Denitrification of Chuquicamata Copper Solutions Citation Hoeppel, Raymond Winfield (1931) Denitrification of Chuquicamata Copper Solutions. Master's thesis, California Institute of Technology. doi:10.7907/sc6j-cf26. https://resolver.caltech.edu/CaltechTHESIS:09032024-210245990 Abstract No abstract. Item Type: Thesis (Master's thesis) Subject Keywords: (Chemical Engineering) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemical Engineering Thesis Availability: Public (worldwide access) Research Advisor(s): Lacey, William Noble Thesis Committee: Unknown, Unknown Defense Date: 1 January 1931 Funders: Funding Agency Grant Number Chile Exploration Company of Chuquicamata, Chile UNSPECIFIED Record Number: CaltechTHESIS:09032024-210245990 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:09032024-210245990 DOI: 10.7907/sc6j-cf26 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16691 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 04 Sep 2024 16:23 Last Modified: 04 Sep 2024 16:23 Thesis Files PDF - Final Version See Usage Policy. 10MB Repository Staff Only: item control page

Deni trifica tion of Ohuquicrunata Copp er Solutions Thesis by Raymond W~ B:oeppe1 In partial Fulfillment for the Degl"ee of Master of Soienoe Oa11fo:rn1a Institute of Technology Pasadena., California 1931 -1- At· OhUq't,li<:amata, Ohile •is l ocated · one of the world• a larg-@st . • coppef mines. The o·:re ii? Braehantiie, and · ca,:rries a n average · of 1 .fl~ oop:pe~. mainly as anso,,. 3 Ou(OE:)'1. • The ol"e is orushed •an d: _leached with epent ele.o trolyte, v1hich oomes from eleot-:rolytio t anks~ Thf $'b:t"ong 80lut1on ( t.e .• solution strong in copper) :trom the lea ching vat-s ie d.eohloridized with cement copper and sent to the eleet:ro.... lytto tal'lks where the coppe:r is removed by passing eleetricity thru the aolutton by means of insoluble lea.d... ant1mony anodes. Theepent eleot:rolyte. ·which contains a :fai.t amount of free aul phu~ic acid.) is sent baok: to the leaching vats. The copper cathodes are later melted and ca.st. Thus it is seen that the process is a oyolioal one. It is not necessary to e.dd sulphuric acid during the leaching; in fact, during each eycle, an a.mount of sulphu:rie acid is gained which is equivalent to the oopper sulphate content of the ore. Hence , during _each cycle, a certain amount of spent electrolyte is rep1¢a oed. by vrater, and , in this way, various othe:t salts which occur in the ore i,,:re kep t from accumulating in the solutions. Of the 1mpuri tte.s • the lee,st desi:rJable are chloride, fe:r r ic, mo1ybate and nit:rate_ions. Chlorides are deoomposed into chlorine in the electrolytic tank thu~ lowering the current eff iciency , and, moreover; are deposited upon the cathode as eupr,,us chlori de, which , d1:1ring the subsequent melting of the cathodes , vo1 atilizes, re su.lt- ing in a copper loas. Ferrio ion lowers the current efficiency by beooming :reduced. at the eathode, migrating over to the anode, and. / then being o·xidized. :Molybdate¢ ion tends to catalyze the reaction between :ferrous ion. and nitric acid, thus oxidizing the iron and henee lowering the Glurrent effieiency. Because of its ability to exist 1n two valence atates it also may act similar to the iron. Nit:rio aei.<i• especially 1n the p:resenoe of nitric oxide, attacks lead. readily at the anode. By constructing the a.nodes of an alloy oonta.ini~g 50 %lead and 50% antimony• this effect ts greatly 1 d1min1ahed.t l;>ut the anodes have the disadvantage of being both brittle and costly~ FJy disoal"d1ng portions of the spent eleotio- lyte, as was mentioned above,. otber ions in solution. nat:n$lY -;-rt N'a. •• + • K, + • ..:rra,n· d. A ·1+++ • ean ""e man 1 t a i ne d a t· ooneen t ra. ti;ona As• Mg -i.. whioh ate not harmful~ Cblotide ion haa in the past been temoved by passing the soluti on over a oopper surfaoe, which precipitates cuprous ohlo,:.... ide according to the rea.c ti&tu -\"T Ou + 201+cu.~OU2,Cl1,. By this means the chloride ion eoneentration oan be deorea,si- ed. to 0.2 g/1. Recently~ a method known as "The L1merook Prooess" has be en developed on a semi-oommeteial soa.le at Ohuquica.ma.ta in order to partially J:"emove molybd.enuni and il"on from the strong solution ( i _e. the solution f:i:-om the lea.ohing vats). In order that t.hie process be eeonomica11y feasible it i11as neces sal'y to develop a di:ffetent method of leaching in o:rde:u to obtain strong solutions whi~h carried a small pe:r:-ee:nt of free sulphu.rie acid. In the pa.st a system kno~ as the normal lea.oh was carried out which developed a strong soluti9n carrying 40 g/1 of oopp-e:r and 35 g/1 of acid, from a spent eleotrolyte ca.wrying 15 gji of ooppe~ and 70 g/1 of aoid.. ne new counter ourrent lea.eh, using the same spent eleo..,. tl"olyte,. develoi,s a strong eolntion catry1ng SO g/1 of copnet a.nd 8 g/1 of acid. In the lattew solution,a.11 of the other elements ?