Researches on Vanadium.-Part II .
By HENRY E. ROSCOE , B.A. , Ph. D. , F.R.S. Received June 16 , -Read June 17 , 1869 .
On the Chlorides of Vanadium and Metallic Vanadium .
IN the first part of these researches* I stated that the nitride of vanadium forms the point from which the study of the vanadium compounds , properly so called , must be commenced .
I now proceed to describe the composition and mode of preparation of the compounds of the metal with chlorine , and also the properties of the metal itself .
The process for preparing vanadium mononitride which I described in the last communication , was that proposed by BERZELIUS for obtaining the substance which he conceived to be metal .
It consists in the action of dry ammoniacal gas upon vanadyl trichloride ( VO Cl3 ) .
Owing , however , to the difficulty of properly regulating the very violent action which occurs when the two bodies come into contact , and the consequent loss of trichloride , this method was unsuited for the preparation of the nitride in quantity .
Seeking for an easier and more economical method , I found that if the black residue left on calcining ammonium metavanadate in the air ( probably an oxynitride ) be heated to whiteness for a sufficiently long time in a current of dry ammoniacal gas , pure vanadium mononitride is left behind .
A portion of ammonium-salt was placed in a platinum boat inside a porcelain tube , through which a stream of dry ammonia was passed .
The tube was then heated to whiteness for three days .
Analysis 1.-Determination of Vanadium .
Weight of Nitride employed ... 0-1705 .
Constant weight of vanadium pentoxide obtained by roasting ... .
0'2415 .
Hence the nitride contained 79 6 per cent. of vanadium .
The formula requires 78*6 per cent. Determination of Nitrogen according to SIMPSON'S method .
Weight of nitride taken ... 0-2070 .
Observed vol. of nitrogen Pressure .
Temp. Vol. at 0 ?
and 760 ( measured moist ) .
millims. o millims. 111-2 vol. 746-2 17'0 100-5 ( 1 vol. =-0331 cub. centim. ) This corresponds to 20'2 per cent. of nitrogen .
Vanadium mononitride contains 21-4 per cent. The substance used in the following analysis was prepared by heating about 3 to 4 grms. of the ammonia-salt in a platinum tube to whiteness in a current of dry ammonia .
The tube was imbedded in a chalk crucible , and the central portions were heated to whiteness by the blast of a powerful blowpipe until no alteration of weight was observed .
In this way , after a long and tedious series of operations , no less than 40 grms. of pure nitride ( see analysis No. 2 ) was prepared .
Analysis No. 2.-Determination of Nitrogen .
Weight of nitride taken. .
0-2250 .
Observed vol. of nitrogen Pressure .
Temp. Vol. at 0 ?
C. and 760 ( measured moist ) .
millims. 0 millims. 123*3 740'5 21-5 108'5 ( 1 vol. =0-331 cub. centim. ) This corresponds to 20'1 per cent. of nitrogen .
A third and still more simple mode of obtaining the nitride is that of exposing the trioxide to the action of ammonia gas at a white heat .
It is a remarkable fact , and one of which we know of but two other examples , that in presence of nitrogen and hydrogen the whole oxygen of an oxide is displaced by nitrogen .
THE CHLORIDES OF VANADIUM .
I have succeeded in preparing three chlorides of vanadium , viz. ( 1 ) vanadium tetrachloride , VC14 , ( 2 ) vanadium trichloride , VC13 , ( 3 ) vanadium dichloride , VC12 .
( 1 ) Vanadium Tetrachloride , VCl4 , molecular weight =193'3 , vapour-density =-96-6 ( H=-1).-This chloride is formed as a dark reddish-brown volatile liquid when metallic vanadium or the mononitride is burnt in excess of chlorine .
Method 1.--The first method adopted for the preparation of the tetrachloride was by passing dry chlorine gas over the mononitride heated to redness ; the whole of the nitride volatilizes , and a reddish-brown liquid is obtained .
In one operation 16-96 grms. of the mononitride yielded , on treatment with excess of chlorine , 44-3 grms. of the crude tetrachloride , the theoretical yield being 50'1 grms. Owing to the inevitable presence of traces of oxide in the nitride , the first few drops of the distillate consisted of the more volatile light-yellow oxytrichloride , and this portion was collected separately before much of the dark tetrachloride had come over .
