1322 MITCHELL STUDIES ON CXLII.-&udies o n Hypophosphorous Acid. Part 11. Its Reaction with Iodine. By ALE'C DUNCAN MITCHELL. IN eat.imating hypolphoaphoiroas acid by mea'ns of merouric chloride aomrding to then met'hod given by Treadwdl (" Qua,nt,it,ative Ana(lysis," p. 2S9) i t was found tha.t the1 oxida.tion procmedeld rela.dily to pholsphorous acid but rellaLivdy slo,wly t o phosphomric acid. It was tjhelrefo,re de.cide,d t,ol investigate the1 progress of the i-elact'ion and c.ert.ain f a.c8ts werei netted in t,ha firstl s8tla.ge-the oLxidatbon t'ol phosphoro'us acid-which arel b'rielfly : (1) The coiioetnt'r adiojn o.f mercuric c,hlo,ride a,ppeiarecl trot ha,vel no inflimnoel 0111 t'he reectdoln vellocity eixcept whe'n velry dilutei. (2) The reactiotn was a,cceleratsd b.y the1 addition of hydrolchlosic acid.(3) It was the'retforei autowat,alyt\io owing t,o tlhel hydrochloric acid proiducetd. (4) The initia(1 vellolcity wa,s approxima,teily pro~portional t,oi the pro'ducti olf t$e concentra,tliom oC the1 hydrogeln ions and the1 hypo-phosphoroas a'cid. From them fact,s it wa,s conoluded t,hat the oixidat'ioii t.0 phosphorolus aaid involved tlwol suceeIssive reia,ctions the! first, of melasura8blel vellocity in which t,hhe> mercuric chloride1 t-oak no1 patr.t, a,nd t'hel secolnd o,f relat'ively p a t ,ve,loclit,y in which the1 melrcuric cihloiridei f unct'io'ned . No coinventiona,l hypot.helsis appeared adequa.telly t'o explain them facts. I n o'rder ta investiga,te certain points inolre fully an ana.logous reaction was soughtl in which the analytica,l rneltho'd used would be of grelat'es a,pplicab,i!ity.The. relactio'n with iodine was found to be acc,ura,tely colrnpa,rable and is here delscrihd. That wit$ rnelrouric chlojridel will be communicateld la.ter but itl may bet melntdoneld that compara8tivs e,:iperimelnts with iofdinel and meircuric chloride!, respeot.ively showeld that the melasurable8 velocity was identical in the two ca.ses t,hus lending a.dditiona1 support tot the! vielw that, only hypolphosphorous acid functions in the slolwer raction. Stredel (T. 1907 9 1 1641) ha8d already invc8stigateid the reaction svit,h iodinel and o'btaineld the1 same1 gelneral res,ult.s a.s alsoive. For the1 relactio8n H,PO + I + II,Q = H,PQ + 2131 he 0Eere.d t,hel follo3w-ing ex-planabion the slo'w stage is represented by which procselds libelrating free1 ionia chargee if thejrel is also preselntl BYPOPHOSPHOROUS ACID.PART II. 1323 substance capable of utilising them as rapidly as t,hey are formed: (2) 1,+2(')=21'. He poinb out that several hypoithetical sche~~es may be substitutd for (1) in orde'r to1 a.ccount for tha product,iojn 0.f free electrons but that int8elrmediate compounds cont'aining iodine cannot be formed in t'his st,age. Any equilibrium between hypophosphorous acid a,nd a,n " active" form would noh agree with t'iie fa& if it be assumed t'hat equilibrium is in&antaneo,usly reatmed; but i t is now shown t'hah if equilibrium is reststred ah ;I finite ra,te the existence of an inklrmediate " ac,tive" form adequatedy agrees with the facts and renders unnecessary the a.bolve unoonveatioaal hypolthwis.The relaction has now been studield under more diverse1 conclitiioins, and Gelriain nelw aspecix have been investigated. Stelele ignoreld two factors which are here shown to have an important bearing on the problem: (1) The deprelssioln by minerd acids of the1 iolnisation of