$main in the same coneentration except molybdenum which builds ) -~ up to three times the eoncent:ration of that element in the normal leach solutions. The oounte:r current strong solution is treated wtth potassium chle.ra:te to oxidize all of the iro-n and then with liii:erock t1nti1 an a.1nount bas been added. which will just precipitate Fe(OH) 3 but not yet 01100'>. ?his end point is determined by noting when a brown ring of f e:rrio h.ydroxide :f'orme on e. filter pape:r thru which 5c. o. of the s·olution ha.e passed. At thiEJ time1 gypsum is added for a fi.1 ter a1d 1 and the preeipita:tes of Fe(OH)~ , Mo( co~)) washed and discarded. and. Oa.80-t is fil te:red, The exoess waah is stri.p:)ed of 1ts o-op.p~r by adding an ex<:,-e.s s of limerook and filtering of'f' the basic ·carbonate precipitate of ooppe:r ~1th an Oliver filter; the filtrate is dis. carded, This preoipi tate is later added to the coutrter curr ent strong solution just before adding the first limerock. The filtrate from the iron and molybdenu111 preoipitation is sent to a blending tank where it :ta mixed with strong solution ftom the normal lea.oh; the tota.l solution 18 deehloridi.zed and sent to the eleot:rolytio tanlts. The spent electrolyte ie l."ecireulated, in ao- oorda.nee with the ~eed, to each of the two leaching vats. The nor- mal lea.oh handles 62}'% ot the total ore while the countex- current lea.oh handles the rebind.er. Thu.a fa:r, no system ha.s been developed at Chuquioama,ta for the :fem(l)va.1 of nitrie acid from solution. Sinoe, in the counter cut.rent leach, it ia neoessary to oxidize ferrous iron with limerook in otde:t to preeipita.te ferric hyd;roxide. it was suggested that 1t would be most eoonomio·a l to deni t:tify the counter cur r ent strong solution by oaueing the :f'er:rous ion presettt to reduce the nittie partially deni t:rify 1:fhe solutton a:nd oxidize the :tron at the srune tlme. \hus $avtng 'the e~ense of oxidiztng the solution with a cost- ir $alt such as l)Otaestum ohlo1rate. Sinoe not enough ferrous ion ia present to oomplet.e ly deni trtfy the solution} it was suggested. th.3,t the fet:rie ion, · i,eaul ting from the fi:rat reduction of ni t:tic acid• miv,ht bc. . . be reduoed with ooppe:r: and again" al lowed to fur·ther deni trify the a-olut:l.on. Thie pt·o oess mif;ht be repeated. until all of the n1 tric acid. had been removed, If it were not possible to denitrify this strong eo1ution 1 it was 1"eco1mn<;mded that attempts be m~.i.e to denitrify eithel' the spent electrolyte o:r the normal strong solution and thus remove ni trio acid f:roni the s ys tem as a whole .• Five thousand pounds of o:te 1 in thirty Si.J<. saoki; , were shipped. from the mine to the Oalifoamia !nsti tute of Teohnology . Nine sacks contained high nitrate o:re 1 rtin.e others high molybdenum ore, and the remaining eighteen oonta1ned l'egulai ore. Analysis of corner; ohlor.... 1.de. nitrate, and moly'bdenurt1. were s u.bmi tted with the shipment. Final results ought to be obtained. 1ronnsolut!ons aotua11.y deriv-ed from the ore 'by leaohing, although p1'e11mina:ry experiments might be made upon synthetic solutions resembling actual Ohuquicama ta solutions and. made up a.a follows:.: (See page 5) Oount:er 0111~rent strong Solu.tton OU as cuse,, e.o Worm.a l Soent Elect~o1yte 40 15 Strong Sol 'l.rt1on 70 H,. S6c., 8 Fe &$ Feso ,.. 1.5 1.6 1.5 a s As ).O~ 0,2 0,2 0.2 Al /0,$ Al .i_{ so '{h 8.0 2 .. 0 Mg &$ Mg86'"' 0. 5 0.5 Na as '.M a,_Sl.Yir ,. MaOt , &N a.N03 11 . 0 11 . 0 l( 8.6 I{ l $0'1 2 1!'5 2 .5 Cl as NaCl 0.4 0.4 0.2 HN03 as NaN03 4.0 4.0 4.0 Mc as ~io03 0.5 0 .15 0 .15 AS 0.5 In studying the r eaction bet ween ni trio aoi d a,nd f er ~cous iron * and also other reactions ap')e.aring 1a ter in this pap ,~r 1 a solution ( tiesigna ted hence a.s the test s olution ) 1Na.s made up whiob conta.ined 4. 0 g/1 of n1t:ra.te., 1.5 g/1 of fe.r rot1a. tton e.nd 70 g/1 of sulphu ll'io acid. :tt i s to be noted th ai;t this solution contains t hose s alts of the ::roent electrolyte which ar e a ctive in the ab ove reaction . E,;cep-b for the ac tual J.ea ching Bolu tions, all reagents and synthetic solut.i ons were made from O.P . ehemioals, us i ng distilled water as the s olvent. The quantita tive system of a.na.lysis usecl was th at ~dopted at the plant . Because C?f the gr eat numbel' of analyses to be run, no effo rt wa.s made t o at t a in better than 1:t a.c cu.