of the presence of the higher membersof the series. These gases readily decompose at temperatures below600°, depositing carbon ; whilst ethylene yields very little carbonon decomposition at this temperat.ure.The general trend of the results is in agreement with our own,although it is impossible to make any direct comparison owing to"Illurninants" ............. 5.3 5.7 3.8 4'3 4'7 4.3 4'9ThTHE VOLATILE CONSTITIJENTS OF COAL.1921differences in the manner of distillation. We are inclined, however,to doubt the complete absence of hydrogen in the gases distilled at500°, and we cannot regard the methods of gas analysis employedby Porter and Ovitz as being entirely satisfactory.EX PER IMENTAL.The Coal Samples.-The validity of comparison between onedistillation and another depends to a very great extent on theobtaining of a uniform and representative sample of coal. Thesamples were procured in each case by pulverising about 150 kilo-grams of screened nut coal in a special form of disintegrator, andcollecting during the operation about 1 kilogram of the fine dustformed.This dust was then passed through a sieve with 240 meshesper linear inch, and stored in screw-top bottles. About 75 per cent.of the pulverised coal passed through such a mesh.The Method of Distillation.-Two grams of coal, dried at 107O,are intimately mixed with 3 grams of white sand, which has pre-FIG. 1.Platinum yetort am? cowexions.viously been ignited. The mixture is placed in a thin platinumboat (B, Fig, 1) 13 cm. long, which slides easily into a retort, R,which consists of it platinum tube 21.5 cm. long and of 1'7 cm.internal diameter.This tube is silver-soldered into a gun-metal collar with a wideflange, whilst a similar flange carrying a short length of gun-metalleading-tube ( L ) of 1 cm. bore is bolted on to the retort by sixsmall screws through the flanges.TKe face connected with theleading-tube has a projecting ring midway between the centre andthe circumference, which is pressed into a corresponding sunk ringon the retort face, using a washer of asbestos and graphite. Aperfect vacuum-tight connexion can be made in this manner.Refore this connexion is made, a tar-scrubber (S) is fitted intothe mouth of the retort. This scrubber consists of a platinumtube packed with ignited asbestos fibre and open at both ends. Itis 1.65 cm. in diameter for 8 cm. of its length, and then narrowsinto a tube 0.8 cm.in diameter and 5.5 cm. long. These dimensionsalIow of its occupying the position shown in Fig. 1, the wide portionmaking a good sliding fit inside the retort1922 BURGESS AND WHEELER :The connexion to the retort having been made, the gun-metalleading-tube is joined by stout rubber pressure tubing to a mercurymanometer and a 2-litre gas-holder. The retort and connexions arethen exhausted of air through the glass taps t , T, t’, the tap tclosed, and T, which is a three-way tap, turned so as to makeconnexion with the gas-holder and the retort as soon as t isopened (Fig. 2).The whole arrangement is mounted on a wooden support, andruns on wheels so as to allow of the retort being quickly pushedFIG. 2.Distillation upparutus.into the furnace, which has been previously brought t o the experi-men t.al temperature.As soon as the pressure of the gases evolved is equal to theatmospheric pressure, the tap t is opened, and the gas at once passesinto the holder and is collected.Heating is continued for a definiteperiod, usually seventy-five minutes, and the retort then withdrawnand allowed to cool. The gases remainiqg in the retort areexhausted by means of a mercury pump, and added to the mainbulk in the gas-holder. The gun-rnetal joint is then disconnected,the tar-scrubber and the boat removed and weighed, and the gasesanalysed.I n this manner the following data are obtained: (1) The rateof evolution of gas; (2) the volume of gas evolved; (3) the cornTHE VOLATILE CONSTITUENTS OF COAL.1923position of this gas; (4) the quantity of tar formed; and (5) thetotal loss in weight of the coal, that is, the total volatile matter.The Distillation Furmace.