B Y E. J . J O h i
Introduction
ACTION OF METALS ON NITRIC ACID
1223
Gay-Lussac concluded that the experiment proved simply that “hyponitric acid” (nitrogen dioxide) is less stable than nitric acid. His work, like Millon’s, was mainly qualitative. In 187j Acworthl made a series of quantitative experiments with the view of determining the proportions of nitric oxide, nitrous oxide, and nitrogen which are formed a t different stages of the solution of copper in nitric acid; but no attempt was made to estimate the trioxide and dioxide of nitrogen produced, nor was any investigation undertaken of the effect of temperatures, above the ordinary. The conclusions of this investigator were: I. Cold dilute nitric acid, acting on copper, gives almost pure nitric oxide, the gas evolved containing 90-95 % of that substance, but with increase of copper nitrate in the solution, the per cent of nitric oxide diminishes. 2. If the nitric acid is nearly saturated with cupric nitrate before the csperiment, the action on copper yields as much as 85% of nitrous oxide with some nirric oxide and nitrogen. 3. Potassium nitrate in solution has little or no effect. 4. Ammonium nitrate added to the solution gives rise to the evolution, chiefly, of nitrogen and nitrous oxide. 3. Zinc, mercury, and iron, when treated with nitric acid and ammonium nitrate, evolve nitrogen chiefly. Acworth recorded an experiment in which copper acting on a nitric acid solution almost saturated with copper nitrate gave a mixture of gases analyzing 8 of nitrous oxide, 4% nitric oxide, and 10% Nz. In testing the action of copper on a solution of nitric acid almost saturated with cupric nitrate, Acworth and Armstrong2 record an experiment in which 0.4j9 grams of copper are dissolved, and the analysis of the gases formed show 8857 nitric oxide, 1oy0 nitrous oxide, and 2y0 of nitrogen. This is only a slight decrease in nitric oxide yield (IO%), as compared to an experiment in which only a small amount of cupric nitrate is added. These two sets of data a t first glance seem to be a t variance with each other, but when the experimental conditions are investigated the differences become quite clear. Acworth used 15 grams of copper in his experiment, added successive portions of nitric acid, and allowed the reacting mixture t o stand in contact with the reacting gases over night. Since the solutions in each case were almost saturated with cupric nitrate the only differences between the experiments are the quantities of reduction products in solution (only gaseous products were determined), and the time of contact of the resulting gaseous products with the solution covering the metallic reducing agent. It has been shown by Sabatier and Senderens3 that moist metallic reducing agents will reduce nitric oxide t o nitrous oxide, With the ordinary conception of the action of a moist solid and a gas, as consisting of an intermediate solution of the gas, it becomes quite clear why Acworth obtained such a large yield of nitrous oxide.
js
I
J. Chem. SOC., 28, 828 (1875:. (1877)~Exp. XVII. Compt. rend., 114, 1476 (1892).
* J. Chem. SOC.,32, 67
I221
E . J . JOYS
ACTION O F METALS ON N I T R I C ACID
I225
“The gases which are evolved on dissolving metals in nitric acid result, we believe, from the decomposition of these reduction products and from their action upon each other. “The nitric oxide is chiefly, if not entirely, formed by the decomposition of nitrous acid, in accordance with the equation 3
”02
= 2
NO
+ HKOs + Hz0
“The nitrous oxide is in all probability chiefly the product of two distinct changes, viz., . . . 2 HNO = XzO HzO NHzOH HN02 = Nz0 2 Hz0
+
+
+
“Finally rye regard the nitrogen as chiefly the product of the action of nitrous acid on ammonia: NH3 HNOz = P\’z HzO,
+
+
but it is perhaps in some cases in part, and in others entirely, the product of the action of nitric action on nitrosidic acid.” Bancroftl pointed out that the part of the theory of Acworth and Armstrong that deals with the solution of the metal, is nothing more than a special case of the electrolytic theory of corrosion, as developed by Rhitney? in 1903. Thorpe3 found that the copper-zinc couple will reduce nitrates to ammonia. This result was confirmed by Gladstone and Tribe14who proved the action to be electrolytic in nature, the reduction taking place a t the cathode metal, that is, the copper. In 1883 Divers5 showed that metals which are acted on by nitric acid may be divided into two classes. The metals of the first class (copper, silver, bismuth, and mercury), when treated with nitric acid, form neither ammonia or hydroxylamine, while those of the second class (including tin, magnesium, zinc, cadmium, aluminum, and iron) produce ammonia or hydroxylamine, or both. Divers advances a theory for the action of the two classes of metals: “there seems to be but one way of interpreting the action of the silver class of metals upon nitric acid, and that is to recognize that these metals combine with the nitrogen of the acid, and do not decompose its hydroxyl. . , . The metals of the tin-zinc class, characterized by forming ammonia with nitric acid, have certainly the power to displace hydrogen of the hydroxyl of such acids as sulphuric acid, and thus may directly form nitrate by displacing the hydrogen of nitric acid.” The conclusions Divers draws do not seem to follow as a logical consequence of his results. He gives no reason why a given metal should fall in one class rather than the other. J. Phvs. Chem., 28, 478 (1924).
* J. Am. Chem. SOC.,25, 394 (1903). J. Chem. SOC., 26, 541 (1873). J. Chem. SOC.,33, 139 (18781. J. Chem. SOC.,43, 455 (1883).
1226
E . J . JOSS
ACTION OF METALS ON NITRIC ACID
1227
of the so-caiied ‘excess voltage’ a t the cathode; and in the electrolytic process we can obtain a more or less quantitative measurement of the phenomenon thought we are still far from knowing the cause of it. “A less simple case is that of copper in dilute nitric acid. CopFer reacts chemically with dilute nitric acid, setting free nitric oxide. The formation of copper nitrate must be the anode process and the reduction of the nitric acid the cathode process. . . , “These experiments were performed to prove that the difference betm-een the electrochemical and the chemical corrosion of copper by nitric acid was an apparent one only and due t o an unsuspected difference in the conditions. In addition they illustrate the zupei-iorflexibility of the electrochemical method over the chemical meihod. I n the electrochemical method there is no difficulty in varying the concentration of the copper salt a t the cathode between any desired limits, while this is very difficult t o do in the case of the chemical method. This is in addition to the advantage, which the electrochemical method always has, of permitting a wide variation in the rate of reaction for constant temperature and constant concentration, If we are ever to have a thorough knowledge of the chemical reactions between nitric acid and the metals we must study the problem electrochemically.” Later Bancroftl gave the scheme of attack which has been follou-ed in this problem : “On the assumption that all corrosion is electrolytic in nature, it is evident that a t least three independent factors must be taken into account in the reduction of nitric acid by a metal: the specific reducing power of the metal which may be measured by the hydrogen over-voltage in a corresponding sulphuric acid s’olution; the catalytic action of the metal on the various reduction products; and the catalytic action of the metallic nitrate on the various reduction products. This analysis of the problem is of no value unless we can devise methods of distinguishing the effects due t o these three causes. Fortunately that is a very simple matter. We can determine the reduction products on adding different, dissolved, reducing agents having no catalytic action, which will give us the reduction products due t o different electromotive forces. Of course it is possible that any given reducing agent may exert a catalytic action itself or through one of its oxidation products. The catalytic action can be recognized and eliminated by using other reducing agents of about the same reducing power or by adding the same reducing agent t o solutions of the different reduction products of nitric acid, This last method will be effective only in case the catalytic action is on one of the first reduction products of nitric acid, such as nitrous acid. There is no reason a t present to suppose that there will be any special difficulty in distinguishing between reducing action and catalytic action. “Elect’rolysis of a nitric acid solution with the metal under consideration as cathode will give the combined effect of the specific reducing action of the cathode plus the possible catalytic action of the cathode metal. If the elecJ. Phys. Chem.,
28, 479 (1924).