The crude product thus obtained was first saturated with chlorine , and then distilled upwards for some hours in a current of dry carbonic acid .
On fractionating , the liquid was found to begin to boil at 148 ?
, slowly rising to 154 ?
, between which points the whole of the liquid came over .
A considerable quantity of a solid mass of peachblossom-coloured crystals , consisting of vanadium trichloride , remains behind in the bulb after each distillation .
A second preparation of 50 grms. of the pure tetrachloride gave a constant boiling-point at 152 ?
--154 ? .
Method 2.-When vanadium oxytrichloride is prepared according to BERZELIUS'S reaction by passing dry chlorine over a mixture of the trioxide and charcoal heated to redness , the distillate possesses a port-wine colour , and it is only by frequent rectifications over sodium that the canary-yellow colour of the pure oxychloride is obtained .
This coloration , though ascribed by SCHAFARIK to the presence of traces of vanadic acid , is really due , as I have formerly pointed out , to the presence of a dark-coloured vanadium chloride produced in the reaction and decomposed on rectification over sodium .
This observation , coupled with the fact already made use of to demonstrate the existence of oxygen in the oxytrichloride , viz. that this substance yields up oxygen to red-hot charcoal , points the way to the second method for the preparation of the tetrachloride .
This consists in passing the vapour of the oxytrichloride , together with an excess of dry chlorine , slowly over a long column of pure sugar-charcoal heated to dull redness .
This operation has to be repeated four or five times before the last traces of oxygen are withdrawn .
Even after one distillation the product attains a dark reddish-brown tint , and boils at 152 ?
; but it still contains 1 per cent. of oxygen , and needs three or four very slow repetitions of the process in order to obtain it perfectly pure .
The solid trichloride is always found in the bulb from which the tetrachloride is distilled ; but this , if carefully heated , slowly burns away in the excess of chlorine .
The portions of substance used for analyses 1 , 2 , and 3 were prepared from nitride , those in Nos. 4 , 5 , and 6 were obtained directly from the oxytrichloride .
The chloride in analysis 4 was distilled three times over charcoal , and still contained a little oxygen ; after another distillation ( analysis 5 ) the oxygen was reduced to 0*2 per cent. , and after a fifth distillation ( analysis 6 ) the whole of the oxygen had been removed .
The determinations were made by breaking a bulb containing a weighed quantity of substance in a closely stoppered bottle containing water ; the chlorine was then estimated as silver-salt , nitric acid being added after the nitrate of silver , whilst the vanadium , either in the filtrate or in another portion of chloride , was weighed as pentoxide .
No. Weight of tetrachloride found .
0.4272 1 0-6645 0-9644 2 1-0362 1.0438 3 .
0-9875 0.8935 .
1-1465 0'2940 5 .
0-4950 6 .
0-4080 Silver chloride found .
1'2555 2'8560 3-0441 2'6266 0'8667 1.5760 1-2086 Vanadium pentoxide found .
0-3195 0'4890 0-4632 0'5440 0'2370 0'1960 Percentages ( ill..~ , _ of chlorine . ?
of vanadium .
72-68 27-02 73.24 26-52 72-1426-36 72-7326-65 72-95 72-83 26-89 73'28 27-05 68l Hence we have for the composition of this chloride : Found .
r---~-- &amp; --- ... ... ... .A ... - .
Mean of Calculated .
( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) 1,2,5 &amp; 6 .
V 51'3 26-54 27'02 26'52 26-36 26-65 26-89 27-05 26-87 C14= 142-0 73-46 72-68 73-24 72-14 72-73 72-89 73-28 73-02 193-3 100-00 99-70 99-76 98-50 99'38 99-78 100-33 99-89 Vapour-density of Vanadium Tetrachloride.-Owing to the ease with which the tetrachloride decomposes into the trichloride and free chlorine , a solid residue is always left in the bulb when the determination is made by DUMAS'S method in the ordinary way , and the experimental number is rather too high , as the following numbers show .
No. 1.-Weight of bulb and air at 10 ?
