摘要:

The next Ordinary Meeting, on Tuesday, May 3rst, will be held in the rooms of the Chemical Society, Burlington House, W.

ISSN:0014-7672    

DOI:10.1039/TF911060X001

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

THE LATE DR. RICHARD ABEGG. An international Memorial Fund is being raised in memory of Dr. Richard Abegg, who lost his life in a balloon accident on April 3rd last. The memorial will take the form of '' Abegg Scholarships " in Chemistry and Metallurgy, to be available at the Technical High School and the University of Breslau. As the subscription list is about to close, admirers and friends of Dr. Abegg are requested to send in their donations without delay to the Ihsiuricr Dtckontobank, h're&iri I . Kriig jo. Envelopes should be marked Abegg - Skytui~g,

ISSN:0014-7672    

DOI:10.1039/TF911060X002

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

OF F O U N D E D 1 9 0 3 . TO PROM3TE THE STUDY OF ELECTROCHEMISTRY, ELECTROMETALLURQV, CHEMICAL PHYSICS, METALLOGRAPHY, AND KINDRED 8UBJECT8, I-OL. YI. FEBRUARY, 191 I . PARTS 2 AND 3. SOME PRACTICAL EXPERIENCE OF THE S H EKARDISING PROCESS. BY J. W. HINCHLEY, A.R.S.M., \IYH~T.SCH. The sherardising process (a method of dry galvanising in which iron or stcel articles are coated with zinc by being heated in zinc dust in closed receptacles), which was discovered about nine years ago by Sherard Cowper- Coles and has been exploited commercially for the past seven years, is of great interest on account of the fact that no adequate scientific explanation has appeared. The general theory given by the inventor, and upon which he and others have proposed analogous processes, is that the vapour of zinc is the active agent in the deposition of the coating. One inventor has suggested that the zinc in zinc dust is in an allotropic state-an unstable form giving off vapour explosively, but the process which he has patented seems to be based on the simple vapour theory.Although I am unable to offer 3 complete theory of sherardising, I hope to be able to show that the vapour theory is unsound and to lay some foundation to a theory which is iiot in disagreement with any of the facts at disposal. In the early days of the process the vapour theory being unquestionably assumed, the only factors controlled were the temperature and the time of operation, and it was expected that a batch of zinc dust could be used without addition until exhausted of metallic zinc.On account of the volume of dust a t disposal diminishing, the practice arose of adding new dust after each opera- tion to maintain the stock. Most careful attempts were made to maintain perfectly uniform temperatures, but to the despair of the commercial men controlling the business, a proportion of bad work was often done and not discovered until too late. The workmen, of course, surrounded the process with mystery and gained, as is usual, some support for their position, that they alone could manage the process if the commercial people would let them have a good time.I34 SOME PRACTICAL EXPERIENCE OF Systematic scientific experiment soon established the fact that if the percentage of metallic zinc in the dust be kept constant similar kesults were always obtained between the same temperature limits, and these limits were much wider than had been suspected.This fact at once destroys the simple vapour theory. I t was also proved that the thickness of coating did not depend on the temperature of the dust, but upon the temperature of the article under treatment. The thinnest articles or parts of articles always received a thicker deposit-a result which is again opposed to the vapour theory, for it would be expected that the vapour would travel to the cooler parts of the drum, i.e., to the large articles, and condense upon them, pro- ducing a thick deposit. With rich zinc dust (80 per cent. metallic zinc) the process is coin- mercially uncontrollable, and with poor zinc dust (5 per cent. metallic zinc) it is almost impossible to obtain a commercial deposit at any sherardising temperature, i.e., a t any temperature below the melting-point of metallic zinc.By tests on a commercial scale carried on during several years it is found that for general jobbing work a dust containing 18 per cent. of metallic zinc uniformly distributed throughout its bulk gives practically perfect even coatings through a wide temperature range. The deposition of zinc on the articles takes place at a rate depending upon the strength of the zinc dust, the temperature of the article, and upon the time of treatment. The last factor can always be got rid of by exceeding the time for deposition upon the heaviest articles. After a certain time no further deposition of any practical account takes place.I am referring in all cases to sherardising temperatures (temperatures below 420° C.), and restrict the term “sherardising” to the coating of iron and steel by the process ; in the coating of other metals and alloys by the method the process is not comparable. Experimental results in the laboratory to determine the relations between temperature, strength of zinc dust, and time were too erratic for the purpose, but the final results were easily controlled and definitely obtained. The temperatures used upon a commercial scale vary from 250” C. to 400’ C., the lower temperature being aimed at for springs and similar tempered articles and the higher temperature for cabt iron and similar goods. Now at 250’ C., a temperature not much exceeded in best work, the vapour tension of zinc by the usual method of comparison cannot much exceed that of mercury at oo C., viz., about mm.Hg, a pressure quite incapable of determining the condensation 011 the articles of I 02. of zinc per square foot in one hour. Generally speaking, the basis of the commercial process consists in making the strength of the zinc dust such that when exhausted within one-twentieth inch of the surface of the article (a range which is controlled mainly by temperature conditions) the coating shall be of the required thickness. A large number of laboratory experiments have been made with a view to determining the changes that occur. Upon heating a bright steel article in a closed vessel half full of zinc dust, and arranged so that the article is half embedded in the dust, one is compelled to observe the following facts :- I.As long as the air-exposed portion of the article-remains bright no sherardising takes place on the part which is embedded in the dust. 2 , At the temperature at which a film of magnetic oxide of iron appears on the exposed portion sherardising is found to have begun on the embedded portion. 3. The growth of the coating does not proceed uniformly after once removing the article and exposing it to the air.THE SHERARDISING PROCESS I35 4. At the surface of the dust and exactly level with it a black band, about & in. broad, is formed upon the article, separating the air-exposed portion covered with a magnetic oxide film and the embedded portion which is sherardised. 5. If a cavity be made about i& in.larger than the article, sherardising only takes place at the areas of contact. COMPOSITION OF THE COATING. The coating always contains iron, whatever the purity of the zinc dust, and from a number of analyses it seems impossible to set a limit to the com- position. The iron appears to be in solid solution in the zinc, and near the surface of the article may amount to nearly 10 per cent. of the coating. It has generally been understood that the coating penetrates into the iron, and from chemical tests this appears to be the case ; but I have not been able to detect this layer of zinc-iron alloy by other methods, although the iron-zinc alloy immediately external to the surface is very clear. This penetration must be much smaller than was formerly supposed.Upon dissolving the coating in caustic soda solution a black slime is found upon the surface of the article. This slime, after washing and drying at the ordinary temperature, brings about spontaneous combustion of the filter paper, and the mass is found to consist of iron oxide. The black band formed on the partly embedded article at the surface of the dust rusts rapidly on exposure to the air, and appears to consist of finely divided iron, and in some cases of magnetic oxide of iron in an amorphous form. In the factory, black spots and areas, called by the workmen ‘ I blows,” are occasionally found, especially on hollow articles at places where a current of expelled air is produced during the process, and these spots are similar in character to the black band referred to.I t appears that a film of magnetic oxide of iron is first produced upon the surface of the article which is reduced by the zinc, in contact, to metallic iron which alloys with the excess of zinc present to form a film replacing exactly the oxide film previously formed. If this theory be correct, iron oxidising with difficulty should sherardise with difficulty, and this is found to be the case ; cast iron needs a temperature of over 200° C. before sherardising begins, while soft steel begins to sherardise at 160° C., and some other steels, nickel steels in particular, need still higher temperatures or do not sherardise at all. On account of these differences cast-iron goods are always treated apart from steel goods. There is no difficulty in sherardising,blued springs or steel already covered with a layer of magnetic oxide of iron, provided that thc thickness of the film be not too great.On this account annealed sheets may be sherardised without pickling, but if the thickness of the film be great the surface of the sheet will be rough and irregular, and in extreme cases the film may not be penetrated. Again, any device which facilitates the formation of the magnetic oxide film ought to facilitate the sherardising process. This is found to be the case, and in the early days of the process the workmen accidentally found that if the goods were introduced into the dust in a wet condition bright coatings were more readily obtained. This practice became very persistent and was difficult to stamp out. The brightness of the coatings was often in these cases due to its thinness, and in any case the wastage of zinc dust was very great.The presence, in small quantity, of oil on the goods has been stated to be unohjectionable, and it has also been suggested that carbon should be added,I 36 SOME PRACTICAL EXPERIENCE OF the bulb, its weight being determined 'by ,transferring it from a weighed bottle. The tip of the knife is freed from any traces of dust t presumably to reduce zinc oxide to metallic zinc. Xow oil in any form ruins the colour of the work, and in many cases acts like water i n impoverishing