RICHARDSON: ACTION O F LIGHT ON PURE ETHER, ETC. 51 TTI.-Action of Light on Pure Ether i?z Presence of Moist Oxygen.. By A. RICHARDSON, PhD. THE formation of an oxidising body in ether which has been exposed' to the influence o€ light and air, was first, observed by Schobein (JOUY. Chenz. Xoc., 4, 134), a i d subscqueiitly this was identified with hydrogen peroxide, first by Dunstan and Uymond (Pharm. J., 17, 841), and afterwards by Berthelot (Compt. Tend., 108, 543) ; and also by Poleclr and Thumel (Ber., 22, 2863)-and myself (PYOC. Chem. rs'oc., 1889, 134). Recently, however, Dunstan and Dymond have published a paper (Trans., 1890, 57, 574), in which they state that pure ether is not acted on by light when exposed in a moist atmo- sphere containing oxygen, and they consider that the formatiori of the peroxide is due to impurities containeci in the liquid.I n the paper above referred to, an abstract o ~ l y of which wits published (Proc. Chenz. SOC., 1889, 134), Dunstan and Dymond discuss my o w n experi- ments on the subject ; tlhe ether used in these experiments was ob- tained from so-called pure ether (prepared from rectified spirits) ; it was repeatedly shaken with water until the last traces of alcohol had been removed, then dried over cslciiim chloride, and repeatedly dis- tilled from sodium until the metal was no longer acted on. Dunstan and Dymond, however, not having seen the original paper, consider that the means adopted to purify the ether used by me, as well as hy other observers, were insufficient t o preclnde the possibility of the formation of hydrogen peroxide from impurities contained in the ether, rather than from the ether itself.I have, therefore, repeated my former experiments, using ether obtained from different sources, and purified by some of the methods suggested by thcru. E 252 RICHARDSON: L u x I o N OF L~GHT ON PURE ETHER Experiments with Corn?nercial Ether. Commercially pure ether was taken, and all the alcohol removed by shaking about 30 timee with water, the process being r+epeated iintil the washings failed t o give the iodoforin reaction. The ether was then shaken with a 4 per cent. solution of potassium dichro- mate acidified with sulphuric acid : after washing with water, it was treated with a solution of sodium hydrogen sulphite, shaken with potassium hydrate, and, lastly, washed with water and distilled ; before exposure to light, the ether so prepared was without action on potassium iodide.Two quantities of this ether, of about 100 C.C. each, were exposed to light in bottles of colourless glass, containing also pure oxygen and water. After 11 days (July 13-24), the ether wqs tested with a solution of potassium iodide, to which it imparted zt deep yellow coloration; shaken with a solution of potassium di- chromate, the ether was coloured blue, thus proving the presence of hydrogen peroxide. Dunstan and Dymond, on the other hand, found that even methylated ether which had been purified in the manner above described was not affected by light, even after an ex- posure of five months. Expeyiments with Ether pp-epnred f r o m Pure Alcohol and Pure 8ulp l~uric Acid.The ether prepared from the pure reagents was first shaken re- peatedly with a solution of potassium hydrate and then with water till the washings failed to give the iodoform reaction. The ether so purified was treated as follows :- (1.) About 200 C.C. of the pure substance was shaken with its own 1-olume of potassium dichromate, every precaution being taken t o follow exactly the method described by Dnnstan and Dymond, and was distilled before being used. Samples of this ether, placed i n bottles together with water and oxygen, were exposed to the influ- ence of sunlight in a glass tank containing water, the temperature of which varied between 13.3" and 26.1". In one experiment, air was substituted for oxygen above the liquid, the conditions being other- wise the same.Another bottle, containing ether and moist oxygen, was protected from the light by a covering of tin-foil, and was placed with the others in the tank, so that the heating effects were in all cases the same. After an exposure t o light of 16 days (August 1-19), ihe contents of the bottles were examined ; the ether which had been exposed to light gave witlh potassium iodide and with a solution of titanic acid it deep yellow colour, and potassium dichromate colouredIN PRESENCE OF MOIST OXYGEN. 