--Tn order to make sure that the retortshall be heated evenly throughout its length, the tube furnaceemployed is a platinum-wound electric resistance furnace, which,with a current of a little more than 1 ampere at 200 volts, can attaina temperature of 1400O. The furnace was tested by means of athermo-couple for each centimetre of its length to make sure thatthe coal would be in the zone of highest and of uniform tem-perature when the retort was inserted.The temperatures are measured by means of a platinum andplatinum-rhodium thermo-couple, which runs through the length ofthe furnace (the junction being in the centre), and is insulated byan unbroken length of silica quill tubing.The Gas Analyses.-The gases are collected over a mixture ofequal parts (by volume) of glycerol and water previously saturatedwith coal gas.The gases do not dissolve in such a mixture to anyappreciable extent, and its use is more convenient than that ofmercury.The gas analyses have been carried out with a modification ofthe Bone and Wheeler apparatus (J. Soc. Chem. Znd., 1908, 27,10). With this apparatus the absorptions are carried out overmercury in one absorption vessel with a comparatively small volumeof the particular reagent, which is always used fresh and is at oncediscarded after use, the absorption vessel being rinsed out withdilute sulphuric acid before the next reagent is used.The reagents that we have employed for the different constituentsare as follows, the absorptions being made in the order named:Gas.Ammonia .........................Benzene ................-. .........Carbon dioxide .................Oxygen ...........................Carbon monoxide ...............Hydrogen sulphide ............Acetylene ........................Ethylene. ..........................Beagent.Dilute sulphuric acid (10 per cent.).Concentrated sulphuric acid (D 1.9).Acidified solution of copper sulphate.Potassium hydroxide solution.Strongly alkaline pyrogallic acid.Ammoniacal silver chloride solution.Bromine water with potassium bromide.Ammoniacal cuprous chloride solution.The gas remaining after these absorptions is passed into a setof exhausted glass bulbs containing ‘‘ oxidised ” palladium pre-cipitate heated in a water-bath to 90°.Heating is continued duringten minutes, the bulbs allowed to cool, and the residual gases with-drawn by means of a mercury pump and measured. The changein volume observed is taken to be due to removal of hydrogen bythe palladium.An explosion analysis is then made in the usual manner1924 BURGESS AND WHEELER :The details of experiments with four samples of coal are given inthe tables following.The distillation temperatures recorded are the retort tem-peratures, not those existing in the coal. The temperature in thecoal ihelf reaches the retort temperature at the end of two minutes;thus, in a special experiment made to test this point, in which athermo-couple was embedded in the coal itself, the temperaturesrecorded immediately after the insertion of the retort in the furnacewere as follows :Retort. In coal.At beginning .....................830" 400"After half a minute ............ 790 580After on0 minute ................. 815 780After one and a-half minutes.. 835 835After two minutes ............... 