E. J. 5088
I228
trolytic reduction agrees with those ohtainetl hy the corresponding cheniirnl reduction,l the cathode metnl has no catalytic action. If there is a clift'ercncc hetween tlic sets of result$. the diflerence gives 11s mliialjle infornintion in regard t o the cntalJ-tic action of t h e cathode metal. It will pro1):;l)ly not I)(> difficult then to formulate the cntnlj-tic action of the iiietnl c1e:irIy n r i t l (lefiriit c ~ J - . "Thr. ctitnlytic action of the iiietnllic nitrate can I)e cictcctctl 1)y :itl(liiig the snlt to nit.ric ncicl n n r l then re1:enting tlie esl:erinicrits with thr3 (lissolvtvl rediicirig agents. These results will \)e supplemented 11)- rcpentirig the c>lr>t*trolytic reductions i n presence of the niet:illic nitrate, :iCter which thc1-c p l i o t i l t l he no clifficulty in :iccoiinting fur d l the protlucts o1)tninetl ~ v l i ~thtl n iiiot:il reacts tlirect n-ith tlie nitric ncicl under the conclitiori:: of ortlinnr!- ttorrosiori." Bnncroft accepts ;Ilxistronp's v i e w on the g ~ o i reduction i ~ protliicts 0 1 nitric acid nnc! ad(ls to tlicni sewral of his ami. "It is e\-itlent from n-lint hw alrencly 1)ecn tloiic t h t .Iriiisti~)iigivn; riglit ) ? , S H , ( )FI. in saying that tlic tlircct rctliiction protlucts iirr [-IS( ).?. H2S& ant1 S H . ; , while S() ? . SO. S20:inti S?are foriiictl 1))- scwintl:iry wnrtioni. very 1irolxil)lj. as follow :I-ISOj H S O , = 2 Si)? 1320 BSO.!= 2 so HSO4 Ij.0 H?S?Oa 2 SO f H?O H?S?O, = S?O H.20 HSO.: SH?OH = S.0 2 I{!() 2 HSO,; 2 SH2OI-I = 2 HSO? A- SZO 3 IHZO HZSZO? 2 S H ? O H = 2 S.2 4 H?O H S O ? S H : j = S,: 2 13.10
+
Iryt>r:.Aliil,q 175, 141 ~ 1 8 ~ ~ 1 .
. : i ~ i i ~r!~(lii(,-
c ~ o i i r a iliavc. ~
ACTION O F METALS ON N I T R I C ACID
1229
“It is usually assumed that nitrogen comes from the interaction of nitrous acid and ammonia, apparently because ammonium nitrite will decompose into nitrogen and water under certain conditions. This seems to me inadequate because we get nitrogen evolved in cases where there is no reason t o suppose that ammonia is formed a t all. We must therefore postulate a reaction between hyponitrous acid and hydroxylamine. HZNz02
+
2
‘NHZOH
= 2
N2
+ 4 HzO
Oesterheldl was forced t o the same conclusion when he found that the electroly tic oxidation of a n alkaline hydroxylamine solution a t a platinum anode gave a mixture of gases approximating 51% K2and 4976 NzO. ‘The formation of nitrous oxide is undoubtedly due to the oxidation of hydroxylamine to hyponitrous acid which then breaks down into nitrous oxide and water:
+ +
+
2
NH2OH 0 = NOH H20 KOH ---+ H&z02 +N20 H2O
“It is impossible, however, that the nitrogen formed from the hydroxylamine can be a decomposition product of ammonium nitrite as was assumed t o be the case when studying the oxidation of ammonia. I n the first place there is no ammonia present and, in the second place, there can be no oxidation of hydroxylamine to nitrous acid a t the low anode potential which was maintained. The simplest explanation is that the nitrogen is due t o the reaction between hydroxylamine and nitroxyl which is the first oxidation product, 0:”
+ NHzOH = + JS2
2
HzO
A. Angeli has observed the same reaction when nitroxyl splits off from benzene sulphohydroxamic acid in presence of hydroxylamine. While two molecules of nitroxyl may combine to form hyponitrous acid or nitrous oxide and mater, yet there is a quantitative evolution of nitrogen so long as there is a sufficient excess of hydroxylamine. “It is also possible that nitrogen may be formed simultaneously with nitrous oxide by a different decomposition of hyponitrous acid,2 perhaps 5 H2N202 = 4 HzO
+ HK03 + 4 N2
“While it is perfectly certain that nitrogen peroxide, nitric oxide, nitrous oxide, and nitrogen are due t o secondary reactions and are not direct reduction products of nitric acid, this does not mean that hydroxylamine and ammonia may not be formed by reduction of the oxides of nitrogen. I n fact, we know that under certain conditions this may happen. The simplest supposition to make would be that the stages were NO2 ----t NO +NzO (4 N2) + NHzOH +S H s ;but there seems to be good reason t o believe that matters are not so simple as this. It seems probable that hydroxylamine is not always a preliminary stage t o the formation of ammonia, that nitrous oxide is not always a preliminary stage in the formation of hydroxylamine or ammonia Z. anorg. Chem., 86, 129 (1914). Berthelot and Ogier: Compt. rend., 96, 30, 84 (1883); Hantzsch and Kaufmmn: Ann 232, 317 (1896); Ray and Ganguli: J. Chem. SOC.,91, 1866 (1907). 2
.
