36 VELEY ON SOME HIGHER OXIDES XI.-On some Higher Oxides of Maizgnnese and their Hydrates. Part 11. By V. H. VELEY B.A. F.I.C. Christ Church Laboratory Oxford. P~el iminary Discussion. IT seemed t o the author that a continuation of the investigations on the higher oxides of manganese and their hydrates (cf. this Journal, 1880 581-592) might throw some fnrhher light on the constitution of these compounds. Experiments have recently been published on the oxides of manganese b.y Wright associated with others (this Journal 1880 775-785) Moissan (Arm. Chim. Phys. [ 5 ] 21 199-255) and Pickering (Chem. News 43 189 et sep.) ; these researches will be alluded to in the course of the present communication. Purity of the Oxides. As considerable stress was laid in the formep communication on th O F JIhSGAXESE ASD THEIR HYDRATES.57 probable difference of behaviour of the higher oxides when pure and when contaminated with alkali a spectroscopic analysis was made of the samples used. The method of procedure was as follows :-A strong solution was prepared by dissolving 1.358 grams of the hydrated oxide in 20 C.C. concentrated hydrochloric acid. A drop of this solution introduced on a platinum wire into a Bunsen burner, and the flame viewed through a double prism spectroscope gave a spectrum in which the presence of no metals other than manganese and sodium was indicated. I n order to determine the maximum amount of impurity of potassium two flasks A and B were taken in the former of which were placed 10 C.C. concentrated hydrochloric acid (prepared from ammonium chloride and sulphuric acid) and in the latter the same volume of hydrochloric acid containing a known weight of the hydrated oxide in solution.To the former mas added a dilute solution of potassium chloride a cubic centimeter at a time, until a drop taken up on a platinum wire gave distinctly the red line Ka,. The experiment was then repeated with flask B and it was found to require exactly the same number of C.C. before the line Kaz became visible. Hence the amount of impurity of potassium is less than the amount contained in 1 C.C. of the dilute potassium chloride solution. By this process it was shown that the manganese chloride solution contained less than 1 part of 1,000 of potassium chloride. (Further corroborative evidence of the absence of potash was afforded by the readiness with which the oxides were reduced in hydrogen and the freedom from so-called spontaneous oxidation of the manganese monoxide formed,) The experiment described above was repeated a calcium chloride (from Iceland spar) being substituted for the potas-sium chloride solution.Owing to the close proximity of the red and green lines Ca and Cag with Mn and Mng observations were made with the yellow calcium lines on either side of the sodium line ; or on moistening the wire after the experiment with dilute hydrochloric acid the calcium spectrum could be obtained free from that of man-ganese. (This is probably due to the retention of the lime by the oxide of mangauese as calcium manganate which is less readily dis-solved by the hydrochloric acid and so is volatilised after the greater part of the manganese passed off.) By this i t was shown that tllc chloride of manganese contained less than 1 in 6,800 of calcium chloride.The apparatus and the methods of analysis employed have already been described (vide s z y m ) . Changes produced when the Higher Oxides are EeateJ i i ~ Xitroym. Nitrogen was chosen in order to study the dehydration of tho per-oxides a t increased tempcratnres in an atmosphere free from oxjgen 58 VELET ON som HLQHER OXIDES Mean of values obtained. I whereby the loss of water is not accompanied with or affected by. any secondary chemicai reaction. The nitrogen gas was prepared by aspiring over red-hot metallic copper air purified from carbonic acid, and mixed with ammonia by passing through a strong solution of the gas.By so regulating the air and ammonia current that only on(' part of the copper is oxidised while the other part is unaltered ti),. nitrogen is obtained free from oxygen and hydrogen with one or other