840 840The records of the ra.te of evolution of gas are taken from themoment that the pressure in the apparatus reaches atmosphericpressure. The volumes are not corrected for variation in tem-perature and pressure.The total quantity of gas evolved is calculated per gram of ash-free dry coal = "nitrogen-free" gas at Oo and 760 mm.The different constituents of the gas mixtures are calculated aspercentages of the nitrogen-free mixture.From 1 to 4 per cent. ofnitrogen is usually found in the mixtures, but since this maypartly be due to the presence of traces of air in the retort con-nexions, or to error in analysis, it is thought best for the purposeof comparison to assume that the gases are free from nitrogen.The quantities of tarry matter and of total volatile matter arecalculated as percentages of the ash-free dry coal.Coal A (Bituminous).The pulverised nut coa1,after passing through a 240-mesh sieve, had the following ultimateanalysis :Coal from the Altofts Silkstone Seam.Carbon ..................80.50Oxygen .................. 9-70 Per cent. of ash-free dry coal,Nitrogen ............... 1'42Hydrogen ...............Sulphur ............... 2.93 "3 and it contained 5-51 per cent. of ashTHE VOLATILE CONSTITUENTS OF COAL. 1925TABLE I.Rate of Evolution of Gas at Different Temperatures.Total gas from 2 grams of coal; measured at atmospheric tem-perature and pressure.Distillation teiiiperature./ ..600" 700 750 800 900 1000 1050 1100c.c C.C. C.C. C.C. C.C. C.C. C.C. C.C. ..............211d ................. 25 25 45 75 125 145 90 5502f,&~hne 1st niinute 40 50 90 145 225 300 3754th ................. 10 15 15 25 25 25 15 105th ................. 5 10 10 20 20 5 10 5Total first 5 minutes.. 95 125 185 300 445 505 515 5651 3rd ...............15 25 25 xi 50 30 25--- -- -- -Next 5 minutes ...... 22 30 30 60 50 60 25 30,, 5 ,, ...... 13 15 25 25 15 15 15 5,, 15 ,, ...... 25 35 35 27 20 25 15 25 ,, 15 7 7 ...... 15 10 15 13 15 16 15 15,, 30 ,, ...... 10 10 5 10 10 25 15 5With distillations at 450° and 500°, 25 C.C. and 60 C.C. respectivelywere evolved, the total duration of heating being two hours.TABLE 11.Volatile Constituents Evolved at Different Temperatures.Per cent. ash-free, dry coal.Distillation temperature.450" 500 600 700 750 800 900 1000 1050 1100Total volatile matter :9.10 18.79 28'37 32'30 34'04 36'30 38-05 38.30 38'80 38.83Tarry matter :4.29 9.05 13.66 14'08 16.20 13.50 12% 10'40 10'90 9-00The cocdensed products at distillation temperatures of 450° and500° were light-coloured oils.The coke remaining after all distillations up to and includingthat at 800° was dun black in colwr, and had a tarry odour.Above this temperature the coke was greyish-white and lustrous,and free from any odour of tar1926 BURGESS AND WHEELER :TABLE 111.Percentage Composition of Gas Evolved at Diferent Temperatures.Calculated as " nitrogen-free " gas.,l ___ Distillation temperature.%50° 500 600 700 750 800 900 1000 1050 1100''NH, ...4-70 1-35 1-40 "O0 '*O0 3.65 3.55 3-00C,H,.. 8'60 4-85 5'20 i'::} 5m20 { 3-65 3.30)COO .... 10.95 3'60 3'50 4'05 3'30 1.70 1-70 1-65 1-65 1.70C,H,.. nil 0'35 - 0;40 - I - 0.50 - nilCO ..... 8-76 6-45 7-10 7.90 9'40 11.85 13.65 15.10 14'80 15.85H, ..... '7.00 16.60 26-60 32.70 41.65 48-55 55-70 56.40 56.55 56.65CH,.., 25.00 37-55 35-20 34-60 29.00 26'10 16-95 17.55 17.05 17'60C,H,..34.10 27*60 19-20 14.30 9-80 6.25 6-00 3.55 4.85 3-40C2H4.. 0.85 1.65 1-80 1-05 0.75 0.90 1.40 1.65 1.55 1.85TABLE IV.Tot& Volume of Gas Evolved per Gram of Ash-free, Drp Coal atDifferent Temperatures, a d Volumes of Principal Constituelzts.Distillation temperature.4cOo 500 600 700 750 800 900 1000 1050 1100--\ ~ _ ~ _ _ ~ _ _ _ _ -C.C. C.C. C.C. C.C. C.C. C.C. C.C. C.C. C.C. C.C.Total gas. 12-00 29.90 99.00 124'00 154.00 218'00 268.0 305'0 315.0 327'0H2 .......... 