from nitrogen Feroxide or nitric oxide, antl that the reduction of nitrogcn peroxide does not necessarily give rise to nitric oxide as an interinetliatc stzgc." K i t h this line of attack in niintl, we can norv avail ourselve3 not oniy of the (lata on the chemicnl action of iiietals on nitric acid, hut d s o of thow 011 the electrochemical rctliiction. keeping always in niind the chemist ~ j of- oiir pro1)leni. that is. the interaction of the various constituents present in L: given mixture. Points of extreme imixjrtance in this consickration arc : The retliiction scheme of nitric acitl consists of direct reduction reactions antl of sccontl:ir\reactions. The quantity of the various protliicts we shall olitnin in n given case t1ei:encls. therefore. ~vhollyon the various renction velocities. If the seconclary reaction has a high velocity n-hile the direct reaction hns a low one. the products of the secondary reaction will predominate. If the rt1ver.e i, true the protliicts of the direct reaction will predominate. There ran lie :ill intermctliate gradations of these two The velocity of the direct re:ic>tion~may lie controlled in t h e electrochemical process 11y cc;iitrolling the current density, while. 2s 1)cjii:trtl out lie Bancroft' the chemical process is not so flexible. This may l i e cffectctl to SOIIL' tlegree hon-cwr by controlling the concentration of reducing :!gent. Thc velocities of the second:.i.y reactions may dci:cn(l on :'. niultituc!c' of fnctors; catalytic effect of the rediicing :tgent. of the products of its orit1:ition. of the retliiction prctliicts of the :icitl itself (no cs of this havc h w n foiltic! J . temperature, solvent, etc. Practicnlly all t1:c i,eactions in the rp(1uction wheiiic of nitric) :icitl :ire reversililc, hence the time factor must lie consitleretl in n-a)-s other' t h i n :i.i:I of gn~cwiis coiiil:onent of re:iction velocity. Fs;Peci:illy is this true in the protlucts. T h c st1uly of the electroly of solutions of nitric ::(>it1 liw rcccivcxtl iiiii('1i :ittention for 11i::ny ye:iri;. I: (1:~y:~ has show1 t l x t n-iicn y t r m g nitrir :icitI is clcctrol~-zctl,:IO hy:lrogen ap1:ears :it the cntiiotlc, h i t x j-cllow o r i~ccl~lisli c o l o i , thought to lie t l : ~to nitrow zcitl. ii; seen. X tliffercnt result ~v ever. o1)tninrvl on mixing tiic acid with an equal or gr(':itc.r voliuiic o for tbcn hyt!rogen w:'s protlucetl, the :>.i:ionnt of the gas (I(~pcn(1ing liotli iipon the concentration of the :wid and upon the current employed. To w e 1;:iixclay'^ on'n nortls, "that acitl from n-hich no gas separntd at the cathotle with a n-e:?li 7-oltaic lxittery, did evolve gas there with a stronge which evolved no gas there with a strong acid, (lit1 cniisc cwliition :\-it11 :in acitl more dilute." little later Schiinliein4made a wries of determinations of the elcctroij-tic reduction of nitric ccitl at :t plntinuni cathode, with :>.rids ranging f'roiii I .A(J t o I , 70 in specific gr:i\-ity, his oliiect lieing to clctermine the contlitions iintler wliich all t h e hytiroyen is ckpolarizetl.
ACTION OF METALS ON NITRIC ACID
1231
,
In 1871 Bourgoinl found that electrolysis of a very dilute nitric acid solution causes no reduction of the acid, hydrogen being evolved a t the cathode. He found that with an acid “ 0 , 6.25 aq., hydrogen was first evolved, this was followed by nitrogen, while finally, ammonia is produced, and with ”03 19 aq., hydrogen and subsequently nitrogen are obtained; the latter gradually increasing in quantity, reaching a maximum and, in the end, disappearing. No nitrous acid could be detected. When, in subsequent experiments, a more concentrated solution was employed, no hydrogen was evolved a t first, however, a blue coloration, “probably due to nitrous acid,” was observed around the cathode; finally, the electrolysis of “ 0 3 1% aq. produced pure nitric oxide for a time and, afterwards, a mixture of nitric oxide hydrogen. Gladstone and Tribe2 electrolyzed acid of specific gravities 1.40, 1.20, end I . 13, respectively, in a Hofmann apparatus with presumably platinum electrodes, employing various current densities. They found that from the first concentration, neither ammonia nor hydrogen was obtained with any current employed; with the second neither of the above compounds was evolved. When the current from a single Grove cell was employed, but with increased current density hydrogen was obtained. The results of Gladstone and Tribe ere not surprising when the work of Ihle3 is considered. This author showed that for a given concentration of acid a definite current density was required to effect reduction to ammonia, and that this value increased with increasing concentration of acid. Gladstone and Tribe, in their work, also confirmed the report of Bourgoin and of Schonbein regarding the ceasing of the mysterious initial evolution of hydrogen. iiIn the first, second, third, and fourth experiments with the I:I acid, the evolution of hydrogen gas a t the cathode ceased almost entirely and quite suddenly in three minutes, and with the first experiment with the 1:2 acid in 40 minutes.” This result was accounted for as follows: “We , . . have traced it to the presence of nitrous acid, which we find almost entirely prevents the escape of hydrogen.” Their conclusions are quite correct but no satisfactory explanation of the effect of nitrous acid is given. This will be discussed later under nitrous acid, Freer and Higley had made by far the most exhaustive study to that time. Their results, as well as all those of other workers in the field, could not be explained on the basis of the theories advanced, simply because they assumed the action of metals on nitric acid dependad on too few factors, or upon factors that had only an apparent affect. A summary of their wo:k is given in their final article.4
+
+
+
J. Pharm., (4), 13, 266-270 (1871).
* J. Chem. Soc., 35, 172 (1879).