of which it would be contaminated if the contents of the heated tube were completely oxidised or completely reduced. (This ready methocl of obtaining pure nitrogen is not described in any of the more recent text-books and was 5rst) adopted by Harcourt.) Before the gas entered the drying and purifying apparatus it was passed through a considerable length of red-hot combustion tubing filled with copper foil and cupric oxide and then through a washing bottle containing a1 kaline pyrogdlate.Before commencing the experiments fresh analyses were made OF t,he hydrated peroxide used in the former publication the method of preparation of which has already been described a t length. Values obtained in former experiments. Manganese monoxide Oxygen Water . 75 *34 75 *37 14-28 14 '06 10.40 10.54 160.02 99 -97 ___-___--These analyses are fairly accordant and agree with the hypothesis t,hat it is an oxide of formula Mn60, of not very definite hydration. A series of experiments were then made to ascertain whether the composition of the oxide when heated in nitrogen from 60-200°, remains unchanged so far as regards the relation of manganese mon-oxide to available oxygen or whether the loss in weight is solely attributable to the loss of water of hydration.Reinadcs ON Table I. I. The temperatures a t which stable hydrates seem to be formed arc' 90-95" and 170-180". 11. Throughout the experiments there is no marked diffei*enw between the loss in weight of the oxide and the water collected in the drying tube. This shows that the oxide when heated in nitrogen from 180-200" loses only water of hydration aiid not available oxygeli. (In the experiments marked $ there mas slightly more water collectct . . . . . . . . . . . . . i + -0 0 0 0 0 0 3 0 0 0 3 0 0 0 0 0 0 0 . . . . . . . . . . . . . 3 A 0 0 0 0 3 0 0 0 0 0 0 0 0 0 . . . . . . . . . . . . . . s E H m 8 c;l YELEP OX SOME HIGHER OXIDES . . 0s 0 3 03 0 0 0 0 0 A+ I m u+ d .r( 4.2 .- a 6 I . . . . . . ci H H w L: 4 m t-i H I I I I I 00% 0 O r 0 03 0 0 0 0 0 0 0 0 A .. . . . . . $ . . . t D . . & . 5 . 4 . o . $ . ? : .- - .+ . . % * . o * . s . c 3 d c 3 : .5 .z :.5 OF MASGANESE AljD THEIR HYDRATES. 61 Values obtained. -78 $7 6 *78 14 '60-14 '53 than loss in weight of oxide; these differences amounting to 0.0065 gram may be due either to the absorption of oxygen by the oxide when cooling before weighing or t o Graces of water not derived from the oxide passing into the drying tube. In order to confirm these experiments and to obviate the first-named source of error the oxide aas.al1owed to cool to the temperature of the room in the current of nitrogen before it was taken out of the apparatus. By adopting this precaution the gain in weight of tho drying tube and the loss of the oxide corresponded exactly one with the other.A third series were made in order to establish the constitution of the oxide when heated to 200" and to corroborate the results detailed in the two tables above (I 11). From analyses of the hydrated oxides which remained constant in weight wBen heated to 90-95" 160-18U0 and 200" C. in Tables I-111 confirmed by others their constitntion was determined. Mesn. 78 -67 6 '78 14.57 Analysis of Oxide wJhen Dried in Nitrogen at 90-95". 3langanese monoxide . Osygen (by titration). . Water - 1 100*02 78-58 14 *77 6 -65 loo '00 --Analysis of Oxide when Dried i i z Nitrogen at 160-180". Valucs obtained. Mean. I I Calculated for Mn60,1H20.Manganese monoxide Oxygen (by titration). . Water . 81.60-81 -18 81 *31 15.21 1 li5fi 3 *44-3 '46 81 *2S 15 '28 3 *44 VOL. XLI. 62 VELEY ON SOME HIGHER OXIDES Analysis of Oziitle when Dried in Nitrogen at 300'. Manganese monoxide I Oxygen 15.66 Water . 