C. and under 763-7 millims. .
11-1507 grms. Weight of bulb and vapour at 219 ?
and under 763'7 millims. 11-5900 grms. Capacity of bulb 118 cub. centims. , vol. of residual air 0-0 cub. centim. Hence the vapour-density is found to be 99-06 ( H=1 ) or 6-86 ( air=l ) , the calculated density being .
96-6 ( H=l ) or 6-69 ( air=l ) .
In order to avoid the error arising from the deposition of a solid residue , an arrangement shown in fig. 1 was used .
The tetrachloride was first placed in the small bulb ( a ) ; the large bulb ( b ) was then submerged , and the oil heated above the boilingFig .
1 .
point of the chloride .
The chloride was then boiled and the vapour passed through the bulb ( b ) , and the tube sealed at c. The bulb was then turned round in the oil so as to bring the point ( c ) below the surface , and the temperature of the bath allowed to rise above 200 ?
C. ; and when the temperature had become constant , the tube was again sealed at d. No. 2.-Weight of bulb and air at 150 C. and under 751 millims ... 19-8945 grms. Weight of bulb and vapour at 205 ?
and under 758-1 millims. 20-3015 grms. Capacity of bulb 112 cub. centims. , residual air 2 cub. centims. No solid residue was left in the bulb .
Hence the vapour-density is found to be 96-6 ( H=l ) , or 6-69 ( air =1 ) .
NTo .
3.-Another determination made at a higher temperature gave a somewhat lower number , indicating that a portion of the tetrachloride had been decomposed .
682 Weight of bulb and air 9 ?
C. and under 762 millims ... .
24-4722 grms. Weight of bulb and vapour at 215 ?
C. and under 762 millims. 250102 grms. Capacity of bulb 169-5 cub. centims. No solid residue left in the bulb .
Volume of mercury entering the bulb 157 cub. centims. Hence the vapour-density is found to be 93-3 ( H=l ) or 6'48 ( air =1 ) .
Vanadium tetrachloride is a dark brownish-red , thickish liquid , which evolves white fumes when exposed to moist air .
Its specific gravity was carefully determined at three temperatures , and found to be : at 00 C. 1-8584 at 8 ?
C ... 1-8363 at 32C ... 18159 .
It does not solidify at temperatures above -18 ?
C. ; nor does it at this , or any higher temperature , undergo any change of properties on treatment with chlorine .
The boilingpoint of vanadium tetrachloride is 154 ?
C. ( corrected ) under a pressure of 760 millims. Vanadium tetrachloride not only decomposes ( as has been stated ) on boiling into the trichloride and free chlorine , but the same decomposition takes place slowly at the ordinary temperature of the air .
The liquid tetrachloride , sealed up in glass tubes and exposed to the light , was found on standing for some months to be changed to a dark blackish powder , whilst free chlorine was liberated in such quantity as not only to exhibit its characteristic greenish-yellow colour , but in several cases by its pressure to have burst the tubes .
This powder on analysis was found to consist of the trichloride ( see analysis , p. 685 ) moistened with tetrachloride ; on removing this latter liquid in a current of dry carbonic acid at 1700 , the pure violet solid trichloride was left behind .
Thrown into water the tetrachloride is at once decomposed , yielding a blue solution identical in colour with the liquids obtained by the action of oxalic , sulphurous , or sulphydric acids on vanadic acid in solution , and containing a vanadous salt ( derived from the tetroxide , V2 04 ) .
In order to prove that a vanadous salt is formed when the tetrachloride is decomposed by water , the quantity of oxygen which this salt absorbs in conversion into vanadic acid was determined with a standard permanganate solution according to the method described in Part I. of these researches ( Phil. Trans. 1868 , p. 17 ) .
Action of Potassium Permanganate on the aqueous solution of Vanadium Tetrachloride .
( 1 ) ( 2 ) ( 3 ) ( 4 ) Weight of V C4 taken ... . .
0-2375 0-5353 0-4628 0-5687 Cub. centim. of permanganate solution ) required ( 1 cub. centim. = 0-00066 15'0 33-2 28-8 35-4 grm. oxygen ) ... .
Percentage gain of oxygen found .
4-17 4-09 4'10 4-09 683 The calculated percentage gain of oxygen according to the equation 2V Cl4 +0+ 4H 0= V2 05 + 8H C1 is 4-14 .
