摘要:

246 SHENSTONE AND EVANS : OBSERVATIONS ON THE XX1.-Observations on the Injhence of the Silent Discharge on Atmospheric Air. By WILLIAM ASHWELL SHENSTONE and WILLIAM T. EVANB. NEARLY forty years ago, the late Professor Andrews observed that when air is exposed to the action of the silent discharge of electricity it contracts to a certain extent, that if i t be then left in contact with oil of vitriol for a few hours i t undergoes further contraction, and that if the residue be afterwards again submitted to the discharge its volume may be yet further diminished. He accounted for these phenomena by supposing that one or more of the higher oxides of nitrogen had been formed from the air. A t the same time, Andrews called atten- tion to the I fact that it is impossible to generate ozone in oxygen which contains small quantities of nitrogen if the mixture has recently been exposed to the action of disruptive discharge.This he also attributed to the “presence of a trace of hyponitric acid gas pro- duced by the electrical sparks ” (Trans. Roy. SOC., 1860, p. 127). Berthelot (Compt. vend., 1881,92,82), and Hautefeuille and Chappuis (Compt. rend., 1881, 92, 80 and 134) have also recognised oxides of nitrogen among the products of the action of the silent discharge on atmospheric air, and most of Andrews’ successors in this field of work have taken pains to employ very carefully purified oxygen for their experiments. Our knowledge, however, of the phenomena described by Andrews still stands almost a t the stage at which he left it ; and, therefore, as the subject is both interesting from the scientific point * The oil of vitriol is not mentioned in Andrews’ descriptio~l of his experiment, but it was present in the manometers he used.INFLUENCE OF THE STLENT DISCHARGE ON ATMOSPHERIC AIR.247 of view, and of some importance t o those who seek t o apply ozone medically, and for other technical purposes, we have recently spent some time in an attempt t o follow u p this part of Andrews’ beautiful researches. W e are conscious that the work described in this paper by no means exhausts our subject, but we venture t o present the results already obtained to the Society at this stage, since we hope they may be found interesting and useful to other workers, and because t h e difficult character of the experiments still before us may make our future progress rather slow.I. A Method of Detecting Nitric Peroxide in tile Presence of Ozone. The oxide of nitrogen produced by the action of the silent discharge on air always exists i n t h e presence of excess of oxygen or ozone ; therefore, as there is no reason to suspect that nitrous oxide is formed by the discharge, nitric peroxide is the only oxide of nitrogen which is at all likely to be met with ; in order to detect and estimate this in t h e presence of ozone, we draw the gas to be tested through a very dilute solution of soda, and then determine the nitrite formed by means of Riegler’s reagent (Abstr., 1807, ii, 464). As the action of the latter seems t o depend in t h e first instance on the forming of a-diazo-naphthylenesulphonic acid, by the interaction of nitrous acid with 1 : 4-naphthylaminesulphonic acid, and on its subsequent con- version into a red dye, its usefulness, as might be expected, is not affected by the presence of ozone, and we believe i t to be very suitable for the purpose to which we have applied it.11. Influence of the Silent DiecJLurge o n Moist Air. Expeviment I,-An ozone generator provided with a manometer (Trans., 1893, 63, p. 943, Fig. 3) was filled with atmospheric air saturated with water at Oo, from which the ammonia and carbon dioxide had previously been removed ; the cont,ents of the apparatus were then stibmitted t o the discharge at Oo, as described i n previous accounts of similar experiments on oxygen (Zoc. cit., p. 938). The difference of potential employed corresponded t o 44.4 C.G. 5. units (electrostatic), and great care was taken not to raise t h e tempera- ture of the gas by employing too rapid a succession of discharges. As soon as the discharge was applied t o the air, it condensed rapidly and t o a remarkable extent, presently the rate at which it contracted became slower, and then, suddenly, it commenced t o re-expand ; the latter change ceased whenever t h e discharge was discontinued, but occurred again as often ns the sparking was recommenced. When a large part of the ozone formed during tbe first stage had evidently been248 SHENSTONE AND EVANS : OBSERVATIONS ON THE destrojecl, but while a little still remained, a hole was pierced in the bottom of the ozone generator, another at the end, 2 (Zoc.cit.), of the manometer, and its contents were drawn through cold water, which was afterwards examined for nitrites in the manner described above. Unmistakable indications of the presence of nitric peroxide were obtained. Owing to the suddenness with which the secondary action of the discharge set in, we are unable to state exactly the maximum amount of ozone formed. We satisfied ourselves, however, from the measure- ments taken, that not less than 91 per cent. of the oxygen present had been converted into ozone ; we believe this proportion has never been exceeded except in one of our own experiments." Having now obtained a general idea of the mode of action of the silent discharge on air, we arranged our subsequent experiments largely for the purpose of gaining answers to the following questions.a. What is the highest yield of ozone that can be obtained from air under the most favourable conditions 1 b. At what stage is nitric peroxide first formed, and what is its subsequent history ? c. What influence has the presence of moisture on the various phenomena under investigation 1 In all the experiments described in this paper, me employed the ozone generator formerly used by one of us for examining the influence of the silent discharge on moist, and on carefully dried oxygen. The present work on air is therefore strictly comparable with the former work on oxygen (Trans., 1897, '71, p. 471). h small tail was added to the bottom of the ozone generator (Zoc. cit., Fig. 3), in order that we might attach an absorption tube to it by a paraffin joint when it was necessary to withdraw its contents in order to test for nitric peroxide.111. The Maximum Yield of Ozone produced by the Action of the Silent Dischw~ge on Air at 0". From the results of a number of experiments, we found that, in order to produce the largest proportion of ozone from air, i t is necessary that moist air should be employed, also that great care be taken not t o apply the discharges in too rapid succession, and that frequent intervals be allowed for cooling towards the end of the operation. It is rather difficiilt to hit the point a t which the condensation of * Rrodie (Trans. Roy. Soc., 1874, 164, p. 101) suhmitted carbon dioxide to the action of the silent discharge and found that as much as 85 per cent.of the oxygen liberated might consist of ozone. Only a few C.C. of ozone weye obtained, however, from a considerable volume of carbon dioxide in Brodie's experimeiit,INPLIJENCE OF THE SILENT DISCHARGE ox ATMOSPHERIC AIR. 249 the air attains its maximum, because immediately this point is reached the contents of the ozoniser re-expand rapidly. Ozonised air, moreover, appears to be far more sensitive t o rise of temperature, such as may be caused by the use of a too rapid succession of discharges, than is the case with ozonised oxygen. The effect of this sensitiveness is well illustrated by the results of our three earliest experiments. Experiment 1.-This was the preliminary experiment described above. The discharges were delivered at a moderate rate with frequent inter- vals for cooling, air saturated with water at 0" being used.91 per cent. of the oxygen present was converted into ozone. Experiment 11.-Air saturated with water a t 0' was used, the succession of discharges was more rapid, but the intervals for cooling were the same as in the previous experiment. Only 82.1 per cent. of the oxygen was converted into ozone. Expewhent 111.-The dynamo did not work well during this experi- ment, and consequently the succession of discharges was very slow ; moist air was used, and the intervals for cooling were the same as in the other two experiments. 98 per cent. of the oxygen present was converted into ozone. If the last of these results be compared with what has previously been done, it will be seen that not only does the proportion of the oxygen which was ozonised far exceed anything previously obtained, but also that the charge of ozone carried by the gas is exceedingly high, much higher, for example, than was obtained by one of us, by means of the same ozone generator, and under the most favourable con- ditions from moist oxygen, namely, 13.6 per cent.(Trans., 1897, '71, p. 475). IV. Experiments to asceytain the InJEuence of Moisture on the Formation of Ozone from A&*. It has been shown by one of us, in a previous paper, that well-dried oxygen yields little or no ozone when it is subjected to the action of the silent discharge (Trans., 1897,71, p. 479) ; a single experiment such as was then performed occupied many months, and during its progress the ozone generator was not available for other work.To avoid such cause of delay on this occasion, we dried our air in a more simple manner, partly because it seemed almost a foregone conclusion that very highly- dried air would, like oxygen, be practically unaffected by the dis- charge, and partly because the progress of the research as a whole was more important a t the time than the making of a single experiment to minutely investigate one particular point. Before we made the experiments with dried air, the ozone generator was dried by heating it, when exhausted, with the little furnace previously desoribed (Zoa. cit., 478); it was then filled with dried air250 SHENSTONE AND EVANS : OBSERVATIONS ON THE re-exhausted and re-filled. The air was dried by passing it slowly over a column of purified phosphoric anhydride (Trans., 1893, 63, p.475) which had previously been freed from carbon dioxide. It should be mentioned that the air used in VI was more nearly dry than that used in IV and V, as it was purposely passed a t a slower rate over the drying material. A.-Resnlts obtained with wet oxygen. Experiment I. Experiment 11. 82.1 9 , 9 9 ? 9 Experiment 111. 98 ?, > 9 Y , B.-Results with dried oxygen. Experiment IV. Experiment V. 67.0 9 , ? I 9 , Experiment TI. 60.0 99 9 ) 9 9 Similar results were obtained in the course of two other experiments with moist and dried air ; in these, the air WAS rapidly ozonised for the purpose of examining its composition a t later stages, and consequently the maximum proportion of oxygen ozonised was smaller than before, but the results tell the same tale.91 per cent. of the oxygen present mas ozonised. 66-7 per cent. of the ozygen present was ozonised. A.-Dried air B.-Moist air 89.0 7 , 7 9 63-5 per cent. of the oxygen was ozonised. V. Expriments to T+ace the Formation of Nitric Peroxide from Air by means of the Silent Discharge. We have seen from Experiment I that air which has been highly charged with ozone, and subsequently partly reduced by the continued action of the silent discharge, contains nitric peroxide. I n order to trace the formation of the latter substance at various stages of an experiment and, if possible, to learn in what manner its appearance is connected with the destruction of the ozone which follows the continued application of the discharge, we made the following experiments.Experiment VI1.-The ozon0 generator was filled with dried air, and its contents submitted to the action of the silent discharge until the rate at which the gas contracted and the proportion of oxygen ozonised, 63.5 per cent., indicated that the point of maximum con- traction was nearly reached ; the contents of the ozone generator mere then drawn off and examined for nitric peroxide. No nitric peroxide was found. Experiment VII1.-This was a repetition of the previous experiment, except that the action of the discharge was not carried quite so f a r , only 61.S per cent, of the oxygen being ozonised. In this case, also,INFLUENCE OF THE SILENT DISCHARGE ON ATMOSPHERIC AIR. 251 no nitric peroxide could be detected, although so small a quantity as 0*000001 gram of nitrous anhydride can be recognised by Riegler's reagent in 100 C.C.of water. Experiment 1X.-The ozone generator having been recharged with dried air, its contents were submitted to the silent discharge until a portion of the ozone first formed, 67 per cent., was again destroyed. The residue was then tested for nitric peroxide. No less than 0.041 C.C. was found." The observed increase of the volume of the gas in this experiment corresponded to a reduction of the percentage of oxygen ozonised from the maximum 67 to 54 per cent. ; the actual proportion of ozone remain- ing was, however, somewhat less than 54 per cent,, because some nitric peroxide had been simultaneously formed. Similar experiments were next made upon moist air. Experiment X.-Moist air was submitted to the discharge until 80 per cent.of its oxygen was ozonised; as the rate at which it con- tracted had then become slow, the discharge was stopped and the contents of the apparatus were tested as before. No nitric peroxide was detected. Experiment XI.-This experiment was a repetition of X, and gave a similar result. Experiment XI1.-Moist air was submitted to the discharge until some of the ozone first formed had been destroyed again, but the destruction of the ozone was not carried so far as in the corresponding experiment with dried air. The residue was found to contain 0.020 C.C. of nitric peroxide, The expansion at the second stage corresponded in this case to a reduction of the ozone from the maximum 82.1 to 72 per cent., but allowance must be made for the nitric peroxide simultaneously formed.From these results, it is evident that nitric peroxide is not produced by the action of the silent discharge on air, either moist or when dried, at Oo, until a large proportion of the oxygen present has been con- verted into ozone; and this fact, together with the rapid destruction of ozone which accompanies the appearance of the nitric peroxide, suggests the idea that the latter may be formed at the expense of the * These volumes are calculated on the assumption that the formula of nitric per- oxide is N,04. There is reason to believe, however, that, under the conditions of the experiment, this gas is partly dissociated. It isof course possible that the presenceof ozone with the moisture, and nitric peroxide, &c., may favour the formation of nitric acid when the iiiixed gases are drawn through a dilute alkaline solution, and thus lead one to underestiiiiate the amount of nitric peroxide present.This, however, if trne, wonltl rather increase the significance of the results obtained in expcrimelits IX and XII, and tliesc alone are likely to have been aflected by such a disturbing cause. N,O,."252 SHENYTONE AND EVANS : OBSERVATIONS ON THE ozone, and not by the combining of nitrogen and ordinary oxygen. This view of the matter is supported by the result of Experiment I11 (p. 249), which shows that only a few tenths of a per cent. of oxygen were present in the gas at the moment when the formation of the nitric peroxide probably began, and is further upheld by the important fact that the formation of nitric peroxide from air under the influence of the silent discharge is distinctly retarded by the presence of water vapour and promoted by dryness.(Compare IX with X and XI. Compare also the results of I, I1 and 111, in which 82 to 98 per cent. of oxygen was probably ozonised before sensible quantities of nitric per- oxide were formed, with those of IV, V, VI, in which only 60-67 per cent. of oxygen was ozonised when its destruction, presumably due to nitric peroxide, set in.) For it has previously been shown by one of us (Trans., 1897, 71, Zoc. cit.) that the stability of ozone is increased by the presence of moisture, which might lead us to expect its power as an oxidising agent to be lower in the pre- sence of water than when dry.Ordinary oxygen, on the other hand, is well known to oxidise best in many cases in the presence of water. But whatever may be the truth in regard to this matter, i t is at any rake very interesting to meet with a fresh case of chemical change which is retarded by the presence of water vapour (see also Trans., 1897, 71, Zoc. cit.). I n connection with the above, it must be remembered that although the presence of a large proportion of ozone is necessary for the forma- tion of nitric peroxide from air by tbe influence of the silent discharge at Oo, yet a rapid destruction of ozone sets in simultaneously, or almost simultaneously, with the formation of nitric peroxide. The latter change is accompanied by the destruction of a great part of the nitric peroxide first formed.The following examples will make this clear. A. I n Experiment IX, some dry air was ozonised, then a part of the ozone was destroyed; the residual contents of the ozoniser were found to contain 0,041 C.C. of nitric peroxide. B. Some moist air was similarly treated (Experiment X I I ) ; i t yielded 0*020 C.C. of peroxide, N204.* C. Some dry air was ozonised and then re-sparked until the final volume showed that all but a trace of the ozone was destroyed; it afterwards yielded only 0,007 C.C. of peroxide, N20,. D. Another specimen of dry air was similarly treated ; it yielded 0.006 C.C. of peroxide, N204. E. Some moist air was ozonised and then submitted to the continued * The difference in the hygrometric state of the air may possibly account for the difference between the results of A and B, since the presence of moisture retards the formation of nitric peroxide.INFLUENCE OF THE SILENT DISCRARGE ON ATMOSPHERIC AIR.253 action of the discharge until only a trace of ozone remained; it was found to contain only 0*008 C.C. of peroxide, N,O,. These last three results do not, it is true, agree veryclosely, but the difference between the proportions of nitric peroxide present in A and B, when most of the ozone still remained undestroyed, and that found in C, D, and E, after the ozone was destroyed, are sufficiently marked. On the other hand, ozone and nitric oxide do not thus destroy each other a t Oo, except under the influence of the silent discharge, if moist.For example, on one occasion, 8.405 C.C. of moist gas containing 0.824 C.C. of ozone and 0.020 of nitric peroxide, or an equivalent volume of the products of its dissociation, were preserved for more than half an hour without any sensible change of volume taking place, and other similar observations have been made. We have reason to suspect, however, that above Oo, or if dry, these gases interact far more readily. Finally, our experience entirely confirms the statement made by Andrews, that oxygen cannot be converted into ozone if a trace of nitric peroxide be present. In one experiment made to test this point, 8*74dc.c. (at 760 mm. and 0') of dried air was ozonised, and the product, 8.35 c.c., was then submitted to the continued action of the discharge until it ceased to expand ; its volume WAS then 8.72 C.C.After this it was submitted to the silent discharge for 75 minutes, but it re- mained quite unaffected, and at the end its volume was still 8.72 C.C. Experiments on this subject are being carried out. Summary. 1. Oxygen, diluted by nitrogen, yields a higher proportion of ozone when submitted to the influence of the silent discharge, under given conditions, than pure oxygen ; the proportion of oxygen ozonised may be as high as 98 per cent. of the oxygen submitted to the dis- charge. This fact deserves the notice of those who are interested in the technical applications of ozone. 2. If the process of ozonising air be not pressed too far, no peroxide of nitrogen will make its appearance. 3. The presence of water vapour is very favourable to the pro- duction of a high yield of ozone, and retards the appearance of nitric peroxide. 4. A t a certain stage in the process of ozonising the oxygen of the air, which depends on the amount of vapour present, and probably also on the temperature of the gas, nitric peroxide is formed. Its appearance is immediately, or almost immediately, followed by a rapid disappear- ance of the ozone, and this in its turn results in the destruction of most oE the nitric peroxide. 6 . That, as stated by Andrew, the presence of a trace of nitric per254 PREPARATION OF ORTHOCHLOROBROMOBENZENE. oxide renders it impossible to convert oxygen into ozone by means of the silent discharge. 6. That nitric peroxide and ozone when moist do not mutually destroy one another at Oo, or do so at a very slow rate, unless they are under the influence of the silent discharge. CLIFTON COLLEGE. February, 1898.