0.84 5-00 26'4 40.55 64.20 105'80 149'3 172.0 178'0 185.2CH, ...... 3.00 11.25 34.9 42'90 46.05 56.90 45'5 53.5 53.6 57.5C,H6 ...... 4'10 8-25 19.0 17.70 15-10 13-30 16-1 10.8 15'3 11'1CO ......,..1-04 1-95 7.0 9-80 14'45 25-80 36.6 46-0 46.6 51.8C,H, ...... 0.10 0-45 1.7 2-00 1-15 1-95 3.7 5-0 4.9 6'0The numbers for hydrogen, methane, ethane, and carbonmonoxide are shown graphically in Fig. 3.All the data obtained indicate that there is a '' critical " periodin the decomposition of the coal between 700° and 800O. Therate of evolution of gas and the total quantity evolved per gramare nearly double at 800° what they are at 700O; the curve fortotal volatile matter shows a marked change in character, becomingsteeper at the point given by the distillation at 750O; the quantityof tarry products reaches a. maximum at 750O; whilst the totalquantity of hydrogen evolved is more than doubled between 700°and 800°, the change in the character of the curve being verymarked (Fig.3).A noteworthy fact regarding the composition of the gases is thepresence of a high percentage of ethane in the gases evolved atlow temperatures, the percentage decreasing progressively witTHE VOLATILE CONSTITUENTS OF COAL. 1927higher temperatures of distillation.* It will be observed, however,that the actual quantity of ethane varies very little with thetemperature of distillation after 600° ; and it would appear probablethat such quantities as are found to exist in the products of thehigh-temperature distillations are evolved only during the initialperiod of slow heating up of the coal which necessarily takes placeFIG. 3.COAL A.150h & 10050500" 600" 700" 800" 900" 1000" 1100"Tenzperatzcres.when the retort is first pushed into the furnace; and that theirappearance in the final products is due to their having been sweptout of the retort as the rate of evolution of gas increased, and thus* It may here he remarked that ethane exists, in quantity up to 5 per cent., inmany samples of ordinary lighting gas ; a fact which appears to have escaped generalnotice, and may account, to a certain extent, for tho discrepancy so oftenobserved between the calorific value of lighting gas as determined directly and ascalculated from analysis1928 BURGESS AND WHEELER :having escaped decomposition.For, according to Bone and Coward(Trans., 1909, 95, l a l l ) , the thermal decomposition of ethane isfairly rapid at 6'75O, and at 1000° it is practically instantaneous.The greatest interest attaches to the quantities of hydrogen andof methane found in the gases.A t a temperature of 750° thereis a sudden increase in the total quantity of hydrogen evolved, andthe rate of increase is maintained fairly regularly up to 900°.The question arises as to the nature of the reaction responsiblefor the evolution of this hydrogen. It cannot be, as has beensuggested, that hydrogen is produced at the expense of methane;for the total quantity of the latter gas evolved undergoes verylittle variation, as is shown in table IV, and graphically in Fig. 3.It might, however, be contended that the fact that the totalquantity of methane evolved' undergoes little or no variation after adistillation temperature of 750° has been attained, is due to thefact that, at each of the higher distillation temperatures, a givenquantity is evolved during the initial stage of heating up of thecoal; and that as soon as the temperature in the coal rises above acertain point, any further methane evolved is at once resolved intoits elements.Bone and Coward have shown (Zoc.cit., p. 1206), however, thatmethane is comparatively stable when heated in a porcelain tubeat temperatures below l l O O o , only about 50 per cent. being decom-posed