the dust. The finely divided carbon produced is disseminated through the porous coating, and gives a dirty appearance which condemns the work. Carbon does not, and cannot, reduce zinc oxide at the temperature of slierardising.Mistakes in carrying out the commercial process by which impurities are introduced in the dust generally necessitate the replacement of the whole of the dust in use by a new stock-a very costly proceeding. In starting new works or in replacing old stocks, the strength of com- mercial zinc dust may be reduced for use to 18 per cent. metallic zinc by means of any inert dust, such as silica, zinc oxide, &c. The presence of hygroscopic substances in the dust must be most carefully avoided, although a small amount of zinc chloride is always present in com- mercial dust. Practically all impurities affect the process, and are generally found in the coating itself. The pickling process should be carried out with sulphuric acid (of suitable strength for the work) otily.Other acids invariably lead to the deterioration of the dust and the necessity for greater additions of new dust than the work should need.THE SHERARDISING PROCESS I37 deposit appears to determine its growth by promoting the adhesion of further zinc particles. The oxides, indeed, seem to behave as fluxes. After a time the growth ceases, but if the temperature be raised, a small further growth may take place. On examination after removal of the article it is found that the dust immediately in contact with its surface is almost free from metallic zinc, while the rest of the dust has lost a proportion of its original percentage. The general lowering of the percentage of metallic zinc other than due to the amount deposited on the article is exactly equivalent to the air and moisture present. T H E VALUE OF THE DEPOSIT.The laboratory tests of the deposit give superior results to those obtained Glass vessels in the laboratory are not equivalent to the diverse in practice. contact materials of practical engineering. ATMOSPHERIC CORROSION. The coating resists atmospheric corrosion two or three times as well as ordinary galvanising of the same thickness with an unbroken surface. In structural work, in which bare iron and galvanised surfaces are often in contact, a sherardised surface has over five times the life of a galvanised surface. In cranes and similar machinery in use outdoors, the life of a well- sherardised machine is practically as great as that of the machine when used indoors.The explanation of the long life is no doubt due to the fdrmatioii of a protective mass of zinc oxide held together by a spongy network of metallic zinc. The first action which takes place on the exposure of a sherardised surface to the atmosphere is the formation of a yellowish deposit, which develops to its greatest extent in the first two or three days. The deposit consists of zinc oxide coloured by the trace of iron in the coating. It does not adhere to the surface, and is washed off by the first shower of rain. No further formation of such a deposit takes place, and the surface gradually darkens, ultimately becoming almost black. This condition is practically a permanent one, pro- vided that the original coating contained + oz.zinc per square foot. With coatings of the order of oz. zinc per square foot-a deposit which is excel- lent for all indoor iron fittings-one year’s exposure out of doors is the usual life, Generally speaking, a sherardised coating containing Q oz. zinc per square foot of surface is as satisfactory as a galvanised coating containing I+ oz. zinc on the same area. The coating, being porous, is perfectly adapted for the application of tar or paint. The life of the coating is much increased by dipping it in a weak solution of paraffin in petroleum. CORROSION IN LIQUIDS. The action of liquids on a sherardised coating is more severe, but in this case, in nearly neutral liquids the life of the coating is much longer than that obtained by galvanising. In acid liquids the life of any zinc coating, espe- cially if in contact with a dissimilar metal, depends mainly on the amount of zinc present.Local action is, however, largely absent, and the solution of the coating takes place at a very even rate. The electrical potential of the deposit is nearer that of iron than is that of pure zinc, and that part of the coating which is nearest the surface of the iron is still nearer, so that138 SOME PRACTICAL EXPERIENCE OF the action of acid liquids upon the deposit is selective, attacking the thicker deposits first. Owing to its porous character newly sherardised coatings do not stand the action of liquids so well as older deposits, but paraffin-dipped coatings are extremely resisted to liquid corrosion. GHARCINC d DISCHARGING APPARATUS FOR SHE R A R Dl2 INC DRUMS TESTING THE COATING.I’reece’s test for galvaiiised coatings is unsuitable for examining sherardised articles on account of the adhesion of the deposited copper to the porous coating. It is better to dissolve the zinc from a definite area of the coating. This is easily done by supporting a hard rubber tube138 SOME PRACTICAL EXPERIENCE OF the action of acid liquids upon the deposit is selective, attacking the thicker deposits first. Owing to its porous character newly sherardised coatings do not stand the action of liquids so well as older deposits, but paraffin-dipped coatings are extremely resisted to liquid corrosion. GHARCINC d DISCHARGING APPARATUS FOR SHE R A R Dl2 INC DRUMS TESTING THE COATING. I’reece’s test for galvaiiised coatings is unsuitable for examining sherardised articles on account of the adhesion of the deposited copper to the porous coating.It is better to dissolve the zinc from a definite area of the coating. This is easily done by supporting a hard rubber tubeT H E SHERARDISING PROCESS I39 in contact with the surface. alkali solution. amount of zinc dctermined. The tube is filled with weak acid or caustic After the action is over the solution is examined and the It is easy for an experienced person to detect I 1 I I I I I I 1 I I J faulty work by the appearance of the coating. The differences are, however, almost impossible to describe. Illustrations of the first plants erected in this country for the treatment of small articles and goods whose dimensions are not extreme have already1 40 SOME PRACTICAL EXPERIENCE OF been published.Tubes up to a length of 14 ft. 6 inches are iibw being sherardised with great success, and Fig. 2 is a diagrammatic view of a recent plant for charging and discharging the long drums used in this process. Fig. 3 is a view of the actual plant. Fig. 4 is a drawing and Fig. 5 a view of the latest type of furnace for heating such drums. The burners, using producer gas, are arranged along one side of the furnace so that direct impingement of the flames on the drums is prevented. The waste gases from the furnace are drawn off by chimney draught through openings on the floor of the furnace ; tiles are used to adjust these openings to maintain the draught even throughout the length of the furnace.In working The gear for rotating the drums is plainly shown. /-> /--- I such furnaces, on account of the bad heat conductivity of the zinc dust, great Care is needed to ensure an equal temperature throughout the drum. The best practical method consists in using a very high temperature for about one-third of the time of treatment until the shell of the drum has reached the correct temperature, and, then, in reducing the amount of gas burned to maintain this temperature for the rest of the time. As a rule the temperature of the goods at the centre of the drum is about 4oOC. lower than the temperature of those near the shell. The difference in thickness of deposit on this account, provided the strength of zinc dust is correct for the temperature used, need not exceed & 02.per sq. ft. Fig. 6 is a view of a device I have used for sherardising sheets which1 40 SOME PRACTICAL EXPERIENCE OF been published. Tubes up to a length of 14 ft. 6 inches are iibw being sherardised with great success, and Fig. 2 is a diagrammatic view of a recent plant for charging and discharging the long drums used in this process. Fig. 3 is a view of the actual plant. Fig. 4 is a drawing and Fig. 5 a view of the latest type of furnace for heating such drums. The burners, using producer gas, are arranged along one side of the furnace so that direct impingement of the flames on the drums is prevented. The waste gases from the furnace are drawn off by chimney draught through openings on the floor of the furnace ; tiles are used to adjust these openings to maintain the draught even throughout the length of the furnace.In working The gear for rotating the drums is plainly shown. /-> /--- I such furnaces, on account of the bad heat conductivity of the zinc dust, great Care is needed to ensure an equal temperature throughout the drum. The best practical method consists in using a very high temperature for about one-third of the time of treatment until the shell of the drum has reached the correct temperature, and, then, in reducing the amount of gas burned to maintain this temperature for the rest of the time. As a rule the temperature of the goods at the centre of the drum is about 4oOC. lower than the temperature of those near the shell. The difference in thickness of deposit on this account, provided the strength of zinc dust is correct for the temperature used, need not exceed & 02. per sq. ft. Fig. 6 is a view of a device I have used for sherardising sheets whichTHE SHERARDISING PROCESS 141 enables such goods to be treated at an extremely low cost. The sheets are packed alternately with '' combs " made of sheet steel of a thickness equal to the layer of zinc dust required between each sheet. The packet of sheets and combs are placed in a cast-iron box or sagger and the whole placed on a shaking-table. The combs are provided with lugs so that every alternate or third or fourth one may be raised while the dust is being shaken in to fill up the space of the combs as they are removed. In this way all the combs are removed, leaving the plates evenly packed in zinc dust and ready for furnace treatment. The sheets packed by this method are remarkably even in thick- ness of deposit, and, on account of the regularity of the thickness of the zinc dust between each sheet, its strength may be increased, 3 larger number of sheets packed in a given space, and the time of operation reduced. Fig. 7 is an outside view of' a small oven for sherardising in stationary boxes. A gas-ring below is the source of heat and the body of the furnace, which is capable of being raised for the removal of the boxes, forms a suitable hot-air bath for the process.