-ra cy. fer:rot1:s i:rton was tlet e rmined by tit:r ation in a sulphUl' iC a cid Scolution 1;11 th stanc..1ard potas Gium per man ganate solution. 0:tten occurr ed due to tedu.ced molybdem:un and a rsenic . small eri·o~s Mi tric a cid was dete:rminEtd. by :1:ntroduoi:ng the s ample. a ftew just ox:i,dizing it with potassium permangana te, into 25 e.c. of oonoentrated aulphurio a.oid and ti t:r:ating with a sulphuric aoid solution of fe:r:rous su.1pha;te, the end. p Oil)t be i ng ta.ken when the chara;oteristic brown color of the +-'T FeNO eomplex is noted .. Sulphuric aoid was det e:rmined by titrating the sample, after dilution wt th water, w1 th standard. sodium carbonate solution.,. using mein-,1 or a,nge as an indioator. The well known iodide method was used in analyzing for coppe:r,after making a. few alt erations. The sample is first oxidized with hydrogen peroxide., evapora,ted. ta dryness to expel nitric acid, :redissolved in water, neutralized with ao!ium oa.r bona te and aoidif'ied with aeetie ~.oid; sodium fluo:ride is a,dded to repress the itmizatton of iton. Potas @i.um iodide is then a,dcled and the iodine libe::rrated is titrated with sodium Thioeulpha.te stanclasd solution in the usual wa.y. For determining chloride, the Volhard method was used, exoept that the silver chloride prteipitate was first filtered out of solution 'befol'e titrating with standard ammoni'lll'.r1 thiocyanate; thie 2.1 teration was neces s ary in order to ob- tain a. permanent end point. No effort was made to analyze for othe:r constituents. During the greater part of the time, experiments were made in order to study the reaction of nit:i:-io acid and ferrou s iron in dilute sulphu;irio acid eolution. Solutions of f'r ,)m five to ten 1n111ili ters in volume were introd.ueed into te st tubes 1ind t h ese in turn were immersed in a, constant temperature water bath which oould be maint e,ined at a.ny temper a:ture betw,een thirty and eighty de grees oentigra,de, with a fluctuation in temperature of les s than one half degree, oould acoomodate fifteen test tubes. tes ts were madet va,rying in • b The b ath More than t wo thous and separ ate eae . oa.se the conditions a.ffeeting the ;re~otiott. ( i "'e. oatalyete, acid oono·entratton, temperature, s al ta, ete.) Reat:rH.~ et Nitrie Acid with :F.erroua Iron The reaction studied is represented by the following equatiorti ( 1) -r+ .. -t . 1"t-t + N0)+4H~3'.fe + N0+2H.i_O l3 Fe The following mechanism tends to hold ni trie oxid.e in solution: ( 2) Fe+-++ wo~-rer. ro+-t++ Thia FelfO . . . t,., complex is easily broken up on passing air thru :u.e solu... tion; and moreover, when the par·ti a.1 pressure of nitric oxide is lowered sufficiently. it diffuses out of solution and the eomplet slowly spontaneously decomposes. In studytng the r ate of r eaction ( 1), the time we,s noted :for t h e fir s t ap 1eara.nce of a. brown color ( the Feno-t+ complex ) and also for it s fi nal d t s appe o,r ance when all of the f e.r rous iron had been oxidized to the ferric st ate. AOOELERATING FACTORS: The re action was :found to be accelerated by: (1) An increase in temperature, (2) An increase in hydro gen ion ooncentration, (3) An increase in the concentration of both Fe-,.-+ & N◊-3 0 - 1" t (4) An Inorease in the weight r atio of N0 3 to Fe (5) Various oa t alys.ts, Of the oe.talysts I the bes t was found to be nitric oxide, one of the :r ea.otion products; henee the reaotion i,s autoc at alyzed.. The r ate of reaction inol.'eaees a.s the oonoent:ration of nitrio oxide is in.... creased. Silica gel was found to be the best of the insoluble ca t al yst s . no doubt due t o it s abil ity to adsor·b l a.t ge quantities of ni:lirio oxide upon its su:rfaoe and reta i .c. it ev·e n a ft er washing . Hence the rea ction read.ilvJ t ak e~ pl ac e uuon tt s surface due to the hisi:h .i,; ,/, '-"' oonoentr a:tion o:f the t:rue cat alys t , ni trie o:.iti de, W h ich exi s ts .t here . The action of the gel Ls a.cJe·e ler&,ted if, before use;. it is s at tir ated. with nitric oxide; tt impwove.s in 11se wh en i :n conta ct with sollltions containing thi s gas .. However on continued wash t ng or drying; it loses much of its !:>Ower as a cat alyst• due p.e:rhaps to the :re~noval of ni tr.ie oxide, bu t it can be revtved by eont aot vii th the gas. Porous pl a te, eruahied tif\1a.rtz, and ciiatomaceous earth behave simil a r l y to t he gel but muoh les B eff eotively1 d.ue no doubt to their smaller surfac e of oont aet * Molybda te ion somewha.t a ocele~e.tee the ini ti a.1 forma,tton of nitric oxide in an unfouled solutioil (i . e. a solution devoid. of nitric oxide), especially upon long standing, but it tends to retard the completion of the :reaction . Solutionsj .if they contain this ion, tend to foul within 8; short time due to n itric d ecomposition . Ferric ion alse sligh tly ca t alyzes the r ea,otion . Oh1orid.e ion s~riousl y ret ai:de rea ction (1) (by 51~ even in an all ooneentrations of 0 . 