53 the ether blue, whilst tlie ether which had been protected from the light was without action on these reagents. (2.) The second method of purification consisted in the treatment of the ether with hydriodk acid; by this, as b y the dichromate method, Dunstan and Dymond state that they mere enabled t o purify methylated ether so completely that after five months exposure to light it, was wit,hout action on potassium iodide solution.A portioc of the ether (about 100 c.c.) prepared from pure materials was, there- fore, shaken repeatedly with about its own volunie of a 3-4 per cent. Folution of hydriodic acid, t'he method described by Dunstan and Dyinond being again carefully followed. The ether, after distillation, was exposed to light i n presence of moist oxygen, the bottle, as before, being imrnersed in water, the temperature of which ranged between 13.3" and 23". A.fter five days exposure (July 24-29), the ether was tested with a solution of potassium iodide, and gave a, yellow colour, which grew darker on standing; it also imparted a yellow colour to a solution of titanic acid.Bther exposctl under the same conditions, except that it was protected from the light, gave no reaction with these reagents. ( 3 . ) The ether from the last experiments, amounting to about, 200 c.c., containing hydrogen peroxide, was collected and again shaken with potassium dichromate, the process being repeated five or six times; the ether, which was colourecl blue, was washed with sodium hydrogen sulphite and with potassium hydrate as before : after distillation, it was exposed with oxygen and water to a north light, so as to avoid direct sunlight, the temperature varying from 10" to 17.7'. Another sample was conipletely protected from the light, but otherwise exposed to similar conditions.After seven days (August 20-27), the ether was tested; that which had been ex- posed to light gave a, yellow coloration 130th with potassium oxide and with titanic acid, whilst these reagents were not acted on by the ether which had been protected from the light. Experiments with Ethel- p e p a r e d froin CL Second Xaiqde of Pure Alcolzol. It was thought desirable to prepare a fresh quantity of ether from a new sample of absolute alcohol, which before use was allowed to stand over calcium oxide for 12 hours and then distilled ; the ether obtained by the action of pure sulphuric acid on this alcohol was, as in the preceding experiment, shaken with potash and with water till all alcohol had been removed ; it was then treated with potassium dichromate, dried over calcium chloride, and lastly over metallic sodium.The boiling point of this ether, amounting to 300 c.c., was54 RICHARDSON: ACTION OF LIGIIT ON PURE ETHER determined with a thermometer graduated to 0.1" ; it was constant throughout at 346' under a pressure of 760 Em. corrected to Oo. This result is identical with that found by Perkin and by Ramsay and Young, whilst the ether used by DunGtan and Dyrnond boiled at 35", a somewhat higher tlsmperature. A portion of this ether was exposed to a north light as in the preceding experiments, the tem- perature varying between '7.2" and 15.5". After seven days expo- sure (September 29-October 6), the ether liberated iodine from potassiuni iodide, and a titanic acid solution was coloured yellow.An experiment was next made in which pure ether was exposed to the action of light at a low temperature, so that the heating effects might, as far as possible, be eliminated ; this was done by surround- ing the bottle containing ether and oxygen with water, which by the addition o f ice was never allowed to rise above 2". After exposure to light for four days, the ether was found to contain hydrogen per- oxide. It appears, therefoi-e, that the formation of this substance is brought about by light aloue, independently of any heating effects. From these experiments it will be seen that hydrogen peroxide was formed when ether, water, and oxygen were together exposed to light when the specimens of ether were prepared-(1) from the corn- mercial " pure '' product subscquently treated with potassium di- chromate ; (2) from pure alcohol and pure sulphuric acid, the prz- duct being afterwai-ds treated with potassium dicliromate or hydriodic acid; (3) from pure ether as in (2), in which hydrogen peroxide had been formed, but had been removed by further agitation with potassium dichromate ; (4) from an entirely different specimen of pure alcohol and pure sulphuric acid, the ether being further purified with potassium dichromate, its purity being confirmed by the agree- ment of its boiling point with that obtained by other observers.Dunstan and Dymoiicl state that the formation of hydrogen per- oxide in ether is