ISSN:0368-1645    

DOI:10.1039/CT8987300246

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

254 PREPARATION OF ORTHOCHLOROBROMOBENZENE. XXII.--Pyepayation and Properties of OTthoclzloro- bromobenxene. By JAMES J. DOBBIE, M. A,, D.Sc,, and FRED, MARSDEN, M.Sc., Ph.D. IN the course of an investigation on the halogen derivatives of benzene, we have had occasion to prepare in quantity all the modifi- cations of its di-derivatives containing chlorine and bromine. We find that orthochlorobromobenzene has not hitherto been described, and the object of the present communication is merely to place its properties on record, The starting point of its preparation was orthonitrobromobenzene ; this was made by nitrating bromo- benzene in the cold in the manner described -by Coste and Parry (Bey., 1896, 29, i, 788); this gave very satisfactory results, but we found it most convenient to separate the ortho- and para-compounds by grinding them up in a mortar with small quantities of cold methylated spirits and filtering, repeating the operation until the filtrate became colourless.The orthonitrobromobenzene dissolved, and after the solvent had evaporated was obtained in long needles which melted at 42' after recrystallisation. On reducing the orthonitrobromobenzene to the corresponding bromaniline, Hubner and Alsberg (Anmlen, 1870, 158, 316), and later Fittig and Mager (Bey., 1874, 7, 1179), found great difficulty in obtaining the latter in the solid state; we experienced the same difficulty, which was completely overcome, however, when we added the nitrobromobenzene in alcoholic solution, instead of in the solid state, to the slightly warmed reducing mixture of stannous chloride and hydrochloric acid.On shaking vigorously, the reduction pro- ceeded with development of heat, and after cooling, adding excess of caustic soda and distilling with steam, the oil which came over solidified in the receiver ; after recrystallising, it melted at 32'. To convert the orthobromaniline into orthobromochlorobenzene, it was diazotised in the usual manner, and the solution of the diazo- compound added to a warm solution of cuprous chloride. The oil which separated was then distilled over with steam, the distillatePREPARATION OF DRY HYDROGEN CYANIDE, ETC. 255 C,H,Br, ............... C,H,ClBr ............ ............... C,H,CI, extracted with ether, the extract dried over calcium chloride, and after driving off the ether the residue was fractionally distilled.The greater part came over between 200' and 202'. The halogens were determined by Carius' method. 0.302 gave 0.5168 mixed silver haloids. AgBr = 0.2948; AgCl= 0.222. 223'8" 219.4" 219" 1.977 1'955 - 204 196 196.3 1'6555 1'6274 179 172 172 1'3254 1.307 Br = 41.52 per cent. } Found C1= 18.18 per cent. } Calculated. Orthochlorobromobenzene is a clear, straw-coloured liquid having a strong aromatic odour. It boils constantly at 204" (mercury column in vapour, pressure 765 mm.), and does not solidify a t - 10'. Its sp. gr. = 1.6555 at 1 2 * 5 O , and pD = 1.583 at 15'. For purposes of comparison, we determined the sp. gr. and refractive index of metachlorobromobenzene and found its sp. gr. = 1.62'74 at 14*, and pD = 1.578 at 15'. The following table gives some of the constants of the chlorine and bromine di-derivatives of benzene. ~~ Sp. gr. I Boiling points. I Ortho. 1 Meta. 1 Para. I Ortho, I Meta. 1 Para. I- I- I- 1-1-1- UNIVERSITY COLLEGE, BANGOR.

ISSN:0368-1645    

DOI:10.1039/CT8987300254

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

PREPARATION OF DRY HYDROGEN CYANIDE, ETC. 255 XXIK-Prepuration of Dry Hydrogen Cyanide and Carbon Monoxide. By JOHN WADE, B.Sc. and LAURENCE C. PANTING, M.B., B.Ch. IT is obvious that, by the action of sulphuric acid on potassium cyanide, both hydrogen cyanide and carbon monoxide may be formed, but it does not seem to be generally known that by suitably varying the concentration of the acid, a practically quantitative yield of either product can be obtained. The alkali cyanide is of course decomposed by sulphuric acid of all concentrations, and on distillation with the dilute acid yields dilute hydrocyanic acid; with a less dilute acid, however, less water passes over, and with a mixture of equal volumes of sulphuric acid and water, the hydrogen cyanide contains only such small quautities of aqueous256 WADE AND PANTING: PREPARATION OF DRY vapour as are readily removed by passing it over warm calcium chloride.When a larger proportion of sulphuric acid is used, a certain amount of carbon monoxide is formed, and as the concentration of the acid is increased the volume of gas increases, whilst the amount of hydrogen cyanide diminishes ; and finally, when ordinary concentrated sulphuric acid is allowed to act on the cyanide, nearly pure carbon monoxide is evolved in almost theoretical quantity. Prepuration @ Dry Hydrogen Cyanide. I n preparing hydrogen cyanide free from water, ordinary 98 per cent. lump cyanide (100 grams), in pieces of the size of a hazel nut, is placed in a capacious flask, provided with a drop-funnel and delivery tube, the latter being connected through two U-tubes of calcium chloride with a series of two Y-tubes (U-tubes furnished with tubuluses), the stems of which pass through the necks of inverted bell-jars into two receiving bottles.The drying tubes are immersed in a vessel of water at about 35", and are filled respectively with pieces of fused calcium chloride and fragments of the well-dried porous material. The condensing tubes are cooled, the first with ice and water, and enough salt to reduce the temperature to about - lo', the second with ice and salt at about - 20'; it is not advisable to cool the first tube much below the temperature Specified, or the tubulus may become choked with crystals of the frozen acid. The receiving bottles are cooled with ice, and closed with corks through which the tubuluses pass, The great bulk of the acid collects in the first receiver, only 2 or 3 per cent.passing on to the second, and practically none escapes condensation. A current of air having been blown through the drying and con- densing tubes, to dry the latter, and the cooling and warming baths brought to the requisite temperatures, a cold mixture of equal volumes of sulphuric acid and water (100 C.C. of each) is allowed to drop on to the cyanide at such a rate that about one drop of hydrogen cyanide falIs into the first receiver per second. As each drop of acid reaches the cyanide, there is a brisk effervescence and frothing, with develop- ment of heat, the mixture becoming sufficiently hot to retain the potassium hydrogen sulphate in solution. At the end of the action, the vapour in the flask and solution is expelled by heating the latter to incipient ebullition.In this way, with the quantities specified, about 40 grams (58 c.c.) of practically pure hydrogen cyanide is obtained, which on rectification from a little phosphorus pentoxide over a bath of warm water, passes over entirely between 26.2' and 26.3'. The average yield of the pureHYDROGEN CYANIDE AND CARBON MONOXIDE. 257 substance is 38.5 grams, the calculated amount being 40.8 grams; the maximum yield obtainable from 100 grams of potassium ferrocyanide is 15.3 grams. Preparation of Cadon Monoxide. I n preparing carbon monoxide from potassium cyanide, the same generating apparatus is used, cold concentrated sulphuric acid being substituted for the diluted acid.The gas is purified from accompany- ing hydrogen cyanide vapour, and in certain eventualities from carbon and sulphur dioxides, by washing twice with strong potash solution ; as a rule, i t is quite free from carbon dioxide, but commercial lump cyanide sometimes contains a little carbonate. If the acid is allowed to flow too fast on to the cyanide, so that the temperature becomes unduly high, the sulphuric acid is reduced, and sulphur and carbon dioxides are formed; but this does not occur if the action is kept within moderate bounds. Even with a slow delivery of the acid, the evolution of the gas is very rapid, and several litres may be prepared in a few minutes. The yield of pure carbon monoxide, completely soluble in cuprous chloride solution, obtained in this way from 50 grams of cyanide, varies from 14 to 16 litres, the calculated quantity a t 15' and 760 mm.being 17.4 litres, moreover, the process is quite as convenient as the formic acid process, and much more economical. The action of concentrated sulphuric acid on anhydrous hydrogen cyanide is similar. The mixture soon becomes hot, but although some of the cyanide is of course volatilised, the greater part enters into com- bination, and on applying heat abundance of carbon monoxide is evolved. If the liquid is overbeated, or even in the cold, if sufficient time is allowed, carbon and sulphur dioxides are formed, as with the alkali cyanide. The Mechanism of the Carbon Monoxide Reaction. It would appear at first sight that the quantitative conversion of the carbon of potassium cyanide into carbon monoxide is effected by ordinary hydrolysis and dehydration, the hydrogen cyanide being con- verted into formic acid, and this into carbon monoxide.Whether hydrogen cyanide be formonitrile, H-CiN, or what may be termed carbamine, H*N:C, the same products of hydrolysis will be formed : from the nitrile a carboxylic acid (formic acid) and ammonia, and from the carbamine, an amine (ammonia) and formic acid. Sufficient water for the purpose is indeed present in ordinary concentrated sulphuric acid, assuming the complete action to be represented by some such equation as KCN + 2H,S04 + H,O = KHSO, + NH4*HS0, + CO. VOL. LXXIII. S258 RICE : MANGANIC SALTS. But the change is equally well brought about by a mixture of two volumes of concentrated sulphuric acid with one of Nordhausen acid, in which water can hardly be present, and it is therefore probable that part, at all events, of the latter is derived from the acid sulphate.This point would scarcely be worth attention, were it not thak potassium hydrogen sulphate itself behaves in a closely parallel manner. On heating an intimate mixture of the dry salt with dry potassium cyanide in an air bath a t 230-250°, anhydrous hydrogen cyanide distils over, and may be collected in the apparatus described above, the yield corresponding with about 55 per cent. of the total cyanogen ; this, of course, is the action which would naturally be expected, But a large amount of carbon monoxide is also evolved at all stages of the action, and the residue left in the flask after the action has ceased smells strongly of ammonia, and dissolves in water, forming a strongly alkaline solution.As this carbon monoxide and ammonia are formed by the interaction of anhydrous substances, the hydrolytic water is necessarily derived from the potassium hydrogen sulphate, in accordance with such equations as the following : KCN + 3KHS0, = K,SO, + R,S,O7 + CO + NH3 ; 2KCN + 8KHS0, = 3K2S0, + 2K,S,07 + 2CO + (NH,),SO,. KCN + KHSO, = K,SO, + HCN. It would thus appear that both sulphuric acid and potassium hydrogen sulphate can behave at the same instant and under the same conditions both as hydrolytic, and as dehydrating agents. Anhydrous potassium ferrocyanide, it may be mentioned, interacts with potassium hydrogen sulphate in a similar manner, about 40 per cent. of the cyanogen being converted into hydrogen cyanide, and about 10 per cent. into carbon monoxide and ammonia, the residual 50 per cent. remaining in combination as potassium ferrous ferro- cyanide, as in the ordinary method of preparing dilute hydrocyanic acid. With excess of acid sulphate and a t a high temperature, carbon and sulphur dioxides are evolved in abundance, as with the simple cyanides. CHEMICAL LAEORATORY, GUY’S HOSPITAL, S.E.

ISSN:0368-1645    

DOI:10.1039/CT8987300255

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

258 RICE : MANGANIC SALTS. XXIV.-Manganic Salts. By CHARLES EMMANUEL RICE, B.A. SEVERAL salts of manganese derived from the oxide Mn20, have been described, but they are all unstable, and decomposed by contact withRICE : MANGANIC SALTS. 259 water, hydrated higher oxides of manganese being precipitated ; the most stable salt seems to be the acetate, Mn(C2H30,), + 2H,O, obtained by Christensen ( J . p, Chem., 1883, [ ii], 28, 1). The salmon-coloured meta- phosphate, Mn(P03), + H20, has been obtained by Hermann (Ann. Phys. Chem., 1848, 74, 303), and the pyrophosphate, MnHP,07, by Lefhvre (Compt. rend,, 1890,110, 405) ; Carius (Annalen, 1856,98, 5 3 ) has obtained the sulphate Mn,(SO,),. The potassium and ammonium alums are also well known, but they are decomposed by contact with much water ; a black double oxalate, KOMn,(C,O,)O + 6H,O, has been obtained by Kehrmann (Berichte, 20, 1594) ; the nitrate appears to bo unknown, but the fluoride, MnF3+3H,0, has been described by Christensen (J.pr. Chem., 1887, 3!5, 57) together with double salts of the type 2M'F,MnF,, but so far, no other haloid compound OF manganese in which the metal is more than bivalent. Nickles (Ann. China. Phye., [iv], 5, 161) claimed to have produced MnCl,, but there is really no evidence for the existence of such a compound. The action of hydrochloric acid on the higher oxides of manganese has attracted a good deal of attention in this connection, but thc higher chloride of manganese which is supposed to exist in the dark- coloured solution has not so far been isolated. W.W. Fisher (Trans., 1878, 33, 409) on diluting the liquid with water and weighing the precipitated peroxide, found that for each atom of manganese precipitated, two atoms of available chlorine (chlorine that will liberatc iodine from potassium iodide) disappeared from the solution ; from this he inferred the presence of a compound, MnCl,. Pickering (Trans., 1879, 35, 1, 654) showed that this could be equally well accounted for by the existence of a compound, Mn2Cl,; and he offered further evi- dence in support of this formula by showing that when manganeso dioxide is dissolved in presence of manganous chloride, the amount of dioxide precipitated on diluting with water increases, the increase being in a greater proportion up to the addition of one molecule of MnCI,.This would seem to show that the chloride combines with the two atoms of chlorine set free when 2Mn0, dissolves, forming a second molecule of Mn,Cl,. He also observed that Mn,O, dissolves in cold hydrochloric acid without effervescence through evolution of chlorine, whereas a marked effervescence always accompanies the dissolution of MnO,. This has been repeatedly confirmed by me with acid satu- rated at - 15O, and it would appear that when MnO, is acted on by hydrochloric acid, one atom of chlorine is a t once set free for each atom of manganese present, and another subsequently on heating the liquid, due to decomposition of the manganic chloride in solution. Pickering's experiments made it very probable that a compound, Mn,CI, or MnC'l,, yielding chlorine and MnCl, by dissociation, exists in the dark-coloured solution.He failed, however, to show s 2260 RICE : MANGANIC SALTS. I. That the reaction is reversible, 'and, therefore, that a definite 11. That this compound can be obtained in the solid state. The following is a record of an experimental investigation of these problems. I. Crystals of chlorine hydrate were sealed up in a tube with some hydrochloric acid solution of mangarious chloride, the tube being allowed to remain at the ordinary temperature. No apparent change occurred for some hours, but gradually, in the course of a few days, the manganous chloride solution became dark-coloured owing to the synthesis of MnC13, the depth of colour eventually reaching a maximum. Evidently the velocity of the reaction MnCI, + C1= MnC1, is very slow.In the absence of acid, a deposit of higher oxide of manganese is formed. The same changes occur if a cooled solution of manganous chloride be saturated with chlorine at the atmospheric pressure and left for some days. 11. I have succeeded in obtaining two double chlorides containing the manganic chloride fairly pure and in well-defined crystals. About 50 grams of a higher oxide of manganese (Mn,03, Mn304, MnO,, or a mixture of any of these) was placed in a flask with 250 C.C. of cold con- centrated hydrochloric acid, and the flask, placed in a freezing mixture of ice and salt, was saturated with dry hydrogen chloride. After the mixture had been left to stand in the freezing mixture for half-an- hour to allow particles of undissolved oxide to settle, the fairly clear upper half was decanted, and a few C.C.of a slightly warmed saturated acidified solution of ammonium chloride was added to it drop by drop, The liquid was then replaced in the freezing mixture, resaturated with hydrogen chloride, and, after the lapse of about half an hour, filtered through a cooled Gooch crucible, The mass of small, dark, lustrous crystals left on the filter was washed with a few drops of cold saturated hydrochloric acid, drained on a porous plate, and left in a desiccator over soda-lime for some days under reduced pressure. The crystals are perfectly transparent under the microscope, trans- mitting a ruby-coloured light. They have a slightly pungent odour, and may be raised to a temperature of 100" without perceptible change ; above that temperature, however, they evolve chlorine and water, and, if heated in a current of air, leave a white residue of man.ganous chloride and ammonium chloride. They dissolve in hydro- chloric acid, yielding a liquid resembling that from which they were obtained. Water at once decomposes them, about half the total manganese separating as hydrated higher oxide, and half remaining in solution as manganous chloride. Analysis gave results whiah correspond with the formula MnC13,2NH,Cl + H,O. chlorinated compound really is formed and dissociated,RICE : MANGANIC SALTS. 261 hIn. c1. NH4. H,O (by difference). Calculated .. . . . 19.1 61.9 12.6 6.3 Found .,..... ... 19.1 62.1 12.2 6.6 The ratio of Mn to C1 atoms is 1 : 5 ; from the formula one would expect one of these atoms, or 12.4 per cent.of chlorine, to be ‘‘ loose,” o r available for oxidising purposes. It was found, on treatment with potassium iodide, that 12.6 per cent. available chlorine was present in the crystals. The potassium salt, RlnC1,,2KCl + H,O, was obtained in the same way, and all its properties were found t o be very similar. The water was determined directly in this salt, whereas in the ammonium com- pound it mas estimated by difference. The analytical results, the average of three separate analyses, are as follows. Btn. c1. K. H,O. C1 (loose). Calculated.. . 16.7 54-0 23.8 5 5 10.8 Found ........ 16.8 53.7 23.6 5.9 10.7 Several attempts have been made t o isolate the manganic chloride itself, but so far without success.When heated in a current of gaseous hydrochloride or chlorine, the double salts decompose, yielding water and chlorine, and leaving a residue of manganoua chloride and alkali chloride. With cold sulphuric acid, they evolve hydrogen chloride and chlorine, manganous sulphate being deposited ; on mixing them with phosphoric anhydride, it manganic phosphate resembling per- manganic acid in colour is produced, The above salts are apparently isomorphous with the corresponding compounds of ferric chloride, FeC13,2NH,C1 + H,O and FeCl,,ZKCl+ H,O, which form monoclinic crystals with triangular faces, resembling those of octahedra ; they help, therefore, to draw closer the relationship between iron and manganese. From the position of manganese in the series, whether it be placed with the iron group or with the halogens, we should hardly expect it to form quadrivalent compounds of the type MnCl,. The dark, chlorine-generating solution is a solution of manganic chloride, MnCl,, and there is no evidence for the existence of any other compound, besides the dioxide, in which the manganese is in the quadrivalent state. The author wishes t o express his indebtedness to Professor Tilden, under whose superintendence these experiments have been carried out. ROYAL COLLEGE OF SCIENCE, LONDON.