hypo-phosphorous a>aid which he regarded as '' prolbably slight." Taking this into account by utilising data obtaine8d in the anihor's paper on tlhe subjeot (this vol. p. 957) it is shown clelarly t'hat t,he effect is considerable and that hypophospholroas a,cid f unct,ions as undis-solcia,ted molleaulea. Steelle obtained higher c1onstant.s in stroager acid solutions than in weaker but insbead of a,thribut,ing this to the inoreased propolrtion of mojleculea he ascribed it to the fact that he was using rno,re dilute hypophosphorous acid solutions the beha,viour of which he thought ammalous and suggested hhe idela thak its ions welre the aotive partl od the acid.(2) The past playsd by the ioldine which is appamntly negligible o'velr small ranges a t modera,t,e co~ncentration b,ut becornea. relat6vely grea't 8,s the conc,entration decreases. Stelele suggested tha,t it played no past in t,he reection untJl a.bout ninei-tenths of it had bees used up but tha,t it' theln tolok part in a.n appamntly bimole-ouhr reachion. Also he was led t40 a,scribe cert'ain cliscre1pa.nciea to1 the abotvementioned anolmaloas behaviour of hypophosphorous acid in dilut,e solution, but thew are now shown to be reg-ulasiseid when the function of t,lm iodine is colnsideired.From the datla obt'ained from abboat thirty expetriments nolw deacribeld actaal rates of de'cretase (dsldt) of hypopholsphorous a,oid were calcula4teld a,t the beginning middle asd end of e,ach elxpelri-m a t 4 and when these were divided by the concentrations at time t of hypo~pho~spho8rolus acid I?, and hydroge'n ioas ht and by the propoation of undissocia.te,d hypophosphorous acid mollec.ules 1 - at, ninety va\luea were obtaineld for the expressioln (ds / d t ) / htZt( 1 - at> which we,rel a.lm& const'ant wheln the corresponding concentratioas This is nojw sholwn tot be elrroneous. 3 c 1324 MITCHELL STUDIES ON of iodine eKceeded N / 5 0 but which decreased more and more rapidly with decreasing concentrations of iodine.This behaviour was ultimately found to be repreaented by the formula [1 + r ~ t / a t ] ( ~ / d t ) = k U t ( 1 - at), where at repreisents the concentration of the iofdine molecules and r and k are constants. It will be noltieed that this equation takes the form & / d t =kZ,(l -at)htat/(at+rht) which the reaction would give if i t werel simply bimolecul~r between hypophosphorous acid molecules and iodine being accelerated by hydrogen ions and retarded by a function at+rht of the iodine. Meymhoffer (Zeitsch. physikal. Chem. 1888 2 585) found that iodine had a retarding effect proportional to1 its concentration plus a constant on tIhe reaction HBr03+ 6 H I = HBr + 3H,O + 31 ; but the slightly modifield form melntioned above has not been regarded as a valid explanation oif the prelsent reaction since it is difficult t o understand why iodine should elxert a reltarding effect when its coacentxation is zero and its acceptance would imply that the hydrogen ioas had simultlaneously both awderating and inhibiting etff acts.The following hypothe& is found to require the same mathe-matical expression and to fulfil all the other requiremelnts previously postulated. Hypophosphorous acid molecules are originally in equilibrium with a very small quantity y of an " active" foirm, say H,PO in which the action of the hydrogen ions is to amelelrate reestablishment of equilibrium : . . . . . . H,PO,+H',O f H,PO (3) H,P03+I,' -+ H,PO,+BH'+31 (4) ds/dt = k,y,a .. . . . . . ( 5 ) kZ,(l -a,)h,=kly,h (6) and the iodine relacting its shown later as the I,' ion takea part : . . . . . The rate of formation of phosphorous acid according to (4) will be : 12 being a very large reaction velocity-constant . in (3) is k and k being the reaction velocity-constants in the1 two directions, k being large but not infinite and the) amount yt being so small as noit appreciably to affect the amount of hypophosphorous acid. Then the rate of inerelase in y owing to the tendency to restore equilibrium is k,?,(l- at)ht - klytht from ( 6 ) and its rate of decrease is d s l d t ; hence The equilibrium . . . . . . dyldt = klt(l - ut)hc - kly,h - ds/dt . . . (7 HYPOPHOSPROROUS ACID. PART 11. 1326 Eliminating yt from (5) and (7) we have (d”/dP)/k + (ds/dt)2/k2c + [a + k,h+/k,]ds/dt = ka,Z,(I - a,)h (8) This elxpression can bei simplified for most’ relactions as follows.Since k2 is very largel and kl moderateily large compared with k, and d%/dt2 and- ( c l s / d t ) ~ arel of a lower ordelr of magnitude than the othelr terms and moreover are of opposite1 sign (except in certlain cases) the f i r s t two terms may be mglected as a first approixiniation and one1 has or [l +k,h,/k,a,]ds]dt=kl,~l -a,)h . . . (9) which is identical with the formula obtained experimentally when the constantl k,/k2 is replaced by the1 constantl T . As the best valuel of r is found to1 bet 0*012 this explains why the iodine appears not to function until it becomes dilute when at is small.Now It=Z-s at=a-s and ht=h+ms since1 ht is foand to be a linear funotion of s. If therefolrre a is a simple function of s, thel expression can be integrate’d. I n some1 cases it is a linear function co that’ a t = a -ns. Letters without suffixes denotg initJal valnes and m and n are constants f o r any particular expelriment. The1 expression simplifies to k . dt = ds[ I/( 1 - s)(h + ms)( 1 - a + ns) + r / ( I - s)(a - s)( 1 - a t ns)] (9a) and one obtains o a integration, [a + k,h,/k,]ds/dt = ka,7,( 1 - a,)h, + m( 1 - a + 1 7 2 ) log ___ h 1 -++Is 1 - a n( h + mZ) log . 0.01 2 a -_-( l - ~ ) ( l -a+In>(l a - s I 1 - (1 -a+na) log - +n(Z-a) log I - s I n cases whetre at is not sufficiently nelarly a linear function of s i t is be81,telr t’o put n=O regarding a as constant, and subselquently to emploiy the average value1 of (1 - a) for the1 period concerned in obtaining the value of the constant.Thus one has (1 -n)k~-[logz-z-s+log - h+ms]/(h+?nl)+ h The elxaeptiolnal caees mentioneid when simplifying (8) to (9) are those in whioh the initial hydrogen-ion concentration is relativel 1326 MITCHELL STUDIES ON small and the rea,ction veiloIc&y increla.ses a t first? tot a maximum so tha.t @s/dtz is not nelgativel until t8hatl ma,ximum is passed. These cases give! results in fair agreejmeat8 with the1 ot8helrs when formula , (11) is usetd so that elvein whejn both the1 ne,gleiot<eid twms in (8) are poaitivet thelir combined effectl is very slight!. Itl is int)eirezsting to note that if hydrogen ions a.ctually tlook part in t8hel react,io*n (3) in olnet direlctioa only and were not metrely oa;tlalytsic h wojuld disappear from the secolnd telrm on t,he left oif qua,tion (9) and the1 thelory wo8uld require resnlta not consisteint with pra.ctqicel.I n the calculations itl is assumed that a fo,r the hypolphmpholrous a.cid delpelnds oinly on the1 initial amount present1 and on tlhe con-centratJon of the hydrogen iolng; and is nolt a'ffecteld by the replace-melnt of hypophoaphorous acid by phosphorous a.cid. This assumption is justified because the ionisine; powers of the two aclids a,ra ve'ry similar a.nd the1 total of the t l w s acids remains uncha.ngeld so th& acc.ordinq t'o the1 (' total ion " hpot.heeis (Arrhelnius Zeitsch,. plhysilcal. Chem,.1888 2 285 ; 1899 31 218 ; Bray and WLmtl. J . Rm'er. Ghe!m3. S'olc. 1911 33 781; and Mitchelll, this vol. p'. 