Z. phgsik. Chem., 19, 572 (1896). 4.4m. Chem. J., 21, 329 (1e99).
1232
E. J . JOSS
“Silver, copper, iron. and electrolytic hydrogen h a w nearly the same effect on concentrated nitric acid, nearly pure nitrogen tlioside being evolved.” “In respect, t o their action on dilute nitric acid, t h e order of the metals silver. copper, lead and iron is that of the potentin1 series. ”The activity of electrolytic hydrogen clepentls, in a niarketl tlcgrcc, upon the charactel. of the electrode upon n-how surface it is lilicratetl. “Electrolytic hydrogen, in contact n i t h cathotles of copper and of silvcr, effect reduction of nitric acid in proportion t o the current tlcnsity. “Mlien nitric acid having n specificgravityof 1.0issiil~,iectet~toelc(~trolysis in cells having cathmles of silwr a n d lend, respecti\-ely. on the surface of nhose cathmles there is eqiinl current dcnsity. the ret1iictic)n prodiicts arc thc .same, i. e., nitrogen and ammonia. in approsinintely the smie proportions. “Mver and lead when tlissolvetl in acitl oi >l:ccific gravity 1.05. yic.l(l recliiction protlucts nhich are ivitlely tlifl’ercnt. lmth qu:dit:itivcly ant1 qiinntiintirely. since from the silver there is ohtninetl only nitric osiclc antl nitrogen tliositle equi\-alent r c q m tively to g5‘ and of thcl nietnl tlissolvctl. whilr~ with lead the products :ire nitric cisitle. nitrogen tliositle. and nitrous oritlti, ccluiralcnt in the ortlf>rnniiied. to ?..;> 7 . 7 . ant1 90( of iiic>tnltlissolretl. ”Lead retluce:: n;tric acitl (if specific gr:ivitJ- I . o j much 111orc cmiipic~tc~l\t h a n copper. yielding almiit five times niore nitrous ositle thnn the latjtr)r. On the other hantl, copper, n-hen iiintle the cathoc!f. cit’ ;in elcrtrolytir wll containing the same strength of a r i t l , h:is :t niiich gi,e:-!tt~ e n t ~ g i z i n gcfiwt upon the hydrogen tlepositctl ulmn its siirf:we than is >hoivn hy lent1 iint1c.r t he sniiie circiiiiistanccs, nntl prodiices nlmiit thrcc times a s niiich aiiiiiionia 3 s the latter. Thjs conclusion is iiiiich inore cniphaticnlly iniprc~~wctl on tis 1 ~ the y compnrison of the results with lcntl m t l silx-ci.. “Pintxe in the clectroly of nitric ?.rid with :I cwist:int ciirrcnt, ihc :itation of the niet:dlic ion. i q esclutlctl, nntl the reduction iniist l)c rffcvtctl hy tho hytlrogcn ions’ in contact with thc ciirfncc of the tlifiercnt rn(>t;ils,m i l sine(., t h c r e f o ~ e the . hydrogen ion. in cont:ict n-ith the rcywciiw nictds, is the sole rctliic’ing agent. it follow that the widely tliffcrcnt jn.oclucts olit:i.inctl ~ \ - t t ~ i lex1 ani1 silver tlissolvc in nitrir :Ic.itl cmnot 1 : nttrihiitctl ~ t o t h e :iction of tht. liytlrogen ctoiiis alone, lmt niust also involve the c1ircc.t tlcositlizing cffrct of the iiictals themsclrcs. I-ntloiihtctlly with the conccntrntetl. ant1 hence not ionized, ncid. the iiirtals d o n e e:?iise the change. ))ut x i t h grcn,tcr cliliition lmth iiietnls :inti hytlrogen iinite in cniising retliiction. Tlic theory ot’ n:’.scrnt hydrogen cannot therefore l:e lieltl in regartl t o t h e stronpc.r acids. in nhich only nitrogen dioside and nitric oside arc given oft’, \nit, siillject t o the niotlifications given ahove. is iintloiibtetlly vxlitl in relation t o tliliitcl : i t i ( l . ” Let 11s first consitler the case of Ieatl. silver m d coplier. Thc protlucts of retluct’ion are the ssnie whrn the reduction is electroclier~ii~:~l, hence there is no catalytic effect of the metal itdeli. h i t the quaiitit)- of aiiinionia prodiicc:l at the lead cathode is higher. This at once tells i i that ~ the specific retlucing poiver of copper antl of silver is the ~ ~ i i iwhile e , that of lead is greater. This H t w t h r author 0l)viously niems ~ ! i a WR t rioiv know
:1.5
monntomic hydrogen.
ACTION O F METALS ON N I T R I C ACID
I233
is confirmed by measurements of their hydrogen overvoltages1 under comparable conditions. Lead was found to be highest, while silver and copper have approximately the same value. The difference be‘ween the electrochemical process and the chemical process can be due only to the metal ions in solution. When lead reacts with nitric acid, under the conditions stated, the chief product is nitrous oxide, hence the nitrous oxide.is the prorluct of a secondary reaction, and this reaction is catalyzed by the lead ions present a t the place of reaction. Another way of saying the same thing is: the velocity of the secondary reaction is so great as compared t o that of the direct reaction, that the intermediate products yield nitrous oxide almost entirely. This point is confirmed by the work of AlexeeP who found that ammonium sulphate upon oxidation a t a lead peroxide anode gives chiefly nitrous oxide and nitrogen. The nitrogen is undoubtedly the product of the interaction of hydroxylamine and hyponitrous acid. He also notes that in an excess of ammonia, the oxidation goes entirely to nitrogen while with an excess of sulphuric acid yields only nitrous oxide. This parallels the observed effect of acid on reducing agents, that is their reducing power decreases, hence the oxidation can be carried farther in their presence. K h e n silver and copper, respectively, react with nitric acid, under the stated conditions, the chief product in each case is nitric oxide, K i t h the same reasoning as above, we can say that the secondary reaction yielding nitric oxide is catalyzed by silver and copper ions respectively. The case of copper was treated very thoroughly by T ~ r r e n t i n ewho , ~ proved the catalytic effect of the copper ions on the reaction by carrying out the chemical reaction under such conditions that no appreciable concentration of copper ions was present. Under these conditions ammonia was formed. In the analytical methods for the reduction products of nitric acid develoFed by M i l l i g a ~ nitric ~ , ~ oxide is determined by a,bsorption in a solution of two per cent nitric acid in concentrated sulphuric acid, the njtric oxide being determined by estimation of the nitrosylsulphuric with permanganate titration. This method of procedure a t once causes one t o wonder what would happen t o the nitric acid if such a solution were to act on copper. The nitric oxide would not be evolved because by definition it is absorbed, and hence, we have a product in solution (a more correct term would be, in combination), capable of further reduction, in contact with copper, a metal with a reducing potential high enough to cause ammonia formation. The answer is obvious. Ammonia should be formed. Foerster: “Elektrochemie wasseriger Losungen,” 310 (1922). J. Chem. SOC.98 11, 98 (1910). 3 Trans. Am. Electrochem. Soc., 10,49, (1906). See also Bancroft: 9, 13 ( I Q O ~ ) ,J. Phys. Chem., 12, 118 (1908); 28, 479, 798 (1924). J. Phys. Chem., 25, 544 (1924).
* J. Russ. Chem. SOC.,41, 11.55 (1909); abstracted
1'34
E . J . JOCS
ACTION O F METALS O N N I T R I C ACID
1235
(shown by a deep color with an almost colorless solution outside) gases were evolved rapidly. A sample of the gases evolved analyzed ninety one per cent nitric oxide. The concentration of copper ions necessary for the production of nitric oxide is fairly sharp. This is probably one of the reasons why the analytical methods were a+ all useful. A further point to be settled is why sulphuric acid is desirable in these methods. It has been shown that sulphuric acid increases the solubility of nitric oxide in nitric acid. Thus any sulphuric acid in the mixture decreases the tendency for nitric oxide t o be evolved. I n connection with the above effect of the sulphuric acid, a study was made of the evolution of nitric oxide from solutions of nitric and sulphuric acids, of varying concentrations of the latter acid, when acted on by copper. The addition of sulphuric acid should causle a slight retardation of the time necessary for the first bubbles to appear. A solution of oxide-free nitric acid (17%) that would caupe the evolution of nitric oxide in ten minutes, when acted rjn by copper was chosen. Solutions of this Concentration of nitric acid with varying concentrations of sulphuric acid were allowed to attack copper and the time noted when bubbles fi-5: began to form. As was expected a retardation did take place, a solution containing one per cent sulphuric acid required twenty-one minutes while three per cent sulphuric acid required forty-six minutes. However, higher concentrations than this decreased the time necessary, until a t eleven per cent sulphuric acid the copper was attacked faster than with nitric acid alone. This is explained on the basis of increased acidity causing increased corrosion. The reason we get a maximum in the first part of the curve is due t o the facto- of increased solubility first being the determining factor, while later the increased acidity becomes the determining factor. The curve s-ould approach infinity a t very high concentrations of sulphuric acid, as has been shown by the experiment in which ammonia was formed by the action of copper on a mixture of nit-ic and sulphuric acids. The accelerating effect of sulphuric acid is apparent in the action of ferrous sulphate on oxide-free nitric acid. Here the retardation cannot be noted for ferrous sulphate will not attack pure nitric acid in a reasciiable length of time.’ This will be discussed later under nitrous acid. This leaves the problem of which ion of copper, the cuprous or cupric, is the catalyst. At first thought one would think that at no time was there an appreciable amount of cuprous ion in the solution to came a catalytic effect. This is quite true, but the concentration we must consider is the concentration a t the point of reduction. This of course is a t the metal surface where the solution of the copper is taking place in two stages, first the oxidation of metal to cuprous ion, and secondly the oxidation of cuprous to cupric ion. This is confirmed by Mellor: “Modern Inorganic Chemistry,” 519 (19121.