ISSN:0368-1645    

DOI:10.1039/CT8987300258

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

262 REYNOLDS : CHENICAL PROPERTIES OF CONCENTRATED XXV.-Chenzical Pyopeyties o f Concentrated Solutions P a r t Ii Double Potassium Cw- of certuin Salts. bonates. By WILLIAM COLEBROOK REYNOLDS, A.R.C.S. IN the course of some experiments on the preparation of potassium percarbonate recently discovered by Messrs. Constam and Von Haussen (Zed. fur Elektrochem., 1897, p. 455, and Moniiew Sc., August, 1897), I noticed that copper sulphate would dissolve in the concentrated solution of potassium carbonate employed, yielding a deep blue liquid ; as nothing appeared to be known about the compound thus formed, attempts were made to isolate it.* It mas soon found that the salts of several other metals will also dissolve in this solution to a greater or less extent, and that from such solutions crystalline compounds could in most cases be obtained which on analysis proved to be double carbonates.Most of the information we possess about the double carbonates we owe to Doville (Ann. Chim. Phys., [iii], 33), who found that by digesting metallic nitrates with a saturated .solution of potassium nydrogen carbonate a t the ordinary temperature, compounds of the type MKH(CO,), were formed, where M is cobalt, nickel, or magnesium, and that if the solutions are heated for some hours they passed into MK,(CO,),; he also obtained the sodium double salts of the latter type with cobalt, nickel, copper, and magnesium. As the result of a large number of experiments, I find that the best general method of preparing the compounds is the following. To the concentrated solution of potassium carbonate (sp.gr. about 1 *55) the finely powdered acetate is added and the solution set aside ; the crystals which form are then separated from the mother liquor by draining on a disc of toughened filter paper placed on a Buchner filter plate, employing LZ water pump to diminish the pressure beneath. The crystals are washed with a mixture consisting of two volumes of alcohol and one of glycerol (which floats on the salt solution and does not mix with it) until all the mother liquor is displaced, when the washing is completed with alcohol only until all the glycerol is removed. The crystals can be dried in a vacuum at the ordinary temperature over aulphuric acid. The acetates are most convenient as, owing to the solubility of potassium acetate, the deposition of crystals of other potassium salts * The solutions of copper potassium carbonate introduced by Soldaini (Gnzzetta, 1876, 6, 322) and afterwards modified by 0 s t (Ber., 1890, 23, 1035 and 3003) for use in sugar analysis contain potasium hydrogen carbonate.SOLUTIONS OF CERTAIN SALTS.PART I. 263 can be avoided. The freshly precipitated carbonates and hydrated basic carbonates have been used in several cases, but their employ- ment is attended with some difficulty. These compounds are decomposed by water alone, more or less rapidly, and alcohol alone causes the deposition of potassium carbonate upon the surface of the crystals of the new salt ; hence the necessity for the use of the mixture already referred to. I n one case, however, namely, that of the ferric salt, this mixturewas found to decompose the compound and petroleum was substituted for it, as, notwithstanding its lower density and immiscibility with aqueous liquids, it served to displace the mother liquid.Copper Potassium Curbonate.-This is very soluble in the strong solution, and a considerable quantity of copper acetate has to be added to get any crystals. The salt crystallises in three ways, namely, in the anhydrous form, with 1H,O and with 4H,O ; the three forms can be obtained from the same solution, provided the right proportions are employed. Thus, a solution made by adding to 70 C.C. of potassium carbonate solution of sp. gr. 1.35, 13 grams of finely powdered copper acetate, a t 65O, filtering and cooling, deposited, after 48 hours, (a) dark, blue six-sided plates of the anhydrous salt, ( b ) light blue, silky, clustered needles of the salt with lH,O, and ( c ) large greenish-blue, square tables of the salt with 4H20.The last could be picked out by hand and freed from the liquid by means of filter paper. The com- pounds (a) and (6) appear to be deposited indifferently from solutions of the same strength and temperature, and I have not been able to find the conditions under which the individual compound could be obtained with certainty. To prepare (a) or (b), the solution of potassium carbonate (sp. gr. 153) is added to the finely powdered copper acetate in a mortar, and after being triturated a few seconds to break up any lumps, when practically all dissolves, it is filtered rapidly through glass wool and set aside, A mixture of 36 grams of the acetate and 100 C.C.of the solution begins to crystallise in 2 or 3 minutes, and after about 10 minutes those crystals which have been formed should be separated and the remainder in two or three fractions, to prevent the solution becoming clogged. Thirty grams and 100 C.C. and separation in two instalments; or 17 grams and 100 C.C. all a t once, are convenient quantities. The proportion of 12 grams to 100 C.C. generally requires some hours before crystals appear. Analysis of the anhydrous compound (a dark blue, crystalline powder). co,. CUO. K,O. H,O. Found ... . . . . . . .. . . . . . . . .. . . . , . .. ... 32.92 31.54 35-39 0.06 Calculated for CuK,(CO,),. .. . , .33-67 30.31 36.02 nil.264 REYNOLDS : CHEMICAL PROPERTIES OF CONCENTRATED Analysis of the light blue, silky needles.Found. ................................ ..30*77 29.93 32.70 6.60 Calculated for CuK,(CO,), + H20.. .31-45 28.41 33.70 6.44 CO,. CuO. K,O. H,O(by diff.). (Specimens (a) and (b) probably were contaminated with a little hydrated copper carbonate.) Analysis of the greenish-blue tables. CO,. CuO. K20. H,O(bydiff.). Found .................................... 26.24 23.14 28.84 21.78 Calculated for CuK,(CO:,), + 4H,O 26.39 23.78 28.26 21.57 The copper solution decomposes slowly when heated to 65O, and rather quickly at 85O, cupric oxide being precipitated; if, however, a large quantity of the acetate is added, no black precipitate forms even at looo, but a green powder of the composition Cu,(OH),C03.Cobalt Potassium Carbonate.-A cobalt salt dissolves in the warm solution of potassium carbonate and small six sided rose-coloured crystals soon separate on cooling. CO,. COO. K,O. H,O. Found .................................... 26.30 22.78 28.70 22.10 Calculated for CoK,(CO,), + 4H20 26.72 22.78 28.68 21-86 much smaller quantity, and pale green crystals soon form on cooling. CO,. NiO. K20. H20. Found .................................... 26.68 22.64 28.82 21.86 Calculated for NiK2(C03), + 4H20 26.72 22.78 28.64 21.86 Nickel Potassium Carbonate.-A nickel salt dissolves as above, but in Magnesium Potassium Cadonate.-The magnesium compound is obtained in like manner, by adding magnesium acetate t o the solution, crystals appear in a few minutes.Found .................................... 29-74 14.15 32.02 24.09 Calculated for MgK,(CO,), + 4H,O 29.86 13*.72 32.00 24.42 CO,. MgO. K20. H,O (by diff.). The last three are identical with the compounds obtained by Devillc by prolonged heating with a saturated solution of potassium hydro- gen carbonate. Manganese Potassium Carbonate.-The last three salts are best pre- pared from warm solutions, but this one must be prepared at the ordinary temperature, otherwise much oxidation occurs. The acetate was powdered in a mortar, and the solution added and stirred for a few seconds; the crystals separate out very quickly, as the salt is scarcely soluble in the solution.SOLUTIONS OF CERTAIN SALTS. PART I. 265 CO,. MnO. K,O. H20. Found .................................... 26.86 21.80 29.20 22.14 Calculated for MnK2(C0,), + 4H20 27.03 21.82 29.03 22.12 Ferrous Potassium Carbonate .-In order to prevent oxidation,this salt must be prepared by a slight modification of the process. The follow- ing answered very well.A bottle containing about five-sixths of its volume of the solution was filled up with a saturated solution of ferrous chloride and stoppered; on mixing the two liquids by shaking, a white precipitate of FeCO,, is at first formed, but this soon dissolves yielding a greenish solution, from which the compound FeK,(CO,), + 4H20 crystallises out in nearly white scales after about 10 minutes. The salt can be separated and purified in the manner already described. GO,. FeO. K,O. H,O (by diff.). Found .................................... 26.80 21.93 28.60 22.67 Calculated for FeK,(CO,), + 4H20 26.99 22.06 28.89 22.06 Calcium Potassium Carbonate.-This can be easily prepared by pouring the solution on to the finely powdered acetate in a mortar and t r i t u - rating for some time.The crystals are anhydrous, GuK2(CO& and are quite different in form from the last five compounds, which con- tain water. CO,. CuO. K20. Found .................................... 36.55 22.40 38.94 Calculated for CuK,(CO,), ............ 37.00 23.50 39.50 (The salt probably retained about 2 per cent. moisture.) Silver Potassium Carbonate.-This can be prepared by adding powdered silver nitrate to the solution, but I found it better to add the saturated solution of the nitrate to a large excess of the potassium carbonate solution, as in the former case particles of the nitrate which have become coated over with the compound are liable to be mixed with it.The yellow precipitate at first formed soon changes into the white double salt, AgKCO,. I f exposed to light during the washing, superficial reduction takes place, due to the glycerol. GO, Ag,O. R,O. Found .................................... 20.58 54.50 23.90 Calculated for AgKCO, ................ 21.23 56.01 22.76 This salt has been already obtained by De Schulten (Conapt. Tend., 1887, 105, 811). Bismuth Potassium Carbonate.-When powdered bismuth nitrate is added to the solution, some effervescence occurs, but a considerable quantity slowly but completely dissolves ; the crystallinecompound that separates after a short time is mixed with crystals of potassium266 PROPERTIES OF CONCENTRATED SOLUTIONS OF SALTS.nitrate and hydrogen carbonate from which it cannot be separated. It was found better, therefore, t o employ the subnitrate. The crystals, which are microscopic, when prepared at the ordinary temperature are of an entirely different form from those obtained by adding the powder to the solution warmed to about 75O; they have? however, the same composition. The commercial subnitrate contains a small quantity of a constituent which is not acted on by the solution. Most of this impurity can be got rid of by pouring the solution, after the addition of the subnitrate, into a tall cylinder and decanting from the impurity which settles down more quickly than the fine crystals.A purer pro- duct can be obtained by preparing the subnitrate from the nitrate clsing a small quantity of water, draining with the aid of the filter pump, and adding the precipitated salt to the concentrated solution heated to about 70". Found ................................ 55.00 21.86 20.26 2.88 Calculated for Bi,OK,(CO,), + H,O 54.90 22.24 20.73 2.13 in well-defined crystaIline form. MnK,(CO,), + 4H20 ...................... 1 Not previously known, CuK,(CO,),. CuK,(CO,), + H,O CuK2(C0,), + 4H20 Obtained by Deville by heating the bicarbonate. FeK2(C03)2 + 4H,O CaK,(C0,)2 Bi,OK,( CO,), + H20 CoK,(CO,), + 4H20 NiK,(CO,), + 4H.,O.. MgK2(C0,), + 4H,O ....................... AgK(C0,) ................................. Obtained by De Schulten. In addition to these, I have obtained compounds containing lead, zinc, iron (ferric) chromium, mercury, and cadmium, but not crystalline or of definite composition.It is noticeable that the chemical behaviour of potassium carbonate in the form of a concentrated solution is quite different from that of the same salt in a dilute solution. I n the latter case, normal or basic carbonates are formed, whereas, in the former, double salts are pro- duced which may be soluble in the liquid. The solutions and the compounds in the solid state are decomposed when diluted or brought into contact with water ; in this respect, they are similar to the double sulphates, which are all decomposed by water, I n the case of such compounds as CaNa,(SO,), and CaK,(SO,),, this is The crystals separate almost immediately. Bi,O,.K,O. CO,. II,O (by diff.). The following list contains all the salts I have been able to isolate ................................ ........................ ........................ ........................ ............................... ..................... ........................ I I ......................THE COLOURING MATTERS OF INDIAN DYE STUFF. 267 evident, since calcium sulphate is precipitated, but in the case of the soluble salts, although not so obvious, the fact of decomposition has been proved by Raoult (Compt. rend., 1884, 99, 914), and also from thermochemical data. It is noticeable that the ratio of K to CO, is unity in all the above compounds, and this is also the case with the double sulphates K : SO,, and double oxalates K : (CO,),, and it is possible to write the constitu- tion of these three classes of double salts thus. Potass. mangan. Potass. mangan. Potass. mangan. carbonate. sulphate. oxalate. A concentrated solution of potassium succinate behaves very like that of potassium carbonate ; although the succinates of the metals, except the alkalis, are almost insoluble, yet on adding their salts to a concentrated solution of potassium succinate many of them dissolve, and crystalline double compounds are obtainable. I propose to give an account of these in a future communication. The above investigation was carried out under the supervision of Professor Tilden, to whom I am much indebted for his advice and encouragement. ROYAL COLLEGE OF SCIENCE, LONDON.