ISSN:0014-7672    

DOI:10.1039/TF9110600133

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

THE SHERARDISING PROCESS DISCUSSION, The Chairman asked whether the sherardising process had been adopted for metals other than iron, such as copper, aluminium, &c. He was aware that Mr. Cowper-Coles had experimented with aluminium some time ago, but he believed without success. It would be a great gain if aluminium could be sherardised ; it would be a means of electroplating it. Had the process any influence on the tensile strength of material treated as compared with ordinary zincing processes ? The author thought that an alloy was formed as a result of the process, and this, of course, involved some penetration of the zinc into the iron. But he gathered from the paper that there was no such penetration. Presumably the penetration, if any, would depend on the time of treatment.What was the maximum depth reached after a considerable time ? He was interested to learn to what an extent the process was used in America ; that was evidence of commercial success in that country, and it was, if so, surprising that it had not been taken up more energetically in this country. Dr. H. Borns : I do not wish to be critical, but, so far as I remember, some statements made by Mr. Hinchley would not appear to agree with what Mr. Cowper-Coles said when presenting a paper on the same subject to the Society of Chemical Industry a year ago. Mr. Hinchley did not especially refer to that paper, but may we regard his own conimunication as in a certain sense supplementary or corrective? Mr. Hinchley may be quite right in accentuating the contact theory, but if actual contact is necessary, how can a fine screw be coated quite true, as Mr.Cowper-Coles claimed ? The latter also said that the iron drum did not become coated on the inside ; that is difficult to understand on the contact theory, though any zinc coating formed might be decomposed again and oxidised by the subsequent heating. Mr. Cowper-Coles distinguished sherardising and vapour-galvanising ; in the latter process the object is suspended over the heated zinc dust, with which it is not in contact. I should also like to ask whether the sherardised sheets really can bear any mechanical maltreatment. Dr. F. Mollwo Perkin expressed his great interest in the process, with which he had some acquaintance, at least on a laboratory scale. He dis-142 SOME PRACTICAL EXPERIENCE OF sented somewhat from the theory put forward by the author. There is no doubt zinc dust must have an appreciable vapour-pressure even at low temperatures, and it therefore seemed to him that the process could be carried out even without actual contact.Of course, it will be necessary for the zinc to be very close to the article, because the vapour will not travel any great distance ; each particle of zinc will be, SO to say, enveloped in zinc vapour. Dr. J. A. Harker agreed with Dr. Perkin regarding the possibility of the reaction being a vapour one. It was of interest in this connection to mention that at 750°C. silver, which boils at 210o0C., has a sufficient vapour tension to spoil electric insulation. Dr. N. T. M. Wilsmore thought that, as the particles of zinc dust were always coated with a film of oside, direct metallic contact was hardly sufli- cient to account for the coating, especially in the stationary process.On the other hand, a very small vapour-pressure of the zinc would be sufficient to produce the results, seeing that the distance through which the vapour had to diffuse was also small, except where cavities occurred. The characteristic smell of zinc seemed to indicate that the vapour-pressure of the metal was finite even at ordinary temperatures. Mr. H. R. Ellis asked whether the process might not be due to some electrolytic action, such as took place when a mixture of zinc dust, ZnSO,, magnesium powder, and an ammonia salt is rubbed 011 brass, copper, or iron. Under such conditions a deposit of zinc is readily formed.Mr. J. W. Hinchley, in reply, said that he retained the term sherardising for iron and steel only. He believed that the coating of other metals by the process was neither practical nor useful. Samples of sherardised aluminium had been shown to him which proved to be merely oxidised. The addi- tion of other materials to the zinc dust contaminates the coatings obtained, and if the materials be metallic oxides capable of reduction by zinc, the metal is found in the coating. This occurs with copper and one or two other metals, but no practical process is as yet founded upon the facts. The coatings obtained by the use of other metallic materials free from zinc were not good, and, as a rule, very high temperatures were necessary for their production.He had never obtained a deposit of zinc and copper the colour of brass ; the colour of the compound deposit is either that of copper or of zinc. The effect of the process on the tensile strength is determined by the entire removal of all mechanical stresses, and consequently of any weakness or strength due to them. In the case of wire and sheet this strength can be largely returned by drawing or rolling after sherardising, choosing, of course, the proper tem- perature for the operation. In the case of chain and forged materials, generally, there is a distinct improvement in strength, due to the removal of the irregular stresses produced in forging. The process detects flaws in materials-defective welds and cracks in forging are revealed with certainty.Spun articles are often full of such cracks and break up in sherardising ; such articles are thoroughly bad, and would be condemned if the cracks were evident at first. He did not state that there was no penetration into the iron, but that it was not easily detected and was much smaller than formerly supposed ; he had no doubt that such penetration did take place. Replying to Dr. Borns, he considered that hlr. Cowper-Coles’ conclusions, in the paper referred to, were unsound and unsupported by accurate observation of the facts. The threads of screws are not evenly coated. The deposit is always lighter in the hollows and heavier on the ridges of the thread than the average deposit on any plane surface of the screw. The drums used for sherardising are covered with aTHE SHERARDTSING PROCESS 1.13 thick layer of rust, which prevents the formation of an adherent bright coating, but a heavy scale is formed, which is very troublesome to remove. The distinction drawn by Mr.Cowper-Coles between sherardising and vapour-galvanising he believed to be only a matter of method and plant. The vapour coatings he had produced are oiily obtained at high temperatures and are of no utility whatever-they are neither adherent nor protective. At sherardising temperatures no zinc 'deposit whatever takes place. The com- mercial success of the sherardisiiig process was only obtained after the abandonment of the vapour theory. Sherardised sheet iron can be worked with ease, although not with the ease of tin-plate; the thickness of the coating should be small for thin sheets, and machine pressing should be done at about 15o'C. A sheet having a deposit of + oz. zinc per square foot was readily worked. He suggested that a few calculations from the actual facts might dispel the credence which Drs. Perkin, Harker, and Wilsmore were inclined to give to the vapour theory. He did not suggest that the vapour of zinc plays no part in the process, but that the deposit itself was formed by contact, and not by condensation of vapour. The sense of smell was a very unreliable guide to quantity of material present. There was no doubt that some electrolytic action such as Mr. Ellis suggested took place, since iron oxide was reduced to metallic iron and formed part of the coating.

ISSN:0014-7672    

DOI:10.1039/TF9110600141

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

THERMIC REACTIONS IN VACUO, BY FRANK E. WESTON, RSc., AND H. RUSSELL ELLIS, BSc. PART I. (A Paper read before tltc Fnrnday Society, Tuesday, Afny 31, 1910, Mr. W, MURRAY &IOR~ISOX ill the Chair.) In a paper read before the International Congress of Applied Chemistry, in June, 1909, the authors gave a synopsis of a number of reactions similar in nature to the well-known “ Thermite” reaction, and also an account of experiments carried out with these mixtures i n vacuo. As the paper appears only in abstract in the Proceedings of the Congress, and as the work has been further extended since that time, it was thought advisable to give further details. The authors are only able to find reference to two attempts where similar reactions have bcen tried in vacuo, Pringj’ heated A1 and C in an electric vacuum furnace and found that at 640° C., the melting-point of Al, combination begins to take place and that at 1,400~ C.it proceeds rapidly. From the description of the experiment it appears that the reaction once started does not proceed throughout the mass, whereas, as shown by the authors, a mixture of 4Al and 3C when ignited with a fuse commences to react and proceeds throughout the mass, when the reaction is carried out in a Hessian crucible in the air. Watts and Breckenridget state that they were able partially to reduce silica by means of an alloy of Mg-Ca-A1 when the mixture was heated in a steel vacuum tube to a high temperature ; also that a mixture of MnO, and SiO, when heated with the same alloy in the steel vacuum tube produced an alloy of Si-Mn (22 per cent.of the theoretical yield). It thus appears from these isolated cases that certain “Thermitic ” reactions which take place readily in the air are only partially reproduced in vacuo, and only then when continuously heated to a high temperature. Experimental Methods. The following methods have been used by the authors with varying degrees of success. Method I.-The mixture is placed in a mild steel test-tube, 18.iin. long by 4 in. diameter; the tube is connected both to a mercury manometer and a Topler pump. It is heated in a Fletcher muffle furnace to a temperature of 1,100’ C. The projecting part of the test-tube is cooled by cold water in order to preserve the connection between the steel tube and the glass * J. C. S., 554, p.2108, ~908. Trans. Amev. Elect. SOC., xiii., 1908, p. 103. I44THERMIC REACTIONS tube ; the cementing material consisted of asbestos wool, litharge, and water glass (see I. Mixtures of Mg and A1,0, when heated explosive violence. IN VACUO I45 an intimate mixture of fine Fig. I). in this apparatus reacted with FIG. I. 2 . Mixtures of Mg and CaO when heated in these tubes did not react, but the Mg distilled from the mixtures and condensed in the cool part of the tube. After several trials with different mixtures it was found that the FIG. 2 . method of heating the mixtures was unsatisfactory and so the method was abandoned. Method 2.-The mixture, contained in a porcelain boat, was placed in a Jena glass tube 12 in. long by I in. internal diameter (see Fig.2 ) . A coil of thin platinum wire lying in the mixture was connected to two stout copper wires, which were embedded in glass tubes to the length of about 6 in. by means of146 THERMIC REACTIONS IN VACUO Faraday’s cement: these tubes passed through a rubber cork and could be connected to a source of electric current. The other end of the Jena tube was connected to a mercury manometer and the Topler pump. It was very soon ascertained that the heat produced in the P t wire by the passage of the electric current was insufficient to start the reaction of the mixtures, even when the Pt was raised to its melting-point (1,780~ C.), except FIG. 3 . in a very few cases, and then the tube invariably broke tlirough the sudden great evolution of heat, e.g.- I. Mixtures of Mg and CaO were repelled from the heated Pt wire, and even when the Pt wire was well embedded in the mixture, a channel was immediately formed on heating the wire.2 . Thermite reacted in this apparatus, but invariably broke the tube. 3. All mixtures of A1 and Na,O, tried reacted with separation of metallic Na, but invariably broke the tube, MefJzod 3.--After several trials with small bottles, using Pt wire and Fe wire as the initial source of heating, it was found that certain fuses couldTHERMIC REACTIONS IN VACUO I47 he easily fired by Fe wire heated by an electric current, and finally, the following apparatus was designed and found to give the best results. A wide-mouth bottle of three litres capacity was fitted with a rubber cork carrying three glass tubes ; two of these tubes carried stout copper wires, which were embedded in Faraday cement for a length of 6 in.and at the top end in 2 in. of paraffin wax. The third tube was connected to a Topler pump and a mercury manometer (see Fig. 3). The bottom of the bottle was covered with a layer of sand mixed with small lumps of quicklime-this mixture had been previously heated in a muffle furnace and finally cooled in a desiccator. A Hessian crucible holding from 20 to 30 grams of any of the mixtures was embedded in the sand, and on top of the mixture was placed ‘I gram of the fuse contained in a very thin piece of tissue paper ; on inserting the cork, a loop of very thin iron wire, connected to the two copper leads, just lay embedded in the fuse. The jar was first exhausted with a water-pump, and then finally by a Topler pump; then the fuse was fired by passing an electric current through the iron wire from onc to two seconds. The reactions that took place always did so quietly and steadily. In no case did the crucible crack and a very small amount of the mixture was always left unchanged. The pressure in t h e apparatus, after the reaction was completed, was usually from 50 to 80 mm.- due to combustion of the piece of paper used with the fuse.