I
236
E. J . JOSS
lit tlc 11111c11
111l l ( ~ l 1
IIttlP -0111c .(IlllC
I10
t lace I10 110
little littlv 110 I10
110
tr:\cc
.kcitls ijriite tliliitc
110 110
110 110
I10 110 110
I10
€ctrotlc. \Yith plntinimi. no :miiiioiii:i or
nnl:?l p:1111:it et I poli-hctl
ACTION OF ?dETALS ON N I T R I C ACID
I239
Tafel is interested primarily in the yield of hydroxylamine. I n a few of his experiments he reports also ammonia yield and quantities of the nitrogenous gases evolved. The gases were not analyzed however. The amounts were always small. Table I1 shows the general order of his results. One must not be misled by the apparent excellence of his analytical results. The data are not reproducible within ten per cent. In the experiments shown the conditions were: “0.4 grams nitric acid, 2 0 ccm. j o y 0 sulphuric acid, I O sq. cm.,’ cathode sud”ce, 2.4 amp., cooled with ice.” The amount of hydroxylamine produced is greater as the solution of nitric acid is more dilute; in ordcr to obtain the largest proportion, a very dilute solution of nitric acid should be ueed, and more added as the reduction proceeds. When the amount of sulphuric acid present falls below forty per cent, the quantity of hydroxylamine formed decreases appreciably, but otherwise the concentration of the sulphuric acid exerts no influence on the reaction. Tafel performed a few experiments on the action of various metals on a mixture of sulphuric and nitric acids. He finds that with mercury or copper neither hydroxylamine nor ammonia is produced, and with lead, reduction to nitrous acid takes place. I n the production of hydroxylamine, the current yield is greatest for lead, slightly lees for amalgamated lead, and decidedly less for copper. The presence of nitrous acid is said to exert no appreciable influence on the reduction of nitric acid a t cathodes of these metals. Incidentally he finds that hydroxylamine sulphate is very stable in the presence of sulphuric acid, and the solution may be warmed to about 40’ without any decomposition taking place. Tafel believed the electrolytic reduction of nitric acid would prove the best method for the prsduction of hydroxylzmine. This method is being ueed for its commercial production a t present.2 He confirms the work of D i ~ e r s on , ~ the effect of sulphuric and hydrochloric acid concentration on hydroxylemine yield. They both find that with increasing sulphuric acid or hydrochloric acid concentration the hydrosylamine yield increapes, although Divers finds that in the caee of hydrochloric acid the yield passes through a maximum. A point of much interest in connection with the effect of sulphuric acid is noted in a paper by Patten4 “Thus, the conclueion of Tafel is confirmed, that the ammonia yield increases with increased concentration of sulphuric acid at a copper cathode, and that the hydroxylamine, on the other hand, forms in greatest quantity in dilute eulphuric acid.” 1
The author gives this as “qdm.” which is doubtless a misprint. J. Am. Chem. Soc., 38, 2042 (1916). J. Chem. Soc., 43, 45.5 (1883. Trans Am. Electrochem. Sop. 1 2 , 3 2 j (1907).
* Srhoch and Pritchett: 3 4
I240
E . J . JOhS
ACTION O F METALS ON NITRIC ACID
I 241
decomposition curves and time current curves a t constant cathode potential, to follow the reduction, instead of analyzing for the products formed, He finds that hydroxylamine lowers the decomposition voltage of twice normal sulphuric acid a t a mercury cathode only slightly, while twice normal nitric acid lowers it almost 0.4 volt. This indicates that nitric acid is a better oxidizing agent than hydroxylamine under these conditions, hence, when undergoing reduction, will be used up instead of the hydroxylamine. This of course leaves a loophole, for we have said nothing about the oxidizing power of nitrous acid. Later in this paper it will be shown that nitrous acid is a slightly less powerful agent than is nitric acid, which clears up this point. There is much of Flaschner’s uTork that does not agree with Tafel’s. We could not expect a very close agreement however, for Tafel used principally fifty per cent sulphuric acid solutions while Flaschner’s highest concentration of sulphuric acid was about ten per cent. Flaschner finds that hydroxylamine and nitric acid also, have approximately the same effect on the decomposition voltage of a twice normal sulphuric solution a t a copper cathode, as was found with mercury, except that the hydroxylamine was a better depolarizer a t the copper cathode. This does not a t all agree with the fact that Tafel could get no reduction of hydroxylamine a t a copper cathode, while it was reduced a t an amalgamated cathode. Flawhner was able t o reduce hydroxylamine t o ammonia a t a copper cathode. Tafel and Bahl’ pointed out, that Flaschner did not reduce hydroxylamine under the conditions stated by Tafel, in the experiment in which hydroxylamine mas not reduced by electrolysis a t a copper cathode. They say that addition of water causes hydrolysis of the hydroxylamine salt and it is probably this hydrolyzed part that undergoes reduction. If Tafel’s results are right, and quite likely they are, it means that in fifty per cent sulphuric acid, hydroxylamine does not depolarize the decomposition of sulphuric acid a t a copper cathode. That the depolarizing power of hydroxylamine does bear a queer relationship t o the acid concentration, is borne out by the report of Muller2 who finds that the oxidizing power of hydroxylamine is greater than either nitrate or nitrite in alkaline solution. Muller observes that this is paralleled by the fact that he has never been able t o obtain any appreciable yield of hydroxylamine in the electrolysis of neutral solutions. The reason why Tafel was able to get reduction of hydroxylamine to ammonia a t an amalgamated cathode while he was unable t o do so a t the copper cathode is without doubt due t o the high hydrogen overvoltage of the former. Frankly, we do not thoroughly understand the relation of sulphuric acid to hydroxylamine yield. The reduction of hydroxylamine t o ammonia a t a copper cathode, under Tafel’s conditions, should be tried, and if no ammonia is obtained the del