ISSN:0368-1645    

DOI:10.1039/CT8987300262

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

THE COLOURING MATTERS OF INDIAN DYE STUFF. 267 XXVI.-The Colouring Matters of the Indian Dye Stuf As b ary, Delphinium xal il. By ARTHUR GEORGE PERKIN, F.R.S.E., and JULIUS ALDRED PILGRIM. ASBARCI consists of the dried flowers and flowering stems of Delphinium xalil, a perennial, herbaceous plant belonging to the Ranuncukacece, which is found in great quantity in Afghanistan. Dr. Aitchison says of it (Watts’ Dictionary of the Economic Products of India, 1890, 3, p. 70) : (‘ This plant forms a great portion of the rolling downs of the Badghis; in the vicinity of Gulran it was in great abundance and when in blossom gave a wondrous golden hue to the pastures; in many localities in Khorassan of about 2000 feet altitude, it is equally common.’’ The dried fragments and flowering stems are taken to Multan and other Punjab towns, from which they are conveyed all over India.It is much used in silk dyeing for the production of a sulphur-yellow colour known as ‘‘ gandhaki,” and together with Batisca cannabina to268 PERKIN AND PILGRIM : THE COLOURING MATTERS OF THE obtain a similar shade on alum mordanted silk; it is also used in calico printing. The flowers, which are bitter, are likewise employed medicinally as a febrifuge. No examination of the colouring matter of this dye stuff appears to have been previously made, although an account of its tinctorial pro- perties has been communicated to the Society of Chemical Industry (J. Xoc. Ch. Ind., 1895, 14, 458) by J. J. Hummel and one of us. Our thanks are due to the authorities of the Imperial Institute, who, on the application of Professor Hummel, were good enough to procure a considerable quantity of this material for our investigation.EXPERIMENTAL PART. At the commencement of this investigation, the flowers and flowering stems of the plant were examined together, but the latter, being prac- tically devoid of dyeing property, were subsequently discarded, an economy of labour being thus effected. Various methods suggested themselves for the isolation of the colouring matter, which exists here entirely as a glucoside, and ultimately the following was adopted. The dye stuff was extracted with ten times its weight of boiling water, the mixture strained through calico, and the filtrate, after treatment with a little sulphuric acid, was again boiled for 15 minutes.The decomposition of the glucosides is very readily effected, and the action should not be prolonged, otherwise a tarry product is formed, which is a great hindrance in the later processes. On cooling, a brownish-yellow powder separated, which was collected, drained upon a porous tile, and when dry digested with boiling alcohol which dissolved the colouring matter, leaving a considerable residue of calcium sul- phate. The dark coloured solution, after being evaporated to a small bulk, was poured into a large volume of ether, the mixture washed with water until the washings were colourless, and then agitated with dilute alkali to free the substance from a wax which alone remained in the ether after this treatment. On acidifying the alkaline liquid, it deposited yellow flocks of the colouring matter still contaminated with some impurity of an acid nature; to remove this, excess of sodium hydrogen carbonate was added to the mixture, and the product ex- tracted with ether; this now dissolved nothing but tho colouring matter, and the latter, on evaporation, was left of a pure yellow colour.Xpwingly soluble Colozcring Matter. It was evident on examination that the product thus obtained con- tained two colouring matters at least, one of which was distinguished by its sparing solubility in alcohol. After being isolated by this means it was converted icto its acetyl derivative, which was recrystal-INDIAN DYE STUFF ASBARG DELPHINIUM ZALIL. 269 lised, and then decomposed in the usual manner, the regenerated colouring matter being finally crystallised from acetic acid.0.1164 gave 0,2598 GO, and 0.0400 H,O. C= 60.86 ; H= 3.81. C1,Hl,O7 requires C = 60.76 ; H = 3.79 per cent. It formed a glistening mass of yellow needles, resembling rhamnetin in appearance, very sparingly soluble in boiling alcohol or acetic acid. With lead acetate in alcoholic solution, an orange-red precipitate was formed, whilst ferric chloride gave a greenish-black coloration. With the haloid acids, in the presence of acetic acid, no compounds of the acids were produced; sulphuric acid also reacted with difficulty, and owing to the small yield and the instability of the product thus ob- tained, it was not further examined. Fusion, with Alkali.-The colouring matter was digested with very concentrated potassium hydroxide solution at 200--220° for 1 hour, and the brown melt dissolved in water; the green solution was then neutralised with acid, extracted with ether, and the extract evaporated.The residue, after being dissolved in water, was treated with lead acetate, and the yellowish-white precipitate thus produced was collected, washed, and decomposed with dilute sulphuric acid. From the clear liquid, ether extracted a crystalline acid, soluble in dilute alkali t o form a green liquid which became brown on exposure to air. The acid was now dissolved in water, and the solution saturated with salt, causing the separation of a trace of a viscous product which was removed by filtration. The filtrate was extracted with ether, and the residue obtained on evaporation purified by crystallisation from water.With dilute alkali, this product now yielded a colourless solution, and examination showed that the green coloration previously obtained in this manner was due to the impurity removed by the salt treatment. 0.1135 gave 0.2275 CO, and 0*0400 H,O. C = 54.66 ; H= 3.91. C,H,O, requires C = 54.54 ; H = 3.90 per cent. It cryetallised in colourless needles melting at 194-196', gave a green coloration with aqueous ferric chloride, and evidently consisted of protocatechuic acid. The filtrate from the lead precipitate was treated with sulphuric acid, filtered from lead sulphate, neutralised with sodium hydrogen carbonate, and extracted with ether. The crystalline residue left on evaporation, when purified, melted at 210°, and was found to be phloroglucinol.The substance which gave the green coloration when the melt was dissolved in water could not be isolated in a pure condition, the quantity present being very small. In a second experiment, em- ploying a new preparation of the colouring matter, this substance was270 PERKIN AND PILGRIM : THE COLOURING MATTERS OF THE not detected, and it was thought that the green coloration had been due t o an impurity. It appeared subsequently, however, that the condition necessary for its production was the use a t the first of a comparatively dilute caustic alkali solution, preferably four parts of alkali in one of water. It is evident, therefore, that the reaction is characteristic of this colouring matter. Action of llydriodic Acid.-The similarity of this substance to rhamnetin, especially in its behaviour towards mineral acids, sug- gested that it contained a methoxy-group. Examination by Zeisel’s method proved this t o be the case.0.1750 gave 0.1310 AgI. CH,=4*77. C15H90,(OCH,) requires CH3 = 4.74 per cent. To the hydriodic acid residue, after dilution with water, sodium hydrogen sulphite solution was added, and the yellow, flocculent pro- duct collected and crystallised from dilute alcohol. 0.1021 gave 0.2242 CO, and 0.0345 H,O. The product from a second experiment was acetylised, and the colourless substance, which melted a t 189-19lo, was analysed. C = 59.88 ; H= 3.75. C15H,,07 requires C = 59.60 ; H = 3.31 per cent. 0.1 165 gave 0.2494 CO, and 0,0462 H,O. C,,H507(C,H,0), requires C = 58.59 ; H = 3-90 per cent.The above results, and the fact that the methyl ether when fused with alkali gives phloroglucinol and protocatechuic acid, indicated that this substance, C15H1007, was quercetin. An examination of its general properties corroborated this view. It is thus evident that the sparingly soluble colouring matter of Delphinium xalil is a puer- cetin monomethyl ether. Acetyl Compound.-To confirm the above results, and to determine if this colouring matter was identical with either of the two known methyl ethers of quercetin (rhamnetin and iso-rhamnetin), it was acetylised in the usual manner, and the product crystallised from alcohol. C = 58.38 ; H = 4.40. 0.1213 gave 0.2650 CO, and 0.0445 H,O. C1,H,O7(C,H,O), requires C = 59.50 ; H = 4.13 per cent.A determination of the methoxy-group gave the following result. 0.2076 gave 0.0880 AgI. CH, = 2.70. Theory requires CH, = 3.09 per cent. Liebermann’s method was employed for the estimation of the acetyl C = 59.57 ; H = 4.07. groups.INDIAN DYE STUFF ASBARG DELPIIINIUM ZALIL. 271 1.0740 gave 0.7047 C1,H1,07. Found 65.6 1. The theory for four acetyl groups requires C,,HI2O7= 65.29 per cent. It formed a glistening mass of colourless, hair-like needles, melting at 195--196O, identical in appearance and general properties with acetyliso-rhamnetin. The sparingly soluble colouring matter of the Delphinium xalil is, therefore, iso-rhamnetin, a substance but recently isolated for the first time (Trans., 1896, 69, 1650) from the petals of the yellow wallflower (Cheiranthus Cheiri).The difficulty of obtaining a suEcient supply of raw material did not allow at that time of the determination of the position of the methoxyl group in this colouring matter. Experiments were, therefore, now instituted with this object. Methylation of 1so-rhamnetin.-As in the case of rhamnazin (quercetin dimethyl ether) (Trans., 1897, 81 8), this reaction was studied, for should the methoxy-group be present in the ortho-position relatively to the carbonyl group, a quercetin pentamethyl ether might be produced, and not the tetramethyl compound, which is always formed when quercetin itself is so treated. Iso-rhamnetin (1 mol.) dissolved in a solution of potassium hydroxide (Pmols.) in methylic alcohol was digested with excess of methylic iodide for 24 hours. After removal of the unattacked iodide and excess of alcohol, the residue was dissolved in ether, and the solution washed with dilute alkali and evaporated; the product, after crys- tallisation from acetone, formed pale yellow needles melting at 154-1556O.0.1170 gave 0.2745 CO, and 0.0543 H,O. C = 63.98 ; H= 5.15. C,,H,O,(OCH,), requires C = 63.69 ; H = 5.03 per cent. It was found to consist of quercetin tetramethyl ether. Oxidation of 1so-rhamnetin.-The previous experiment having given a negative result, it was necessary to determine if the methoxyl group was present in the catechol nucleus. For this purpose, air was aspi- rated through a dilute alkaline solution of the colouring matter until this, which was at first yellow, had become brown, and on treatment with acids no longer yielded a precipitate ; the liquid was then neu- tralised with acid, treated with excess of sodium hydrogen carbonate, extracted with ether (A), and again acidified and extracted with ether (B).The latter extract, on evaporation, deposited crystals which were purified by recrystallisation from water until colourless ; a lustrous mass of needles was thus obtained, which dissolved sparingly in cold water, gave no reaction with ferric chloride, and melted at 206-207O. This substance was vanillic acid, and the locality of the methoxg-group in iso-rhamnetin is thus evident. On evaporating extract (A), a residue was left having the reactions of phloroglucinol.272 PERKIN AND PILGRIM : THE COLOURING MATTERS OF THE Aluminium. I Employing the constitution assigned to quercetin by Herzig, that of iso-rhamnetin may be expressed as follows. OCH, OH/\/ O \-/--\OH I I /I \--’ Tin.Chromium. Iron. Consequently rhamnazin (Zoc. cit.), a quercetin dimethyl ether, is closely related to iso-rhamnetin and to rhamnetin, it being a mono- methyl ether of both these colouring matters. Rhamnazin. Rhamnetin. Dyeing Properties.-A comparison was made of the tinctorial pro- perties of iso-rhamnetin and quercetin, woollen cloth mordanted with aluminium, tin, chromium, and iron being employed for this purpose. The shades obtained were as follows. Iso-rhamnetin Quercetin . . . Lemon yellow. Orange yellow. Orange brown. Pale brown o h 6 . Brown orange, Bright orange. Red brown. Green black. inclining to yellow. I I I I These results are interesting, as they indicate the effect of neutral- ising the two ortho-hydroxyls of quercetin by the conversion of one of them into a methoxy-group.As previously discussed (Trans., 1897, 71, 818), the presence of these hydroxyls in quercetin enhances its tinctorial value, although iso-rhamnetin is again an instance that such a property is not essential to the dyes of this group. The shades produced are weaker than those given by quercetin, but considerably stronger than those given by rhamnazin, which is a very feeble dye ; they somewhat closely resemble, especially in the yellow tint with aluminium mordant, that yielded by chrysin and apigenin in a similar way. Unlike these latter, however, it is capable of dyeing with a tin mordant. It is interesting to note that little evidence is now required for a complete study of the shade effect caused by the various hydroxyls existing in quercetin,INDIAN DYE STUFF ASBARG DELPHINIUM ZALIL.273 Readily 80luble Colouring Matter. The alcoholic mother liquors obtained during the isolation and purification of the above colouring matter were evaporated to half their bulk, and the small quantity of impure iso-rhamnetin which separated on long standing removed by filtration, and the filtrate treated with boiling water. On cooling, a yellow, crystalline product was deposited, which was collected and a portion acetylated, in the hope that this would lead to its identification. The colourless needles thus obtained had, however, no definite melting point, indicating a mixture, for they softened below loo", and did not fuse completely until about 160".It has been recently shown in a communication to the Society (Proc., 1898, 56), that various yellow colouring matters in alcoholic solution decompose potassium acetate with the production of insoluble salts, although this behaviour is not common to d l of them. Fractional crystallisation having failed to effect a separation of these colouring matters, recourse was had to this reaction, in the hope that one only would form an insoluble product. A hot concentrated alcoholic solution of the substance was, therefore, boiled with a small quantity of potassium acetate for a few minutes, and the crystals which gradually separated were collected, washed with alcohol, and decom- posed with acid.The regenerated colouring matter was further purified by suspending it in acetic acid and adding sulphuric acid, and the deposited acid compound was then collected, washed with acetic acid, and decomposed in the usual manner. 0.1200 gave 0.2620 CO, and 0.0410 H,O. C,,H,,07 requires C = 59.60 ; H = 3.31 per cent. It formed glistening needles, an alcoholic solution of which gave an orange-red precipitate with lead acetate, and with ferric chloride a dark green coloration, The acetyl compound wag obtained as colour- less needles melting a t 189-191'. C=59*54 ; H=3*79. 0.1182 gave 0,2518 CO, and 0.0425 H,O. As fusion with alkali yielded protocatechuic acid and phloroglucinol, this colouring was evidently quercetin. To obtain, if possible, some clue as to the nature of the substance previously associated with the quercetin and which did not react with potassium acetate, the filtrate from the potaseium quercetin was first saturated with the acetate, any deposited salt thus produced being removed by filtration; the filtrate was then acidified, treated with boiling water, and the crystals which separated collected, washed, and dried, Experiments with this product gave results which indicated that C = 58.09 ; H = 3.99.CI,H,07(C2H,0), requires C = 58.59 ; H = 3.90 per cent. VOL, LXXIII. 'C274 THE COLOURING MATTERS OF ASBARG DELPHINIUM ZALIL. Aluminium. it was not homogeneous, but as the quantity available was very small no further attempts a t purification could be carried out with advan- tage. On analysis, it yielded numbers C = 59.8, H = 3.90, somewhat similar to those required for quercetin, C = 59.60, H = 3.31.Its acetyl derivative, which formed colourless needles, behaved peculiarly when heated, for it fused entirely a t 118-120°, then gradually solidified, and melted again a t 179-180°. A portion which had melted a t the lower temperature and again Holidified, after recrystallisation, behaved in an identical manner. Analysis gave C = 58.54, H = 4.24. Acetyl- quercetin requires C = 58-59, K = 3.90 per cent. On fusion with alkali, it yielded phloroglucinol and protocatechzcic acid. These results indicate a close similarity between this colouring matter and quercetin, and it would appear that i t was merely the latter in an impure condition. Should this be the case, it is difficult to understand that so small an amount of impurity could account for its not giving an insoluble potassium salt when treated with potassium acetate, and also for the melting point of its acetyl derivative.Further study of this substance must be discontinued for the present, owing to the exhaustion of our supply of material. Chromium. Iron. Tin. The &ping P?*operties of Asbarg. Asbarg .... . . .. . ... Quercitron bark I n the previous communication on this subject (Zoc. cit.), it was shown that a close resemblance exists in this respect between asbarg and quercitron bark, although the former yields, with aluminium mordant, a purer or less orange yellow. It is, however, a much weaker dye stuff, having but 35 per cent. the dyeing power of quercitron bark. As the above experiments were carried out with the material in the condition received from the Imperial Institute, further trials were instituted with a sample from which the worthless flowering stalks had been removed. The shades obtained with this product on wool mordanted in the usual manner are given below ; they still indicated a lack of dyeing power as compared with quercitron bark. Golden yellow. Brownish orange. Brown olive. Bright orange. Brown yellow. Deep brown orange. Olive black. I I ” -1 I I I Percentage of Colouring Matter.-An examination of asbarg, minus the flowering stalks, showed that it contained 3.47 per cent. of colour- ing matter (not as glucoside). The estimation was carried out by a method we have lately devised for this purpose, and which we areRESEARCHES ON TERPENES. Ii. ON OXIDATION OF FENCHENE. 275 applying to the estimation of colouring matter in dye stuffs of the quercitron class. The details of the method are not given here, for we hope to make a communication on this subject a t a later date. An examination of the blue flowers of BeZphinium consolida has shown that these contain, curiously enough, some quantity of yellow colouring matter; this, however, is not identical with that present in Delphinium xalil. The study of this product, which is at present delayed owing to lack of raw material, will be shortly continued. CLOTHWORKERS’ RESEARCH LABORATORY, DYEING DEPARTMENT, Y OKKSHIRE COLLEGE.