957) for any definite hydrogein-ioln concent,ratioa tlhe,re is a definite valuel ocf a for each initdal co'ncelnt.ra;t.ioa od hypo-phospho'rous acid. Also it' does not appe8a4r t4hat the1 undistsolciateld mo81ec,uleB oQ the mineral aaids have1 any appreciable1 catt,a81ytic influe8nce in the concmt.ra,tio,ns used, I n several eixperimeliits (Nos. I 11 and V) po,saible oxida.tlon to phosphoric acid was checked alkalimetrically (see p. 1332) and found t.01 be1 negligible.. It1 has indeed been shown to1 be inhibited in an acid medium (Royer and Ba,uzil J . Ph,mm. Chim. 1918, [vii] 18 321) a'nd Fede'rlin (Zeitsch,. physikal. Chem,. 1902 41, 565) c.ould only olb8t'aiii melasurab3e vellocities for t,he re:a.dtion beltlwee8n phosphorsolus a.cid and iodine by using oonc,e,ntaateld solu tions .I n olna phase1 olf this work in ordelr to obtain small yet constant coaceintlraftioins ,oQ hydropn ions at '' regulat,olr " solutIon was use8d, consisting of equimo~lsoular proportions of phosphoric acid a.nd potassium dihydrogen phosphattel so tlha,t in the1 rea.otion-mixture elaBch warns prelselnt in mo1a.r coinc!eint'ration. Since the colncrelntra,t,ioln olf hydrogea ions was Coiist8antl the degrelei of iolnisatqio,n of hypol-phosphorous acid was a h const8ant a.nd f ojr these e8xpelrimelnts, (9) beaomes where T I and am constlank rl being assigned the1 value 2.0 as giving most coasistent remlts. Attempts t,ol relate r' to rh and li' to Ic(l-a)h whioh they replacel are co,mplic.a;ted by lasok of (1 +r'/a,)d~/dt = Elt .. . . (12 HYPOPHOSPHOROUS ACID. PART 11. 1327 sufficient data for such eolncentrated solutions but h is of the order N / 5 0 and (1 -a) about 0.4 so that it. is concludeid that both k / and rI are1 sevelral times greater than would bet expecteld. Owing to the) form of equations (11) and ( l a ) in many cases they gavel extrernelly cloncordant value6 of a constant based on the equation &/dt =7;'/Ztda; which was found not to hold when applield to ioldinei conoeintrations of a differeintl order. I n experiments XI XX and XXV tlhis constant is given as well as k o r k' as an interesting example of the danger of relying oln what happens to be1 melrelly a matThelmatical coinclidence .Tho integrarteld form od (12) is . . (13) E't =log - I + 2.0(log - a - log-) I I - s 1-a a-s I - s The 1,' Ion.-Considering t.he equilihrium Tti/ I + 1' investi-ga.ted by Jakowkin (Zeitsch. physi,kaZ. Chewt.. 1896 20 19) it' can I w slmwn thatl folr any st'age .of€ t.he relaction, [I.] = q a ' - s) / ( p + s - a') to a c10,se degree of a.pprolximatioa where pi is the molar mncen-tmtion of t'he potassium iodidel (assumed compleltely ionised) and tlhat [I3/] = (a - s)[l - K / (p + s - a)] sinm p increlasea by 2s and n delcreesels by s whe'ii s molecules o t iodine1 have beeln cha'ngeld to iodids ions. At' the! beginning olf an etxperimentl wheln s is zelro, [I,] = K a / ( p - a) and is thesefolrel independeint olf the co'ncelnt'ra-tio.11.It wa.s a