1.75 82 04.2-82.56 -Values obtained. M-ean. I I 82 *49 15-66 1 - 7 5 99.90 I--I- -Both +b,e latter partially dehydrated oxides readily took up oxygen iit 100" C. and slowly even a t 6OO.X Repetition of Experiments. I n order t o confirm the former experiments it was considered desirable to repeat the whole oi the research ab initio. A different sample of manganesg chloride was purified by the processes described before and the oxide precipitated from the acetate by a current of chlorine a t 52-54" and finally dried a t 60-80".The composition of the new sample? was found to be very nearly but not exactly the same as that of the former sample the one being expressible by a formula Mn,,O4,8H2O the other by a formula Mn24041.SH20. This slight difference may be due to a rather greater oxidation by a longer continued action of chlorine or t o an absorption of oxygen during a protracted heating of the oxide in the air-bath. It is here only neces-sary to state that on repetition the former observations were confirmed in every respect. The sample B when heated in air went through the same cycle of changes ; there were the same stopping points a t which the oxide neither gained oxygen nor lost water; the absorption of oxygen began at the same (temperature and took place within the same limits ; and the composition of the oxide ultimately obtained after heating to 200" was tlie same as that obtained by the same pro-cesses in the former experiments, * Pickering has observed (Zoc.cit. supra) a similar absorption of oxygen a t 100' C. in the case of oxides prepared by different processes. H e considers that the tem-perature at which absorption begins probably depends on the physical condition of the oxide the quantity of so-called available oxygen t l e water of hydration and the condition to which it is exposed when heated. t This sample will hereafter be designated sample B OF MAKGASESE AND THEIR HYDRATES. Analysis of Oxides Dried in A ~ T at 200". 63 Mean values of second sample.--Manganese monoxide,. - . . Oxygen Water . Mean values of Calculated first sample. for Mn,,O,,H,O. 81 *68 16.75 1.65 99 -88 - -I--/ 81 '48 81 '44 16 -79 16 *84 1 '69 1-72 99 *96 100 .oo ---Condit.ion. Action of laryhogen. The behaviour of the oxides of manganese when heated in a current of hydrogen has lately been made the subject of careful investigation by Wright associated with others (this Journal 1878 Trans. 512-527 ; 1880 799-781). The author's experiments would not here be adduced did they not offer an independent confirmation of Wright's researches though the original object in view was the possible prepa-ration of hydrated peroxides containing less so-called available oxygen than the oxide MnGOli. Time.TABLE IV.-Contiwous Series iu Hydrogen. Heated in a elow current of hydrogen. Ditto,. Ditto . . . . . . . . Ditto . . . . . . . . Ditto Ditto . . . . . . . . Ditto Ditto . . . . . . . . Ditto Ditto . . . . . . . . Ditto . . . . . . 2 hrs. . . . . . . . . . . . . . . . . . . . . . . . . I - -Temp. 100" 110 120 130 140 152 180 190 200 200 60-17( Weight )f oxide. Gram. -0 *4192 0 *4062 0 -4040 0 *4010 0 *3970 0.3970 0.3880 0 '3792 0 '3710 0,3530 0.3525 0 *3525 Loss. Gram. -0.0130 0 .oil22 0 '0030 0 '0040 nil 0.0090 0 '0038 0 -0082 0 .Olt30 0 *0005 nil 8 (Mn,C4) 3H20 Remarlis 012 the Table above. I. The reduction or loss of available oxygen begins at a temperature of 130" when there is a marked difference between the percentage loss in hydrogen and the corresponding loss at the same temperature in a i r ; but this change is more marked a t 150".F 64 VELEY ON SO3E HIGHER OXIDES Mean* 11. The oxide constant in weight at 200" has the characteristic colour of the red oxide ; the hue was peculiarly bright on first taking the oxide out of the heating apparatus but on exposure to the air the bright surface was immediately dimmed. This outward change is probably accompanied with a.n absorption of oxygen for the oxide on keeping gained weight and had a composition only approximate to Mn304. Cnlculated for 8(Mn20,)311,0. Analysis of Oxide when Heated to 200" in a Current of Hydroqen. Manganese monoxide . . .Oxygen . . . . . . . . . . . . . . . . Water (by difference) . . . . Values obtained. 89 -90-89 90 7 '23- 7 -10 -I-- I 89 -80 90 *35 '7.16 1 6 .78 3 .O& 2 -87 100 *oo I 100 *oo In the oxide obtained the ratio of oxygen to the monoxide is '*'* x 100 = 7.97 ; in Mn304 it is 15.96 = 7.52. 