ISSN:0368-1645    

DOI:10.1039/CT8987300267

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

RESEARCHES ON TERPENES. Ii. ON OXIDATION OF FENCHENE. 275 XXVK-Researches o n the Teypenes. 11. On the Oxida- tion of Fenchene. By JOHN ADDYMAN GARDNER, M.A., and GEORGE BERTRAM COCKBURN, B. A. FENCHENE, which was first prepared and studied by Wallach (Annulen, 1891,263,130), is related to fenchone in the same way that camphene is to camphor. The derivatives of the fenchone and the camphor series show many points of analogy (;bid., 269, p. 326, p. 347; ibid., 2’75, p. 157; ibid., 276, p. 315), and, a t the same time, camphor is evideiitly closely related to turpentine hydrochloride (Armstrong, Trans., 1896, 69, p. 1397). The present investigation was undertaken with the hope of deciding more definitely the relationship between these various compounds, and with this object we propose to study the oxidation of fenchone, feuchene, and turpentine hydrochloride by nitric acid, under conditions somewhat analogous to those employed by Marsh and Gardner in the oxidation of camphene (Trans., 1896, 69, p.74). I n the present paper, we give some account of the oxidation products of feuchene, and of turpentine hydrochloride, from the former of whicli we have obtained, as a main product, camphopyric acid, and from the latter, carnphoric and camphopyric acids. We mainly followed Wallach’s methods in preparing fenchene from fenchone, but our experiences differ in several points from his, and we have introduced several modifications which we think it of interest! to place on record. Fenchone.-This was obtained by rectifying a fraction of fennel oil boiling at 170--230°, distilled for US by Schimmel and Co., and con- taining 40-50 per cent.of fenchone. The fraction boiling a t 190-1 95’ obtained from this sample by careful fractional distillation, and consisting mainly of fenchone, was purified by boiling with strong ‘r 2276 GARDNER AND COCKBURN: RESEARCHES ON THE nitric acid (sp. gr. 1 *42) as Wallach directs. The product thus obtained, after being dried and distilled, was finally purified by freezing. It boiled constantly at 191-192', melted a t So, and had a specific rotatory power [a],= +61' 58' (not in solution). This method of purification, although i t yields very pure fenchone, is somewhat wasteful, as the fenchone is attacked slowly by the strong nitric acid. FenchgZ Alcohol. -Wallach prepared the alcohol by the reduction of a solution of fenchone in absolute alcohol, by means of metallic sodium (Zoc.cit., p.' 143), but although we carefully followed his instructions, we did not get very satisfactory results, the reduction being slower and less complete than his account led us to expect. By using amylic alcohol instead of ethylic alcohol, the reduction took place much more com- pletely, and we were able to operate easily on comparatively large quantities of material. Two hundred grams of fenchone, dissolved in 1200 grams of amylic alcohol, was heated on the water bath, and 120 grams of sodium gradually added, a layer of sand having been placed at the bottom of the flask to prevent the melted sodium cracking the glass. When all the sodium had dissolved, which took from four to five hours, about 14 volumes of water were added, and the supernatant layer of fenchyl and amylic alcohols was separated and distilled until the thermometer rose to 140O; the residue was then distilled in a current of steam, and the oil which came over dried with fused potassium carbonate and fractionated. The fractions boiling a t 197-200", which solidified in the condenser, consisted of pure fenchyl alcohol.It forms hard, white crystals, which melt at 45" and have a specific rotatory power [a],= - 13.37' (in alcohol). FenchyZ Chloride.-Wallach (Zoc. cit., p. 148) prepared this compound by the action of phosphorus pentachloride on fenchyl alcohol in pet,roleum or chloroform solution, removing the chlorides of phos- phorus, &c., by distillation under diminished pressure on the water bath, distilling the residue with steam, and finally fractionating under diminished pressure.We found it advantageous t o remove the oxy- chloride of phosphorus by mixing the product with a large quantity of ice, and allowing it to stand, instead of distilling under diminished pressure, &c. The oil was then separated, dried, and distilled under a pressure of 20 mm., when the bulk of the liquid, consisting of fenchyl chloride, passed over a t 85-90'. Feenc7Lene.-Fenchene was prepared from the chloride by Wallach's method, namely, heating with aniiine. The fenchene obtained, after careful fractionation, distilled mostly at 150-1 52', with very small fractions 152-154", 154-156', 156-15S0, j158-160'.The fraction 150-152', on analysis, gave the following figures, The yield was S3-84 per cent.TERPENES. 11. ON THE OXIDATION OF FENCHENE. 277 0.216 gave 0.6961 CO, and 0.2316 H,O. Found C = 87.89 ; H = 11.9. Fenchene requires C = 88.21 ; H = 11.76 per cent. A fraction boiling at 150-154" had a specific gravity 0.8667 at 1 So, and a specific rotatory power [.ID = - 6.46' (not in solution). P~~enylfee.lzchyZuminA.-After the fenchene obtained by the action of aniline on the chloride had been distilled over with'steam, a thick black oil was left in the distilling flask* ; as it did not solidify, it was submitted to careful fractional distillatioil iinder a pressure of 13 mm., when the following fractions were obtained : below 169O, 169-17 lo, l71-173", 173-175", and 175-180°, of which 171-173" was by far the largest; the latter, on analysis, gave the following numbers.0.2054 gave 0.6305 CO, and 0.1919 H,O. 0.3033 ,, 16.4 C.C. nitrogen a t 20" and 760 mm. N = 6.18. 0.2920 ,, 16.5 C.C. ,, 23" ,, 767 mm. N=6*42. C,,H,7*NH-C,H, requires C = 83-84 ; H = 10.04 ; N = 6.12 per cent. This substance was a pale yellow, very viscous oil, which turned black on exposure to the air, but remained clear and transparent in a vacuous tube, and did not solidify when kept for many months, or by long standing in a freezing mixture of ice and salt. It seems to be isomeric with Wallach's substance. Action of Acetic Chloride on this Oil.-Phenylfenchylamine, when mixed with an excess of acetic chloride, became hot; the mixture was heated for some time a t loo", the excess of acetic chloride distilled off, and the product distilled under a pressure of 24 mm., when most of it passed over at 190-193'.This acetyl derivative, Cl,H17*NPhAc, was a very viscous, yellow oil, which did not solidify on long standing, or in a freezing mixture of ice and salt. It turned dark brown on exposure to air, but remained unchanged in a vacuous tube. It gsre the following numbers on analysis. C = 83-71 ; H = 10.38. 0.1974 gave 0.5777 CO, and 0.1657 H,O. C = 79.80; H = 9.32. C,,H,,NO requires C = 79.7 ; H= 9-22 per cent. Oxidation of Fenchene. When fenchene, in portions of 20 grams, was heated on the water bath in a long-necked flask with 100 C.C. of water, and 100 C.C. of concen- trated nitric acid added, a fairly vigorous reaction took place ; after the action had considerably abated, 200 C.C.more nitric acid was added in quantities of 50 C.C. at a time whenever the action became very slow. After heating for 3 days on the water bath, the fenchene had entirely * In Wallach's experiments, this oil partially solidified, and the solid after repeated cryatallisation from alcohol, melted at 94". Analysis showed this to be fenchylphenylamine (Zoc. cit.) p. 150).278 GARDNER AND COCKBURN: RESEARCHES ON THE disappeared. The contents of the Bask were then thoroughly distilled in a current of steam ; the distillate contained a small quantity of an oil which solidified on standing, but which we have not yet obtained in sufficient quantity for investigation, and there were some acids in solu- tion.The acid liquid was first neutralised with sodium carbonate, then made acid again by oxalic acid and distilled; acetic acid was recognised in the distillate by means of its silver salt. The non-volatile residue left on distilling the oxidation product in steam was evaporated to a small bulk and left for a few days, when the yellow liquid deposited crystals ; after the adhering oil had been re- moved by washing with chloroform, the crystals, which are insoluble in it, were purified by several recrystallisations from strong nitric acid and froin water; they melted at 207", and proved to be cis- caiiiphopyric acid. 0.2102 gave 0.4465 CO, and 0,1460 H,O. The oil remaining after the, crystals of camphopyric acid had been separated was distilled under diminished pressure ; a corisiderable amount of charring took place, and an oil and a solid substance distilled over.The former was an acid, but the latter neutral, so that they mere easily separated by treatment with sodium carbonate solution ; the solid was recrystallised from alcohol, and gave the characteristic crystals of cis-camphopyric anhydride melting a t 178'. C = 57-93 ; H = 7.71. C,H,,O, requires C = 58.06 ; H = '7.59 per cent. On combustion, it gave the following figures. 0.2834 gave 0.5253 CO, and 0.1437 H,O. The yield of camphopyric acid obtained in this way was just over 20 per cent. We have not as yet been able to purify the acid oil which distilled over, but after neutralising with ammonia it gave a white, insoluble lead salt with lead acetate; this lead salt contained 60.7 per cent.of lead. c! = 64.12 ; H = 7.14. C,H,,O, requires C = 64.28 ; H = '7.14 per cent. Oxidation of Turpentine Hydrochlode. Turpentine hydrochloride, when oxidised on the water bath with nitric acid diluted with half its volume of water, was very slowly attacked ; the greater portion slowly dissolved, leaving a heavy oil at the bottom of the flask, which was separated and oxidised by further treatment with nitric acid. On distilling the oxidised liquid in steam, acetic acid passed over, which was recognised by an analysis of its silver salt, whilst in the retort there remained, besides nitric acid, carnphoric, camphopyric or camphoic and other acids.TERPENES. 11. ON THE OXIDATION OF FENCHENE. 2'79 The presence of camphoric acid was to be expected, as Armstrong some years ago obtained small quantities of this acid by oxidising turpentine hydrochloride with dilute nitric acid (1 S80 edition of Miller's Chemistry, also Trans., 1896, 69, p.1398). I n order to isolate the camphoric and camphoic or camphopyric acids, and as far as possible determine the quantities present, the above-mentioned residue was evaporat.ed t o dryness to get rid of nitric acid, taken up with ether, and extracted with sodium carbonate solution. The mixture of sodium salts thus obtained after being reconverted into the acids, was neutralised with caustic soda and precipitated with lead nitrate; the insoluble precipitate was then decomposed by sulphuric acid and extracted with ether. This solution, on long standing, gradually deposited a quantity of solid matter, which was crystallised from strong nitric acid, washcd with chloroform, and again recrystallised from water.It melted at 199-200". It was camphoric acid, and on combustion gave the following numbers. 0.2106 gave 0.4591 CO, and 0.1603 H,O. C = 59.46 ; 11 = 8.46. C,,H,,O, requires C = 60. H = 8 per cent. The acid, on treatment with acetyl chloride, was converted into the anhydride, which, after recrystallisation from alcohol and sublimation, melted a t 217'. On analysis, it gave the following figures. 0.1994 gave 0.4804 CO, and 0.1421 H,O. UloH,,O, requires C = 65.92 ; H = 7.7 per cent. This camphoric anhydride was then reconverted into the acid by dissolving it in caustic soda, and reprecipitating by hydrochloric acid.After crystallisation from water, it melted a t 202-203". C = 65-7 ; H = 7.9. 0.2011 gave 0.4414 CO, and 0.1487 H,O. The yield of camphoric acid from 300 grams of turpentine hydro- chloride was 18 grams. The oily mother liquors remaining after the separation of the ciepo\it of camphoric acid, together with some oily matter obtained from the filtrate and wseh waters froin the lead salts, were submitted to distilla- tion under a pressure of 20 mm. A considerable amount of charring took place, and a liquid passed over between 120" and lS0". The higher fraction of this (180-200") for the most part solidified; this solid was purified by sublimation on an oil bath at 170", and after recrystalli- sation from alcohol melted a t 177-178". These crystals had the characteristic form of cis-csmphopyric anhydride. On combustion, the following results were obtained. C = 59-86 ; H= 8-21. CI0Hl6O4 requires C=60; H=8 per cent.280 RUHEMANN AND BROWNING : FORMATION OF ETHYLIC 0-2140 gave 0.5024 CO, and 0.1418 H,O. C = 64.02 ; H = 7.36. 0.1989 ?, 0.4696 CO, ,, 0.1305 H,O. C = 64.39 ; H= 7-29. C,H,,O, requires C = 64.28 ; H = 7.1 per cent. The yield of camphopyric anhydride was 9 grams from 300 grams of turpentine hydrochloride. It is probable, in the light of the work of Gilles and Renwick on the oxidation of ketopinic acid (PYOC. Chem. Soc., 1897, pp. 64, 182) that this camphopyric acid was produced by the decomposition of camphoic acid, although we were unable to separate any camphoic acid from the crystalline part OF the oxidation product. We have as yet been unsuccessful in our attempts a t purifying the accompanying oils. The oxidation of fenchone under similar conditions takes place very much more slowly, and the products of this oxidation are a t present under investigation. CHEMICAL LABORATORY, ST. GEORGE'S HOSPITAL, S.W.

ISSN:0368-1645    

DOI:10.1039/CT8987300275

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

280 RUHEMANN AND BROWNING : FORMATION OF ETHYLIC XXVIII. -Formation o f Ethyl ic Dihy droxydinico tinate from Et h y lie Cycc nacetnte. By SIEGFRIED RUHEMANN, Ph.D., M.A., and K. C. BROWNING, B.A. THE experiments described in this paper have been undertaken with the view of investigating more closely the action of chloroform on ethylic malonate in presence of sodium ethoxide, as the yield of the sodium derivative of ethylic dicarboxyglutaconate thus obtained was found to be far from satisfactory, and in fact never reached that stated by Conrad and Guthzeit (Annulen, 1884, 222, 249), although this compound had been frequently prepared by one of us and his pupils. The mother liquor of this substance which had accumulated in the course of their researches served as material €or this study.The red-brown alcoholic filtrate from e thylic sodiodicarboxyglutaconate which, after concentration, did not yield any further crop of this substance, even when left for several weeks, deposited an oil when diluted with water; this was separated by the use of ether, and distilled under diminished pressure. Although the examination of this oil is not yet complete, it has been ascertained that it consists to a large extent of ethylic malonate. Excess of hydrochloric acid was added to the aqueous layer left after treatment with ether, and the solution again repeatedly shaken with ether ; on distilling off the ether, a redDIHYDROXYDLNICO'l'INATE FROM ETHYLIC CYANACETATE. 281 oil was left which did not deposit any solid on standing. When heated under diminished pressure, only a small quantity of a colourless liquid distilled at about 150°, but at 160" a reaction took place which resulted in the contents of the flask becoming a semi-solid mass of crystals; these were freed from the mother liquor by washing with dilute alcohol.The dark coloured product thus obtained contained inorganic matter and was only sparingly soluble in water or alcohol, but dissolved in cold concentrated hydrochloric acid. On adding water to this solution, an almost colourless, crystalline product was obtained which readily. dissolved in alcohol, and crystallised from it in transparent needles melting at 202O to a red liquid. The compound, on examination, was found t o contain nitrogen and to be the ethylic salt of dihydroxydinicotinic acid, C H COOC,H,- ('fie COOC,H, HO-C C*OH \/ N 0.2040 gave 0.3855 CO, and 0.1000 H,O. C= 51.53 ; H= 5.44.0.2400 ,, 0.4510 CO, ), 0.1130 H,O. C = 51-25 ; H = 5.23. 0.2285 ,, 0.4318 CO, ?, 0.1045 H,O. C=51.54; H=5.08. 0.3084 ,, 14.8 C.C. nitrogen at 19" and 769 mm. ; N=5*58. C,,H,,NO, requires C = 5 1.76 ; H = 5.10 ; N = 5-49 per cent. The compound is identical with that obtained by Guthzeit (Ber., 1893, 28, 2795) from ethylic ethoxy-a-pyronedicarboxylate and with the substance formed by heating ethylic dicarboxyglutaconate with formamide (Ruhemann and Sedgwick, Ber., 1895,28, 825). Guthzeit gives the melting point of this ethereal salt as 199", whilst we found it to be 202' (as compared with 801' stated before, Zoc. cit.). The formation of this pyridine derivative by the interaction of chloroform and ethylic disodiomalonate could only be explained by assuming that the ethylic malonate contained ethylic cyanacetate.Its presence might easily be accounted for by the circumstance that, in the preparation of ethylic malonate from chloracetic acid and potassium cyanide, and the subsequent action of alcohol and hydrogen chloride, a small quantity of the first formed ethylic cyanacetate remained unaltered. With the view of verifying this explanation, me undertook the study of the action of chloroform on ethylic cyanacetate in the presence of sodium ethoxide. It was to be expected that this reaction would take place according to the equation 2CN* CNa,. COOC,H, + CHCI, = COOC,H,* C(CN) : CH. CNa(CN)* COOC,H, + 3NaC1, and lead t o the formation of the sodium compound of ethylic dicyanoglutaconate ,282 HUHEMANN AND BROWNING: FORMATION OF ETHYLIC which in turn might be transformed into ethylic dihydroxydinicotinate. These reactions actually take place.We had completed this enquiry when we noticed Errera's paper (Gazzetta, 1897, 27, ii, 393), which deals with the formation of ethylic dicyanoglutaconate and some of its derivatives. Ethy Zic Bicyanoglutaconate. The sodium derivative of this ethylic salt is prepared by a method similar to that used for ethylic dicarboxyglutaconate. To a solution of 9.2 grams of sodium in 150 grams of alcohol, 22.6 grams of ethylic cyanacetate and then 12 grams of chloroform are added ; on heating the mixture on the water bath, the white precipitate of ethylic sodio- cyanacetate, first formed, becomes yellow, and after an hour's digestion the liquid is no longer alkaline t o litmus.As the alcoholic solutioIi can only with difficulty be freed by filtraticjn from the gelatinous solid, the alcohol is distilled off, and water added to the residue, when the sodium derivative of ethylic dicyanoglntaconate remains undissolved. It is filtered from the dark coloured mother liquor, washed with water in which, in the cold, it is not very soluble, and twice recrystallised from boiling wntor ; the long, yellow needles thus obtained, which, when dried in the air, contaiii 2H20 (as mas alho found by Errern, Zoc. c i t . ) , have the formula COOC,H,. C(CN):CH* CNa(CN)* COOC,H, + 2H,O. C,,H,,NaN,O, + 2H,O requires H,O = 12.24 per cent.0.2460 (air dried) lost 0.0300 a t 100'. H,O= 12-19. 0,2405 (drieJ a t 100') gave 0.0652 Na,SO,. Na = 8.78. 0.2143 ,, 9 ) ,, 20.8 C.C. nitrogen at 19" and 753 mm. N = 11 -06. CllHllNaN20, requires Na = 8-91 ; N = 10.85 per cent. On adding silver nitrate to the aqueous solution of this substance, ayel- lowish precipitate of the corresponding silver compound, C,,HllAgN,O,, is thrown down, which dissolves in water with great difficulty. 0.2745 (dried at 100') gave 0.0865 Ag. Ag= 31.51. Cl,H,,AgN,O, requires Ag = 31 *48 per cent. Most characteristic is the copper derivative of ethylic dicyanoglut- aconate, which is precipitated in red-brown glistening needles on mixing aqueous solutions of the sodium compound and copper sulphate. It dissolves in boiling water and, on cooling, crystallises unchanged. The air-dried crystals have the composition (CllHl,N,O,),Cu + 4H,o.They slowly lose their water of crystallisation at looo, and turn pale brown.DIHTDROXYDINICOTINATE FROM ETHYLIC CYANACETATE. 283 0,3190 (air-dried) lost 0.0385 a t 100'. H20 = 12.06. (C11HllN20,),Cu,4H,0 requires H20 = 11 *90 per cent. 0.2635 (dried a t 100') gave 22.5 C.C. nitrogen at 10' and 766 mm. N = 10.3. 0.2800 ,, ,, ,, 0.0412 CUO. CU= 11.73. (C11H,1N20,)2Cu requires N = 10.50 ; Cu = 11.82 per cent. In order to isolate ethylic dicyanoglutaconate, the sodium compound is dissolved in warm water, an excess of hydrochloric acid gradually added to the solution, and the yellow ethylic salt which soon separates is collected and washed with water until the washings are no longer acid.The pale yellow filtrate, after some days, loses its colour and acquires a beantiful blue fluorescence, whilst colourless needles are deposited ; the small amount of the latter precluded us from examining it further. Ethylic dicyanoglutaconate dissolves readily in hot acetone, and crystallises from it in yellow, glittering plates which melt and decompose a t 187-188'. It was found to contain a small quantity of another substance which was not removed by repeated recrystallisa- tion. The following analytical results show that the percentage of carbon in different specimens of the product varies from the theoretical number by one to two per cent. 0.2179 gave 0.4320 CO, and 0.1090 H,O. C = 54-07 ; H = 5.55. 0.2012 ,, 0.4015 CO, ,, 0.0975 H,O.C=54*42 ; H-5.38. 0,2073 ,, 0.4198 CO, ,, 0.1020 H20. C=55*22 ; H=5~46. 0.2258 ,, 0.4525 CO, ,, 0.1103 H,O. C = 54.65 ; H = 5.42. 0.1980 ,, 19.5 C.C. nitrogen a t 17' and 765 mm. N=11.50. 0.2050 ,, 20.5 ,, ,, 16" ,, 763 mm. N = 11.70. C,,H,,N,O, requires C = 55.93 ; H = 5.08 ; N = 11% per cent. A similar observation was made by Errers (Zoc. cit.), who was also unable to obtain the ethylic salt in a pure state by crystallising i t from alcohol, although he states that he succeeded in removing the impurity by benzene, and that the compound thus purified melts at 178-179". The question arose whether ethy Iic dicyanoglutaconate under the influence of ammonia undergoes a decomposition analogous to that of ethylic dicarboxyglutac,onate (Ruhemann and Morrell, Trans., 189 1 , 59, 743).Experiment, however, showed t h a t such a reaction does not take place either on leaving the ethereal salt in contact with aqueous ammonia at the ordinary temperature or on heating it with the reagent at 100'. The ethereal salt dissolves, and on concentrating the yellow solution, almost colourless, silky needles of the ammonium compound of ethylic dicyanoglutaconate crystallise ; this dissolves fairly easily in cold water, and with the greatest ease on boiling; it melts and decomposes at 162-1 63'.284 RUHEMANN AND BROWNING: FORMATIOX OF ETHYLIC The following analytical data point t o the view than this substance, when dried at loo', contains +H,O. 0.2397 (dried at 100') gave 0.4455 CO, and 0.1132 H,O. C = 50.71 ; H = 6.17.0-2396 >> ,, 33.5 c.c. nitrogen at 22' and 772 mm. N = 16.08. 0.2267 9 , ,, 31 C.C. ? 7 20' ,, 775 mm. N = 15-95, CllHI,N,O,,NH, c hH,O requires C = 50.38 ; H = 6.10 ; N = 16.03 per cent, On heating the ammonium compound t o 110", it turns yellow, and most probably suffers partial dissociation into ammonia and the ethylic salt. We have, as stated above, undertaken the preparation of ethylic dicyanoglutaconate with a view of ascertaining whether it can readily be transformed into ethylic dihydroxydinicotinate. This change mag indeed be effected, and this fact affords an explanation of the forma- tion of the pyridine derivative from the product contained in the mother liquor of ethylic sodiodicarboxyglutaconate. Ethylic Dihyd?.ox?/dinicotinate. If ethylic dicyanoglutaconate, or the sodium compound of the ethereal salt, is boiled with dilute hydrochloric acid, the substance after a short time enters into solution, and in the course of a few minutes a solid separates, which rapidly increases in quantity.When cold, this is collected, washed, and dried, and then treated with cold concentrated hydrochloric acid, which dissolves most of i t ; on adding water, a precipitate is thrown down which readily dissolves in chloroform, and crystallises from alcohol in colourless needles. This compound is characterised as ethylic dihydroxydinicotinate by the melting point (202'), the reddish-violet coloration produced on adding ferric chloride to its alcoholic solution, and by a nitrogen determination. 0.2025 gave 10 c c. nitrogen a t 20' and 772 mm.N = 5-73. CllHl,NO, requires N = 5.49 per cent. The formation of ethylic dihydroxydinicotinate is undoubtedly preceded by that of the diamide of ethylic dicarboxyglutaconate, which then loses ammonia and condenses to the pyridine derivative, as illustrated by the following symbols. CH CH /\ /\. COOC,H,*y f]H*COOC,H, 3 COOC,H,*y FH* COOC,H,-+ CN CN NH,*CO CO-NH, Ethylic dicyanglutaconate.DIHYDROXYDINICOTINATE FROM ETHYLIC CYANACETATE. 285 CH COOC,H,*Y g*COOC,H, OH*C C-OH N /\ \/ Ethylic dihydroxydinicotinate. On prolonged boiling with hydrochloric acid, ethylic dihydroxydini- cotinate decomposes, losing carbon dioxide, and on evaporating the solution on the water bath, the residue gives a reddish-violet coloration with ferric chloride, and strongly reduces silver nitrate and potassium permanganate.The investigation of this compound is not yet complete; there can, however, be no doubt that it is aa'-dihy- droxypyridine, which could not be isolated from the product of the interaction of ammonia and ethylic glutaconate, on account of the rapidity with which the dihydroxypyridine is oxidised by the oxygen of the air in ammoniacal solution (see Ruhemann and Morrell, Trans., 1891, 50, 745). Ethylic dihydroxydinicotinate not only forms a sodium and an ammonium derivative, but with phenylhydrazine also yields a com pound which is thrown down on adding the hydrazine to a warm alcoholic solution of the ethylic salt. It dissolves but sparingly in boiling alcohol, and, on cooling, crystallises in groups of slender, colourless needles which decompose at 198'.0.2130 gave 0.4365 CO, and 0.1112 H,O. 0.2278 ,, 22.5 C.C. nitrogen at 19' and 769 mm. N = 11.49 Cl,Hl,N06,NH,*NH*c6H, requires C = 56.19; H = 5.78 ; N = 11-57 p. c. The hydroxyl groups in ethylic dihydroxydinicotinate have the same relative position in the pyridine ring as in citrazinamide ; the latter, as shown some time ago (Ruhemann, Be?*., 1887, 20, 3369), may be transformed into trichlorocitrazinamide and the corresponding bromo- derivative, by the action of chlorine or bromine on the solution of the amide in hydrochloric acid. It seemed to be of some interest to subject ethylic dihydroxydinicotinate to a similar treatment with the view of arriving at analogous halogen derivatives. The ethylic salt, as has been stated before (Ruhemann and Sedgwick, Ber., 1895, 28, 825), and as mentioned in this paper, has feeble acid properties; it dissolves in concentrated hydrochloric acid, but the hydrochloride thus formed dissociates on adding water.This behaviour of ethylic dihydroxydi- nicotinate has been made use of for its purification. On adding bromine to the solution of the ethylic salt in hydrochloric acid, an unstable additive product is formed (Ruhemann and Sedgwick, Zoc. cit.) ; chlorine however, we find, behaves differently.? C =55*89 ; H =5-80. VOL. LXXIII. U286 FORMATION OF DIHYDROXYDINICOTINATE. EtiLylic Di~ydroxydic~lororzicotinccte. On saturating the:solution of ethylic dihydroxydinicotinate in concen- trated hydrochloric acid with chlorine, a viscid mass separates and ad- heres to the sides of the vessel ; this is washed, dissolved in dilute potash, and the solution acidified with hydrochloric acid, when a white precipi- tate is obtained which is sparingly soluble in boiling alcohol, and on cooling crystallises in colourless, shiny plates.These begin to darken a t about 238O, and are completely decomposed a t 248O ; ferric chloride gives a reddish-violet coloration with their alcoholic solution. On analysis, numbers were obtained corresponding with the formula 0.2090 gave 0.2930 CO, and 0.0528 H,O. C = 38-23 ; H = 2.80. 0,2318 ,, 0.2636 AgCI. C1= 28.14. C8H,N0,CI, requires C = 38.10 ; H = 2.78 ; C1= 28.17 per cent. The mode of formation of the compound and the analytical results indicate that it is ethylic dichlorodihydroxynicotinate j its constitution may, clpiori, be expressed by one of the formula. C8H$!?O*Cl,. C(COOEt)*C(OH) C(COOEt)*C(OH) I. CH*CCI,-- I I 'I* 8CI*CCl=C(OH)~N The first formula, I, would correspond with the structure of tri- chlorocitrazinamide and to that of the dichloro-derivative formed from methyl dihydroxypyridine (Ruhemann, Ber., 1894,!2'7, 1271), but the stability of ethylic dichlorodihydroxynicotinate compared with that of the others would indicate that the formula I1 has to be assigned to it. This formula is, moreover, supported by the behaviour of the substance towards phenylhydrazine. Whilst the former chloro- derivatives, on treatment with the base, are transformed into phenyl- hydrazones, with removal of halogen, the latter yields merely an additive product. This readily dissolves in alcohol, and on treatment with dilute hydrochloric acid the chloro-derivative is precipitated unchanged. In conclusion, we express our best thanks to Mr. 0. Reinherz, of Trinity College, for help afforded US in the course of this work, GONVILLE AND CAIUS COLLEGE, CAMBRIDGE.