ISSN:0014-7672    

DOI:10.1039/TF9110600144

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

THEKMIC REACTIONS I N VACUO. PART 11. EXPERIMENTS WITH MIXTURES OF ALUMINIUM AND SODIUM PEROXIDE. BY FRANK €3. WESTON, BSc., AND H. RUSSELL ELLIS, B.Sc. Mixtures of A1 powder and Na,O, react vigorously and with great heat evolution in the air, when treated with a very small amount of water ; this is especially the case with the mixture made in the proportion of 4Al to 3Na,O, ; one drop of water allowed to fall on a few grams of this mixture produces a violent reaction, accompanied with a vivid flash of light. When such mixtures, however, are treated with water in vacuo no such reaction takes place, but only a slow evolution of gas is observed. In a partialvacuum it is found that the reaction takes place, but always then much quieter than in the air. The experiments described below were carried out on the mixture 4Al and 3Na,O,.The apparatus used was similar to that described in Part I., Fig. 3, excepting that the leading wires were replaced by a tap funnel and the bottles varied from one-half to one litre in capacity. From one to two grams of the mixture was used in each experiment, and was contained in a porcelain crucible. Experirnertt I.-Bottle charged and exhausted on water - pump till manometer read 720 mm. ; barometer reading 750 mm. One drop of water produced a vigorous action : a flash of light, and the bottle smashed. Experiment 2.-New bottle charged and fitted up, but on leaving for two minutes preparatory to exhausting, reaction took place with such violence as to blow out the cork. Experiment 3.-Apparatus as Experiment z ; exhausted on water-pump and then on Topler.One drop of water produced no reaction ; more water (up to ten drops), no reaction ; air admitted, no reaction. Expcrintent 4.-As in Experiment 3, but before exhausted on mercury- pump the mixture reacted with explosive violence as in Experiment 2. Experiment S.-Clean, dry apparatus used and exhausted as in Experi- ment 3. No reaction produced by three drops of water, but on allowing air to enter till manometer stood at 500 mm. reaction started and went smoothly to an end ; pressure reading at end of experiment was 450 mm. on niano- meter. No reaction at first, but after a few minutes a vigorous evolution of gas took place, but no incandescence pro- duced; reaction proceeded thus till manometer stood at 300 mm.From these experiments and a few more of a similar nature not described the following conclusions were arrived at :- I. The apparatus must be perfectly dry and clean before charging with Experiment 6.-As in Experiment 5. 148the mixture. The gases produced in the reaction-steam, H, &c.-must be completely removed before recharging, otherwise the reaction starts appa- rently spontaneously. The sand must be removed and ignited before each experiment. 2. Only one drop of water is required to start the reaction in a norinal manner, but more water prevents the normal reaction and causes a secondary one, which is probably as represented by the following equa- tions :- Na,O, + H,O = 2NaOH + 0 3NaOH + A1 = Na,AlO, + 3H H, + 0 = H,O. 3. Reaction does not start iii absolute "JCICIIO, but the mixture becomes wet, whereas in air the first drop of water does not wet the mixture but causes the reaction ; in dropping water rapidly from a burette into the mixture in air the reaction takes place just as readily as if only one drop is used.Further experiments were carried out in which great care was taken to have reacting bottle aiid materials perfectly dry and clean. The following are a few typical ones :- E.r)eriment ~.--Apparatus as in previous experiments ; barometer, 743 mm. ; manometer, 720 mm. Three drops of water added, no reaction except slight effervescence ; pressure increased till manometer read 600 mni., aiid more water added but without effect. Pressure brought to atmospheric and crucible removed. The mixture was then covered with water, but no reaction took place.The water was poured off and the moist surface removed, and on adding one drop of water to the freshly-exposed dry surface the reaction immediately took place with the normal intensity. E.qheriment 2.-Same as Experiment I , only starting with manometer a t 640 mni. ; no reaction obtained. On removing crucible as in Experiment I the dry powder reacted instantly with one drop of water. Experiment 3.-Same as Experiment 2 , but manometer at joo mm. ; re- action took place on addition of ttvo drops of water; the manometer fell rapidly during the reaction, but filially rose to gro inm., aiid then began to fall very slowly. E.v~crzirzt./zi +-Same as previous, hut manometer at 550 i n i n . ; four drops of water started the reaction.Experimeizt ;.-Same as previous, but a crucible lid used instead of a crucible, and the mixture was well spread out over the lid. Manometer at 550 inin. ; reaction started easily, and manometer rose to 570 mm., and then very slowly fell. Manometer 450 mm. ; no action except slight effervescence ; air admitted till manometer was 570 mm., and a drop of water allowed to fall into a dry part of the powder ; reaction took place immediately. Eq5eriwzent 7. --As in Experiment 6. Manometer 700 inin. ; no reaction ; reaction started when manometer was at 550 mm. From a large number of similar experiments it was found that the reaction could be started when pressure in the apparatus was between 183 mm. and 193 mm. Experiments on other fuses were tried, but it was found that the fuses either would not react at all in air with water, or else reacted with explosive violence. For example :- I . A mixture of Mg and Na,O, exploded when treated with two drops of water . Experimerzt 6.--As in Experiment j. VOL. VI-T61.50 THERMIC REACTIONS IN VACUO 2. A mixture of ZIIg and Na,O, gave no reaction with water except a 3. A mixture of' 33Ig and h'a,O, behaved ;IS in 2. One or two drops of concentrated HCl, however, produced violent ex- plosions with all the mixtures. After many trials it was filially found that the best fuse to use iiz vizcziu was a mixture of 4-41 niid 3Sa,0Z, and that it could best be fired by an iron wire electrically heated. quiet eff ervesceuce.

ISSN:0014-7672    

DOI:10.1039/TF9110600148

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

THERMIC REACTIONS I N VACUO. PART 111. REDUCTION BY MAGNESIUM POWDER OF SILICIC ANHYDRIDE BY H. RUSSELL ELLIS, B.Sc., and FRANK E. WESTON, BSc. Magnesium powder readily reduces silicic anhydride in presence of air without external heating during the reaction ; the products consist of free silicon, one or more of the silicides of magnesium, magnesium oxide, and possibly silicon monoxide. It has been shown previously that in similar actions the air plays an important part, more especially in the initial stages of the reduction. In the experiments now described the interaction has been studied in absence of air, and mixtures containing the following varying quantities of the constituents have been made to react in vacuo and the products examined. ( a ) Equal weights of each. ( b ) In proportions Mg + SiO,.(c) In proportions 2Mg + SiO,. ( d ) In proportions 4Mg + SiO,. The magnesium powder was supplied by Kahlbaum, and was found to contain a small amount of MgCO,, which caused an evolution of gas during the reduction. The silica used was precipitated, and before use was strongly ignited. The fuse used to start the reaction consisted of a mixture of A1 powder The apparatus used was the large bottle of about three litres capacity, fitted and Na,O, in proportion 4Al to 3Na,O,. up as shown in diagram 3, Part I. Analysis of the Mixture resulting from the Reactioii. It was thought that by using excess of Mg it would be possible to use up the whole of the SO,, and by subsequent treatment of the residue with acids to obtain pure silicon.A mixture was therefore made consisting of equal parts of SiO, and Mg ; for the reaction demanded by the equation- zMg + SiO, = 2Mg0 + Si, 48.7 parts Mg are required to 60 parts of SiO,. Twelve reductions werc carried out, each time 20 to 30 grams of the mixture being used ; the contents of the crucible from each reaction were powdered, mixed, and analysed. ‘The products consisted of free Mg, unreduced SO,, two or more silicides of Mg, free Si, MgO, and possibly some silicon monoxide. The complete quantitative analysis of the mixture has not yet been successfully carried out, owing to the difficulties met with. The authors are carrying out experiments to throw light on the formation of ‘‘ monox,” and to find certain reliable tests for this substance. Owing to lack of reliable information with regard to magnesium silicides, no complete estimation of these bodies has yet been made 1.;r152 THERhfIC REACTIONS I N VACUO I .Examiizufion and Esiinzatioit of Silicidc.s.-On treatment with strollg HCl a large amount of spontaneously inflammable gas was evolved, and on washing and drying the residue at rooo C. and then treating with H F more inflammable gas was evolved. It was found impossible to completely decompose the silicides with HCl ; this was either due to ( a ) a protective layer of SiO, formed in tile reduction, ( b ) to a film of SiO, deposited during the action of the acid, (c) to different silicides being present which are iinattacked by HCI. The decomposition by HCl was carried out in an atmosphere of hydrogen or CO, and also i i t vacuo, and in each case a large amount of undecomposcd silicide remained.2 . Decantposition iri CO, OY H.-The weighed powder was placed in t h e middle flask and just covered with water ; H or CO, suitably washed was passed through the apparatus until the whole of the air was expelled; by adjusting the taps shown a very slow stream of gas was then driven through the apparatus during the experiment. Strong HCI was added very slowly, and a vigorous reaction took place, the evolved gases passing through the three test-tubes, which contained AgNO, solution. The AgNO, solution in the first tube becariie black immediately, while if the operation was carefully carried out the solution in the last tube remained clear. The flask was warmed when the reaction had bccome slow, and H or CO, passed until the whole of the SiH, was driven over.‘The deposited silver was dissolved in HNO, after filtering and washing, and estimated by standard NH,CNS. The mixture in the flask still contained undecomposed silicides. If the reactiotis are- (u) Mg,Si + +HCl= SiH, + .91gC1,, ( b ) SiH, + 4AgN03 = Si + +lg + 4HN03, then 108 granis Ag =_ to xcj grams Mg,Si. I n two estimations- I gram mixture uscd and 5.9 C.C. of ,098 N. NH,CNS. I gram mixture used and 6.35 C.C. of -098 N. NH,CNS. corresponding to about 1.2 per cent. hlg,Si.THERMIC REACTIONS I N VACUO I5 3 The estimation in this manner was abandoned after many attempts. 3. Reacfzort between HCI arid the Mixture in Vaczio.--The gas was generated by adding concentrated HC1 very slowly and carefully to the mixture after exhausting the apparatus (see Fig.j) ; the gas evolved was finally driven over into a mercury gas holder by means of freshly boiled distilled water, and was stored out of contact with water. The gas when first made was spontaneously inflammable, but after storing for- some time it lost this power and consisted of almost pure hydrogen, giving no deposit of SiO, w h p burnt. (The mixture, after keeping for some time and decomposed irt uacxo gave 110 spontaneously inflammable gas with HC1, although with H F this was given off rapidly.) The gas evolved was estimated as shown later. The residue in the flask, although it showed no reaction on warming with HC1, readily evolved spontaneously inflammable gas with HF. FIG.j. The authors were not able to estimate the total silicides by this method hut consider that at least two silicides were present. Nethod of Ataalysiiig ilre Gas. A small volume of about 10 C.C. of the gas was transferred very slowly into an explosion pipette containing a known volume of pure oxygen. As soon as a bubble of the gas came into the oxygen a sharp explosion took place and a cloud of SiO, was produced. This continued until the whole of the gas was transferred. A slight black deposit formed on the bulb, probably due to a small amount of the gas being decomposed by the high temperature of the explosion. The diminution in volume was observed and the residual gas sparked, the volume again noted, and finally the gas was mixed with electrolytic gas and again sparked.Knowing the volume of gas used and the contraction after explosion, and assuming that the gas consisted of SiH, and H, only, the amount of each can be calculated. To illustrate the method of calculation an experiment is given fully. 2 grams mixture used. 247-3 C.C. gas at 19' C. and 760 mm.I54 THERMIC REACTIONS IN VACUO Analysis of gas : 14.5 C.C. of gas used and passed into 49'9 c,c. of Contraction after mixing = 20 C.C. Further loss on mixing with electrolytic gas = 1.7. Total loss of oxygen = 7.4 C.C. oxygen . Equation representing reaction between the gases and oxygeii is- x SiH, + yH, + 2x0, +? 0, = x SO, + (2% + y)H,O, where x and y are volumes of SiH, and H,. :. x + y = 14'5 and 2 z + - = 7.4 and y = 14.4 ( 9 :. x = 0'1 The total loss on explosion was 21.7 C.C.For above composition should be 21.9 C.C. A large number of analyses gave the same result, showing the gas to consist of almost pure H mixed with a very small quantity of hydride of silicon, and that, moreover, a very small quantity of this gas is sufficient to inflame the hydrogen when mixed with air or oxygen. This analysis enables the percentage of free Mg to be found ; this was 12-27. The action of the gas on KOH was examined, but the changes in volume were too small to be measured, showing the existence of traces onlv of gases other than hydrogen. Estimation qf Total Silicon in the Mixture. It was thought possible to estimate the Si and SiO, by (a) dissolving i n boiling HC1 and weighing the residue left ; ( b ) dissolving the Si in KOH and weighing the SiO, left ; (c) fusing with KOH and estimating the total silica. Concordant results could not be obtained, and it was found that the iinely divided SiO, dissolved freely in both HCl and KOH, and that any treatment to render SiO, insoluble would also convert some of the Si into SO,. The total silicon was found correctly by fbising with KOH, &c. ; i t was always necessary to evaporate HC1 solutions to dryness and ignite, 21s the finely divided silica readily dissolved. Percentage Si found = 22'124. Estimatioti of' Free Si. The mixture was first heated with concentrated HCI and then witli H F The residue obtained completely combined with CI gas, The percentages of Si found by this method were con- for several days. forming SiC1,. cordant and equalled 12'75. Reszilt oj- A nalyses. I. Percentage of free Mg = 12-27 2. Percentage of free Si = 12-75 3. Percentage of total Si = 22-124 Hence (22'124-12'75) = 9'374 per cent. of Si is combined as SiO,, SiO, and CHEMICAL DEPARTMENT, THE POLYTECHNIC, REGENT STREET, W. silicides.