2
Z. nnorg. Chem. 56, 375 (1908). 2. nnorg. Chem., 26, I (1900).
I241
E. J. JO5a
ACTION OF METALS ON NITRIC ACID
1243
Tafel showed that mercury acted on nitric acid t o give nitric oxide. “It has long been known that in the presence of very concentrated sulphuric acid mercury reduces nitric acid readily t o nitric oxide, and mercury is used in Lunge’s nitrometer for the quantitative determination of nitric acid. That a more dilute acid acts in the same way and that not a trace of hydroxylamine is formed by the action of mercury on nitric acid, has been proved by me in special experiments so contrived as to duplicate the conditions of electrolysis as closely as possible.” The results of his electrolysis experiments have been given before. These experiments, together with those on copper form the basis for his entire objection t o the theory of the identity of chemical and electrochemical reactions. Wilkinson,l working in this laboratory, straightened out this point in an admirable fashion, using the EiZme method as Turrentine2 used in the case of copper. He repeated that experiment of Tafel, in which a solution of nitric acid and sulphuric acid was electrolysed in a diaphragm call, with a mercury cathode. The only difference between the experirnats of the two investigators was that R7ilkinson covered his cathode with a layer of mercurous sulphate. Under those conditions he obtained sixty per cent of nitric oxide. The electrolysis had taken place in the presence of mercury ions which catalyzed the decomposition of an intermediate product yielding nitric oxide.
To further prove his point he let mercury act on nitric acid in such a way thp-t no high concentration of mercury ions could build up. Under these ccnditions hydroxylamine was obtained. Thc above experiments furnish the proof that when mercury acts on nitric acid in the absence of mercury as ion, the product is hydroxylamine, whether done electrolytically or chemically. Divers3 proved that iron acts upon a mixture of nitric and sulphuric acids to give ammonia. He was not interested in gaseous products, so these were not analyzed. Freer and Higley4 found that iron acts on nitric acid t o give nitric oxide, nitrogen peroxide, and ammonia, the nitrogen peroxide yield increasing and the ammonia yield decreasing with an increase in acid concentration. The nitric oxide yield increases slightly with increasing concentration of acid. K h e n electrolyzed5 a t an iron cathode, nitric acid gives the same products as when acted upon by iron, but the low concentrations give little or no nitric oxide. This indicates that the decomposition yielding nitric oxide is catalyzed to a slight extent by the presence of metallic ions. Trans. Am. Electrochem. Soc., 13, 309 (1908). Soc., 10, 43 (1906). 8 J. Chem. SOC.,43, 453 (1888). Am. Chem. J., 21, 377 (1899). Freer and Higley: Am. Chem. J., 21, 377 (1899).
* Trans. Am. Electrochem.
ACTION OF METALS O N N I T R I C ACID
1245
Assuming that Personne and Maumenk are right regarding the ammonia from the action of nitric acid on antimony, it becomes apparent that the specific reducing powers of antimony, arsenic, cadmium, nickel, and cobalt are sufficient t o reduce nitric acid t o ammonia. With molybdenum and bismuth we have not sufficient data to say with certainty, but from their position in the electrochemical series (both above silver), in seems quite likely that they too have specific reducing power enough t o form ammonia from nitric acid.
Catalytic Action of Nitrous Acid’ In 1842 ;\lillon2 showed that copper reacts with twenty per cent nitric acid a t zoo only when nitrous acid is present in the solution. When no reaction occurs, bubbling nitric oxide through the solution immediately starts the action.3 The reaction is negligible if some substance such as urea, which uses up the nitrous acid as fast as it is formed, is present in the solution. Millon considers that nitrites are the first products formed, and that these are decomposed by nitric acid yielding the metallic nitrate and regenerating the nitrous acid. He worked with copper, mercury and silver. Russell? confirmed Millon’s results, and found also that bubbling a stream of an inert gas through a solution of nitric acid, in contact with a metal, hindered the action of the nitric acid by removing the nitric oxide, and thus keeping the acid free from nitrous acid. He also postulated the intermediate formation of nitrates. TTeley’s5results confirm the results of previous investigators. He interprets his results on the assumption that nitrous acid is reduced t o nitric oxide. “The experiments detailed above for the metals, copper, mercury, and bismuth, to which, according t o Russell’s experiments, silver must also be added have established the following facts:-(I) the primary change is that between the metals and nitrous acid; ( 2 ) no gas is evolved a t first from the surface of the metal; (3) the amount of nitrous acid increases up t o a constant and maximum proportion; and (4) those conditions which increase the amount of metal dissolved per unit time, are equally those conditions which increase this constant proportion of nitrous acid. If then a trace of nitrous acid becomes once formed, and if, also, the quantity of nitric acid is present in very considerable excess, it would appear that these results may be explained as follows:-(I) the metal dissolves in nitrous acid t o form the metallic nitrite and nitric oxide; ( 2 ) the nitrite formed is decomposed by the excess of nitric acid t o reproduce the nitrous acid; and (3) the nitric oxide formed in ( I ) is not For more detailed account of early work and his formulation of the problem, see Bancroft: J. Phys. Chem. 28, 973 (1924). * ComDt. rend., 14, 904 (1842). , . 3 This was noted also in the case of iron by Young and Hogg: J. Phys. Chem. 19, 647 (1915). J. Chem. SOC.,24, 8 (1874). Phil. Trans. 182 A, 279 (1891). See also Bancroft: J. Phys. Chem., 28, 973 (1924.). I
I 236
E . J. J O S S
evolved as such. hut reduces the nitric x i t l or the nitrnte t o proclucr. a trirthcr quantity of nitrous acid. T h e v changes may he presented thiic, taking coplwr for es:?mple :
“-It this point the n i t r ~ t i sw i i l is tlecoinposeti :is fast as i t i$ foixi(v1, t h o aiiioiints of tlie reactions ( 3 ~anti ) (41 per unit tiiiic Iwing tloul)t,lcus tl(~!iciitIc~iit as in otlicr sin:ilar ca , I I ~ I ) I It lie relntive 111: c a of the nit~,oiiqnncl n i t r i c . acitls. i i p r i n tlie tcniperatiire ant1 ot1:er contiitions of t h e c~s1wriii;eiit.
