82 BAGSTER THE REACTION BETWEEN VIII.-The Reaction between Nitric Acid and Copper. By LANCELOT SALISBURY BAGSTER. ON attempting to measure the volume of nitric oxide produced by the action of 5N-nitric acid on copper under certain conditions, it was found that when the gaseous products were evolved and passed into sodium hydroxide solution in a vacuum complete absorption took place although a considerable volume of nitric oxide should have been obtained (Higley Amer. Chem. J. 1905, 3.7 IS). This reaction was investigated in the hope of throwing further light on the reaction between nitric acid and copper. Veley (P/~il. Trans. 1891 182 279) showed that the nitrous acid present was the cause of solution in dilute acid and proposed a series of reactions involving the reduction of the nitrous acid to nitric oxide followed by the oxidation of the nitric oxide to nitrous acid by the nitric acid.The work of Lewis and Edgar ( J . Amer. Chem. SOC. 1911 33 292) shows that there is an equilibrium between nitric acid nitric oxide and nitrous acid and that this equilibrium is butl slowly attained. It seems unlikely therefore, that in such case the nitric oxide would be completely and instantly oxidised as would1 be necessary to secure compleite absorption of the product in the experiment just described. The follolwing explanation of the procms in terms of well-known electrochemical theory is suggested (1) oxidation of the hydrogen film on the copper by the! nitrous acid which itself is reduced to hyponitroas acid the copper passing into solutioln to replace the hydrogen removed ; (2) oxidation of the hyponitrous acid to nitrous acid by nitric acid which itself is reduced to nitrous acid.I t may be assumed thatl the first reaction will pro+ duce hyponitrous acid in unimolecular form (H2+ HNO + M,O+ HNO') and as Divers (T. 1889 75 112) has sholwn that hyponitrous acid has the double molecule i t may be considered that a t the moment of formation it' will be in a reactive state corresponding with nascent hydrogen and completely oxidised by nitric acid. It will bei shown that the product of reaction with dilute acid is nitrous acid produced in quantity necessary to satisfy the above scheme. It might be expected that sufficiently dilute nitric acid would fail to oxidise completely the hyponitrous acid formed and in such case nitrous oxide the well-known decom-position product of hyponitrous acid should appear as a product NITRIC ACID AND COPPER.83 This has been found to be the case by Higley (Zoc. cit.) using 3N and more dilute acid. It would not be expected that much nitrous oxide should be obtained; i f the nitric acid were so weak that itl failed to oxidise most of the hyponitrous acid the reaction would cease when tho nitrous acid originally present was exhausted. This has been found to be the case with N/Z-nitric acid. Reaction started by addition of sodium nitrite ceased when the nitrous acid was destroyed but could be restarted by the addition of more nitrite. Divers (Zoc. cit.) has shown that nitrous acid oxidises hype nibrous acid with the formation of nitric oxide and water and in the course of the present work it was found that nitric oxide is the only gaseous product of reaction between copper and an excess of N / 3-nitrous acid prepared by adding sodium nitrite solution to excess of dilute hydrochloric acid.Here hyponitrous acid formed would in the absence osf nitric acid be oxidised by the excess of nitrous acid. As in the case of very dilute nitric acid, it would be expected that very dilute nitrous acid would also produce some nitrous oxide. This has been verified in the case of N/50-nitrous acid prepared as above reacting with an excess of copper. Nitric oxide was produced but at the end of the reac-tion the gas in solution was collected and found to consist largely of nitrous oxide.The nitrous oxide was probably formed when the nitrous acid had become nearly exhausted. It is unlikely that the hyponitrous acid will be oxidised by nitrous acid in the presence of much nitric acid as the nitrous acid in the neighbour-hood of the copper will react with it and conditions will con-sequently be favourable for oxidation of the product by the nitric acid if present# in quant<ity. Further Peters (Zeitsch. nnorg Chem. 1919 107 313) states that in the presence of carbon dioxide a 5 per cent. solution of sodium nitrite reacts with copper producing nitric and nitrous oxides nitrate being formed in solution. Known constants for nitrous and carbonia acids show that about' 0.25 per centl. of the nitrite will be present as free acid which a t this very small con-centration will react with the copper in the manner already dis-cussed the same products being obtained.The nitrate would be formeid by direct decomposition of the nitrous acid. Reference should finally be made to the work of Ackworth and Armstrong (T. 1877 32 54) where it is shown that copper salts in solution give rise to an increased yield of nitrous oxide. These authors used a small vcllume of nitric acid and itl is probable that as the acid became usedl up the nitrous acid accumulated in solu-tion would react with the hyponitrous acid formed. A deep blu 84 BABSTER THE REACTION BETWEEN collour characteristic of complex copper salts was noticed when copper was dissolved in nitrous acid during the present work. This complex in the case of Armstroag's wotrk would diminish the concentration of free nitrous acid thus leading to the form-ation of a greater quantity of nitrous oxide.It has been shown (Ihle Zedtsch. physdkal. C'hem. 1895 19 577) that nitric acid above 35 per cent. by volume will react in the absence of nitrous acid. This concentration corresponds with just under 8A7. Reference t c the table (p. 86) shows that such acid yields nitrogen peroxide as well as trioxide. As a secondary reaction is not apparent in the experiments described with dilute acid it is probable that the products obtained from stronger acid nearly represent primary ones. The nitrogen peroxide may be regarded as the product