ISSN:0014-7672    

DOI:10.1039/TF9110600151

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

RE LAT I 0 N S BET \V E EX C R I T I C A L '1'E hf I? E KA'I' Li RE, ROILING-POINT, AND EXPANSION COEFFICIEKT OF PHOSPHORUS PENTACHLORIDE. BY E. B. K. PKIDEAUX, MA., BSc. ('4 Pnfxr read bcfuue the Fcircirln\' Socicty O I L Tiicsdqs, .21a-y 31, 1910, During a recent investigation with the thermal expansion of liquid PCl, the critical temperature was also determined, since it was thought that a comparison of the relations between critical temperature, boiling-point, and expansion with the corresponding relations for other compounds might have some interest in view of the dissociation which takes place in PCI, vapour at low pressures and high temperatures. The method of determining the volumes of liquid PCI, has been already described.hk The expansion measured was that of the liquid confined under the pressure of its own vapour in sealed glass tubes. It was found that the orthobaric volumes could be expressed by the forniula- where Vt is the volume at any temperature not too far removed from the interval 160"-1g0° and V160 the volume at 160~.In the above-mentioned paper it was shown that the temperature at which the pressure of PCI5 vapour = 760 mm. was 160 with a possible error of 19 Mr. W. MURRAY ~IORRISON irz the Chnir-.) vt=vIh TI + 0~00107 (t - I~o)], C ri f ical Temperature. A suitable quantity of PCl, which had been previously purified b y re- distillation in a current of dry chlorine was sealed into glass tubes. 'I'hese were hung close to a thermometer in a gas-heated air-bath provided with mica windows. The temperature was raised gradually, and readings taken ~~~ Experiment.I 2 3 4 4 Meniscus Disappeared. 370° 375 370 378 374 374 Mean 373 - Meniscus Appeared. 376 371 371 Mean 372 -- both with rising and falling temperatures. Two values found for the critical temperature of ethyl alcohol with the same apparatus were 24ga and 251' C. * T r a m Chcirz. SOL, 91, 1715 (1907). I55156 CRITICAL TEMPERATUI<E, BOILING-POINT, The critical temperature of alcohol, as givcn in the tables, ranges between 234' and 259'. There is no noticeable change in the pale yellow colour of the PCl, at higher temperatures, and the appcarance and disappearance of the nieniscus takes place in thc usual way. Each of the experiments tabulated ahovc was carried out with a different tube and a fresli quantity of KI5. Rdairoii behieen Ci ztzcnl Tenrteiatzirc (TR) mid Boiling Poriit (Tb).It was first pointed out by Guldbergr that the ratio T b / T R for all substances is approximately constant =0'64. A survey of the lists of this ratio published by Heilbronn 1 in Laiidolt and Bornstein's tables and the literature generally makes it clear that there is much variation from the mean value, the extremes being 0.55 (N,O) and 0.72 (acetal). In connection with this Young states that in any homologous series the ratios increase with rise of nioleculat weight, and that in the case of isomers Ihe ratios decrease in passing from normal to secondary and secondary to tertiary compounds . Since the boiling-point rises with increase of molecular weight and falls with increase of number of side chains (for isomers), it is clear that in these cases the ratio Th/Tx will increase with rise of boiling-point.Quite generally, also, for all substances the ratio does, on the whole, iiicrease with rise of boiling points, and by taking this fact into consideration average value3 of the ratio can be calculated for particular temperature intervals, which will allow of the prediction of Th wif h a greater probable accuracy than will the total average ratio 0.64. Since individual variations from the mean are con- siderable it is of no advantage to choose small temperature intervals. The following are the mean values of TI,/Tk for a representative and neai I? complete collection of elements arid compounds for which expel 1- nientally determined values of T k are availablc.I. Boiling-point below 273" absolate, twenty-five substances. Mean value : 2 . Boiling-point '73' to 3 7 3 O , forty-six 5ubstances. Meaii- 3. Boiling-poii\t above 37jo, twenty-three 5ubstances. hlean- In the case of €U5- l'liis is to be compred with the mean iatio for substances boiling above 373', i e . , 0.68, and shows that dissociation brings nl-out no abnormality in t h i s relation. In the case of another dissociating suhstance, N,O,, TA = '*) = 0.67, 'L,& 444 wiiich is rather high, mean value for group I above being 0.61.AND EXPANSION COEFFICIENT OF PCl, 157 E.r,baitsion ,!I odu I if s ii i i d Cri fical Tenipera t w e . If V, D, &c., are the specific volumes and densities of a liquid at tempera- tures TI, &c., then it was shown by hIendelejefY':: that the tnodulus of exDansioii is constant for iiiaiiy substances over a wide range of temperature.In a recent paper on the relations between various physical constants, Waldent finds that ( I ) a single constant k expresses accurately the expansion of liquids boiling bclow IOO', including liquefied gases, notably 0, arid N, ; ( 2 ) a single constant K expresses the expansion of high boiling liquids with an accuracy of a few units in the fourth decimal place between oo and p", and often over a much larger range. It was further proved (Zoc. czt.), that if H - COOH, CH,OH * CH,OH, H,SO,, and other liquids known to be associated are excluded, the product kTR is constant for all other elements and com- pounds. I n the following table are values of I;Taas tabulated by Walden, together with others calculated from the data by the formula mentioned above, the temperatures being expressed on the absolute scale.The relation between R calculated from degrees absolute and degrees Centigrade is- C6H6 ... h', ... 0, ... H Br CH,( c.5 H 5jz' ' CroHs ... C6H5NH, ... n.C6HIJ ... CS, ... CCI, ... PCl, ... B 1- ... YiCI, ... SnCl, ... so, ... ? = Expansion Modu I us. *00091 '00392 *oo30h '00 I 34 .oooj9 -00068 j '0~9.53 *00089 *0009 I 6 *oo08g4 *OOIog '001 I6 '000677 '0010 '000879 1; c;ilculated between T," - T2. 2'73 - 3 1 9 O 182 - 204 327 - 426 400 - 450 283 - 393 288 - 308 273 - 323 273 - 353 '73 - 353 273 - 333 273 - 333 68 - 89 68 - 89 273 - 383 243 - 263 I: x T,e. '494 ' 9 9 '498 '482 'SO3 '520 -500 Mean = -494 - I t should perhaps be pointed out here that the relation k x TR = a con- stant (A) appears to be the same as that discovered by Thorpe and Riicker,: Z.".-r 58 C K I T I C A L T E M P E RAT U K E , B 0 I LI N G- PO I N 'I', where D,, &c., are densities at temperatures TI, &c., and A' is another constant.The expression on the right is evidently the same as -f_, where k is the A'k modulus of expansion as defined above. But- k T k = A. :. A (Walden) = 2 (Thorpe and Riicker). A' Thus, taking the case of C6H5CI- k (0' - 132O Cent.) = '000764 :. A = -484 and A' = 2.065. T k = 634 The mean value of A' found for twenty-four substances by Thorpe and Rucker's formula is 1-974, and :, A = 0.506 instead of 0.49, the mean value for the substances tabulated above.The difference is due to a different choice of substances, and perhaps, to a small extcnt, to the choice of a different temperature interval for the calculation of " k." " K " for PCI,, calculated from the expansion between 1 6 0 O and 190°, is- 1.0321 - 1'000 - - 0.0007 I 3, 463 x 1.032 -433 TR = 645 ; kTk = 0'46. One other substance showing marked dissociation for which accurate If V at 0" = 1-000 :I: ,, 21.6 = 1.03196 :. K = 0~00103 values of Tb, Tk, and k are available is N,O,-Tb= 294-6. 'rk = 444 kTR = 0'458. As previously mentioned, Walden has excluded associated liquids from his tabulated values of k . Tk. I n view of the marked deviation in the case of dissociated liquids, it is of some interest to tabulate the values of k . TR for substances known to be associated.The following are typical :- H,O, D at 4" = ~ - m o o ; D at 100' = 0'9584- k = 0-000387 ; TR = 638O. CH;COOH, V at oo = 0.9348 ; V at 120°= 1.0682- k = 0*038102 ; TR = 595.t CH,OH, D at 0" = 0.8142 ; D at 65" = 0.7476 1- k = 0.0~9367 ; TR = 51 I. C,H,OH, D at oo=o.8063 ; D at 131~=0-6796!j - k = 0~0~91 ; TA = 523. * Thorpe, Trans. Chenz. SOC. 37 (1880). t Ramsay and Young, Trans. Chcin. SOC., xlix., 792. Beilstein. 5 Mendelejeff, J . F . Pvalzt. c'licm. 14, 20.AND EXPANSION COEFFICIENT OF PCl, I59 CH3CN, k (Cent. = 0'001334 %:- k (abs.) = o*o,q7S ; T R = 543. C,H,CN, k (Cent.) =0*00125- k (abs.) = 0.0~932 ; Tk = 559. C6H,CN, k (Cent.) = 0.0~847- k (abs.) = 0~0,688 ; TR = 699. Values of kTa fos Abiaormal Liquids. Associated. HZO . . . . . . 0'247 CH,COOH ... 0.482 CH30H . . . . . . 0.478 C,H,OH . . . . . . 0.476 . . . . . . CH,CN 0'531 C,H,CN . . . . . . 0.521 C6H,CN . . . . . . 0.480 Dissociated. NZO, . . . . . . 0.46 I PCl, . . . . . . . . . 0.46 It is evident that such typically associated liquids as the alcohols and fatty acids do not give abnormal values. On the other hand the product for the lower iiitrils is distinctly high, becoining normal again as the molecular weight of the hydrocarbon radical increases. So far as the evidence goes at present, then, it may be said that this relation holds with an accuracy of 2 per cent. for all normal substances considered except SnCI, and for many associated liquids, while for certain associated liquids the value tends to be high, and for dissociated liquids is low. * Walden, hi. cif.