t firstl tqhought thatl this woluld amount' for the peculiar belha,viour of the iodine in appareat'ly not a,ffecting t'he velocity olf the rea.eti0.n; but [I2] decrelasm molrel rapidly than ai- s with inorease of s sol tlhatl i f the iodine reacteld as such its appa.rent active maSs should fall o,ff morel rapidly than a - s whelrema.s adually it chaagee fa.r less ra,pidly. More,ov.er a colmparison wae made in expelrimeinta XI1 and I which were ideintdcal elxcept t<ha.t in XI1 pota,ssium iodide was a,dded so tlha,tl p wa,s equa.1 t'o 9*78a whelrelas in I it was 3.76a; if molle8culas iodine1 welrel the active1 faotolr the initial resot.ioln velocit8ies shoald be in the1 ra,t,io 2.76 t'o 8.78 whelrelas they were a81mo.st identioal. Errpeil-ime'nts XXV a.nd XXVI a'lso illustrate this in presence of the " regula,t,or " solution.The iodine therefore does nolt reactl as molecules. Thus it is t<he I,' ion whioh funchioins a'nd using t'hel const'ant 0.00135 giveln by Jalrowkin for the1 equilibrium i t is found tlha,t this ion constit,utes 99 per cent. of the available iodine in N/10-s,olut,iolns b,utq is le'ss in wela,kker solutions bsing olnly a'boiut 65 pe'r aent.. in N/400-solutio~ns a t the beginning of reactions a.nd increas-ing somewha,t' as the! relactio'n pro,aeeds. This change in t,hel rellatsive conmatmtion of tlhe I,' ion has been ignolreid t,hro,ughouh a.s th 1328 MITCHELL STUDIES ON correictiion involvetd only affect?s the smaller term and is more? o w r less at the elnd of a reaction than a t the1 beginning since the increase of iodide1 ions outweighs the! effect olf dilution and thus increases the proportion of 1,’ ions.Itl would unnecessarily complicate equations which are evidelntly sufficiently accurate to demonstrate the validity of the hypothelsis. I n conclusion it may be said that whatever the nature1 of the supposed active! ’’ form of hypophosphorous acid ita existence could probably nob bet detected by ohemical melans and as thel indications are that it is less than 1 per cent. of (and in constant proportion to() the1 acid mollecules its dchection by physical means would ba difficult. It is also possible! that another hypothesis coluld be1 found which would relquirs similar mathematical expression but if so i t is certainly more complicated than that devellopeld herel.E X P E R X M E N T A L . The reactions were carried out in stoppereid brown glass bobtlea in a theamoatat at 25+0*05O. The1 iodine solution and hydro(-chloric acid or “ regulator ” soJiitions. when used were made) up t o 400 c.c. and when they had acquired the1 temperaturel of the thermostat< 1.013 2.000 or 4.000 C.C. of a concentrated solution of hypopholsphorous acid were added from stlandardised pipettes. Atl definite times quantitieis were withdrawn and run into1 a large volume1 of watelr containing the volume1 of standard sodium thio-sulphatlei solution estimateid to1 bet necessary and the1 final adjustr ment made ati once. Check experiments showed that the hypo-phosphorous acid did not1 affect the1 titrations.The1 stlock iodine1 solution was accuratetly decinormal. As it was subsequently required t o know the potassium iodide content of this solution i t was found tot be 31.2 grams per litrel (or 3.76 mole\-mles per mdeculei of iodine) by an adaptatioin of the method with potassium iodate1 described by Sutton (“ Volumetrio Analysis,” p. 133). The stock hypophospholrous acid solution was a