the result of some unknown impurity, which can, however, be removed from even the inethylated product by treatment with potassium dichromate ; it might, thcrefore, be supposed t h a t ether prepared with the utmost care from pure materials would after such treatment be freed from this substance, but, as has been shown, hydrogen peroxide is still formed in such a liquid after expo- sure to light, provided oxygen and water are present.Even on the supposition that the impui-ity escaped decomposition in the first instance, a second application of pot aasium dichroniate should be sufficient to completely rid the ether of this body, but again it is found that hydrogen peroxide is formed in the ether after a few days exposure to light. In 110 case have I been able to obtain a specimen of ether in whichIN PBESENCE OF MOIST OXYGEN. 55 hydrogen peroxidc is not formed after short exposure to light under tbhe conditions above mentioned ; and it, would seem that the explanation of the discordant results obtained by Dunstan and Dymond and by myself is to be found, not in the supposition that my ether was impure, but in the conditions under which the ether was exposed to light in the two cases.I n the first place, one is led to enquire whether Dullstan and Dymond really had any oxygen above the ether in the bottle during cxposure to light. For, although they describe minntely the nature and size of the bottles and the quantities of ether used, no allusion is made in their published paper to any precautions they may have taken to secure an atmosphere of oxygen or air in the space above the liquid. When ether is poured into a bottle, the vapour is evolved in such laiege quantities that the air is, to a great extent,, expelled, and it is therefore open to doubt whether, if special precautions were not taken, sufficient oxygen was really present.I n one experiment made with the object of ascertaining whether oxygen has any action on ether (Trans., 1890, 57, 577), these ob- servers passed 200 litres of oxygen in the course of two days through ether heated nearly to its boiling point, and they remark that, although the ether was exposed t o light during this period, no hydrogen peroxide was formed; two days exposure t o light in a London laboratory would, however, hardly be sufficient to bring about the formation of a recopisable quantity of peroxide, even when oxygen is used.In my own experiments, oxygen or air was in every case passed into the bottles immediately after the ether had been added; the stoppers were then inseyted, and the bottles exposed to light in an inverted position, so that no gas could escape without first expelling thc whole of the liquid. Again, it is to be noticed that Dunstan and DTmond used bottles of faintly greenish glass, which absorbs, to a considerable extent, the rays most influential in bringing about the formation of hydrogen peroxidc. Lastly, it does not appear that they distilled the ether after treatment with potassium dichromate and sodium hydrogen sulphite, o r after purification with hydriodk acid and sodium thiosnlphate ; traces of any of these reagents would be sufficient to decompose the minute quantities of hydrogen per- oxide formed under the most favourable conditions.Oiie is, there- €ore, led to suppose that the results obtained by Dunstan and Dymond, seemingly SO diametrically opposed t o my own, may possibly be due to one or more of the following circumstances :- (1.) To the presence of ether vapour instead of oxygen above the liquid.56 RICHARDSON: ACTION OF LIGHT ON PURE ETHER (3.) To the partial absor.ption by the glass of those rays which are (3.) To the presence of traces of reagents which decompose hydro- Taking these points into consideration, it is not, perhaps, remark- able that their results do not agree with those of other observers, and t,hey do riot appear t o be justified in concluding, as they do, that the experiments made by them " conclusively demonstrate that hydrogen peroxide cannot be formed from purified ether by exposing it to light, under ordinary atniospheric conditions in contact with air and water (ZOC.cit., p. 584)."" The experiivents described by me show, on the contrary, that by-drogen peroxide is formed in pure ether, even after treatment with potassium clichromate or hydriodic acid, when exposed to light,, provided- (1.) That the ether vapour in the bottle is replaced by moist air 01- (2.) That the bottles used are made of colourless glass. (3.) That the last traces of the reagents used in its purification are most influential in bringing about