The solution of the tetrachloride in water does not bleach litmus-paper ; but if the vapours of the boiling tetrachloride are condensed in water , a liquid is obtained which contains free chlorine , as it bleaches litmus-paper and liberates iodine from potassium iodide .
At higher temperatures , as , for instance , when the vapours are led through a red-hot tube , much larger quantities of chlorine are evolved .
Vanadium tetrachloride acts violently upon both dry alcohol and ether , giving in the first case a deep green , and in the second a deep red-coloured liquid .
I am at present engaged with the examination of the products of this action .
Action of Bromine on Vanadium Tetrachloride.-Vanadium tetrachloride , heated with excess of bromine in a sealed tube to 180 ?
, yielded a solid mass , which on drying in a stream of carbonic acid at 160 ?
, presented the appearance of the peach-coloured trichloride , and subsequent analysis proved that this body was formed .
It is thus seen that , heated in excess of both chlorine and bromine , the tetrachloride splits up into trichloride and free chlorine , so that vanadium does not readily form a pentad compound with the chlorous elements ' .
( 2 ) Vanadium Trichloride , V Cl3=157'8.-The trichloride is a solid body , crystallizing in splendid peachblossom-coloured shining tables , closely resembling in appearance the crystals of chromium sesquichloride .
It is non-volatile when heated in hydrogen , and decomposes when heated in the air , red fumes of oxytrichloride being given off in small quantity , whilst the solid mass glows with absorption of oxygen , forming the pentoxide .
Heated in hydrogen it loses , first , one atom of chlorine , forming the dichloride ( V Cl2 ) , and afterwards , on exposure to a higher temperature , loses the whole of its chlorine , metallic vanadium being left as a grey lustrous powder .
Vanadium trichloride is an extremely hygroscopic substance ; it instantly changes colour on exposure to moist air , deliquescing to a dark-brown liquid , which on the addition of a drop of hydrochloric acid becomes green , containing a solution of vanadium trioxide ( V2 03 ) , or rather of the corresponding hypovanadic salts .
The trichloride is obtained ( 1 ) by the slow decomposition of the tetrachloride at the ordinary atmospheric temperatures , ( 2 ) by the decompositions of the tetrachloride at its boiling-point , ( 3 ) together with dichloride , when the vapour of the tetrachloride is passed mixed with hydrogen through a red-hot tube .
Method No. 2 is that which yields the trichloride in largest quantity and in the purest state , large quantities of the peachcoloured crystals remaining behind in the bulb-retort ; and these only require heating in a current of carbon dioxide at 160 ?
in order to yield the pure trichloride .
The substance lThe only other case of an element of a decidedly triad character yielding compounds of a tetrad nature , appears to be the nitrosyl-dichloride , NO Cl0 , of GAY-LussAc , and the corresponding bromine compound , NO Br , , discovered by LANDOLT .
684 used in analysis ( 1 ) was obtained by the first method , that used for analyses 2 , 3 , and 4 was prepared by the second process .
Weight of Silver Vanadium Percentages of trichloride chloride pentoxide No. taken .
found .
found .
chlorine .
vanadium .
1 .
1-6615 4-5704 0-9240 68-05 31-25 { 0-2295 -01320 -32-31 2 0-2550 0'6931 67-25 0-1094 0-2993 67-67 0-2188 01290 33-12 Hence we have for the composition of this chloride : Found .
Calculated .
( 1 ) ( 2 ) ( 3 ) ( 4 ) Mean .
V -51-3 32-5 31'25 32-3l 32-59 33-12 32-57 C13=106'5 67-5 68-05 67-25 66-71 67-67 67-42 157-8 100'0 99'30 ' 99-56 99-30 100-79 99-99 When the trichloride is thrown into water it does not at once dissolve , the peachblossom-coloured crystals floating in the liquid ; these , however , gradually dissolve , forming a brown-coloured solution , which on addition of a drop of acid turns to a bright greenl colour , identical in tint with the liquid obtained by reducing vanadic acid in solution with nascent hydrogen evolved by magnesium , and containing a hypovanadic salt .
Like the solution obtained by reduction , the green-coloured solution of the trichloride absorbs oxygen on standing , passing into the blue solution of the vanadous salts .