chemically pure artiole of commerce (D 1.14) which sholwed no impurities other than a little1 phosphorous acid. Its composition was checked by two metholds 25 C.C. of a one-tenth solution gave 1.5124 grams olf Mg,P,Q, therefore H,PQ + H,PO,= 5-45 moles. per litre ; 20 0.0. of cme-fiftieltlh solution required 21.80 C.O. of N/lO-NaQ€I with methyl-orange and 22.60 C.C. with phenolphthalein.Since hypophosphorous acid is monobasic t o both indicators and phw-phorous acid is monobasic to the1 former and dibaaic to the latter, tahe former is 5.25 molar and the latter is 0.20 molar thus givin HYPOPHOSPHOROUS ACID. PART 11. 1329 a total identical with the gravimetrio value. For obtaining the value1 of I the! solution was regardetd as 5*25N and the values of the hydrogen ions were based oa a normality of 5.45 since the seoond hydrogeln ion of phosphorous acid can be ignored. CaJculajtiom of ResztZts.-Thel degrees of dissociation of hydro-chlolrio aoid are taken from the reisults of Bray and Hunt (J. Amer. Chem. Soc. 1911 33 781) and the hydriodic acid is reckoned as hydraohloric aoid for this purposel. The degree of dissociation of the hypophosphorous acid in the presence of the mineral acids is obtained by methods based on the) autholr's recent ciommunication on the subject (this vol.p. 957). From table I givea therein a curve is constructetd showing the relation beltween a and the con-centration of hydrogen ions a/v when no extraneous acid is preeent. From hhis it is possible to obtain a series of curves one for eaoh value of I showing the rellation betwesn a and the con-centIration of extraneous hydrogen ions (El',) as follows for a delfinite concentration of hypophosphorous acid I the value of a when H' is zero can be1 obtainetd at once. A slightly lower value, al is then selected and the concentration of hydrogen ions in equilibrium with the1 acid alone1 a t that delgree of ionisation is found from the! first curve; this must also be1 the concentratioa of hydrogen ions in the1 equilibrium if foreign aoid is present as shown in the paper quoted; from this is thelrefore deducted lal due to the hypophosphorous acid itself and the remainder is H*,.A selries of corre6ponding values of a and H' is obtained in this way for elach value of I and from the resulting curve the value of a oan be obtaineld for a known Hez. This melthod is simpler and of wider applicability than solving the] equation developed in the former communication. I n order to avoid small decimals all conceiitratioiis are multi-plied by 200 and the value of X shoald therefore bet multiplied by 200 in order t o give1 absolute units. Time t is in minutes; a - s shows the iodine concentration in molecules; s shows the! number of moles.reduced ; I - s the hypophosphorous acid remaining. The column HI shows the original hydroehlorio acid (when uwd) pluq 2s the hydriodic acid folrmed. H*=[ are the1 corresponding hydrogen-ion concentrations of the1 mineral acids the degree of ionisation (not recorded) being assumed t o be unaffected by the prese,nm s f the weaker acids. a gives the degree of dissociation of ths hypophosphorous and phosphorous acid and under H' is recordetd la the hydrogen ions deriveld from them the slight correc-tion neceissary f o r the phosphorous acid initially present having been made throughout. Then follow the1 total hydrogen-ion ooln-cwntrat,ion it and 1 -a, the mean value1 o l 1 - a fur the period 3 c 1330 MITUHELL STUDIES ON (or 1-+ the value at the middle of the period) for which the constant k is odoulated.Farmula (11) is used except where otherwise stated and in any one experiment the deviations of k from the mean rarely e x d 3 per oent. The six following experimenta wer0 without original hydroohloric acid. Experiment 1. t. a- s. s. 1 - s. HI. H'H~. 