ISSN:0368-1645    

DOI:10.1039/CT8987300280

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

ACTION OF ALKYL IODIDES ON SILVER MALATE, ETC. 287 XX1X.-The Action of AZkyl Iodides on Silver MaZate and on Silver Lactate. By THOMAS PURDIE, F.R,S., and G. DRUCE LANDER, B.Sc. THE silver salts of organic acids react, as a rule, very readily and com- pletely with the lower alkyl iodichs, and as the chemical chwge usually follows a perfectly normal course, it affords, as is well known, one of the most trustworthy processes for the preparation of ethereal salts, furnishing a pure product and a satisfactory yield in many cases where the commoner processes of etherificationcannot be employed. The method obviously recommends itself for the preparation of the ethereal salts of optically active acids which are liable to have their activity im- paired by the action of mineral acids ; for this reason, it was employed in the preparation of the active ethereal alkyloxy-succinates (Trans., 1895, 67, 970), and it gave satisfactory results. I n applying the method, however, to hydroxy-acids, we find that the reaction does not proceed entirely in the normal direction, and that it cannot be relied ou to produce the ethereal salts of such acids in a state of purity. J.Wallace Walker (Trans., 1895,67,916) prepared ethereal lactates from the silver salt and found them to be more active than specimens prepared by other investigators by the commoner methods. This result suggested a research on the optical activity of ethereal malates and lactates prepared by different methods (Purdie and Williamson, Trans., 1896, 69,818), which showed that in general the ethereal salts obtained by the action of primary iodides on the silver salts did, in fact, exhibit a considerably higher activity than those prepared with the aid of hydrochloric or sulphuric acid, the difference amounting in some cases t o about 20 per cent.of the total activity. The reaction in these cases appeared, nevertheless, to run its normal course ; the results of analysis, the nearly constant boiling points of the products, the yield obtained, and the comparative constancy of the activity of different preparations seemed to preclude the idea that the ethereal salts could be contaminated with substances of higher activity. The obvious alternative explanation, that the lower activity of the ethereal salts made by the commoner methods might be due t o racemisation, was not supported by experiment, as racemic compounds could not be detected in the products of hydrolysis in such quantity as would account for the defect in activity.No explanation of the apparent anomaly could be given. In attempting to prepare isopropylic malate (Zoc. cit.), it was found that the action of isopropylic iodide on silver malate was quite abnormal. The yield of ethereal salt obtained was very small, the activity of different preparations varied considerably, and was nearly two and a half times as great as that of normal propylic malate, an n 2288 PURDIE AND LANDER: THE ACTION OF ALKYL observation which is quite at variance with what is known of the relative influence of the two propyl groups on optical activity. The results of analysis and experiments on the hydrolysis of the ethereal salt led the authors to conclude that the product WAS a mixture of iso- propylic malate with some more carbonaceous and much more optically active substance.The object of the present investigation was to ascertain the cause of the anomalous results referred to. We find that the high activity of the product of the action of isopropylic iodide on silver malate is due to the presence in it of a considerable amount, more than 20 per cent., of isopropylic isopropoxy- or propoxy-succinate, the propyl group of the iodide having replaced, not only the silver of the salt, but also to some extent the hydrogen of the alcoholic hydroxyl, and that an analogous reaction occurs between isopropylic iodide and silver lactate.We have succeeded also in establishing the fact that ethylic malate and ethylic lactate, made by the method referred to, notwithstanding the evidence for their purity, mentioned above, are in reality contaminated, although to a much smaller extent, with more highly carbonaceous and more active substances, which, although we could not isolate them in a state of purity, are without doubt the corresponding ethoxy-compounds. Action of Isopropylic Iodide on Silver Malate. In carrying out the reaction, the method pursued was as follows. The finely powdered dry silver malate was added, a small portion at a time, t o excess of isopropylic iodide, the temperature being allowed to rise. The mixture having been heated for an hour on a water bath, was diluted with dry ether and filtered ; the filtrate, being extremely acid, was allowed to stand over potassium carbonate, again filtered, and after removal of the ether distilled under reduced pressure. After the iodide had distilled off, the liquid boiled at a nearly constant tem- perature.We employed 360 grams of silver malate and 600 grams of iodide, and obtained after two distillations 68 grams of a product having the boiling point 163-165" under a pressure of 60 mm. A small quantity of the substance was accidentally lost during distilla- tion, but making allowance for this, the yield amounted to only about a half of that obtained (Zoc. cit.) with primary propylic iodide. The observed rotation of the liquid was a= - 65-92' in a 200 mm. tube at 20°, the rotations found in the previous preparations on a smaller scale being - 69-80' and - 61.95'; that of n-propylic malate made by the same method is - 29.6".Normal propylic and isopropylic malate, prepared by Walden by Anschutz's method (Zeit. physikal. Chem., 1895, 17, 248) showed under similar conditions the rotations - 24.96" and - 23-06' respectively.IODIDES ON SILVER MALATE AND ON SILVER LACTATE. 289 In order to isolate the more active constituent, 66 grams of the mix- ture were shaken in the cold with a 10 per cent. aqueous solution of potassium hydroxide and left in contact with the alkaline solution for 24 hours, previous experiments having shown that by this treatment the less active substance was first hydrolysed. The residual oil, after being washed with water and separated from the latter by the addition of some ether, amounted to 15 grams.It was found that the removal of the less active constituent had raised the rotation from a = - 32.9' t o - 54.95' in a 100 mm. tube. On being distilled under a pressure of about 25 mm., nearly all the liquid passed over a t 148'. The analysis of the liquid gave results agreeing with the formula of propylic propoxysuccinate. 0.0980 gave 0,2155 CO, and 0,0825 H,O. C = 59.97 ; H = 9.35. 0,1670 ,, 0.3660 CO, ,, 0,1385 H,O. cl=59*77; H=9*21. C,,H,,O, requires C = 60.00 ; H = 9-23 per cent. A determination of the specific rotation of the substance gave the following result : t = 18', I = 1, u = - 57.08, d 17*5/4' = 0.9762, [aID = -58.47". Its activity far exceeds that of normal and isopropylic malates, the specific rotations of which are respectively only - 11.6' and - 10.41' (Walden, loc.cit. ; Anschutz, Zeit. pliysiknl. Chem., 1895, 16, 495). The ethereal salt, although not attacked by 10 per cent. aqueous potassium hydroxide in the cold, was readily hydrolysed when heated on the water bath with aqueous alkali to which enough alcohol had been added to bring the oil into solution ; a small portion of it was there- fore hydrolysed in this manner, On adding silver nitrate to the neutralised solution, a silver salt was obtained as a gelatinous, white precipitate, differing essentially in appearance from silver malate, which is a crystalline powder; it was much more diflicult to wash than the latter, and apparently more soluble in water. Unlike silver malate, it turned rapidly brown when dried a t loo', and even at the ordinary tem- perature in a vacuum became dark coloured.Estimations of silver, therefore, did not give satisfactory results ; the numbers obtained with the salt dried in a vacuum and at 100' were 59.30 and 60.81 per cent. respectively, the calculated percentages for silver propoxysuccinate and malate being 55.38 and 62 07. As it seemed possible that potassium hydroxide might have decom- posed the acid, the rest of the ethereal salt was hydrolysed by heating with an aqueous alcoholic solution of barium hydroxide. After re- moving the excess of the latter with carbonic anhydride and evaporat- ing the solution to a small bulk, the barium salt separated as a white, indistinctly crystalline powder of pearly lustre ; this was readily soluble in water, leaving, however, a small residue of a highly in-290 PURDIE AND LANDER: THE ACTION OF ALKYL soluble salt, probably barium malate, which was removed by filtration. AS trustworthy determinations of water of crystallisation in the barium alkyloxysuccinates are dXicult to obtain owing to their hygroscopic character, some of the salt was dried for analysis at 130-140'.The results found on combustion were as follows. I. C = 26-19 ; H = 3-22 ; Ba = 44-19, 11. C: = 26.29, H = 3.26, Ba = 44-41 ; the calculated per- centages for barium propoxysuccinate are C = 27.01 ; H = 3*22 ; Ba=44*05. Estimations of barium gave 43.70 and 43-96 per cent. The low results for carbon shown in the combustions may possibly be due to the difficulty of drying the salt completely without its under- going slight decomposition.The specific rotation of the salt was observed in solutions of different strengths, the concentration of the initial solutions, from which others were made by dilution, being found by evaporating a measured volume and weighing the residue dried at 130-140°. (1) t=lgO, 1=4, ~ = 1 2 * 9 5 , a= -4*11', [ a ] D = - 7.93'. (2) t = 19', I = 4, c = 6.475, U = - 2.72', [a], = - 1050°. (3) t=20', Z=4, ~=6*170, a= -2*72', [a],,= - llg02'. (4) t=20', f?=4, c=3*085, a=-1*50', [ a ] D = -12'16'. The activity of this salt differs widely from that of barium malate, which, according to Schneider (Annalen, 1881, 207, 277), has the specific rotation + 4.69' when c = 5.19. The acid was obtained from the solution of the barium salt by adding rather less than the calculated quantity of sulphuric acid.The solution having been filtered and the filtrate evaporated to dryness, the acid was extracted from the residue with ether, and on evaporat- ing the ether it was left as an oil which quickly solidified to a crys- talline mass. It was, however, very deliquescent, and did not show a sharp melting point. Observations of the activity of its aqueous solu- tion are given below; the concentrations were found in the case of the first three determinations by titrating the initial solution with standard alkali and diluting this to known volumes, and in the case of the fourth by weighing the acid directly. (1) t=lSO, Z=4, c=10*0036, a= -14*51', [aID= -36.26'. (2) t=16', l = 4 , ~=5*0018, U = -7.21', [ U ] D = -36.04'- (3) t=16', 1?=4, ~=2*5009, U = -3.66'9 [a],= -36.59'.(4) t=18', t=4, ~ ~ 3 . 3 5 8 3 , a= -4.74'9 [a],= -35'29'. The calcium salt, prepared by neutralising the acid with calcium carbonate, was only sparingly soluble in cold water, and still less so in hot water ; on slightly warming a cold saturated solution, it became rapidly turbid from the separation of the salt in the form of fine scales, which, on cooling, went again into solution. The salt, dried at llO*,IODIDES ON SILVER MALATE AND ON SILVER LACTATE. 291 was found to contain 18.71 per cent. of calcium, and combustion gave the following results. Found I. C = 38-60 ; H = 4.48 ; Ca = 18.30. ,, 11. C = 38.82 ; H= 4.65 ; Ca= 18.74. Calculated for calcium propoxysuccinate, C7HLo05Ca, C = 39.25 ; H = 4.67 ; Ca = 18.69 per cent.To determine the specific rotation, a solution was made by shaking the salt with cold water, and its concentration was found by evaporat- ing an aliquot portion to dryness and weighing the dried residue; the following result was obtained. t=18", 1=4, ~=1*635, (L= -1*28O, [ a ] D = -19.5'7'. A solution of the potassium hydrogen salt, prepared by exactly neutralising a measured volume of the solution of the acid with standard potassium hydroxide, adding an equal quantity of acid, and making the liquid up to a definite volume, gave the followiag specific rotation. t=18', l = 4 , c=1*6992, a= -2.16'9 [ a ] ~ = -31.78'. A solution of the normal potassium salt was also prepared by exactly neutralising a solution of the acid with standard potassium hydroxide and making up the liquid to a known volume.The specific rotation was as follows. t = 16', I = 4, ~=3*9820, U = - 3.03'7 [a],= - 19.02'. A portion of the solution was evaporated and an estimation of potas- sium made in the salt dried at looo. Found, K = 30.85 ; calculated for C7HI0O5K2, 31.01 per cent. The optical observations and the analyses which have been quoted prove that the abnormally high activity of the product of the action of isopropylic iodide on silver malate is due to the unexpected formation of a considerable amount of isopropylic propoxy- or isopropoxy-succinate. This conclusion is borne out by a comparison of the activity of the sub- stances described with that of the corresponding derivatives of active normal propoxysuccinic acid obtained by the resolution of the inactive acid which was prepared by the addition of normal propylic alcohol to maleic anhydride or propylic fumarate (Trans., 1895, 67, 949).The ethereal salts of this acid have not been prepared, but judging from the active methoxy- and ethoxy-succinates, the observed activity of the compound we have isolated is such as a propylic propoxysuccinate would be expected to exhibit. The specific rotation of propylic methoxysuccinate is 45.2 lo, that of the corresponding ethoxysuccinate 51.25' (Trans., 1895, 67, 979) ; assuming that the rise of activity is nearly constant in ascending the series, the specific rotation of propylic292 PURDIE AND LANDER: THE ACTION OF A L m L n-propoxysuccinate should be about 57', and that of the isopropyl com- pound should not differ much from this value. The specific rotation of our compound is 58-47'.The two acids show nearly the same specific rotation in aqueous solution, and the activity in both cases alters but slightly on dilution. Taking the mean of the observations at different concentrations, the specific rotation of n-propoxysuccinic acid from fumaric acid is 36-23', that of the compound formed from malic acid, as described above, 36.04O. The rotations of the potassium hydrogen salts in dilute solution are also similar, namely, 32.30' and 31-78" respectively. In the case, however, of the normal potassium salt, and of the calcium and barium salts whose specific rotation alters largely with dilution, the salts from malic acid show a somewhat higher activity than the salts from the other source.Thus, the normal potas- sium salts have, in 3 to 4 per cent. solution, the specific rotations 19*02O and 17.26' respectively ; the calcium salts, resembling each other in their sparing solubility in water, and in their greater solubility in cold than in hot water, show in dilute solution the specific rotations 19-57" and 14-18' respectively, The barium salt of our acid shows the same rapid rise of activity with dilution which is exhibited by the barium alkyloxysuccinates in general (Trans,, 1893, 63, 239). The specific rotations of the barium propoxysuccinates from the two sources for about equal concentrations are 12-16' and 10" respectively. Our experiments afford no conclusive evidence as to whether the acid under investigation is a normal propoxy- or an isopropoxy-deriva- tive of succinic acid.The similarity in activity of the acids and of the acid potassium salts favours the former view; on the other hand, the difference exhibited by the other three salts mentioned are such as might be expected if the acids were isomeric. T t may be mentioned that, in the case of the sparingly soluble calcium salts, where the difference is greatest, the solutions examined were so dilute that the observations of activity may be affected with a considerable error. With regard to the composition o€ the mixture of ethereal salts resulting from the interaction of isopropylic iodide and silver malate, the chief constituent is, no doubt, the normal product, namely, isopro- pylic malate.This was confirmed by neutralising the alkaline solu- tion used in effecting the partial hydrolysis, and adding silver nitrate, when silver malate was precipitated; the salt, dried at looo, was found t o contain 62-14 per cent, of silver, the calculated percentage being 62-07, The quantity of malic acid in this solution was esti- mated approximately by weighing the silver malate precipitated from an aliquot part of it. The quantity of isopropylic malate, corre- sponding t o the malic acid found, would amount to about 80 per cent. of the weight of the mixture of ethereal salts, from which it follows that the propoxysuccinate constituted about 20 per cent. of theIODIDES ON SILVER MALATE AND ON SILVER LACTATE.293 mixture ; the quantity actually isolated by the partial hydrolysis was 23 per cent. It should be mentioned, however, that from the optical observations it would appear that either the quantity of this substance separated was much less than that actually formed, a considerable part of it having undergone hydrolysis even in the cold potassium hy- droxide solution, or else there was some other substance more active than isopropylic malate in the mixture. If the product consisted only of isopropylic malate and isopropylic propoxysuccinate, the quantity of the latter, calculated from the rotations of the separate constituents* and of the mixture, should have amounted to more than 40 per cent. of the whole. Action of Ethylic Iodide on Silvev XaZate. As already stated, the ethereal malates from the action of primary iodides on silver malate were found t o be more active than the corre- sponding substances prepared from malic acid and alcohol with the aid of mineral acids.The differences of activity, however, in these cases are small as compared with that observed with iropropylic malate ; the specific rotations found for ethylic malate, for example, from the two different sources were - 12.4" and - 10.3O. Besides this, in the case of the malates of primary alcoholic raclicles, neither boiling point nor analysis indicated the presence of foreign Substances, so that it was only after ascertaining the cause of the abnormal activity of the product obtained from isopropylic iodide that it occurred to us that the malates made from primary iodides probably also owed their higher activity to the formation of alkyloxysuccinic acids.Five per cent. of ethylic ethoxysuccinate would suffice to account for the difference in activity observed in the case of ethylic malate. The substances differ so little in percentage composition that analysis would not betray the contamination, and the boiling points also are so much alike that fractional distillation of small quantities of material might readily fail to detect it. We accordingly prepared ethylic malate in the manner previously described (Zoc. cit.), but on a larger scale. We employed 454 grams of silver malate and 868 grams of ethylic iodide, and obtained 200 grams of ethereal salt, boiling at 130-1 35" under a pressure of 15 mm. and the rotat,ion a = - 13-93' in a 100 mm.tube at 12O ; the rotations observed in previous preparations were - 14.06" and - 14*30°. On being redis- tilled under reduced pressure, the whole liquid boiled within a range of 4', but the activity of the first fraction collected, 86 grams, was found to have risen to - 14.8", and that of the last to have decreased t o - 12.6'. On further repeated distillation, the activity of the fractions * The activity of isopropylic malate ha3 been determined by Walden (Zoc. cit.).294 PURDIE AND LANDER: THE ACTION OF ALKYL of lower boiling point increased continuously, whilst that of the higher boiling fractions decreased, but to a smaller extent. Owing to slight variations of pressure, it was not possible to observe the exact boiling points of the various fractions under identical conditions, but those of the extreme fractions did not differ by more than 3'or 4'.Finally, after prolonged fractionation, 17 grams were collected, having the rotation a = - 17.05, whilst much the larger part of the liquid had ac- cumulated in the fractions of higher boiling point, the rotation of which had decreased to - 12.2'. The maximum rotation had eyidently not been nearly reached, but the quantityof material did not admit of the fractionation being carried further ; the very slow decrease of activity in the case of the fractions of higher boiling point, on the other hand, in- dicated that the minimum rotation had been much more nearly attained, the minimum being evidently that of ethylic malate as prepared by the usual methods, namely, a = - 1107~.That the substances of higher and lower activity were not produced by any chemical change caused by the repeated distillation was proved by an analysis of the latter, which showed that it still retained the composition of ethylic malate. We found on combustion C = 5045 ; H = 7-41, the calculated numbers being 50.53 and 7.37 per cent. The results of the fractional distillation show that the product of the action of ethylic iodide on silver malate is mainly ethylic malate, but that the ethereal salt is contaminated with a small quantity of a much more active substance, probably ethylic ethoxysuccinate. It seemed that it would be impossible to isolate the ethoxysuccinic acid without repeating the experiment on a still larger scale; with the view, however, of finding corroborative evidence of the presence of this acid, we examined the products of hydrolysis of the more active fractions of the ethereal salt. A brief account of the results of this examination may be given.One of the more active fractions was partially hydrolysed by treat- ment with aqueous caustic alkali in the cold, which raised the rotation of the unhydrolysed part considerably. A sparingly soluble calcium salt, prepared from this more active portion by hydrolysing it with calcium hydroxide, was found, on combustion, to have a percentage composition intermediate between that of calcium malate and calcium ethoxysuccinate. The fraction of ethereal salt having the rotation u = - 17-05' was similarly subjected to partial hydrolysis, which raised the rotation of the residual oil only to - 20' ; this method of separating the constituents of the mixture was not so effective as it had proved in the case of isopropylic malate.The hydrolysed part contained chiefly malic acid ; the calcium salt prepared from it, after drying a t 140°, contained 21.30 per cent. of calcium, the percentage for monhydrated calcium malate being 21.05. The residual un-IODIDES ON SILVER MALATE AND ON SILVER LACTATE. 