ISSN:0014-7672    

DOI:10.1039/TF9110600155

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

NOTE ON THE COMPOSITION OF EUTECTIC MIXTURES. BY CECIL H. DESCH, DSc., I-’H.D. (A Pnpcr rend bcforc the Faraday SocicfJ,, Tucsdaj; Maq’ 31, 1910, Mr. IV. MURRAY MORRISON ipL thc Chair.) During the passage of a mixture of two substances from the liquid to the solid state the temperature either decreases progressively or remains constant. The latter condition may present itself in the freezing of :- (a) One oLthe pure components ; (b) A definite compound of the components ; (c) A solid solution occurring at a maximum or minimum 011 the freezing-point curve, or at some other point at which the common tangent to the liquidus and solidus curves is parallel to the axis of concentration ; (d) A eutectic mixture. The frequent occurrence of the last case, even amongst the mixtures first examined by early workers in this subject, caused eutectic mixtures to be regarded as chemical compounds.The constancy of freezing-point, the absence of liquation, and the apparent homogeneity of the solidified mass are all tests frequently employed to determine the individuality of a sub- stance. Hence came the term “ cryohydrate,” applied to a mixture of a salt with water having a minimum freezing-point. Levol’s alloy of silver and copper, formerly regarded as a compound, Ag,Cu,, is another instance of a eutectic mixture assumed to be homogeneous. The assignment of a chemical formula to a mixture occupying a minimum 011 the freezing-point curve is even now occasionally met with, and a number of erroneous formulae have in this wag gained entrance to chemical literature.In the case of cryohydratcs the ratio of water to salt in the eutectic mixture is usually large, and it is therefore easy to find a formula, containing whole numbers of molecules, which will express the composition with sufficient accuracy. Even in alloys and mixtures of salts or organic compounds the approximation to a simple formula is often remarkably close, although such cases are generally regarded as simple coincidences. The present paper is an examination and discussion of certain relations which have been put forward as existing between the molecular composition and freezing-point of eutectic mixtures and the molecular weights and freezing-points of their components. The direct application of the laws of depression of freezing-point in dilute solutions to eutectic mixtures is not legitimate, as the assumptions which underlie those laws are valid only over a small range of concentration.We occasionally find, however, that the molecular depression remains sensibly constant throughout the entire range, so that the freezing-point curve, drawn I00THE COMPOSITION OF EUTEC'I'IC MIXTURES 161 with molecular concentrations as abscissz, consists of two straight lines. few examples are collected in Table I. A TABLE I. System. Gold-t hallimn . . . . . . . . . . . . Sodium ni tra te-lithium .nitrate . . . . . . . . . Sodium nitrate-thallium nitrate . . . . . . . . . Cadmium chloride-cadmium iodide . . . . . Phenol-diphenvlamine . . . . . . . . . . . . . . . Ethylene bromGide-acetic acid . . . . . . .. . Ob5errer. I€. I x y i n , %~.i/.~t.h. nr/or& Chwi., 1905, 45, 31. H . K. Carveth 7 Physical Ciicirr., 1898 2 209. C. van Eyk, Zc&lL. fliysikal. Clrcrra., 18;0,'5, 451 ; K. ZiTacken, Cerrfr. Min., 1907, 301. 1. C. Philip, Trans. Chew. Soc.. 1903, 83, 814. 11. Dahnis,.4/in. PlijJsik, 1895 [ii.], 54, 486. 1893, 11, 328. By adopting a modified method of expressing the depression of freezing- point in concentrated solutions, F. Flawitzky ::: has obtained a rule which may be applied in certain cases to predict the composition of the eutectic mixture. If M, and MI are the molecular weights of the components, to and t, their freezing-points, i, the freezing-point of the eutectic, we may write the total depressions f, - t, = To and t, - f, = T,, and the composition of the eutectic may be expressed by the forniula- M, + IZOMO or &Io + nlM1 where 12, atid 1 2 , are the numbers of molecules, and 1 1 , = I .'10 the ratio of the molecular percentages of the components. depressions thus calculated arc equal, so t!iat- Evidently 12, is Flawitzky finds that for a large number of binary mixtures the two /z~M,T~ = 11 I M ,T , . Remembering that 1 2 , = 1 / r z 0 , we have- from which- n ~ N , T , = RI,T,, This enables us to calculate the molecular ratio of the two components in t h e eutectic mixture. In Table 11. a few of the results obtained by Flawitzky are reproduced, and some. further tests of the method are given in Table III., the sources of the data employed being given in the last column. In the case of metallic alloys, the molecular weights of the metals are taken as being equal to their atomic weights.TABLE 11. System. j 11,. 1 III. 1 'r,. 1 TI, 1 $2, caIc. 1 no found. Water-acetic acid . . . . . . Water-formic acid . . . . . . Mercuric bromide-mercuric iodide . . . . . . . . . . . . Thallium-thallium mercuride.. . Thallium-cadmium . . . . . . I I I I I I 8 I 8 * 7. Russ. PILJLS. Clccrii. SOC., 1905, 37, 862 ; Bull. Sac. cltim., 1906 [iii.], 35, 478.System. LiCl - KCl . . . . . . . . . LiCl - RbCl . . . . . . . . . RbCl - NaCl . . . . . . . . . CsCl - NaCl . . . . . . . . . Phenol-diphenylamine . . . . . . $-toluidine-a-nitrophenol ... Naphthalene-diphenylamine ... Naphthalene-thymol . . . . . . Naphthalene-phenanthrene ... Naphthalene-a-biomocamphor ... ... ...... ... 6 . . ... ... ... ... ... ... ... ... ... ... ... ... ... ... 3 : 4 : 5-tribromotoluene-2 : 4 : 5-tribromoto- luene . . . . . . . . . . . . . . . . . . Gold-thallium ... . . . . . . . . . Bf 0' 42'5 42'5 I2 I 168.3 94 107 I 28 128 I 28 I28 - 197.2 T,. 264 302 I 86 I54 22'4 27'7 20.5 19'5 31 39 27 168 TI. 440 4'4 279 327 34'5 82.5 47'5 49'5 51 36 50 932 'lo calc. 1'71 1-97 0.85 0.85 I *66 1'1 j 1-75 1.72 1.51 1.29 "34 2.30 found 1-50 1.27 1-22 1 '93 2'12 I *08 I *78 2'22 1'22 1-04 1'50 2.70 'Reference. I S. F. Schemtschuschny and F. Rambach, Z ~ i f s c h . nnorg. Chem., 1910, 65, 403. r I 1 J. C. Philip, Trans. Chem. SOL, 1903, 83, 814. 1 M. Roloff, Zeitsch. fihysikal. Chem., 1895, 17, 325. i A. Miolati, Zeitsch. Plzysikal. Chem., 1892, 9, 649. E. Paternb and G.Ampola, Gazzetta, 1897, 27, 481. F. M. Jaeger, Dissert., Leyden, 1903. M. Levin, Zeitsch. anorg. Chem.., 1605, 45, 31.THE COh/lPOSITIOIV OF EUTEC'I'IC MIXI'UKES 163 The agreement between the observed and calculated values of no is fre- quently imperfect, and in many cases the eutectic composition thus predicted differs very widely from that determined experimentally. Flawitzky has sought to explain the deviations by assuming association of one or both the components in solution, the formula becomiiig- if p and q are the factors of association of the molecules M, and M, respec- tively. By assuming arbitrary values for the association, he obtains some show of agreement with the observed values of no in a number of cases, but such a procedure is entirely unjustified, and the theoretical conclusions as to association in binary mixtures which Flawitzky draws from his results are not in accordance with what is known as to the molecular weights of the substances in question.When the eutectic point lies near to one end of the diagram, that is, when no is 2.5 or more, Flawitzky's rule breaks down altogether. It has been shown by I. Schroder ': and H. Le Chatelier 1- that an expres- sion for the whole of a branch of a freezing-point curve may be obtained by making the assumption that the heat of fusion of either component is un- aflected by the presence of the other. The expression in its simplest form, if ?' is the freezing-point of a mixture, To that of the pure component, and x the concentration of that component in the mixture, that is, its molecular percentage + roo, is :- T O 2'l- 1 - * 9 'I' = The application of this formula to the prediction of the form of freezing- point curves has been very fully discussed by Roozeboom, 1 and the principal causes of the experimental deviations from that form have been explained, A discussion of metallic alloys from this standpoint by D.Mazzotto 5 has shown that the introduction of two corrections brings about a very close correspondence between the theoretical and the experimental curves. One of these depends on the heat developed or absorbed on mixing the two fused substances, and the other on the degree of polymerisation of the components in the fused mixture. Both of these magnitudes are assumed to be zero in the simplified formula given above.The correction for the heat of mixture consists in multiplying the expres- A sion for T by van Laar's factor, I + - A is derived from the constants in van der Waals' equation, but this theoretical derivation is incapable of application, and it is necessary to determine A experimentally by direct measurement of the heat of mixture. This has been performed for a small number of pairs of metals by W. Spring I] and D. Mazzotto,lT with the result that the mixing in the liquid