A solution of hypovanadic salt obtained by dissolving 0-293 grm. of trichloride in water required 45 cub. centims. of permanganate solution in order to oxidize it completely ( 1 cub. centim. permanganate=0-00066 grm. oxygen ) .
Hence the quantity of oxygen needed to oxidize the solution was 10-10 per cent. ( on the trichloride ) , whilst that required by the formula 2V C13 + 02 + 3H2 0= V2 05 + 6H Cl is 10-14 per cent. The specific gravity of vanadium trichloride at 18 ?
is 3-00 .
Vanadium trichloride dissolves readily in absolute alcohol , forming a greenish-blue solution ; in ether it also dissolves , giving a green-coloured solution .
( 3 ) Vanadium Dichloride , V C12=122-3.-Vanadium dichloride is a solid body crystallizing in fine bright apple-green plates , having a micaceous lustre and an hexagonal form .
It is prepared by passing the vapour of vanadium tetrachloride mixed with dry and pure hydrogen through a glass tube heated to dull redness .
If the distillation of the tetrachloride be conducted slowly , and the mixture of vapour and hydrogen brought at once into the heated portion of the tube , the pure dichloride is deposited in light pearly scales , which can easily be shaken out of the tube .
If the operation be conducted too quickly , or if the hydrogen be not present in large excess , a considerable quantity of the peach-coloured trichloride is formed in the anterior portion of the tube , and the green crystals of dichloride collect in one spot and are fixed firmly to the glass .
If the temperature be raised to a bright red heat during the preparation , the dichloride is further reduced , a black crystalline powder containing a mixture of lower chlorides and metal being produced , whilst the glass tube becomes opaque with formation of the silicon-vanadium compound .
When heated for a long time in a platinum boat in a current of pure and dry hydrogen the dichloride loses all its chlorine , metallic vanadium being left in the form of bright greyish-white lustrous metallic grains .
The dichloride prepared as above , gave the following analytical results : Weight of dichloride No. taken .
1 .
0-3317 f 02490 2 .
0-2462 Silver chloride found .
0-7693 0-5876 Vanadium pentoxide found .
0'2510 0-1832 Hence we have for the composition of the chloride : Found .
, r. Percentages of chlorine .
vanadium .
57-40 42-53 58-37 41-80 Calculated .
( 1 ) ( 2 ) Mean .
V= 51-3 41-95 42-53 41-80 42-16 C12= 71-0 58-05:57-40 58'37 57'88 122-3 100-00 99-93 100'17 100-04 Vanadium dichloride , when heated in hydrogen or in carbon dioxide , does not volatilize without decomposition .
It is extremely hygroscopic , instantly absorbing moisture from the air , deliquescing to a brown liquid : a portion of this chloride exposed to the air for five minutes gained 4 per cent. in weight , and in standing for sixteen hours the increase amounted to 50 per cent. When thrown into water the dichloride does not at once dissolve , the scales floating on the water without becoming wet ; soon , however , they dissolve , forming a violet-coloured liquid identical in tint with the solution of hypovanadous sulphate obtained by reducing a solution of vanadic acid in sulphuric acid by hydrogen evolved from zinc or sodium .
Like this latter liquid , the solution of the dichloride in water acts as a very powerful reducing agent , bleaching litmus and indigo solutions .
A solution of hypovanadous salt obtained by dissolving 0-2875 grm. of vanadium dichloride in water required 81-8 cub. centims. of permanganate solution completely to oxidize it ( 1 cub. centim.= 000066 grm. oxygen ) .
Hence the percentage ( on the dichloride ) of oxygen needed to oxidize the solution was 18-78 , whereas the formula 2V C12 + 03 + 2H2 0 V2 05 + 4H C1 requires 19-6 per cent. ; the difference between the found and calculated numbers is due to the great difficulty of preventing the solution absorbing some little oxygen during the analysis , which must of course be conducted in an atmosphere of carbon dioxide .
When the dichloride placed in a platinum 686 boat is heated to whiteness in a current of dry ammonia , dense fumes of ammonium chloride are given off , and on cooling after an hour 's heating , the boat contains vanadium mononitride in bronze-coloured pseudomorph crystals , which exhibit by reflected light a bright metallic lustre .
The specific gravity of vanadium dichloride at 18 ?