after one minute's heating;. whilst at a temperature of 785O-the retort temperature in our experiments corresponding with themaximum rate of evolution of hydrogen-10 per cent.only wasdecomposed at the end of an hour's heating.Moreover, the carbon deposited during the course of the thermaldecomposition of methane is a characteristic hard and lustrousvariety, almost metallic in appearance, and can readily be detected.We have been unable to detect the slightest trace of such carbondeposit below 900° distillation-temperature, although a t 1050O aquantity just sufficient to weigh could be brushed off the sides ofthe platinum boat and retort.Hydrocarbons of the ethylene series yield a greater quantity ofmethane than of hydrogen on decomposition below 800°, and inany case the quantity present is insufficient t o account for the largeincrease in hydrogen.We are inclined to believe that, at a temperature of about 750°,one or more of the higher hydrocarbons of the paraffin seriesevolved undergoes rapid decomposition, yielding chiefly hydrogenand carbon ; decomposition at lower temperatures yielding methane,ethane, and hydrogen.We are studying this point in connexionwith another part of this investigationTHE VOLATILE CONSTITUENTS OF COAL. 1929Coal C (Bituminous).The pulverised nut coal,after passing through a 240-mesh sieve, had the following ultimateanalysis :Coal from Abertillery, South Wales.Carbon ............... 85.72Oxygen ............... 7-34 Per c+ of ash-free, dry coalNitrogen ............ 1-09Sulphur ............ 0.92 4'g31 Hydrogen.. ..........and contained 7.65 per cent. of ash.TABLE I.Rate of Evolution of Gas at Different Temperatures.Total gas from 2 grams of coal; measured at atmospheric tem-perature and pressure.Distillation temperature.7---600'C.c.During 1st minute ......,, 2nd ), -,, 3rd ,) -,, 4th ,) -,, 5th ,, ...... -..................-Total first 5 minutes ... 70Next 5 minutes ............ 30 ,, 5 ) ) ............ 20) ) 15 ), ............ 20 ,) 15 ,, ............ 10 ,) 30 ), ............ 10700 800 900 100060 100 200 32035 70 105 13025 40 50 2520 25 15 1510 15 10 10150 250 380 50030 50 50 4525 50 20 1020 30 25 2020 20 5 155 15 20 10C.C. C.C. C.C. C.C.- - - -71100475453055c. c.5604051555With a distillation temperature of 500°, 65 C.C. were evolvedduring two hours.TABLE 11.Volatile Constituents Evolved at Different Temperatures.Per cent.ash-free, dry coal.Distillation temperature. - ,\\ 500" 600 700 800 900 1000 1100Total volatile matter ...... 13.40 21.17 25.68 28.96 30.14 31.36 31.50Tarry matter ............... 7'00 10 02 7.36 10'56 9-67 9-19 8.21The coke remaining after the distillations at 900°, 1000°, and1 looo was greyish-white and lustrous ; that remaining after thedistillation at 500°, 600°, TOOo, and 800° was dark and compact.This coal does not swell so much on coking as does coal A.VOL. XCVKI. 61930 BURGESS AND WHEELER :TABLE 111.Percentage Composition of Gus Evolved at Different Temperatures.Calculated as " nitrogen-free " gas.Distillation tern pera ture.(500" 600 700 800 900 1000 1100 ' -NH, ......CO, ......C2H, ......C2H4 ......co .........H, .........CHIC6H6 ......C2H6 ......2 004 '403 -950-450.904 708'0064.5011.050 *703 *553-200 301 *056'4525.0547-2012.451-102.403.400.300.057 -4534-7546.254 '250.852 -002 '500-050 7 09-8050.8028-604.70nil1.851 *650.051-0511-2557.0521-805 *25nil1-451.200.101 '4513-5559.8019.253.10nil2.251 '400.501 -5513.0060.7018-801'80TABLE IV.Total Volume of Gas Evolved per Gram of Ash-free, Dry Coal atDifferent Temperatures, and Volumes of Principd Constituents.Distillation temperature.5/00" 600 700 800 900 1000 1100Totalgas ...33.50 83'00 135.00 208.00 254-00 296.00 312-00H, ............ 