2 g/~, probably because of a complex compound which this ion forms with nit~ic oxide. thus diminishing the activity of the later . In acoordanee with the B:ronstead theo:ry of the salt effect it might b e exp eoted. fr om the na t ure of t he reacting ions th at the a.-ddi t i on . of sol uble s al ts to the s olution would retar d the r ate of r eatt:ion~ For a J.1ea,e tio1iu wheJte A, B; •a,nd O, are conoe:ntrations of the reacting i ons, Z is the v ale'l'lGH~ charge on. the ion, o is the concentration of an added salt. and K and.~ are constants. H 1 The rate for equation (J.) hence becomes: - f.. R-==- K( Fe ) ( NO,) (H -t) 1 0 -...~f2-t7.-1) J~c.,--z...->: OK + (.. 3 R ::: K( Fe*) ( M~) (H ) 10 Ren.o e as t~e :tonie $trength., _-i.(3~,. oz\ of the solution increases ) t he r a t e , a., deol'easee; the effect is greater 1.:f t he valence of the added ion ts high .. This was f~'tind to b e true . Exeept for t he reacting i ons; all added aal ts retard the · reaction and the ord.et of thei.r eff e(}.t was a.s , - <-+ -t+-t --- . -t + +- t"' follows: (1) 01, (2) MO ,f ( 3 ) Al ,ft) AS-i..0.n (5) cu .( S) Mr{ , (7) K , ( 8) Na: ( 9) NH;,. Chloride ion no doubt exerts some other marked ' influe.n ce upon the reaction. Ferr<61Ca :tron; contra~y to exp eta ti0ns , in small quantities( 2 g/1) oata.1;/zes the :r?ea ction .. Stirring the soluti on a,ocelere.tes the initial formation of Mi t:rio O:tide a$ it establishes 'b.e tter conta ct with the sol id c a t alyst . How- ever, once sufficient ni trio oxide has been formed to ca.use t he :rea,e .... tion. to occur rap i dl 1;7 1 stirring the solution 1-vill ret ard the rate of re a ction bece~us e this p:ro ced.u.re as s :i.sts the escape of the ca t alyst. ni t:rio o:idd.e; and h enc e lowers its eonoentrati on in s olution. Pa.a sing air th:ru the solution has the same effect. PA'.l'ft+ OO!{QERN! nQ ;f!.E;fi.OTI CHf 0 .): The following tabl e serves to i llustrate the effect of oonoen.... tration. t emp~rature, eatalysts, and dissolved salts upon the rate of Solution Time of reaction Temperature 0 . 4.5 c. . 45 0 45° Non.e 81.liea gel 18 Hrs~ Silica gel treated with Nf.ti!o oxide 3 Hrs. 45° As a,bove • but 60" 0 .4 g/1 NO, - 3 Min. ~00 0.4 g/1 N8i. 8 Min. ?O g/1 H,.SO.., 20 llin. Qff~~J1.C_T~tp;s1-:to~ OF REAO!IO~ ,( 1 l: A felt:>rile reaetion ia one which ie autooata.lyzed. It occms tn three stages: (1) an incubation period 1 when the concentration of t he catalyst is slowly being built up ; ( 2) an induction period., during ,;,-vhich enough catalyst has been formed to cause the reaotion to oocur rapidly; (3) a period of extinction., dul'ing which the rate gradually decreases 1 due to a defficienc-y of reagertts. Reaot1on {1) passes th:ru a11 stages but the thi:t:d, in the ea.se where the- reagent in excess is nitric aoid. During the first per:tod, nitric oxide is slowly formed, but it is heJ.d 1m oomple:t _formation by ferrous iron (see rea.etion (2)), add its power as a catalyst is thus reduoed.. Henoe the first period is very long, Wh.en ni trio oxide, in excess of that held by the com:ple:x, hae been formed, the reaction suddenly increases in :rate. Additional ni trio oxide henoe ap~)ea:rs f':to111 two eouroes at once, first, from the decomposition of nit:ll'ie aold., a.nd se,oond from the oomple~ which must give up its ni tr1e oxide because of the oxidation of fer r ous iron~ immediately 1norea.see the :ra,te. This additional gas Thus. within a :few seconds, the great- er part of the reaction occurs at a surprisingly fast i /lj.te which. in- erea.ses as the ferrous iron concentrat ion decreases. The slight cat alytic eff ect of ferric iron also may be a f actor increasi ng the r ate . From this discussion it is obv-ious why a low nitrate to ferrous ion ratio decreases the rate, REACTION (1) j\$ A.PPLIEP TO omrg,uICAMAl'.A SOLUTIONS : In the light of the preeeeding results one se es that, exoept for the test solution, the most favorable solution for denit:rifioation is the spent electrolyte, as it runs highest in acid and lowest in total s alts. Reaction (1) is complete in three hours at s o-oc. in the spent electrolyte in the p:r:esenoe of silica gel as a cat alyst. However by introducing nitric oxide,in