,. 1 his csplarntion
is. lion-ever, only valid pim-iclctl tlint tlic iiia+ of nit IIC acid lie in very large esc over that of the i i i t r o ~ < ncitl. for i t is c\-i(lcnt f i v i i ! t h e cspci.ii~ientstlesci~ii~etl almve. tlxit. if thcro is no vci’y prcnt tlitferc~nc~c 1 ~ tween t h e ~i~:i.:icitl, thereby proclucing nitrous acid. Ahregnrtls t h i a latter point, it is worthy of remark that, of the metals e s p x h e n t e t l n-ith, mercury 11 uswpt ihlc of chemical change th:m cop~:er, nntl copi:er in its turn than liismuth. This order is equ:il t o that of their pro1)alde tleprce of I)uritp. “-4 fen- es1:criments w r e nccorclingly iiintle t o tlcterminc the nmorint of a foreign metal which c ~ ~ i l he ( l ntltleil to pure merciiry without pronioting its reaction with nitric acid. These showed that one part in a thoiia:i.ntl of p i i r ~ copper ant1 one part in eight hundretl of piire silver ntltletl to nicrcwry t l i t l not rentler the latter susceptitile of chmiical change, even with nitric ncitl of 3 0 . 5 per cent concentration, :ind heatecl to 28.;’, provitletl that h h nictnl ant1 acid were stirred continually. If hoth n-ere :it rest, t h c ~as in t h r prevvior:~ experiments, the change coiiirnencetl imrnetliarcly.“
ACTION O F METALS O N N I T R I C ACID
I247
T'eley has given no proof for his explanaticn. His assumption regarding the action on copper is superfluous for what he really assumes is that nitrous acid is the depolarizer. This involves the reduction of the depolarizer to nitric oxide. Armstrong* showed that nitric oxide is not a direct reduction product of nitrous acid, so his views seem quite improbable. If Veley's three equations are grouped the result is CU
+3
"03
=
Cu(NO3)z
+ HNO:! + H:!O
This does not preclude the possibility of nitrous acid forming an intermediate product in the course of its catalytic action, and quite probably it does in some cases as will be shown below; but we shall prove that nitrous acid is not the depolarizer. Burch and Veley2 attempted t o prove that nitrous acid is a stronger oxidizing agent than nitric acid, and would therefore be reduced first, though this was not compatible with the marked building up of nitrous acid concentration. Their general conclusions are as follows:((I. The experiments detailed above show that when the metals, copper, silver, bismuth, and mercury are introduced into purified nitric acid of varying degrees of concentration, and a couple made with platinum, the electromotive force of such a cell increases considerably until it reaches a constant and (in most cases) a maximum value. This rise of electromotive force is attributed t o the production of nitrous acid by the decomposition of the nitric acid, and the final value is considered to be due to the former acid only, while the initial value is due for the most part to the latter acid, though it is affected t o a remarkable degree by the amount of impurity of nitrous acid either initially present or produced by minute and unavoidable uncleanliness of the metallic strips and the containing vessel. ( ( 2 . If nitrous acid has been previously added t o the nitric acid, then the maximum electromotive force is reached a t once. 3. If the conditions, namely, increase of temperature, of impurity, and of concentration of acid, are such as could favour a more rapid production of nitrous acid, then the rise d electromotive force is concomitantly more rapid, ((4. Conversely, if the conditions are unfavourable to tlie production of nitrous acid, the rise of electromotive force is less rapid. ( ( j .If any substance, such as urea, be added, which would tend to destroy the nitrous acid as fast as it may be formed, then the rise of electromotive force is extremely slow, being dependent upon the number of molecular impacts of nitrous acid upon the surface of the metal. Thus the results obtained by the electrometer and of the chemical balance are in every way confirmatory, the one of the other. These results open out the further question as to whether the electromotive force of batteries, in which concentrated nitric acid forms an ingredient, is dependent not upon the nitric acid per se, as hitherto supposed ((
J. Chem. SOC., 32, 57 (1877). Phil. Trans., 132 A, 319 (1891).
ACTION O F METALS ON N I T R I C BCID
I249
small one, and consequently by some chemical change. If the chemical changes take place very slowly, the measurements of electromotive force may be affected seriously thereby.” Bancroft’ pointed out further evidence that contradicts Veley. “ R e can see in another way that there is nothing t o Veley’s assumption that the metal dissolves t o nitrite. If we add nitrous acid to the Grove cell or t o the Smee cell, we increase the depolarizing action of the nitric acid; but no metal dissolves in the nitric acid solution and consequently there is no formation of nitrite. The behavior of these two cells was discussed by Ihle2 in an earlier paper, “If one dilutes the concentrated nitric acid in the Grove cell gradually and measures the electromotive force (of the cell at the same time, one finds that the electromotive force) remains nearly constant until the concentration of the nitric acid drops t o about 38%. At slightly lower concentrations the cell shows, though at first only for a very short time, a changed and much smaller elwtromotive force than the 1.8 volt characteristic of the Grove cell; it corresponds rather to the 0.7 volt of the Smee cell. If one dilutes the nitric acid more and more, the time during which the cell stays at the lower potential increases until a t a nitric acid content of 2 7 - 2 8 5 the element keeps the lorn elecromotive force of the Smee cell permanently. “TOcarry out this experiment, one pours dilute nitric acid into an ordinary battery jar and puts into this two small, porous cups. In one of the cells is placed an amalgamate zinc rod and a zinc sulphate solution while a platinum electrode and the nitric acid to be tested are placed in the other porous cup. The two poles of the cell are connected through a galvanometer. Since the resistance remains the same in all the experiments, changes in the electromotive force of the cell will show as changes in the deflection of the galvanometer. “It seemed probable that the peculiar relation between the concentration of the nitric acid and the electromotive force of the cell was due to nitrous acid present in the nitric acid solution or formed in it while the circuit is closed. This guess has been confirmed. If one adds a small amount of potassium nitrite to :L cell containing less than z8yc nitric acid and therefore giving permanently the potential of the Smee cell, the galvanometer needle swings at once t o the deflection corresponding to the electromotive force of the Grove cell. If one removes the nitrous acid again by adding potassium permanganate,3 hydrogen peroxide, or urea, the galvanometer needle goes back at once to the original value which corresponded to the Smee cell. “.Just as one can make dilute nitric acid active as a depolarizer by means of nitrous acid, so one can also remove the depolarizing power of concentrated nitric acid containing more than 38% “03, by removing continuously the J. Phys. Chem., 28, 978 (1924). Z. Elektrochem., 1, 174 (1895). 3 On adding potassium permanganate there is at first an increase in electromotive force corresponding t o the high potential of permanganic arid; but the galvanometer needle goes
back to the value corresponding to the electromotive force of the Smee cell as soon as the permanganic acid in immediate contact with the platinum electrode is used up.