295 hydrolysed part gave, on hydrolysis with barium hydroxide, two barium salts, one sparingly soluble, the other readily soluble in water. The insoluble salt, which was precipitated as a granular powder on boiling the solution, was approximately pure barium malate. It contained, when dried at 120-130°, 50.51 per cent.of barium, the calculated percentage being 50.9 3. The soluble salt, consisting of crystalline scales, when dried under similar conditions, contained 47.88 per cent. of barium, a result intermediate between the calcu- lated percentages of malate and ethoxysuccinate, but nearer that of the latter, which is 46.13. Polarimetric observations were also in agreement with the supposition that the soluble salt was a mixture of the salts of ordinary malic acid and I-ethoxysuccinic acid. The activity of both barium salts is known to vary much with concentration, the change with dilution being in both cases in the lsvodirection, but for similar concentration the ethoxysuccinate is much less dextrorotatory or much more laevorotatory than the malate.The specific rotations found were + 6*85O for c = 15.92 and + 3.02O for c = 7.96, results intermediate between those which the pure salts would have given under similar conditions. (Annnoclm, 1881, 207, 277 ; Trans., 1893, 83, 235). We prepared a larger quantity of the mixed barium salts from some of the more active fractions of the ethereal salt, and tried to remove the malate by repeated boiling and evaporation, by which treatment it sepamtes in the anhydrous insoluble form, but the sepa- ration was only partial ; analysis and polarimetric examination of the soluble salt showed that the malate had been further eliminated, but that much was still present. Attempts to purify the ethoxysuccinate by conversion into the lead and the calcium salts also failed.Action of Nor& Butylic Iodide on SiZue~ Malate. Anschiitz and Reitter have shown that normal butylic malate, pre- pared by the hydrochloric acid method, has the specific rotation - 10*722O (Zed. physikd, Chrn., 1895, 16, 495) ; the product of the action of normal butylic iodide on silver malate has the specific rotation - 12*20°. Having a small quantity, 12 grams, of the latter substance, we examined the product of its hydrolysis to ascertain whether it con- tained, besides malic acid, the more laevorotatory butoxysuccihic acid. On boiling the solution of the barium salt, obtained by hydrolysing the ethereal salt with barium hydroxide, a large quantity of barium malate was precipitated, but there remained some barium salt readily soluble in water which was not rendered insoluble on prolonged boiling of the solution.An estimation of barium in this gave a result mid- way between that of malate and butoxysuccinate, and a 6 per cent, solution showed only a slight dextrorotation. At this concentration,296 PURDIE AND LANDER: THE ACTION OF ALKYL barium malate is more dextrorotatory, and barium butoxysuccinate is laevorotatory. The quantity of material was too small to attempt a separation. Action of AlkgZ Iodides on Silver Lactate. It has been shown (Zoc. cit.) that ethylic lactate prepared from the silver salt exhibits a higher activity than when prepared by other methods. Klimenko and J. Wallace Walker, using the former method, found the specific rotation to be 14.1 9 O and 14.52' respectively ; Purdie and Williamson, by the same method, obtained a somewhat lower result, 13-46', but still notably higher than the activity found in the case of the ethereal salt prepared with the aid of sulphuric acid, the specific rotation of which was only 10.33'.Examination of the active zinc lactate obtained from the latter ethereal salt proved that its lower activity was not due to racemisation. In view of what has been stated above with regard to the action of isopropylic iodide on silver malate, it seemed probable that the excess of activity exhibited by ethylic lactate prepared from silver lactate was t o be attributed to the pro- duction of a small quantity of ethylic ethoxypropionate. We have been engaged recently in resolving inactive alkyloxypropionic acids into their active components, and find that these compounds possess a very high degree of activity as compared with lactic acid, so that the presence of a small proportion of ethylic ethoxypropionate would suffice t o account for the excess of activity referred to.As the two ethereal salts have practically the same boiling point, the ethoxypropionate would not be removed by fractional distillation, and the presence of such a small quantity of the substance as would suffice to raise the rotation to the degree mentioned would not be readily detected by analysis. Our immediate object in the following experiments being only to ascertain if alkyloxypropionates are produced simultaneously with lactates when alkyl iodides act on silver lactate, inactive lactic acid was employed. Ethylic lactate was prepared from the silver salt in the manner pre- viously described.From 173 grams of lactate and 290 grams of ethylic iodide, we obtained 57 grams of ethereal salt having the boiling point 151-154' under atmospheric pressure ; tho small yield, amounting to only 55 per cent. of the calculated quantity, indicated that the reaction had not proceeded entirely in the normal manner. The distinctive properties of the zinc salts of ethoxypropionic acid and lactic acid presented a method of detecting the former in the presence of a large quantity of the latter, or, possibly, even of separating it in the pure state, zinc ethoxypropionate being a gum soluble in alcohol, zinc lactate, as is well known, a crystalline salt nearly insoluble in that liquid. The ethereal salt, accordingly, was converted into barium salt by boilingIODIDES ON SILVER MALATE AND ON SILVER LACTATE.297 with solution of barium hydroxide and removal of the excess of the latter with carbonic anhydride. The acid was obtained from the barium salt by adding to its solution rather less than the calculated quantity of sulphuric acid, evaporating the filtered liquid, and extracting the residue with ether. Finally, the acid was converted into zinc salt by means of zinc carbonate, and the solution having been evaporated, the dried residue was treated with absolute alcohol. The great bulk of the salt was left undissolved, and on evaporating the filtered liquid a zinc salt was left in the form of a gum. The separation, however, was evidently not complete, as some zinc lactate crystallised from the gum on standing.Zinc lactate, in fact, dissolved in the alcoholic solution of the uncrystallisable zinc salt to a much greater extent than in pure alcohol, and although a considerable amount of the lactate was removed by repeated evaporation and treatment with alcohol, and also by add- ing alcohol to the aqueous solution, the separation was not complete. The salt, dried at 150°, at which temperature it showed no signs of de- composition, was found to contain 23.38 per cent. of zinc, the calculated percentages for zinc lactate and ethoxypropionate being 26.87 and 21.84 respectively. A combustion of a crystalline calcium salt which was made from the uncrystallisable zinc salt gave results for carbon, hydrogen, and calcium intermediate between the calculated numbers for lactate and ethoxypropionate.From 57 grams of ethereal salt, me obtained about 2 grams of the uncrystallisable zinc salt. To ascertain whether ethylic lactate made by other methods gave similar results, we prepared the compound from inactive zinc lactate, alcohol, and sulphuric acid. The same weight of this ethereal salt as had been used in the experiments just described was converted into zinc salt under exactly the same conditions; and as much as possible was separated by crystallisation from the aqueous solution in order that the uncrystallisable salt, if there were any present, might be more certain of detection. The crystaliised salt was pure zinc lactate, being found on analysis to contain 18.06 per cent. of water of crystallisation and 26.89 per cent.of zinc calculated on the anhydrous salt, the cal- culated numbers being 18.16 and 26.87. The residue left on evapor- ating the solution to dryness showed no sign of gum, and the alcoholic extract of it, when evaporated to dryness, left a small quantity of zinc lactate with only a trace of viscid matter. In the preparation of ethylic lactate from zinc lactate, just mentioned, a small yield of ethereal salt was obtained. I n order to satisfy ourselves that the un- etherified acid contained no ethoxypropionic acid, the organic potassium salt, which had been formed on the addition of potamium carbonate to the crude product, was separated from the inorganic salts by alcohol, converted into acid, and then into zinc salt. This salt was crystalline, and proved on analysis to be pure zinc lactate.We could not detect298 PURDIE ABD LANDER: THE ACTION OF ALKYL any ethoxypropionic acid either in the et herified or unetherified acid. We conclude that ethylic lactate made from silver lactate is con- taminated with a small proportion of an ethereal salt of an acid containing more carbon than lactic acid, which, judging from what follows below, is doubtless ethoxypropionic acid, and that it is the presence of this substance which raises the activity of active ethylic lactate made by the method in question. On the other hand, ethylic lactate made by the sulphuric acid method does not contain this im- purity, or, if a t all, only in very minute quantity. Our experiments with silver malate having shown that isopropylic iodide gives a much larger proportion of the alkyloxy-compound than primary iodides, i t seemed likely that the same would hold good for silver lactate.We found that this was the case, and that the salts of isopropoxypropionic acid could consequently be prepared from the product of the reaction without much difficulty. The product was strongly acid; having been diluted with ether, it was accordingly neutralised with dry potassium carbonate, which caused the separation of a yellowish, gummy, potassium salt. The filtered liquid was distilled fractionally under atmospheric pressure, and after several distillations much the greater part boiled a t 165-161O. The yield of ethereal salt was very small, as in the case of the malate; from 235 grams of silver lactate and 358 grams of isopropylic iodide we obtained 40 grams boiling a t the temperature mentioned, that is to say, only about 25 per cent.of the calculated yield. Isopropylic lactate, pre- pared by heating the acid and alcohol, boils, according to Silva (BUZZ. Xoc. Chim., [iii], 17, 97) a t 166-168', and isopropylic isoprop oxypropionate, which mas prepared by Silva, by acting on the lactate with sodium and isopropylic iodide, but was not obtained pure, is stated to have a boiling point a little above that of isopropylic lactate. As it was impossible to separate the ethereal salts by fractional dis- tillation, the method of partial hydrolysis which had proved successful in the case of the malate was also employed here. Twenty grams of the ethereal salt (b.p. 155-161') were shaken with excess of 10 per cent. cold potassium hydroxide solution, and the unaltered oil was separated and washed with water. The alkaline solution, on being acidified with sulphuric acid and extracted with ether, gave lactic acid, as was proved by an analysisof the crystalline zinc salt obtained from it ; we found H,O = 18.31, Zn = 26.72. By boiling the unhydrolysed oil with an aqueous alcoholic solution of barium hydroxide and removal of the excess of the latter with carbonic anhydride, we obtained a barium salt in the form of a gum ; this, when dried a t 140°, was found to contain 34.33 per cent. of barium, the calculated percentage for Isopropylic iodide was found to act readily on silver lactate.IODIDES ON SILVER MALATE AND ON SILVER LACTATE.290 the propoxypropionate being 34.08. The acid, which was obtained from the barium salt by decomposing it with the calculated quantity of sulphuric acid and extracting with ether, was a liquid. The silver salt, prepared from this acid by neutralising it with silver carbonate, crystallised from the cooled aqueous solution as a felted mass of fine needles, not very soluble, and darkening on heating with water. An estimation of silver in the salt dried at 100' gave Ag = 45.32 per cent., andbycombustionwe found C = 29.41,H = 457,Ag = 45-35; C6H,,0,Ag requires C=30*12, H= 4.60, Ag= 45.19. The low result for carbon was probably due to the salt having undergone slight decomposition ; an analysis of another, purer, preparation is given below.Another portion of the residual oil which was unhydrolysable in cold aqueoug caustic alkali was hydrolysed by heating with aqueous alcoholic sodium hydroxide, and the acid obtained in the same manner as from the barium salt above mentioned. The zinc salt made from the acid, which was a gum, like the corresponding ethoxypropionate, mas converted with calcium hydrate into the calcium salt, which crystallised from a concentrated aqueous solution in small needles containing 2H,O. Found, loss of weight a t 115O= 1058 ; Ca in anhydrous salt = 13-31 per cent. Calculated for calcium propoxy- propionate, 2H,O= 10.65 ; Ca= 13-24, As already stated, the crude product of the action of isopropylic iodide on silver lactate contained much free acid. This acid was found to consist chiefly of lactic acid, mixed, however, with a considerable proportion of propoxypropionic acid, The potassium salt, formed on adding potassium carbonate, was separated from the excess of the latter by boiling alcohol, in which it was soluble ; i t was a gum which showed no signs of crystallisation.The acid was obtained from it by acidification with sulphuric acid and extraction with ether. It gave a crystalline zinc salt which analysis showed to be pure zinc lactate, and also an uncrystallisable zinc propoxypropionate which was separated from the lactate by taking advantage of its solubility in alcohol. The calcium salt, procured from the gummy zinc salt, crystallised readily in needles, and was found on analysis to have the same compoaition as the calcium propoxypropionate, an analysis of which is quoted above.The acid, recovered from the calcium salt by acidification with sulphuric acid and extraction with :ether, was con- verted into silver salt, which was now found to be nearly pure, aa is shown by the following analysis. C = 29.70; 0.2860 gave 0.3115 CO,, 0.1175 H,O and 0.1300 Ag. H= 4.56; Ag= 45.45. C,H,,O,Ag requires C = 30.12 ; H = 4.60 ; Ag = 45.1 9 per cent. Our experiments ahow that the high activity of ethereal malates300 ACTION OF ALKYL IODIDES ON SILVER MALATE, ETC. and lactates prepared from the silver salts is due to their being con- taminated with the salts of the much more highly active alkyloxy- succinic and alkyloxypropionic acids. The quantity of the latter produced may amount, when isopropylic iodide is used, to 20-40 per cent..of the whole product ; with primary iodides, on the other hand, the quantity probably does not exceed 5 per cent. The late J. W. Rodger and J. S. S. Brame (Trans., 1898,73,301) found that the ethereal tartrates from silver tartrate have a very ab- normally high rotation. Having received Mr. Brame’s permission, me are investigating this reaction, and we have already obtained some evidence that the abnormal activity is due t o the production of dial kyloxysuccina tes. The silver salts of other hydroxy-acids mill probably be found to react similarly with alkyl iodides, and this method of preparing ethereal salts, a t all events under the conditions observed by us, can- not therefore be used with safety in the case of these acids.The ob- servation was made previously (Zoc. cit.) t h a t whilst the ethereal malates and lactates differed in activity according to their method of preparation, the acidyl derivatives prepared from them had nearly the same rotation ; the explanation no doubt is that, owing to the difference of the boiling points of the alkyloxy-ethereal salts and the acidyl derivatives, the former were eliminated in the course of fractional distillation. With regard to the mechanism of the reaction by which the alkyloxy- acid is produced, we are unable at present to say anything definite. It is well known that alcohols and alkyl haloids interact when heated to a sufficiently high temperature, yielding halogen acid and ether (Annalen, 1864, 131, 5 5 ) , but so far as we know the analogous formation of an alkyloxy-acid by the action of an alkyl halide on a hydroxy-acid or its ethereal salt has not been observed, and it seems unlikely that a direct action of this kind should occur at the temperature we employed.We made some experiments, the results of which tend to negative the idea that the reaction is merely of the nature referred to. Active ethylic lactate made by the sulphuric acid method was heated at looo in a closed tuba with ethylic iodide. The product distilled irregularly and was coloured with iodine, which was removed by shaking the liquid with mercury. The activity had decreased somewhat, from which we conclude that ethoxypropionic acid had not been formed. Another experiment * in which lead oxide was added with the view of remov- ing the hydriodic acid, supposed to be formed, from the sphere of chemical action, gave also a negative result. The fact that isopropylic iodide was more active than the primary iodide in producing the alkyloxy-acids suggested the idea that these acids might possibly be * Ethylic malate was used in this case.OPTICAL ROTATIONS OF METHTLIC AND ETHYLIC TARTRATES. 301 formed by the addition of the alkylene resulting from the decomposi- tion of the iodide, according t o the equation CH,*CH(OH)*COOEt + C,H, = CH,*CH(OEt)*COOEt. The reaction is certainly a very improbable one, and as a matter of fact we found on trial that the activity of etbylic malate was not altered by passing ethylene through it at 100’. We are at present engaged in some experiments which we think may throw some light on the re- action and may perhaps also furnish a method of preparing the active alkyloxy-acids directly from the active hydroxy-acids. I n conclusion, we may draw attention to a fact, established by our experiments, which is of some importance on account of its stereo- chemical bearings. The replacement of the alcoholic hydrogen of malic and lactic acids by alkyl groups produces a very great change of activity. From considerations based on the ionic rotations of malic and ethoxysuccinic acid (Trans., 1895, 67, 982), it was concluded t h a t a change of sign of activity accompanied the substitution in question, but this conclusion now appears to be incorrect. The rotations, of the ethereal salts of the alkyloxy-acids, produced in the reactions described, are in the same sense as those of the ethereal malates and lactates together with which they are formed. UNITED COLLEGE OF ST. LEONARD AND ST. SALVATOR, UNIVERSITY OF ST. ANDREW’S.