C. is 3'23 as a mean of two wellagreeing determinations .
Vanadium dichloride dissolves slowly in alcohol and in ether , giving in the one case a blue and in the other a greenish-yellow solution .
Metallic Vanadium , V=51 3.-From what we now know of the character of vanadium , we learn without surprise that the metal cannot be prepared in the free state by any of the methods given in the books .
BERZELIUS ( POGG .
Ann. vol. xx .
p. 1 ) obtained a brilliant metallic scale by heating the oxytrichlorides , placed in a bulb-tube , in an atmosphere of ammonia ; but this substance , as SCHAFARIK observes , is mononitride , and not metal .
The great Swedish chemist also states that the metal can be prepared in a pulverulent form by exploding fused vanadic acid with potassium .
The black powder which is obtained by lixiviating the fused mass is , however , not metal but an indefinite mixture of oxides .
SCHAFARIK ( Ann. der Ch. und Pharm. vol. cix .
p. 97 ) describes as metal the brownish-yellow lustrous crystals obtained by passing the vapour of the oxytrichloride mixed with hydrogen through a red-hot tube .
The body thus prepared , as I have already shown , is not the metal , but a mixture of the lower oxychlorides of vanadium .
JOHNSTON ( N. Edin .
Journ. of Sc. vol. v. pp. 166 , 318 ) obtained a hard brittle metallic globule by reducing vanadic acid with carbon at a white heat ; this was certainly not metallic vanadium , probably the alloy with silicon .
Although it appeared unlikely that any compound of vanadium containing oxygen would yield the metal by direct reduction , I have nevertheless repeated the above experiments , but without success .
In order to test the second method proposed by BERZELIUS , the fused pentoxide was mixed with excess of sodium , and the mixture heated in a closed wrought-iron crucible ; after the explosion the black mass was washed until free from alkali .
A black powder is thus obtained which glitters when suspended in water , and this powder on oxidation only gained 16 per cent. , showing that it contained even more oxygen than the trioxide .
JOHNSTON'S experiment was also repeated by exposing a mixture of trioxide , charcoal , and oil contained in a graphite crucible for several hours to the heat of a wind-furnace in which manganese can be fused .
Not a trace of either a bead or metallic powder was obtained , and the black powder gained on oxidation only 16 per cent. In order to ascertain whether under other circumstances the trioxide can be reduced , a portion of this oxide was heated to whiteness in a current of hydrogen , both when alone and mixed with excess of sodium ; in neither case was the metal formed , a black residue of oxide remaining behind in the boats .
The trioxide heated in a graphite crucible to bright redness with magnesium also yielded a black powder consisting of mixed oxides .
4v2 687 Another reaction consisted in passing the vapour of oxytrichloride mixed with hydrogen over metallic sodium heated in a porcelain boat ; a black shining powder was thus obtained , which after lixiviation and drying was burnt to vanadium pentoxide , and gained only 11 per cent. in weight .
The only methods which promised possible results were : 1 .
The reduction of a vanadium chloride ( free from oxygen ) in hydrogen gas .
2 .
The reduction of the mononitride in a current of hydrogen .
The first of these methods has been found to be successful , the second does not appear to yield the metal ; for in two experiments in which the mononitride was ignited in a platinum-tube to whiteness in a current of hydrogen , in one case for 3and in another for 2 hours , the loss of weight was only 88 and 8-4 per cent. respectively , whereas the nitride must lose 21'4 per cent. when converted into the metal .
Reduction of the Metal from Vanadium Dichloride in Hydrogen.-The only method by which it has been found possible to obtain metallic vanadium is by the reduction of a chloride free from oxygen , in an atmosphere of perfectly pure hydrogen .
Although this process appears simple enough , yet there is , I believe , no metal more difficult to obtain than vanadium .
This arises from the circumstance that while vanadium appears to be stable at ordinary atmospheric temperatures , it absorbs oxygen at a red heat with the greatest avidity , and that therefore every trace of air or of moisture must be excluded .
Another difficulty consists in the preparation of sufficiently large quantities of the solid chlorides free from oxygen or moisture , as also in the length of time necessary in order to reduce these chlorides in hydrogen at a red heat , during which time diffusion ( which can never be wholly prevented ) brings traces of oxygen in contact with the heated metal .