the f orm of nj_trite ion,into the solution. the ineubation period is eliminated and one may dispense with the gel. In this case the above reaction occurs in twenty 0 . minutes instead. of three hou.r s . . At 75 C this time may be halved. Sinoe the spent eleetrolyte contains 4.0 g/1 of nitrate ion and 1.5 g/1 of ferrous iron. it is seen that with a complete oxidation of the iron only 0.55 g/1 of nitrate are removed. Hence, onoe the reaction has occurred 1 1t becomes neces sary to again reduce the iron in order to remove more nitrate. Oop )er was found to be the 0 best available reducing agent, but it was found that this metal could be used for reducing nitric acid directly rathe;r than tbru the i ntermediary~ ferrous ion. This lead to another phase of the denitrifioa- tion p;r obleut which will be treated below. Reaction of Copper in the Denitrifioation Process - tt -t" ;._ ( 3) 3 CU+ 2 N03 -+ 8 H -= 3 Cu + 2 N0~4 ff1 0 The reaction of copper with nitric acid, like rea.otion (1), also is autocatalyzed by nitric oxi.de. Like a true febrile reaction, it goes thru a.11 o-f the three period,s mentioned. b efore. For this :reaction, theB:ronstead theory predicts a negative s alt eff ect ae in the case of reaction (1 ), and this has been found to be t he case. In all oth er res pects ( i.e .. influence of temperature,, acid, etc.), 1 t is similar to reaotion ( 1) . Unfortuna tely, when copper is introduced. i nto the solution, the following reaction also occurs: ( 4) This reaction appea..r s to be but sli ghtly affeoted by hydr ogen .ion or the presence of nitric oxide. Its r ate inc17eases slightly with temperature and is proportional to the surface of copper in oontRct with the solution. By constan.tJ.y r eforming ferrous ion, re- aotion (4) tends to retard bot.h reactions (1) and (3). In general, reaction (3) will not progress until pr aot1ca11y all of the iron is in the oaidized st ate. Renee. the problem is to adjus t conditions so that reactions (3) and (1) will oc cur more readily than reaction (4). Before copper is introduced, reaction (1) is caused to t a.le e place by pas s ing nitric oxide into the solution and placing it in contact with siliea gel. W ben all of the iron i s in the fer r ic form, o&pper is pla.eed in oontaot with the solution and it at once begins to dis solve, thus deoomposing the nitric acid pres ent~ When t he nitrate concentration has f allen to a low value, r eaction ( :~ ) decrease s in r ate. However reaotion (4) continues undiminished in speed,and soon forms an a;pp:recia.ble a.mount of :ferrous ion, which eventually completely stops all further d.eni tri:fica,tion des-pi te the large a.mount of nitric oxide then present. If the eoppel:' is removed from the solu... tion before the ferrous ion ooncentra.tion has arrived at a high value, reaction ( 1) will slowly oocuri! leaving the fine.1 solution cont a ining only oxidized iron. By d.eoreatting the copper surfaoe>reaotion (3) oa.n be ca.used to bl'ing the :final nitrate oonoent:ration to a lower value, although the rate of reaction is considerably decieased thereby. Also, by incl."easing tbe temperature, the S9.me e:tfe4t is noted because the increase in rate of reaction (3), with :rising temperature. ia greater than tha.'b of :teaotion ( 4). J.?ATA QOWOERNll'.tI(} R~Aq_TIOF,S (3l ,9tt,d ._(4); The following experiments illustrate tbe effect o:f temperature, copper surface, and nit:ric oxide concentration unon reaction (3): Experiment (1). To each of several 10 o.o. portions of oxidized spent electrolyte containing 1. 2 mg. of NO, were added. one foot of 11'18 Ooppe;r wire, and they were placed in various temperatu:r.-e ba.the: Temperature 0 Time :for Reaction Final N~ Cone. 50 0 3 Hrs. 4.0 g/1 so· o 50 Min. 2.3 g/1 70° 0 30 Min. 2.0 g/1 80° 0 30 Min. 1.2 g/l 25 Min. 1.s g/1 . 0 80 C Experim.e nt ( 2). (This soln. was agitated.) W:ttb 100 c.c. portions of oxidized Spent Eleotn- lyte, eaoh ool1'ta.i11.ing 20 mg of I\O ,the· following results were obtained 0 at 80 C: ....,. wire used l'ime for reaotion }"'inal fl°J Final. Fe Aoid Loas 2,0 :ft, 120 Min •. ·O· ., g/ 1 0 . 37 gjl. 1.2 g/1 4,0 ft. 80 Min. 0.4 g/1 0.75 g/1 o.s g/1 8 . 0 ft. 40 Min. 0.8 g/1 0.98 g/1 0 .. 2 Length o:t .fJ,13 Ou oono. ''° oono. g/1 To 10 o.o. portion$ of test solution vary.. Expe:riment ( 3) . 0 ing amounts o:f ni trio oxide were added and they 'l.'lere kept at 60 C. lTO used Time tor Reaction 2 .