2-70 20.80 46'95 105.70 144.95 177.00 189-45CH, .........21 60 39-20 62'45 59'50 65.50 57.25 58-65C,H, ........ 3'70 10.35 5.75 9.80 13-35 9-15 5'60CO ....-.....,. 1-65 5.35 10.05 20.35 28.55 40.15 40.55The general behaviour of this coal on heating is similar to thatof coal A. This is seen most clearly on comparing curves showingthe volumes of the constituent gases evolved at the different tem-peratures of distillation.A sudden increase in the quantity of hydrogen evolved is againapparent at a temperature lying between 700° and 800°, whilstthe quantity of methane evolved remains remarkably constant atall temperatures above 700O. The presence of " methane-carbon "could only be detected after the distillations at 1000° and 1100O.The quantity, and the percentage, of ethane evolved is con-siderably less at all temperatures than in the case of coal A..c.c. C.C. C.C. c. c. c. c. c. c. c. c.Coal D (Semi-bituminous).The pulverised nut coal,after passing through a 240-mesh sieve, had the following ultimateanalysis :Coal from Penrhycyber, South Wales.Carbon ............... 90'72Oxygen ............... 1.25 Per cent. of aah-free, dry coal.Nitrogen ............ 2 '99Sulphur ............... 0.81 4*231 Hydrogen ............and it contained 3.5 per cent, of ashTHE VOLATILE CONSTITUENTS OF COAL, 1931TABLE I.Rate of Evolution of Gas at Different Temperatures.Total gas from 2 grams of coal; measured at atmospheric tem-perature and pressure.Distillation temperature.6 0 " 800 900 1000 1100C.C.- C.C. C.C. C.C. C.C.During 1st minnte ......... - 90 175 300 550), 3rd ,, ......... - 40 60 25 15\), 2nd ,, ......... - 60 120 150 50 .. 4th ,, ......... - 25 30 25 10 ,, 5th ,, ......... - 15 15 10 10Total first 5 minutes ...... 100 230 400 510 635Next 5 minutes ............... 25 55 45 40 20,) 5 ,, ............... 20 30 15 10 10,, 15 ,, ............... 20 50 30 10 10,)15 ,, ............... 10 15 15 5 5 ), 30 ,, ............... 10 10 10 5 5With distillations at 500° and 600°, 30 C.C. and 114 C.C. respec-- - - - - -t,ively were evolved, the total duration of heating being t'wo hours.TABLE 11.VolatSe Constituents Evolved at Different Temperatures.Per cent. ash-free, dry coal.Distillation temperature.500" 600 700 800 900 1000 1100Total volatile matter 3.54 7.71 10.47 14'40 16-75 17'34 19.00Tarry matter .........0.99 2,lO 2.85 2.50 1'73 2.07 notes t irna t edThe coke did not appear to cake at all until a temperature ofA/- 1800° had been reached; it swelled very little.TABLE 111.Percentage Composition of Gas Evolved at Different Temperatures.CalcuJated as nitrogen-free " gas.Distillation temperature.500" 600 700 800 900 1000 1 1 2I ,Nft, ...... 3-10 - 0'80 0.60 0.30 0.20 0.10H,S ..... - - 0.20 0.35 0.30 0-25 0.25CO, ...... 6.85 2-85 1.60 1'40 1.35 1-00 0.20C,H, ...... 0.70 0.90 0.80 nil 0-40 0.25 0.35CO ...... 4-60 3.95 3-85 630 9-30 10'20 11.10CH, ...... 51.50 48.50 45.90 28.35 20-65 14.45 12.85C6H6 ...... 3.70 1.95 1'40 0.75 0.60 0.70 0'45C,H, ......0.05 0.20 0.10 n i l nil nil nilHP ......... 15-70 36.75 43'30 59-65 65.50 69'90 72.90C,H, ...... 13.70 4.65 2.05 2.50 1'60 2.95 1'806 L 1932 BURGESS AND WHEELER :TABLE IV.Total Volume of Gas Evolved per Gram of As?Lfree, Dry Coat atDifferent Temperatures, and Volumes of Principal Constitzcents.500"Total gas ... 13.00TI, ............ 2-05CH, ......... 8 '45C,H6 ......... 1 *80GO ............ 0.60c. c.Distillation temperature.600 700 800 900 1000c. c. c. c. c. c. C.C. c. c.56.00 94.00 193.00 248-00 288.0020.60 40.70 115-20 162.50 201.3027-30 43-20 54.70 51-20 41-602.20 3-60 12-20 23-10 29-40---LA2-60 2-00 4-80 4-00 a,50- 1100333'00242.7042'806 -0036-90c. c.Coal B (Anthracite).The pulverised nut coal,after passing through a 240-mesh sieve, had the following ultimateanalysis :Coal from Pontyberen, South Wales.Carbon ..................92 '66Oxygen .................. 