Then , again , the destructive action of the metal on glass and porcelain prevent these substances being used to contain the metal , whilst tubes of platinum and wrought iron become so porous at a red heat as to admit sufficient oxygen to convert the whole vanadium into trioxide .
The only means which can be adopted is to heat the chloride in a platinum boat placed in a porcelain tube .
A laborious series of preliminary experiments on the reduction of the metal in this way proved that , in order to obtain a substance even approaching to the metal , many precautions must be taken , whilst they showed that , in spite of every care , it is extremely difficult to obtain the metal itself .
Thus no less than ten experiments made by different modifications of this method yielded mixtures of metal with more or less oxide , of which the analyses showed an increase on complete oxidation varying from 60 to 67 per cent. , that for pure metal being 78 , whilst in every case in which the pure metal was prepared a considerable quantity of black sesquioxide was formed at the end of the boat nearest the inflow of hydrogen , showing that the source of error lay in the presence of diffused oxygen .
The apparatus which , after many alterations , was found best adapted for the preparation of the metal is represented on fig. 2 .
688 It consists of a hydrogen generator ( A ) yielding a stream of hydrogen which can be kept constantly passing through the wash-bottles day and night for a week at a time , Fig. 2 .
-+ +C _T .
by occasionally adding fresh acid to the upper bottle and drawing off the zinc-sulphate solution from the lower bottle by the caoutchouc tube .
The first wash-bottle contains a solution of lead acetate , the second silver nitrate , and the other three boiled sulphuric acid .
In order to remove any trace of oxygen which may have accompanied the hydrogen arising either from diffusion or from air absorbed in the dilute acid used , a tube ( C D ) is attached to the end of the last washing bottle ; the first portion of this tube contains a quantity of platinum sponge ( C ) , which is heated to redness during the whole time the hydrogen is passing through the apparatus , whilst the further portion of the tube D is filled with phosphorus pentoxide and plugs of cotton-wool .
The greatest care was taken to have all caoutchouc stoppers and joints made as tight as possible with copper wire and paraffin .
At right angles to the drying-tube ( CD ) is placed the reducing arrangement shown in the lowest part of the drawing .
This consists of a por/ S ... ... ... ... ... ... . .
-- .
689 B+H ri G WI=W ] celain tube ( E E ' ) placed in a Hofmann 's furnace and protected in the central portions , where it is heated , by an outer casing of sheet iron .
The porcelain tube is connected with the hydrogen apparatus by means of the wide glass tube ( F F ' ) provided with the tubulus ( G ) , and narrowed down to join the drying-tube at F ' .
The joint between the porcelain and glass tubes is made of seamless caoutchouc well wired , and covered by an outer short glass cylinder , the space between the tubes and the cylinder being filled either with mercury or fused paraffin , and a similar joint is placed at the further end of the porcelain tube .
The introduction of the anhydrous dichloride without exposure to the air is effected by means of the tubulus ( G ) , the dichloride being contained in the bent tube ( H ) in which it was prepared and sealed up in hydrogen .
After the whole arrangement has been set up , the platinum boat being in position as shown in the figure , hydrogen is allowed to pass through the apparatus for twelve hours to dry it completely and clear out the air ; the caoutchouc stopper of the tubulus is then withdrawn and the end of the tube containing the dichloride cut off , and the tube and stopper quickly replaced , so that the crystals lie in the horizontal portions of the tube .
The bent tube is next so turned in the stopper that the crystals of dichloride fall out and are collected in the platinum boat placed below .
This boat , charged with dichloride , is then drawn into the centre of the porcelain tube by means of the platinum wire , the end of which ( W ) passes through a small hole in the caoutchouc tube at the end of the apparatus .
As soon as the boat is in position , the wire is cut off short at the end of the glass tube , a proper joint made , and an exit-tube attached dipping under sulphuric acid .
Before the porcelain tube is heated , the caoutchouc stopper of the tubulus is surrounded by a bath of paraffin , and the hydrogen is allowed to bubble through for six hours .
The temperature of the porcelain tube is then gradually raised to the highest point ( a bright red heat ) which the Hofmann 's furnace will yield , and kept constant until the reduction is complete .
Torrents of hydrochloric acid gas at once come off , and the process must be continued for some hours after the last trace of acid can be detected in the hydrogen .