• 0 g/1 2.0 Bra, 1.1 g/1 2.6 g/1 1.5 Hrs. 1.1 g/1 2.9 g/1 1.0 His. 1.1 g/1 Final Noj Cone. ~UUE!ARY coiio.EB,;NiliTQ R:tn40T!01l~ (:;) . A)f!) J4): The amount of nitrate removed by reaction (3) varies directly with the temperature and inversely with t};le copper surface. The rate of re action ( :3) varies d.irectly with the temperature, the oop~)el" surface, the conoentra:tion of nitric oxide, hydr~gen ion, and nitrate, &.nd inveisely with the concentration of ferrous ion and dissolved saltsi' Agitating the solution. results in a decrease in both rate and nitrate :removed. ACID LOSSE:$: It was noted, if copper 1s allowed to remain long in oonta.ot with the soluti.on 1 after reaction ( 3) has oea.sed, that more 1Jha.n the stoiohiometric quant:t ty of hydrogen ion is used. This acid loes was found to va:ry directly with the initial acid oonoentration. For in- stance, with a solution containing originally 4.0 g/1 of ni·tra.te, l.5 g/1 of ferrous ion and 70 g/1 o:t eu1phut1o acid, an excess of 60 i of acid ia consumed~ With a. aolution similar to ·this one, ex- cept that it contained zo g/1 of a.oid• ;just the stoichiometric quantity of aoi d was used.. Dis s olved sa lts also tendl to diminish this &Oid t~ss ·~• •·the a~id loss tend13 to dec:tease if the copper surface is kept t o a minimum . Absorp~~on .md Diffusion of Ni trie Oxide . Since nitric oxide plays such a vital :role in both reactions (1) a.nc1 (3), it was thought desirable to get a semi quantit ative idea of its solubility in solution . . . . A g11l...a s tube l·i" I 24° is filled full of . ~ ;s . glass beacts. and solution at 25 a. ,_ tricltled at a. rate of 1 0 c.o. per minute down the .tube oountel'ourr ent to a slowly mov1nr-: stream of nitric oxide gas. It was found that the test s olution absorbed the most nitric oxide, the spent eleat:rolyte abs orbed a little les s , and water' absorbed only a trace of the gas. At 20° C., very little nitric oxide diffuses out of solut ion in two hours; after six to eight hours, however, most of the nitric oxide has disappeared. At 60° O most of the nitric oxide has left the solu- tion in two hours, while at 80° 0., only one hour is requ~red for the saJt.1:e amount of loss. Den1trif1eation by a Continuous Process In order to oxidize the ferrous ion in the untreated. spent electrolyte. the solution at 25° 0. is first s aturated with nitric oxide by being passed down a glass absorption tube simil ar to the one used above 1n the experiments on the absorption of nitric oxide . It is tnen introduced into the bottom of a test tube (voltune 120 o .c.) which is filled with silica gel and immersed. in a constant tempera ture b a th. The solution coming from this tube contains no ferrous iron and is saturated with nitrio oxide. dtte to the initial treatment a.nd to the :nitric o-: d.de to:rm.ed. during the oxidation of the iron; therefore it is in a condtt:ton suitable for teaction ( 3) to oooux. the bottom of a large tes t tube ( volume It is passed into zoo c. c.), o.o ntaining ten :feet of ,U.18 ooppe:t vrl re, where :reaction ( 3) is }71l'tially completed. it is passed i nto the top of another tes t tube (volume Then 200 c.c.) containing five :f'eet 0f ,U,18 copper ,vire, wh ere its nitrate concentra1 tion is :reduced to a minimum . Both of these test tubes a,:re i mmersed in a const ant temp e:rature bath. The nitrie oxide gas formed du.ring these reactions i s collected and utilized in initi ally saturating the solution in the absorbin$ tube. If the solution passes thru the system too fast, not enough nitric acid is :removed; if the movement i$ too slow,. the rate of .r eaction (3) is greatly ret ardedJand e,cid losses .r esult from long contact with ooppe:r. Hence the problem is to adjus t the flo w so that the maximum amount of ni trio a,cicl is ren1oved, without reducing the iron. on :further e~perimentingJI a system pas dis,e-o vered whi ch would oxidize the solution faste,: than had been possible previ ously with the silica gel tube. In this system, the solut ion, fl°a.tu.rated •with nitrio oxide as before, is passed into the bottom of a test tube (volume 120 c.c.) containing solution in which the ferrous to nitrate r atio ie very low.; no aolid catalyst is present . I n1medi a tely upon striking ·the solution the reaction oocurgi and ni trio oxide is formed, ~ rising to the suxfaoe in the form of extremely small bubbles,"tending t o keep the nitrio oxide concentra.tion in solution ver y high; hence _ this condition of the gas favors an increase i n the rate of reaction. Later it was found that once nitric oxide is formed in the reaction it is unnessary to tube, 1/irst satura.t.e the solution with the gas. Instead1 the untreated solution is passed directly into the bottom of a tube i nnnetsed in a. constant temperature bath, where it comes in contact with hot solution containing nitric