ACTION O F METALS ON KITRIC A C I D
1251
reaction must occur to at least a slight extent before a potential is built up on the electrode. There will always be a reaction, and in this instance a polarization, as a product of the depolarization makes its appearance in the slonr building up of the electromotive force. T’ieneg carried out experiments in which the platinum electrode was rotated. This should cause a decrease in the concentration of the nitrous acid at the electrode and, consequently (since the condition of zero flow of current is not experimentally realizable with the apparatus used), a fall in the electromotive force obtained. This is found to be the result with the dilute, but not with the concentrated acids. The reason no decrease is obtained with the latter is the fact that the concentrated acids reach their maximum electromotive force practically instantaneously. These mperiments are confirmed by the viork of Brown1 upon passivity of iron. The electromotive force of a cell, consisting of a platinum and an iron electrode clipping into a solution of nitric acid, was measured n-ith a galvanometer condenser device whereby very little current was used. The electromotive force of the cell could be follomd rapidly. Kitric acid of various strengths were used, and no precautions were takzn to have it free from oxides. The measurements were made with the iron electrode in motion, ~ n t lhen l buddenly brought to rest, the electromotive force being followed throughout the experiment. The values obtained were fairly constant throughout the time of rotation, but upon being brought to rest the electromotive force row immediately to a higher value where it once more became constant. Thiq is exactly the result Vieweg found with a rotating platinum electrode. The rotation prevertecl a building up of the concentration of nitrous acid at the electrodc, and hence polarization prevented the cell from giving its maximum value. The explanation given by Brown is as follow: “That this rise is clue to ferrous nitrgte is shown by the fact that when a quantity of this salt is adtletl to the tlilule acid2 the cell has an electromotive force of 1.j volts from the start, n-hetlier the electrode is in motion or at rest.” I t certainly does not seem probable that the rise obtained by Yieweg upon hinging hi\ platinum electrode to rest, or upon addition of nitrite. ~ v n . due to ferrous ions. The ferrous nitrate added by Brown, being a weak reducing agent, cauwtl the formation of nitrous acid. The rotation of the electrode could not sufficiently dissipate the nitrous acid concentration to allow polarization, when only a small current n-as drann, hence the cell gave the maximum value. The measurement of cells of this type and microscopic examination of the iron electrode, comprise the entire experimental work reported by Brown, and upon these determinations he bases his theory of the passivity of iron. He draws conclusions that active iron (the metal) is ferrous, while passive iron is ferric. This is the theory of S m i t ~ . ~ J. Phys. Chcm., 2 5 , 424 i19zr:. The most dilute acid used was approuimately eleven pw c-nt nitric acid. E. J. J. 3 “The Thcory of ;Illotropy,” 2.42-377 (1922).
2
ACTION O F METALS ON N I T R I C ACID
12.53
equilibrium was especially desirable in cases of this type. “I think that this critical study of the literature shows that nitrous acid is not the Pctual depolarizer but that it acts as a catalytic agent making the nitric acid active. The next question is 2,s to the way in which nitrous acid acts and I think that the answer to this question is suggested by the previously quoted statement hy Burch and T’eleyl that ‘nitrous acid ready-formed is not so active a material as nitrous acid in the alternative process of formation and decomposition.’ If one substitutes nitric acid for nitrous acid in this statement, we have made a real step forward. I have pointed out in another paper2 that any reaction between so-called saturated compounds involves ejther a preliminary dissociation which means the breaking of a regular bond, or a preliminary association which mems the opening of a contravalence or residual valence. In the raee of a solid catalytic agent the thing of real importance is not the formation of 2 definite intermediate compound or the formation of an adsorption complex but is the formation of the free radical, taking the term in its broadest sense-which is the real reacting substance. In so far as me can duplicate photochemically the action of a contact catalyst this must be so, because light can x t i v a t e a substance only by opening a normal vplence or a contravalence. T e can word this in a different way, if it will make it clearer to anybody, by saying with Daly3 that activation means opening up fielle of force, because ‘it is a necessary deduction that the condensing together of the lines of force must result in a decrease of chemical activity, and, indeed, it would seem to follotv that the true chemical affinity of any molecule cannot be exhibited until the condensed systems of force lines within each molecule have been unlocked or opened by some means.’ “If we apply this point of view to the case that interests us a t the moment, we pee that a dynamic equilibrium must represent a more active state then a static one. In the case of the reversible reaction
HXOa
+H N O z S
2
+
S O 2 (or N204) H20,
the nitric acid is continunlly going over t o nitrogen peroxide and being formed from the latter. During the moments of change and nitric acid, the nitrogen atom, or some radical containing nitrogen and oxygen must be in a different state from whst it would be if no nitrous acid were present and if this reaction were not taking place. If the particular intermediate stage involved in this reaction is one which permits more rapid reaction with nascent hydrogen, our problem is solved. Actually. we have merely formulated our problem in a new way.” Bancroft’s summarized etatement is :“It seems probable that the activation of nitric acid depends on the dynamic equilibrium represented by the equation “ 0 3
+ HXOn a
Phil. Trans., 182 A, 330 (1891).
2
* Bancrcft: Ind. Eng. Chem., 16,270 (1924). J. Chem. Sac., 101, 1469, 1475 (1912).
KO2
+ HzO.”
ACTION O F METALS OK XITRIC ACID
I 2 jj
The presence of nitrosic acid in solutions of nitric acid may be explained on the basis of the work of Reynolds and Taylor,’ who found that light de.composes nitric acid vapor according t o the equation: 2
HNOa
= 2
+ HzO + 0
NO:!
The formation of nitrosic acid is nothing more than the intermediate step in one of the reactions for the preparation of nitric acid,*by fixation of atmospheric nitrogen in the arc process. Most reducing agents will not react appreciably with nitric acid in the abFence of nitrosic acid. Ferrous sulphate, sodium sulphite, stannous chloride, and titanous chloride react with pure nitric acid only after long standing, hut uron atltlition of a small amount of a nitrite the reaction occurs with great rapidity. Rourgoin, Schonbein, and Gladstone and Tribe3 noticed, that upon electrolysis, solutions of nitric acid first gave an evolution of hydrogen, which ceased allruptly in from three t o forty minutes. Gladstone and Tribe traced the suc!den stoppage of the evolution t o the presence of nitrous acid in the solution. Thus it is seen, as was shown by Luther and Schilom-4 that the presence of nitrosic acid is essential for the reduction of nitric acid, whether it be by cihemical or electrochemical means. Quite likely the reason T’eley and others felt it incumbent upon them to show the formation of a metallic nitrite, is because in some cases it is poseible to separate the nitrite from qolution. In the case of silver the nitrite is only slightly soluble while the nitrate is very soluble, hence silver nitrite may he crystallized from the reaction mixture. Mercury offers the same problem. This was encountered by RAy.j ‘Werciirous nitrite is the sole product6 other than water, of the immediate action of mercury on dilute nitric acid containing any nitrous acid. I n the absence of the latter acid, the former is quite inactive. The formation of the salt may be expressed by the equation 2
Hg
+ X02H + HO.SO?
=
+
Hg*(SO*)? H:!O,
but it is in part converted by nitric acid into nitrate and nitrous acid. At suitable temperatures and in the presence of the right proportions of water and nitric, acid, a stage is soon reached in which the quantity of nitrite produced by nitrous acid is just balanced by that decomposed by nitric 2cid. There then ensues for some time a steady accumulation of mercurous nitrite, (luring which the production of permanent nitrite is due wholly to nitric acid, as expre-sed by the equation : -IHg
+ 4 HKOs
=
+ Hgz(K0s)z +
Hg*(N02)?
J.