The process lasts from forty to eighty hours , according as the quantity of dichloride employed varies from 1 to 3 or 4 grms. Metallic vanadium prepared by reduction from the dichloride in hydrogen is a light whitish grey-coloured powder , which under the microscope reflects light most powerfully , and appears as a brilliant crystalline metallic mass possessing a silver-white lustre .
It is remarkable that vanadium thus prepared does not oxidize or even tarnish in the air at the ordinary atmospheric temperature , nor does it decompose water at the ordinary temperature , and it may be moistened and dried in vacuo frequently without gaining in weight .
Vanadium is neither volatile nor fusible when heated to redness in hydrogen .
When the powdered metal is thrown into a flame , or rapidly heated in an excess of oxygen , it burns with the most brilliant scintillations , and when slowly heated in a current of air it glows brightly with absorption of oxygen , forming in the first place a brown oxide ( V2 0 or V2 02 ) ; and on further heating this oxide again glows , and passes through the 690 black trioxide and blue tetroxide to pentoxide .
The metal-powder is not magnetic , and it could not be pressed into a coherent form .
Its specific gravity at 15 ?
was found to be 55 as a mean of three determinations .
The metal is not attacked by hydrochloric acid* , either when hot or cold ; neither strong nor dilute sulphuric acid act on the metal in the cold , but when heated with the strong acid the metal slowly dissolves , giving a greenish-yellow solution .
Hydrofluoric acid dissolves the metal slowly with evolution of hydrogen , and formation of a green solution , whilst nitric acid of all strengths oxidizes it with violence , evolving nitrous fumes and forming a blue liquid .
Both hot and cold solutions of caustic soda are without action on the metal , but when fused with the hydroxide , hydrogen is evolved and a vanadate formed .
The metallic powder obtained by reduction contains a quantity of combined or absorbed hydrogen , the amount of which frequently rose up to 1'3 per cent. The weight of this hydrogen must be determined by burning the metal placed in a platinum boat , in a glass tube through which first a stream of dry air , and afterwards of pure oxygen is passed .
A drying-tube placed in front of the combustion-tube gives the weight of water formed .
Analysis No. 1.--01365 grm. of substance , containing 0*0005 grm. of hydrogen , yielded on oxidation 0*2420 grm. of vanadium pentoxide in long clear needle-shaped crystals .
Hence the percentage gain is 77'94 , whereas that calculated from metal is 77'98 .
Analysis No. 2.-0*2305 grm. of metal from a different preparation , containing 0*0022 grm. of hydrogen , yielded 0'3915 grm. V205 .
Hence the percentage gained was 70-8 .
Metallic vanadium burns in excess of dry chlorine , forming the tetrachloride .
A portion of dark reddish-brown chloride , formed by combustion of the metal in chlorine , gave the following results : 0'6958 grm. chloride yielded 1 9939 grm. Ag Cl , and 0'3383 grm. V2 05 .
Hence the chloride contained 70'8 per cent. of chlorine and 27*3 per cent. of vanadium ; the difference between these and the theoretical numbers ( 73-46 and 26-54 ) being explained by the fact that the metal operated upon contained a slight admixture of oxide .
Heated in an atmosphere of pure nitrogen , metallic vanadium at once absorbs this gas and is converted into mononitride .
Vanadium attacks all glass or porcelain in which it is heated , a compound of silicon and the metal being formed .
This compound coats all tubes of glass or porcelain in which the chlorides have been heated and reduced , as a grey lustrous metallic mirror , the formation of which renders the tubes very brittle and liable to crack .
The silicon compound is not attacked by hydrochloric , and only slightly by nitric acid , but it tarnishes after a time on exposure .
The metal also forms an alloy with platinum .
All the platinum boats and tubes in which the reduction has been made are found after the operation to be completely saturated with vanadium , and require to be frequently ignited in Hydrochloric acid readily dissolves the trioxide .
691 the air , and as often boiled out with a solvent for vanadic acid , before they can be freed from the absorbed metal .
If the platinum be frequently used for the reduction without this purification , it becomes brittle and dark-coloured from absorption of metal in quantity .
My best thanks are due to Messrs. OELHOFER and FINKELSTEIN , who have ably assisted me during the progress of the above investigation .