oxide and no :ferrous ion, and where reaotiom (1) readily ooou:ra . At $,0° 0 , with this system, 6 c.c. of epent electroQ lyte can be oxidized. pel' minute while this oap a o1 ty is doubled. a.t 80 O. The silica. gel tube of the same total volume will handle only on e half' this amount of solution. If the solution is first warmed in a heat... ing coil before. its introduction into the reaction tube; t he r ate de ... creases $oraewhat due to less effective absorption of nitric oxide in the hot i ncoming solution. Using this l a tter Eiystem of oxidizing iron, and the previously mentioned scheme of :fu:rthe:c d.enitri f ying the s olution, it wa.s found . possible at 70 c. to reduce the nitrate conc entration in the s pent eleotrolyt!;i'. o g/1 to lee B than 1 , 0 g/1 . The amount of solut i on handled 1,7 :,i:,s one h alf liter per hour. For best resultG with this continuous sys tem of denitrification, it is es sential that none of the solutions bear ing nitric oxide,,..,, be a gi t a:ted, as this -tends to lower the concentra tion of the gas in them¢. Hence gas traps a re used between ea oh rea cti on tub.e in order that no large bubbles a.re sent to the bo ttoin of any of the tubes . Wh a t is desired i s a constant s ·tream of ve-py small bubbles of ni trio oxide and this situation ie found naturally to be the case in solutions iiberein either reaction (l) or (3) is t ak ing place. Continuous Denitrifioa.tion as applied to Spent Electrolyte obtained. from leaching Ore. One month was spent in leaching 240 pounds of Ohuqutcam1-1ta ore ( l e,beled. regu.l a:t) in order to obta in a repre sentative spent eleot1u, ... lyte soh1tion; er~r-nthetic solutions we:re made in order to s t a rt the leaeh. During this time,two leaches (two thirds of a complete leach- -18.. :tng eycle) w,e re completed, t he spent electrolyte after each leach being obta ined by eleotropl a.ti ng the copp er out of :,:,olution, using lead-antimony insoluble anodes, which were furnished by the Chile Explora ti on Co. Part ial analysis of the resulting spent electro- lyte solution snowed: H. 1- SO ·, '1 e• g/1h IDT0 3 7 .4 , g/1 of ' 0 . 24 g/1 ++ ,-, . , ,. . Gu ., JJ'.5.4 g/1 ~ er-I' 0.13 g / 1 1:1. :tt is to be noted th at the acid coneentrati on of this solution is 7 g/1 lower than norr11al, a f a otor which should be unfavor abl e in the demi t:rifica ti ()n pro cess ,. al though this di s advant age is somewhat off-set by t h e l'.'\1lla.tively high ni trio acid concentra tion ( almost t wice normal). Sinee the· iron is in an al most completel y ,reduced s t ate the oxidation reaction ·t11be, .aentioned a:nove, was not used. Instsa.d the solution was warmed il'1 a hea ti ng coil a,n.d introduced immedi ately into the bo ttom of the first tillbe conta ining copper, and from there the arrangement was identical with ths,t desoTibed above . tn order t o st a:rt the r eaction, 1 00 mg of nitrio oxi de wa.s introduced wi th the first 200 c.o. of solution treated . From then on the reaction proce eded well in both tubes containing copper, without the need of an outside source of nitric oxide; in fact a.11 of this gas formed was allowed to elea.pe into the air. ' Solution wa,s handled at the :rr ate of l l/hr. 0 at 70 C, and the nitrate concentration was reduced to 1. ?. g/1. By circul ating the solution more slowly, or by adding additional copper, the ni t .:ra.t e concentrat ion can be lowered to O. 7 g/1 with ee.se . acid loes results. No Thus, a system ha~ been developed which requires tri al on a l a r ge eoale bef ore it s tntrocluotion into th e pre s ent leal!Jhing oycle. It is beli eved that by d.enitrifying the spe.n-t el ectrolyte, the nitrio aoid. oontent of the strong solution -c-vil l be oons.iderably :redueed, and in time, t h e ni t:rate content of the sys tem as a who+e ,vi1 l be lowered to a low value which will insure a longer life to the anodes . Were thi s deni.trifioation ca:rried out in the s t rong solution . troubl e mi ght resul t in the el ectrolytic t anks due to solutions foule d. with nitric oxi cle, unles s s pecial care were t aken to remove the gas. By deni tri- fying the spent eleetrolyte :tnstead, 1-rhich is, in regard to time. further r<~r 1oved fro m the eleotrolytic tan.ks. there is les s ohano'e for foul solutions a t the ta:nka. Also, s:Lnoe the volume of spent M electrolyteAfour fifths ~ tha t of the strong solution, less sol ution will have to be tre~ted for nitrate removal. However ,the chief ad- vantage.. of demi t:ri f ying the s trong solution is the possibility of oxidizing the iron i n t hi s solution and thus eliminating t he nedeas!ty of adding an