the change in the ether. gen peroxide. oxygen in the first instance. removed. Itz$wence of Temperature on the formation of Hydrogerz Peroxide in.Bther. The oxidation of ether at high temperatures has been investigated by Legler (Ber., 14, 602, 18, 3343; AnnaZen, 207, 381) and by Perkin (Trans., 1882, 41, 343), but it does n o t appear clear that hydrogen peroxide was amongst the products of decomposition ob- tained by these experimenters. It ww also observed by Dunstan aud Dgmond (Trans., 1890, 57, 585), that hydrogen peroxide was not formed when a mixture of ether vapour and oxygen was passed Over ~t~rongly heated pumice ; they, however, found that when cold water was in close proximity t o the heated vapours, the peroxide was formed in considerable quantities. Although I have been un- able to detect the prcsence of hydrogen peroxide in ether which has been kept in contact with moist oxygen in the dark, at ordinary tem- peratures, yet I have found that at compam.tively low temperatures * This, as Messrs. Dunstan and Dymoncl have since explained (Proc., 1890-91, p.147), applies to the conditions under which the first series of cxperinients was pel*- fdrmed, namely, diffused suiilight in London, and the electric arc-light. I n the second series (Trans., 1890, 988), with intense sunlight in a clear atmosphcre, hydrogen peroxide was formed in quantity sufficient to be detected by the chromic reaction. [EDITOR.]IN PRESENCE OF MOIST OXPIXEX. 57 this substance was formed. Ether and moist oxygen were heated together in the dark a t 60" for 40 hours; on testing the liquid, it was found that hydrogen peroxide had been formed t o a considerable extent.Again, when a similar mixture was heated in the dark for four days to a temperature ranging between 78" and 88", hydrogen peroxide was detected in the liquid ; in these experiments, the quantities of ether taken were such that it was not entirely converted into vapour. In a tliird experiment, the same quantity of ether as in the last case (contained in a sealed tube) was placed in a flask of 1000 C.C. capacity, which was filled with moist oxygen and sealed ; the tube was then broken, and the mixture of ether and oxygen exposed to a tempera- ture of 75-88'. Under these conditions, the whole of theliquid was converted into vapour ; after four days heating, the contents of the flask were tested for hydrogen peroxide, a.nd it was found that only an extremely minute trace of this substance was formed.The experiments so far described throw no light on the question whether the hydrogen peroxide formed in presence of ether is a direct product of oxidation of water present, or is due to the oxida- tion of the ether. As this point is of considerable interest, the following experiments seem to show that the formation of the per- oxide is due to the direct oxidation of the ether by light, as suggested by Berthelot, and that perfectly dry ether and oxygen give, after exposure, a compound which, on the addition of water, fornis hydro- gen peroxide. Et'her was dried by exposing i t to a large surface of metallic sodium contained in one limb of a bent sealed tube, and after 11 days contact with the metal, a portion of the liquid was distilled into the other limb, which had been previously drawn out. It was then sealed off, slid the distillate so obtained transferred to an outer tube, through which dry oxygen vas drawn for four days; i t was then sealed, and the inner tube broken. The niixture of ether and oxygen was tested after eight days exposure t o light, and gave the hydrogen peroxide reactions with potassium iodide and with titanic acid. In a second experiment, the mixture of dry ether and oxygen was exposed in the dark to a temperature of about 70" for four days ; in this as in the previous case, thc liquid was found t o contain hydrogen peroxide when tested with a solution of titanic acid. If sodium is capable of removing completely the last traces of moisture from ether, the formation of hydrogen peroxide in these experiments is not due to the direct oxidation of water, as 1 had at first supposed ; the results of other experiments, however, which are net yet completed, show that, under proper conditions, water itself is slowly oxidised under the influence of sunlight.38 BRAUSER : VOLLURIETRIC ESTIMATION OF TELLURIUM. Note.-In a subsequent paper, Dunstan and Dymond have published the results of further experiments on this subject (Trans., 1850, 57, SSS), in which they state that hgdrogen peroxide is formed by the action of sunlight on moist ether in presence of oxygen.