9 3 21228 The retention of a greater quantity of water (3.04 per cent.) by the oxide when heated in hydrogen to 200" as compared with that retained by the oxide when heated to the same temperature in air (1.69 per cent.) oxygen (1.77. per cent.) or nitrogen (1.75 per cent.) can be explained thus the oxide though exposed to a current of dry hydro-gen is perpetually surrounded by the vapour of water formed by the reducing action of hydrogen on the oxide. It follows that the oxide will be dehydrated oiily when the tension of vapour of water emitted by it is greater than that of the water in the hydrogen atmo-sphere.Although the oxide was constant in weight after a second heating at 200° yet this constant weight was attained by the removal of the oxygen absorbed by the necessary though brief exposure to air chiring the process of weighing ; the formation of a small quantity of water in the cooled part of the apparatus gave on reheating a further proof of such absorption of oxygen by Mn301 which has been pointed out by Dittmar Wright and others. S$eciJic Grnvitl:ea of the Hydrated Oxides. It seemed of interest to ascertain how far these hydrated peroxides of manganese differing slightly from one mother in the relative pro-portions of constituents differ in their physical properties as apecific gravity OF MANGANESE AND THEIR HYDRATES.65 Pure benzene was carefully dehydrated by keeping for some week$, and subsequently distilling over sodium ; the fraction which passed over between 80-81" (uncorr.) was reserved for the determina-tions. In order to obtain as correct a value as possible for the specific gravity of the benzene two bottles were taken and their weights com-pared when filled to the mark with benzene and with water a t the same temperature (16.5") and the specific gravity of benzene was subsequerit,ly reduced to terms of water at 4". Specific gravity of benzene by bottle A Specific gravity of benzene by bottle B compared with water a t 4' C. 0.8944 cornpared with water a t 4" C.0.8945 Mean 0.89445 The specific gravity of the oxides was ascertained by dropping about 1 gram of the oxide into a specific gravity bottle from a weighing tube which fitted closely to the neck of the bottle to avoid unnecessary exposure to the air. It was theu covered by a lsyer of benzene and warmed in a water-bath to expel air bubbles and allowed to cool iu water to the temperature 16.5". Spea9c Gravity of Sample B ( Mnz10458H:20). Twb determinations were made by the method described above. (1) Specific gravity compared with water a t 4" . . 4.671 > > Y9 4-681 Mean 4.675 - (2) Specific Gravity of the Hydrate constant i ? ~ Air at 200" ( M T ~ ~ O ~ ~ H ~ O ) . (1) Determination compared with water a t 4" (2) $7 >> Y 4,750 Mean 4.775 4-800 Specific gravity of manganese monoxide obtained by igniting the higher oxide for some days in a current of hydrogen:-(1) Specific gravity compared with water at 4" 5.010 Former observations by Rammelsberg 5.000 As several determinations have been published by Playfair aud Joule and by Rammelsberg (Monuisb.Ber. Akad. 1865 110) of' the sesquioxide and red oxide prepared by artificial processes it was thought superfluous to repeat them 66 HOWARD AND HODGKIN ON A Summary of Xesults. I. The higher oxides of manganese when heated in dry nitrogen a t temperatures ranging from 60-200" are simply dehydrated without loss of available oxygen the dehydrated oxide formed readily absorbs. oxygen. 11. The oxides when heated in dry hydrogen are simultaneously dehydrated and reduced ; a hydrate of the red oxide is formed which readily absorbs oxygen. 111. The quantity of' water retained by tho peroxides when heated to 200" in dry hydrogen is greater than that retained by the same oxide when heated to the same temperature in air oxygen or nitrogen. The higher oxides and their hydrates obtained in the course of the former and present research are tabulated below and in order to facilitate comparison and to show their mutual relation they are represented as oxidised produds of a protoxide of' molecular formula AIn2,0,4 :-In conclusion the author wishes to express his best thanks to Mr. Vernon Harcourt for kind assistance and advice during the course of this investigation