state of tin with lead, zinc, bismuth, or mercury, or of lead with mercury, is accompanicd by absorption of heat, whilst lead and bismuth $2. * Zcitscli. phjsikal. Chnit., 1893, 11, 449. t Conrpt. rend., 1894, 118, 638. Heterogae GLeicl~gewrclttc, vol. 3, 5 4, v. $ Nziovo Cinz., 1907 [v.], 13, 80 ; 1908 [v.], 15, 401.k . Aknd. Wcteizscli. Anrsfcrdmn, 1903, 5, 424 ; 6, 21. ~l=BuL/. A C L I ~ . Sci. roy. Kclg., 1886 [iii.], XI, 35;. C; Rcird. last. Lonibwdi, 1888 [ii.], 18, 165 ; Mcm. Irtsf. Louthai-di, 1891, 16, I. See also J. J. van Laar, Proc.164 THE COMPOSITION OF EUTECTIC MIXTURES develop heat on mixing. The fact that ;I lowering of temperature codd be produced by mixing two amalgams had been observed as early as 1824.;:: Other data for the application of this correction are so far' lacking, and the experimental difficulties in the case of metals of high melting-point are evidently very considerable. Mixtures of orgatiic substances, however, afford a better field for measurements of this kind. If polymerisation of the molecides takes place, it is neccssary to replace ,2: by x,, where- I1.V .Y, = )IS + I - .T' IZ being the factor of polymerisation.The latter number may be determilied from the initial portions of the freezing-point curves, that is, from measure- ments of the depression of freezing-point i n dilute solutions. In the case of alloys, closely grouped measurements, of a very high degree of accuracy, have been made by Heycock and Ncville,t and their data were employed by Mazz ot to. If we write A~ for the value assumcrl hy 1 when .I' is I-eplaced by . Y ~ , the full expression beconies- in which form it can be applied for the coiistructioii of fails, however, to give results coinciding closely with inentally if either compounds or solid solutions arc theoretical curves. It those obtained experi- formed by the com- ponents, aiid this point is of importance when the attempt is made to determine the eutcctic point from the intersection of two such theoretical curves .In a single instance the application of the simpler formula of Le Chatelier leads to a freezingpoint diagram coinciding very closely with the real one, in spite of the fact that solid solutions are formed to a limited extent. This is in the system copper-silver, studied by Heycock and Neville, [ and it is uncer- tain how far this may be regarded as typical. The application of the formula for the ideal curve evidently demands a previous knowledge of the thermal behaviour of the components, so that it is of little value for the prediction of eutectic points, but may be employed i n the study of the molecular condition in solution.The discovery of a regularity in the composition of eutectic mixtures in 5eries in which compounds are formed was ;tniiounced by A. Gorboff in a communication to Section IIb. of the International Congress of Applied Chemistry in 1909, aiid the paper has since been published.$ Gorboffs rule applies only to eutectics one constituent phasc of which is a compound liielting without previous decomposition, and therefore indicated by a maxi- mum in the freezing-point curve. From the cxainination of a number of such mixtures, the conclusion is drawn that the two constituent phases of the eutectic are present in such quantities that one component is equally dis- tributed between them. For example, the freezing-point curve of mixtures of nitric acid and water j l has two maxima, corresponding with the two hydrates HN0,,3H20 and HX0,,H20.The eutectic formed by these * J. W. Uijbereiner, Sclawcigg. J., 1824, 42, 182. t Tram. Cheni. Soc., 1889-97. $ Phil. Trans., 1897, I@A, 25. 9: 7. Russ. Phys. Chern. Soc., 1909, 41, 1241. ]I p. W. Kiister and R. Kremann, Zcttsch. nuovg. Chct11., 1904, 41, I .TARLEI IV. System. * A1 - JIg Bi - hlg Hg - T1 ... hlg - Sb .,. * Big - Pb ... Jlg - SI ... b l g - Sll ... Ag - M g A1 - xu ... A1 - c u ... AU - hlb( Au - Sn ... AU - ZII ... Cd - CU ... * Cd - Na ... * CS - Hg ... CLI - hlg ... Hg - h'a... Na - Sn ... * Na - I'b ... Na - T1 ... Pb - I'd ... ' hfn - P ... Ni - Si ... * cu - r' ... . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phenol - b-to!uidine ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... l'henol - 'a-n.~phthylattiine ... Catechol - fi-toluidine ... Guaicol - picric acid . . . . . . ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... Composition of Eutectic. Calculated. __ 'ound. __ 37'6 70'2 82 60 81 88.5 96 88.5 17'3 65 0 59.8 78 5 18 63 292 94 32'9 55 75'7 82.6 21.5 45 15 14.8 38.3 24'2 41'5 62.7 7'2 75'4 40' 5 09 23'5 68 5 2 83.6 J 21 - ZZ.2 64 Reference.G. Grube, Zeitsch. ariorg. Chcui., Iwj, 45, 225. G. Grube, ibid., 1906, 49, 72. N. S. Kurnakoff and X. A. Pushin, ibid., 1902. 30, 86. G. Grube,ibid., 1905, 44, 117. G. Grube, ibid., 1906, 49, 72. R. Vogel, ibid., 1909, 61, 46. N. S. Kurnakoff and N. J. Stepanoff, ibid., 1905, 46, 177. S. I?. Schemtschuschny, Zeitsclt. aitoig. Cltem., 1906, 49, 400. C. T. Heycock and I;. H. Neville, Phil. Traris., r q o , I Q ~ A , 201 H. C. H. Carpenter and C. A. Edwards, Proc. Iizst. Mcclt. Etig., 1907, 57. R. Vogel, ibid., 1905, 43, 60. R. Vogel, ibid., 1906,48, 319. R. Sahmen, ibid., 1ywh,49, 301. N. S. Kurnakoff and A. N. Kusnctzoii. ihid., 1907, 52, 173. N. S. Kurnakoff and G. U. Schukoirsky, Zeitsch. arzovg Chein., 19a7, 52, 416.G. G Urazoff, y. Rim. Plryys. Clicnz. SW., 1 y 7 , 39, 1566. G. G. Urazoff, Zgifsch. ~ / L o Y ~ . C~ICWZ., 1909, 64, 375. A. Schuller, Zeitscli. aiiorg. Cheiii., 1904, 40, 385. C. H. hlathewson, ibid., 1906, 50, 171. C. H. Nathewson, ibitl., Ip6,50, 171. N. S. Kurnakoff and N. A. Pushin, ibitl., 1902, 30, 86. li. A. Pushin and N. Paschsky, 3. h'rra. Pliys. Cliem. Soc., 1908, 40, 826. 15. Heyn and 0. Bauer, Zeitsclr. iiiiorg. Cheni., 1907, 52, 129. S P. Schemtschuschny and N. N. Efremott, y. K i m Phyb. Cliem. Sor., 1907, 39,777. \V. Guertler and G. Tamniann, Zeitsch. aiiovg. Clzenz., I*, 49, 93. 1 J. C. Philip, Trniis. Chefn. Soc, 1903, 83, 814. 1 J C. Philip andS. H. Smith, ibid., Iq35, 87, 1735.166 T H E COMPOSITION OF EUTECTIC MIXTURES hydrates has the composition .+HN03,6H,0, or H N0,,3H20 + 3( HNO,,H,O), the water being equally distributed between the two solid phases.In general, if two components, A and B, form the compounds AB, and AB,, the eutectic mixture of A and AB, has the composition A + AB,, or A,B,, and that of the two compounds, if B is the component which is equally dis- tributed, has the composition yAB,, + .TAB,, or A, + ,,R1.TY. In solutions of acids or salts in water, it is the water which is thus equally distributed between the two phases in the eutectics or cryohydrates, whilst in alloys it is the more electropositive metal. Thus in the system sulphur trioxide-water, the hydrates formed have the formulz S03,5H,0 ; S0,,3H,O ; S03,2H,0 ; S03,H,0 ; and 2SO,,H,O. 'lhe eutectics or cryohydrates correspond very closely in composition with the formulze' : ::: SO,,IOH,O, or SHgO + SO,,~HZO 7S03.20H,0, or ?(S0,,3H20) + 5(S03,2H,0) t 5S03,r2H.0, or 2(S03,3H,0) + 3(S0,,2H20) 3S0,,qH20, or 3SO3,2H,O, or SO,,H,O + 2SO,,H,O SO,,? H,O + 2(SO,,H,O) 3SO,,H,O, or 2SO,,H,O + SO, the water being equally distributed between the two solid phases in each case, with the exception of the last.In Table IV. Gorboff's rule is tested by application to a number of eutectic mixtures, principally of a metallic character. Examples which have appeared in Gorboff's paper nre marked with a :::. The discrepancies are sometimes considerable, but ;i good agreement is obtained in many cases, and the rule may prove of some value 111 predicting the form of the freezing- point diagram. In one instance, that of the alloys of sodium and cadmium, the rule affords a means of deciding between the different results obtained in two investigations. Whilst Kurnaltoff and Kusnctzoff (Zoc. cit., Table IV.) found a compound having the formula NaCd6, Mathewson obtained results pointing to NaCd, as the coriipohition of the compound. Gorboff remarks that the eutectic, having a composition approaching very closely to NaCd,,, or NaCd6 + 6Cd, the evidence for the first of these formuh is strengthened. The agreement is not good iii the case of most mixtures of organic sub- stances, although an examination of the mixtures studied by Philip (Zoc. cit., Table IV.) suggests that the more hasic substance is clistributed almost equally between the twc phases prcsent i n the eutectic. There is no obvious theoretical reason why such a rule should prevail, and some of Gorboff's statements, as, for example, that the space-lattices of substances crystallising together to form a eutectic must be identical or a t least similar, are certainly erroneous. As an empirical rule, however, this relation may prove to be of assjstance in determining the equilibrium of binary systems. M ETALLU RGI CB L DEPART XI E NT, UNIVERSITY OF GL.~SCO~V. * Landolt-Biirnstein Tahellen, 3rd edn., p. 567. t This eutectic is obtained when the formation of the hydrate S0,,3H,O is suppressed by undercooling. Zeitsclz. nnorg. Chew., 1906, 50, 171.