ISSN:0368-1645    

DOI:10.1039/CT8987300287

出版商:RSC    

年代:1898

数据来源: RSC

摘要:

OPTICAL ROTATIONS OF METHTLIC AND ETHYLIC TARTRATES. 301 XXX.-The Optical Rotations of Methylic and Ethytic Tart?-ates. By (the late) JAMES WYLLIE RODGER and J . S. STRAFFORD BRAME. THE intention of the authors at the commencement of this research mas to prepare the ethereal tartrates in a high state of purity and investigate the influence of different alkyl radicles on the optical activity at varying temperatures. In attempting t o prepare dimethyl- and diethyl-tartrates by two distinct methods, it was found that widely different rotations were exhibited by the compounds obtained by saturating an alcoholic solution of tartaric acid with hydrochloric acid, or heating the acid and alcohol in sealed tubes, and by those obtained from the action of the alkyl iodide on silver tartrate, Similar differences have been noticed in the case of lactates by J.Wallace Walker (Trans., 1895, 6’7, 914), and for malates and lactates by Purdie and Williamson (Trans., 1896, 60, 818). Owing to the lamented death of Mr. J. W. Rodger and the appoint- ment of the other author to a different laboratory, it has been VOL. LXXIII. X302 RODGER AND BRAME: THE OPTICAL ROTATIONS impossible to carry out the work as intended, but there is snfficient of interest and importance to justify the publication of what has been done. Preparation. First wthod.-By saturation with gaseous hydrogen chloride. The method of Anschutz and Pictet was used. A saturated solution of tartaric acid in ethylic alcohol, kept cool by ice, was thoroughly saturated with dried hydrogen chloride ; air was then rapidly drawn through it for some time, and finally the excess of hydrogen chloride, alcohol, and water quickly distilled off under reduced pressure.The residue was again dissolved in alcohol and the process repeated, allow- ing the saturated solution to stand five weeks ; finally, the product was fractionated under diminished pressure. After the fractionation had proceeded some time, much charring took place, the pressure rose rapidly, the temperature fell, and a limpid liquid distilled. On ex- amination, this liquid readily gave the iodoform test, and, although not completely purified owing to its small amount, the results of a combustion fully agree with those required for ethylic alcohol ; more- over, its boiling point (78") confirms this supposition. Ethylic tartrate No.1 was prepared by this method. Xecond method.-Freundler's was adopted, which differs from the preceding in evaporating the alcoholic solution of the acid until a crystalline mass (supposed to contain ethylic hydrogen tartrate) is obtained, dissolving this in alcohol and saturating as before. No. 2 specimen of ethylic tartrate was prepared in this way. Third method.-By heating tartaric acid or the alkylic hydrogen tartrate in sealed tubes with alcohol ; the yield, however, is exceedingly small if tartaric acid is merely heated with alcohol in a sealed tube, The method used was to heat the acid with twice its weight of the particular alcohol in a reflux apparatus for 4 hours, and then rapidly distil off the alcohol, and the water formed, under reduced pressure.The residue of mono-substituted ethereal salt was then dissolved in more alcohol, and the mixture heated in sealed tubes for 5 hours a t 150-160°; the contents of tube were transferred to a flask, the excess of alcohol and the water formed quickly removed as before, and the ethereal salt fractionated under diminished pressure. By this method, methylic and ethylic tartrates are readily prepared, and the yield is good. The ethylic tartrates Nos. 3, 4, and methylic tartrates Nos. 1, 2 were prepared in this way. Fourth method.-By the action of an alkyl iodide on silver tartrate. An excess of iodide was boiled in a reflux apparatus, silver tartrate, which had been dried thoroughly in a vacuum, was added in smallOF METHPLIC AND ETHYLIC TARTRATES.303 quantities at a time, and the mixture boiled for 2 hours after the whole of the tartrate had been added ; the excess of iodide was then distilled off, the residue extracted several times with pure dry ether or benzene, the solvent distilled off, and in the case of ethylic tartrate the residue distilled under diminished pressure. For methylic tartrate, benzene was used as the solvent, and after partial evaporation crystals of methylic tartrate were obtained on adding some solid tartrate from another preparation. It is most di5cult without a nucleus to obtain the methylic tartrate in the crystalline form, the liquid being very viscous. It was found necessary to distil the tartrate in every case, for crystallisation, even although carried out several times, did not give a sufficiently clear product for measurements of the rotation.When this method is adopted, i t is essential that the materials used be as dry as possible, otherwise little or no yield is obtained. The ethereal salts prepared by this method were always so coloured with iodine as to render accurate polarimetric measurements im- possible, their solutions were therefore shaken with mercury, filtered, the solvent distilled off, and the residue fractionated under reduced pressure. Ethylic tart,rates Nos. 5, 6, 7, and methylic tartrates Nos. 3, 4, 5, 6, were prepared from silver tartrate. Observation of the Angle of Rotution. A half-shadow instrument by Laurent was used, the vernier giving readings to five seconds. The samples were contained in glass tubes similar to those used by Rodger and Watson (Trans.Roy. Xoc., 1895,186, pt. ii, p. 626), which ensured the tubes being full at any temperature ; the tubes were enclosed in a water jacket, and the observations were always made a t 20° or very close t o that temperature. Every care was taken that the readings should be accurate, the tubes when filled being left for some time in the apparatus to ensure the contents having the proper temperature ; this is important, as W. H. Perkin, sen., has pointed out that the rotation of these ethereal salts is very appreci- ably affected by change of temperature (Trans., 1887, 51, 363). The experience of the authors fully bears this out. A number of observations for each sample were taken, distilling each time until a pure product was obtained.Frankland and-Wharton (Trans., 1896,69,1310) found that methylic tartrate prepared by Freundler's method gave a rotation of 2.74" in a tube 1 dcm. long. It will be noticed at once that the salts prepared by the silver tartrate method invariably give a much higher rotation than when The final results obtained are given in the tables (p. 304). x 2304 RODGER AND BRAME: THE OPTICAL ROTATIONS [a]:' Temperature acid. distillation. Tartaric of 13'32 169-171" 13'47 168-5-170" 13.47 163.5-168" 13-32 164-166" 13'47 165-169" 13'32 159-162.5" Methylic tartrate. Pressure Temperature during of for1 dcm. distillation. observation. 19-20 mm. 20.2" 2.785 17-23 mm. 20 -3 2.804 14-17 mm. 20 3'64 13-5-18 mm. 20'1 3'127 15-19.5 mm. 19'8 3-30 9-18 mm.19'9 4-038 Method of preparation. Sealed tube Silver salt 9 9 9 9 9 9 9 9 9 9 9 ) 9 ) 8s Speci- men. No. 1 2 3 4 5 6 Ethylic turtrate. Method of preparation. Speci- men. No. I- Temperature of observation. for 1 dcm. [ alEO Tartaric acid. Anschutz Freundler Sealed tube Silver salt 9 9 9 9 9 9 9 9 9 J 9 9 13'32 13.32 13'32 13'32 13'32 13-47 13'32 1 2 3 4 5 6 Tem pera ture of distillation. 20'1" 20 20 19.9 20 20 19'8 185-1 77" 185-191" 169-172" 172-178" 161 '5-1 63 *5' 171'5-172" 174-176" 9 *37 9 '31 9.25 9 '22 14.33 14.91 14.37 Pressure during distillation. 31-22 mm. 37-43 mm. 14-17 mm. 22-21 mm. 20 mm. 18-20 mm. 17-20 mm. prepared by either of the other methods; the differences here are, indeed, much greater than those observed by Purdie and Williamson in the case of the methylic and ethylic malates and lactates. Further, it will be seen that the compounds produced by the sealed tube method have nearly the same rotatory power as those produced by saturating with gaseous hydrogen chloride.Again, the rotation for the different samples of methylic and ethylic tartrates produced by the silver method is by no means constant, varying in a manner which cannot easily be accounted for, because, as far as possible, the preparations were similarly conducted. Purdie and Williamson found that the rotation of ethylic lactate prepared from the silver salt differed from that of the same compound prepared in a similar way by J. Wallace Walker, confirming the experience of the authors that the product formed is variable in composition when this method is employed.In order t o determine any difference between the ethereal salts produced by the silver method and those produced by other methods, the following experiments were carried out. 1. 4.5 grams of methylic tartrate (No. 1) and the same weight ofOF METHYLIC AND ETHYLIC TARTRATES. 305 tartrate. Carbon. Hydrogen. No. 1 40.16 5.58 No. 3 40.01 5.55 4050 5.62 (theoretical) tartrate. Carbon. Hydrogeu. No. 1 45.69 6.82 No. 7 45.82 6.83 46.65 6.80 :(theoretical)306 OPTICAL ROTATIONS OF METEYLIC AND ETHYLIC TARTRATES. almost exactly the same rotation. With regard to the third hypo- thesis, Purdie and Lander, in a “Preliminary Note on the Action of Alkyl Iodides on Silver Malate ” (Proc., 1896, 170, 221), ascribe the higher rotation in the case of malates to the presence of “small quantities of the ethereal salts of the highly active alkyloxy-acids.” The same explanation may hold in the case of the alkyl tartrates, the hydrogen of one or both of the alcoholic hydroxyl groups having been replaced by alkyl radicles during the transformation of the silver tartrate.These derivatives of active tartaric acid not yet having been prepared, their activity is unknown ; they must in this case be highly active substances, otherwise their presence in such quantity as to account for the great rise in activity produced would have been de- tected by analysis. On this hypothesis, the identity of the rotations of the products of hydrolysis of the two methylic tartrates must be a co- incidence; a great difference between them was probably not to be expected.Judging from the known optical relations of the malates and monalkyloxysuccinates, it is very likely that the sodium salts of tartaric acid and of the alkylated tartaric acids referred to would show a much smaller difference of activity than the ethereal salts of the cor- respondingacizls. That the rotations of the products of hydrolysis of the ethereal salts prepared by the two different methods should coincide so closely is, however, very remarkable. It has not been possible, so far, to prove the existence of these com- pounds in either of the ethereal salts prepared by the silver process, and the only promising method of doing so appears t o be successive recrystallisation of the methylic tartrate from benzene. As the accom- panying figures show, this recrystallisation lowers the rotation con- siderably. Methylic tartrate. No. 6. Temp. of observation. all 1 st crystallisation 1 9 * 9 O 4.035 2nd Y) 1907~ 3.886 3rd 9, 20*05O 3.561 It is therefore probable that the mother liquor, if examined, would be found to contain the more active constituent. Unfortunately, it has not been possible to resume the experimental work, and any further explanation or proof of the cause of difference will no doubt be supplied by Dr. Purdie and Mr. Lander, who have undertaken to investigate the question further.

ISSN:0368-1645    

DOI:10.1039/CT8987300301

出版商:RSC    

年代:1898

数据来源: RSC