of the direct oxidation process by the nitric acid, the nitrous anhydride also obtained in quantity from 8N- and 1ON-acid being due to the simultaneous progress of both types of reaction.So far discussion has been confined to dilute acid. EXPERIMENTAL. When the experiment described a t the beginning oif this work wa8 carried out quantsitatlively itl was found thatl the absorbing solutions contained nitrite corresponding with from 310 to 320 C.C. of N/lO-permanganate for every gram of copper dissolved. This result wit& obtained with the proiduct of reaction from acid of shength from 5N to 15N. The theoretical quantity of perman-ganate corresponding with the reactions discussed is 318 c.c. being the same whether the product is nitrous anhydride nitrogen tri-oxide or nitrogen peroxide. Nitrous acid i f formed would be expected under the conditions described to distil as anhydride, owing to the equilibrium ZHNO H,O + N,O,.To distinguish between the possible prolducts it was necessary to determine the total nitrogen in the absolrption vessel apart from any that might distil there directly as nitric acid. For this pur-pose the apparatus shown in the figure was constructed. Nitric acid was admitted after exhaustion to the vessel A containing the copper by means of the funnel B. Nitric acid was condensed in C and returned to A . Gaseous products were absorbed in D and) 8 which contained sodium hydroxide solution. Connexion with the mercury pump was closed by a tap during reaction. Tesk with this apparatus in the absence of copper showed that n NITRIC ACID AND COPPER. 85 appreciable quantity of nitric acid distilled through the condenser when mled to loo.In tthe course of preliminary tests using 10N and weaker acid, it was found that when the condenser was cooled ta Oo nitric oxide was produced there the effect being still slightly apparent a t loo witb 6N-acid. The probable explanation is that gaseous trioxide being a mixture a t the lower temperature the nitrogen peroxide is partly coadensed thus allowing the nitric oxide to pass on and throlugh the alkali unabsorbed. This effect would a t once be apparent in the caw of trioxide; if excess of peroxide were also present' greater condensation would have to take place before the composition fell below that' corresponding with trioxide when complete absorption would cease and nitric oxide appwr.The production of nitric oxide will be accompanied by a diminished quant4ity of prolducts in the absorption train. I n carrying olutl an experiment the nitric acid was kept boiling, the temperatares shotwn in the table being the boiling points of the acid solutions in a vacuum. A slow stream of carbon dioxide was passed through the tube F to sweep the gaseous products from the condenser after removal of the acid vapour. For analysis the nitrite in a portion od the absorption soflution was determined with permanganate and the total nitrogen in the residue estimated as ammonia by distillation with Devarda's alloy. A series of results is shown in the following table. In each case 0.250 gra 86 THE REACTI.ON BETWEEN NITRIC ACID AND COPPER. of copper was used and a t least’ ten times the quantity of acid needed for solution.No. 1 2 3 4 6 6 7 8 9 10 Nitric acid. D. 1.40 1-40 1.30 3.30 1.30 1.25 1.20 1.20 1.20 1.17 Nor-mality of acid. 14.5 14.6 10 10 10 8 6 6 6 5 Nitrogen Total Temper- as N,Oa. nitrogen. ature. Gram. Gram. 60” 0.0553 0-0980 60 0.0546 0.1001 45-50 0.0557 0.0616 45-50 0.0560 0.0630 45-50 0.0542 0.0600 35-40 0.0542 0.0582 30 0.0550 0.0610 30 0.0490 0.0490 30 0.0420 0.0423 30 0.0546 0.0595 N,O& Per cent. 77 80 10.2 11.0 10-5 7.2 (10.8) -(9) Temper -Nitrogen ature of =NO. conden-Gram. ser. - 10” - 10 10 10 5 - 10 - 16 0.0045 10 0.0090 7 I 20 - - -The theoretical quantity of trioxide to satisfy the reactions suggested is 0.0555 gram the total nitrogen being the same when trioxide is the product and twice the amonnt if peroxide is the only product.The blank tests showed that in the case of experiments 1 to 6 there was no nitric acid distilled to make the total nitrogen value high. The small loss of nitrogen mentioned as occurring with the weak acid must still be considered as taking place however with the condenser a t 10”. Although as explained this loss will not be apparent in the trioxide value it will introduce an error of a few milligrams in the total nitrogen value. I n the case of experiments 7 and 10 the value for the total nitrogen is higher than it should be; the copper in the case of the 6N-acid relquirsd a considerable time for solution the acid occasionally “boiled with bumping,” and unless the temperature of the condenser was as low as loo nitric acid reached the absorbers by direct distillation.Values such as are shown were repeatedly obtained but it was not possible to secure a more definite result. As already indicated, the fact that the product from 5 N - and GN-acid more easily forms nitric oxide than that from 8N- and 10N-acid shows that it con-tains less peroxide. Considering the facts stated it may be con-cluded that the product from 5N- and 6N-acid is almost entirely nitrogen trioxide (nitrous anhydride). The low values for the products in experiments 8 and 9 will be raised tot the same value as the others if allowance is made for the nitrogen lost owing to the folrmation of nitric oxide this value being nearly that demanded by theory. Allowing for the correction suggested for the prGducts from more concentrated acid the product from 14.5H-acid will contain 85 to 90 per cent. of peroxide the residue being trioxide for 10N FORMATION OF DERIVATIVES OF TETRAHYDRONAPHTHALENE. 87 acid there will be about 15 t o 20 per cent. of peroxide and for the 8N-acid somewhat less. It was not practicable to carry out work with acid more dilute than 5A7 owing to the time taken for solution in a vacuum prob-ably on account of the continuous removal olf nitroius acid which in consequence could not ex& its usual ‘‘ autwatalytic ” effects. THE UNIVERSITY OF QUEENYLAND, BRISBANE. [Received Pebrumy 27th 19201