ISSN:0014-7672    

DOI:10.1039/TF9110600160

出版商:RSC    

年代:1911

数据来源: RSC

摘要:

THE DIGBY-BIGGS CONDUCTIVITY TUBE. A PORTABLE APPARATUS FOR THE RAPID DETERMINA- TION O F THE CONDUCTIVITY OF WATER AND SALINE SOLUTIONS. (Exhibited before the Fnraday Socicfy O I L Tiiesdaj), May 31, 1910, by Mr. F. E. POLLARD, F.I.C. ; Mr. W. MURRAY MORRISON iiz the Chair.) The methods usually employed for ascertaining the conductivity or the resistance of liquids necessitate the use of delicate apparatus, involve calcula- tions, and can only be carried out satisfactorily in the laboratory. The apparatus devised by Mr. W. Pollard Digby and Mr. C. W. V. Biggs is portable, can be used in the boiler-house or in the open air as well as in the physical laboratory, and the “conductance meter ”-which is simply a megger furnished with a scale graduated in reciprocal megohms-gives direct readings of one-tenth of the specific conductivity.This apparatus consists of a glass U-tube having a sectional area of one- tenth the distance between its annular platinum electrodes, the form and arrangement of which are designed to minimise the effects of polarisation and similar difficulties. The inlet from the filling-funnel is at the lowest part of the bend, as also is the outlet for emptying the contents of the tube; whilst near the extremities and above the electrodes overflow tubes are attached to permit a constant flow of liquid through the tube. The limbs are closed by brass covers connected to the terminal screws, and through these covers pass stout glass-covered brass rods from which the electrodes are suspended by platinum wires, and the whole arrangement is mounted on a strong wooden stand carrying a thermometer.The tube having been filled with the water or other liquid, and the tem- perature noted, the twin lead from the conductance meter is joined up to the terminals, the handle of the generator is turned, and a reading is obtained in a few seconds. For solutions containing not more than I per cent. of dissolved saline matter the correction for variation of temperature between 5 O C. and 38“ C. is approximately 2’19 per cent. per I’ C. At higher temperatures the accuracy of the correction is impaired by evolution of dissolved gases and by other complications. I t is well known that very small proportions of impurities in distilled water cause considerable alterations in the conductivity, e.g., Ganot’s Plzysics, tenth edition, states that “ standing in the air for five hours doubles it ; the addition of a millionth part of sulphuric acid-that is, a drop in about seventeen gallons-increases the conductivity tenfold.” The following table shows the specific conductances of extremely dilute solutions of sodium chloride, sodium carbonate, and calcium sulphate as 167168 THE DIGBY-BIGGS CONDUCTIVITY TUBE determined with an earlier pattern of tubc at temperatures from 18' to 20.50 C.:- Parts per ~ I i l l i o n . ( i n Aq. drst.) 0 I 2 4 I 0 2 0 50 Ka,C03. &SO4. 2-86 5 6.9 8-33 62 I 2.83 21.3 34'5 74.1 West London tap water contains about 18.5 grains per gallon, or about 26-5 parts per IOO,OOO, of total solids, and its specific conductivity varies from 355 to 400 reciprocal megohms, being generally about 380 or 390 ; yet the addition of so slight a quantity as 0.1 per cent.of this water to good distilled water raises the conductance very perceptibly. The results givcii below were obtained from mixtures of \Vest London tap water with distilled water having a conductivity of 2-5 reciprocal inegohms at 20' C. :- Percentage of tap water ... o 0.05 0-1 0-2 0.5 0.75 1.0 Conductance (J) ... ... 2.5 3-0 3.3 4-2 7-1 8.3 9.3 The I per cent. mixture contained about 2.65 parts of dissolved saline matter per million. For many purposes-particularly in the operation of steam plant-it is a matter of importance that variation in the water supply in respect of the quantity of dissolved saline matter should be detected. A slight leak in a surface condenser will cause cumulative addition of dissolved solids to the boiler water which may remain undiscovered until priming occurs ; but the conductivity tube constitutes a means of checking this at a much earlier stage-long before the use of a salinometer would be possible.By testing the hot-well water under suitable conditions of load, priming is readily dis- tinguished from condenser leakage ; and the cxtent of either is definitely ascertained by comparison of the hot-well samples with mixtures of varying proportions of the boiler water and distilled water. The conductivity tube affords a convenient and efficient control in the Clarke process of softening hard water. As the lime is added the conductance diminishes, reaching a niininium when just sufficient calcium hydrate has been used, and rises again with excess.I n the working of oil eliminating plants in which chemicals are used, the conductance tester serves as a safeguard against the accidental or wilful employment of too great quantities of saline substances and consequent con- tamination of the boiler feed water ; and in numerous operations this apparatus provides the engineer and others with simple, rapid, and accurate means of control. Neither the discovery of the conductivity tube nor of any new law relating to the properties of solutions is claimed, but merely the devising of a form of tube suitable for use in the ordinary routine of engineering work as well as for scientific investigations. Fig. I is a sketch of the tube.168 THE DIGBY-BIGGS CONDUCTIVITY TUBE determined with an earlier pattern of tubc at temperatures from 18' to 20.50 C. :- Parts per ~ I i l l i o n .( i n Aq. drst.) 0 I 2 4 I 0 2 0 50 Ka,C03. &SO4. 2-86 5 6.9 8-33 62 I 2.83 21.3 34'5 74.1 West London tap water contains about 18.5 grains per gallon, or about 26-5 parts per IOO,OOO, of total solids, and its specific conductivity varies from 355 to 400 reciprocal megohms, being generally about 380 or 390 ; yet the addition of so slight a quantity as 0.1 per cent. of this water to good distilled water raises the conductance very perceptibly. The results givcii below were obtained from mixtures of \Vest London tap water with distilled water having a conductivity of 2-5 reciprocal inegohms at 20' C. :- Percentage of tap water ...o 0.05 0-1 0-2 0.5 0.75 1.0 Conductance (J) ... ... 2.5 3-0 3.3 4-2 7-1 8.3 9.3 The I per cent. mixture contained about 2.65 parts of dissolved saline matter per million. For many purposes-particularly in the operation of steam plant-it is a matter of importance that variation in the water supply in respect of the quantity of dissolved saline matter should be detected. A slight leak in a surface condenser will cause cumulative addition of dissolved solids to the boiler water which may remain undiscovered until priming occurs ; but the conductivity tube constitutes a means of checking this at a much earlier stage-long before the use of a salinometer would be possible. By testing the hot-well water under suitable conditions of load, priming is readily dis- tinguished from condenser leakage ; and the cxtent of either is definitely ascertained by comparison of the hot-well samples with mixtures of varying proportions of the boiler water and distilled water. The conductivity tube affords a convenient and efficient control in the Clarke process of softening hard water.As the lime is added the conductance diminishes, reaching a niininium when just sufficient calcium hydrate has been used, and rises again with excess. I n the working of oil eliminating plants in which chemicals are used, the conductance tester serves as a safeguard against the accidental or wilful employment of too great quantities of saline substances and consequent con- tamination of the boiler feed water ; and in numerous operations this apparatus provides the engineer and others with simple, rapid, and accurate means of control.Neither the discovery of the conductivity tube nor of any new law relating to the properties of solutions is claimed, but merely the devising of a form of tube suitable for use in the ordinary routine of engineering work as well as for scientific investigations. Fig. I is a sketch of the tube.170 T H E DIGBY-BIGGS CONDUCTIVITY TUBE APPARATUS FOR THE DETERMINATION OF THE SPECIFIC RESISTANCE O F OIL. This apparatus consists primarily of two metal discs of even surface, each having an area of 100 sq. cm., arranged perfectly parallel to, but insulated from one another. The distance between these is regulated by a micronieter head on the spindle of the instrument, which spindle is hollow and of 811 insulating material such as ebonite.Through the spindle is carried the lead of the bottom plate, a separate connection being taken to the top plate. A suitable quantity of oil having been placed in a flat glass dish, the test plates are opened to a distance of, say, IOO mils, and are then immersed i n the oil, care being taken to avoid the presence of any air bubbles or foreign matter between the plates. The leads are connected to a megger or other form of ohm-meter. The handle operating the generator of the megger is then turned and the reading noted. Successive readings are also taken at gradually reduced distances between the plates until the megger needle oscillates instead of giving a steady reading or until sparking takes place, when the .series of readings is discontinued.By reason of the fact that a series of such readings, if plotted, gives a curve of hyperbolic shape instead of a straight line (this resembling the curve showing the relation between pressure and distance upon dielectric test of such oil), the mean specific resistance is calculated as the mean of a series of readings. The readings must always be taken in a descending order of value for insulation and of thickness of the film of oil, as when once sparking takes place the specific resistance of the oil is altered and lower values result. The specific resistance per cubic centimctk for aiiy given distance then equals : observed rcsistance x area of plates in sq. cm. x 3()3*7 distance between plates in inils. With a megger having a range of 1,000 megohms at 500 volts a really first-class oil for switch or transformer use has a specific resistance well over 6,500,000 megohms. In fact, several samples recently examined gave readings at, or close to, ' I infinity," down to a distance of about 7 mils. The variation of specific resistance with temperature depends upon the grade of the oil ; and it is desirable that the determination should always be carried out at a definite teinperature, such as 60" F. Fig. a is a half-sectional drawing of the apparatus.

ISSN:0014-7672    

DOI:10.1039/TF9110600167

出版商:RSC    

年代:1911

数据来源: RSC