2-20 Per cent. of ash-free, dry coalNitrogen ............... 0 *99Sulphur.. ................ 1 ."'1 *01Hydrogen ...............and it contained 3.9 per cent. of ash.TABLE I.Rate of Euolution of Gas at Dijferent Temperatures.Total gas from 2 grams of coal; measured at atmospheric tem-perature and pressure.Distillation temperature.700"During 1st minute ...... -), 2nd , -,, 3rd . . . . . . . .,) 4th ), ......, , 5th . , ......Total first 5 minutes.. .c. c.......L--158005025252020c. c.14090012095403015300c.c.-Next 5 minutes ............ 40 50 43 ,, 5 ,, ............ - 20 10 ,, 15 ,, ............ 20 30 20), 30 ,, ............ 6 10 5With a dist,illation temperature of 600°, 50during two hours.1000 iiooc. c. c. c.250 375100 7525 1520 1010 10405 48540 1515 105 255 15- -C.C. were evolveTHE VOLATILE CONSTITUENTS OF COAL. 1933v oFIU. 4.COALS D and B.25020015010050500" 600" 700" 800" 900" 1000" 1100"Temperatures.TABLE 11.daiile Constituents Evolved at Different Temperatures.Per cent. ash-free, dry coal.Distillation tenmeratme. ,%OO" 700 800 900 1000 1100Total volatile matter ...... 3.57 3'76 6.18 7'82 8.77 12.50Tarry matter ...,,.......... 0.22 0'40 0.95 0.80 0.22 not.estimat1934 THE VOLATILE CONSTITUENTS OF COAL.There was no change in the appearance of the coal after heatingexcept at llOOo, at which temperature some of the particles appearedto have caked slightly.FIa. 5.350300250A d& 200 h;8 150 s x cv10050500" 600" 700" 800" 900" 1000" 1100"Temperatures.TABLE 111.Percentage Composition of Gas ~ v b l v e d at Different Temperatures.Calculated as nitrogen-free " gas.Distillation temperature.'600" 700 800 900 1000 1100- nil NH, .........4.35 - I -C,H, ......... 1.20 2 *45 1 -35 1.35 1-55 0.75 co, ......... 6.55 3 -45 1-35 1 -55 0-25 1-45C,H, ......... nil nil nil nil nil 0 10C2H4 ......... 0.10 0 -20 nil 0.25 0.25 nilCO.. ......... 6.65 7-85 6-70 9.35 9.70 14.60H2 ............ 29.35 47'30 68.65 70.50 75.35 74'10CHI ......... 51.70 36.60 21.05 15.95 11.40 8 -85C,H, ......... 1 *oo 2-05 0'90 1.05 1-50 0'1THE VJSCOSlTY OF CERTAIN AMIDES. 1935TABLE IV.Total Volume of Gas Evolved per Gram of Ash-free, Dry Coal atDifferent Temperatures, and Volumes of Principal Constituents.Distillation temperature.600" 700 800 900 1000 1100Total gas ...... 20-5 44'0 133.0 193.0 230.0 272-0#--- .c. c. C.C. c. c. c. c. c. c. c. c.H, ............... 6-0 20.8 91.3 136'0 173.5 201.5CH, ........... 10'6 16'1 28'0 30'8 26 '2 24 -1C,H, ............ 0.2 0.9 1 '2 2 '0 3.4 0'4 co ............. !. 1.3 3 -5 8'9 18'0 22-3 39.8These two anthracitic coals, D and B, serve to show the differencein behaviour from the bituminous as regards the gases evolved atdistillation temperatures above 700O.There is the same abrupt increase in the quantity of hydrogenevolved between 700° and 800° (Fig. 4), but the rate of increaseis maintained up to the highest temperature employed, namely,llOOo, whereas, in the case of bituminous coals, as exemplified bycoals A and C, the rate of increase falls off above 900° (Fig. 5).In the case of coal D (semi-bituminous), the total quantity ofgas evolved by distillation at l l O O o is greater than in the case ofthe two bituminous coals, although at 700° it is less. The firstproducts of distillation of the anthracitic coaIs are probably notsuch as will allow of the recombination to any great extent of freeunsaturated residues to form more complex molecules, as appearsto be the case with bituminous coals yielding greater quantities oftarry matter,ALToFrs

ISSN:0368-1645    

DOI:10.1039/CT9109701917

出版商:RSC    

年代:1910

数据来源: RSC