additional oxidizing r eagent, It i s bel i eved t hat t h e ~ iron i n t he s trong solution me.y be oxidized. by s a tur a.ting thi s solution wt th ni trio oxide, wh ich is obta ined :fxom d.en i -trifying the spent ele.ctrolyte. and allowing it to r emain in cont a.c·t with t a ilings, from the o:re until t he iron i s oxi di zed . It may be n eoeasr::.l'Y to pstss a ir a1so t hr u the solution du.ring thi s t i me in 01?der to insure oxida tion. Heat ing the solution wi 11 be unneces sary, Economi c Dat a Twenty-five to fifty doll ars per ton could be spent for the elimination of nt trio acid fr om leaching solut i ons. Heat is ave,ilable £,rom waste heat boilers in e..n amount of 650 to 700 tons of steam per day a t a cost of 25¢ to 50¢ per ton, neglecting the installation cost of the bol,lers. Ooppef ts a-v..ailabla at les s thm,n ½¢ pe.r pound. At pl'e$ent, 17,000 cubic meters of spent electrolyte are produced pel' day~ • Assuming this e-o lution is heated from 35° o to 70° o and only ~ g/t ot ni t~a.te a::re removed au.ring each deni tr1f1cation ( this is a ve-ry coruH~l:'111!:tti ve figure), an e2tirnate shows tha.t the cost of stea.i"11 ot nit:r:io acid removed pe• ton./\is five doll ars a.nd the cost of oopper 1.s fifteen dolla:rs . .Assunttn.g diredJt heating with stealn and no h(iat loss, 850 tons of steam are requited. to heat 17.000 eubic meters of solution. Hence enough ste-am. ia available, assUI:ning h.ea.t losses, to heat a,t least on.e half o:f the solution, which is more than the amount used in the lime~ rook: process; and the oost of removing ni tri.c acid, a.E? suming only 2 g/1 • removed, is 'bt"1enty dollars per ton as comp ared with an allowable cost of :iwen.ty-.:f'i ve to fifty dollaJts pe: ton. The aotual equipment needed for carrying on denitr:J.fioa.tion would be merely a series of 1 ?.rge tanks composed of a nitric a cid :res1J$t1ng material, 111 which ~oppe;r could be sue:9ended, and a system of prehea:te:rs and heaters for bl"inging the solut:1..on up to temperature. Also , the cost of waste heat boile:rs must be included in the init18.l inste.11ation. No moving pa:rte are required. No data is available on the cost of equipment• out, since this ini t ic1.1 cost is to be divided ·tnruout a period of years. tt seems that the coat of removal of ni trio acid would not exceed thirty-five doll ars nei" ton. Thus the proce.s e seems p1·actioal both economically and technical 1. y. Summary Sheet. A system h~s been developed for the deni tr1:f'ioation of the spent electrolyte de.tived from copper solutions obtained fro m lea chi ng Ohuquieama t.a oop~:i er ores, by means of reducing the :nitric a ci d p resent with metallic copper in the presence of the oa t :9.lyst, ni trio oxide . During d.eni trifice;tion 1 t ··,i s es s ential tha t the iron which the solution cont ains be ke:pt in the fer r ic fotm and that a high eonoen.... tr e.ti on of n i trie oxi de be YJ.6vinta.ined . Thia is a ccomplished by first caus i ng the iron to be oxid.ized with the nitri.c acid present 1b y in... troducing the untreat ed solution into ~J ves sel cont a.ini ng s olution hea ted to S·o· to 80" O.1 1vhich is saturated with nitric oxide a nd. which oonti:dns no f etrous iron . Then the solution is placed in con- ~ • ta.et with copp e r ~ the m.et al dissolves, reduc ing nitric acid to nitric oxide. The ratio of the surface of copper to the volume of solution being trea ted must be adjua·ted.. If thio :ratio iS' too high, ferrous iron will be formed more rapidly than it can be oxidized. by ni trio acid an d consequ.ently th e rea cti on will stop ; if the r atio is too low, not enough nitric a@id wllJ. be ren10ved . With this system, using a OO!'.'_)p e:t to solution r a tio of about , 013 0 reeiprooa.l cent i 11H~ters I at 70 c th e concentra tion of ni trio a.oid ie re c\uoed f:rom 4.0 g/1 to les s than 1. 0 g/1 at a :rat e of about 4 o.e. of solution per minute per squaxe oent emeter of exposed oopp e:r surface. The procel!s is practical both technically and economically. ffi~ • t ttl" of t'ht toret~lng wi&b$$ to e.~ ,;~ss hte: lndebte&Hl$~ to ·tbe <;,"b1'.tt ! ~lo:r~ti.on Q0,~an1 of Ol~ute~ta,Ob11e f:$'1 f1:n -o1ng tb!.$ ,.,~eU<th WOJ'k, a.ntl fo~ the 11tl:ttable 8Ull ~$t'l0l'l:$ ~ , O'bt'H)~•t tloM rei~tll tli 4ell1tnftolltl-ott tf th-e le•a.tb eolut:ton:e nlkd.t\M 'tly u. :ta.t,.rt n d Mi. lritltt of ftlil oo-mpanr.• tl'u~ utter ~•• aolmolfflt~N th~ e'ttl'emel?f ftluab:le eu.1g~1lte);tlS g1V-f fl b.f Ott. w.r . L:St.4 •Y ()f th$ Ouift'lltai&. l~etiwte ot r eohn{f)1o§ am hi.a wlllt~e$$ "t ~ld ln. CV$~Y pt)ll!i!ibl.t . , tbit t')tQpea~ of the te-ata•o:b ..