0 4.99 - 10.52 - -5 4-82 0.17 10.35 0.34 0.34 20 4.34 0.65 9.86 1-31 1-28 50 3.29 1.70 8.82 3.40 3-27 80 1.96 3.03 7.48 6-07 5.77 105 0.98 4.01 6-51 8.02 7.58 125 0.19 4.80 5.72 9.59 9.02 145 0.03 4.96 6.56 9.93 9.34 a. 0.645 0.640 0.625 0-590 0.553 0.530 0.513 0,510 H'P. 7.10 6-99 6.77 6.44 6-00 9-79 5.58 5-57 m = h. 7.10 7.33 8-05 9.71 11.77 1.3.37 14.60 14.91 1.543.l-aM. k x lo5 0.355 118 0.365 114 0-382 112 0.400 117 0.413 118 0.421 120 0.422 113 Mean = 116 - -The data of the other five experiments gave constants of the Expt. a. 1. U. Hey= h. rn. k x lo5. 11. 9-80 20.62 0.552 11-81 1.546 148 IV. 9.92 5-34 0.735 4.07 1.651 134 V. 9.86 10-44 0.645 7.05 1.606 124 VI. 2-43 25-50 0.525 13.91 1-425 139 VII. 2.46 13.00 0.622 8.40 1.607 120 [In experiment VII the complete formula (10) was usad with n.= 0*030.] The following experiment is comparable with I and shows the negligible effect of adding more potassium iodide (see p. 1327) : XII. 4.97 10.44 0.650 7.05 1.506 114 same order of agreement and are here summarissd. The next thirteen experiments were with initial hydroshloric wid.Only experiments I11 and XVI are given in detail the latter showing the accuracy of the constant at very low conmn-trations of iodine. t. 0 6 10 15 20 25 30 35 40 45 50 65 60 Experiment IZI. a - s. s. 1 - s. HI. H'HI- a. H'p. h. 1-aar. k x lo", 4.93 - 10.44 11.30 10.59 0.496 5.38 15.97 - -4-33 0.60 9.84 12.51 11.68 0.485 5-26 16.94 0.510 (139) 3-89 1-04 9.40 13.38 12.47 0.478 6.18 17.65 0.513 126 3.34 1.59 8.85 14.48 13.47 0.470 5.10 18.57 0.517 129 2.80 2.13 8-31 15-56 14.44 0.463 5.02 19.46 0.620 130 2.31 2.62 7.82 16-55 15.35 0.45% 4.95 20.30 0.524 129 1.81 3.12 7.32 17.55 16.24 0.450 4.88 21.12 0.527 130 1-40 3.53 6.91 18.37 16.97 0.445 4.82 21.79 0.530 127 0.95 3.98 6.46 19-25 17.75 0.439 4.76 22.51 0.533 128 0.57 4.36 6.08 20.02 18.44 0.434 4.70 23.14 0.535 128 0.28 4.65 5-79 20.60 18.95 0.430 4.66 23.58 0.537 128 0.11 4.82 562 20.94 19.26 0.428 4.64 23.90 0.538 128 0.04 4.91 5-53 21.12 19-41 0.426 4.62 24.03 0.639 130 nz = 1.644.Mean = 12 t . 0 7 12 16 22 38 44 9 I a - s. 0.499 0.460 0.389 0.319 0.267 0.196 0-141 0.043 HYPOPHOSPHOROUS ACID. PART 11. S. -0.039 0.110 0.180 0.232 0.303 0.358 0.456 Expt. a. XIV. ...... 9.76 XV. ...... 0.250 XVlI. ...... 0.250 XVIII. ...... 0.986 XIX. ...... 9.58 XX. ...... 0.985 XXI. ... . .. 0.985 XXII. ...... 0.493 VIII. ...... 4.97 IX. ...... 2.485 XI. ...... 2.485 1 - s. 2.670 2.631 2.560 2.480 2.438 2.367 2.312 2.214 m. 14.10 14-18 14.32 14.46 14.57 14.71 14-82 16.01 1.2.670 2-670 1-040 2.080 2.080 1.040 2.67 2.67 2.67 10.30 20.20 H‘HI. 13-13 13-20 13.33 13.47 13-56 13.68 13.79 13.96 a. H‘p. 0.515 1-43 - 1.42 - 1.42 - 1.42 - 1.41 - 1-40 - 1.40 0.503 1.39 412 H.I. 11.10 28-20 14-10 2 8.0 10-86 28.0 11.20 28.0 113.0 56.5 113.0 H’HI. 10-41 25-66 13.13 25.5 10-19 25.5 10.47 25.5 96.0 51-0 96.0 a. 0.498 0-409 0.515 0.415 0.459 0.412 0.550 0.41 5 0-25 0.30 0.25 I&. 14.56 14.62 14.75 14.89 14.97 15.08 3.5.19 15.35 = 1.73. H’P. 5.33 1.13 1.43 0.45 9.64 0.9 1.20 0.44 ---1331 1-ua,71. k x 10s - -0.486 (135) - 116 - 116 - 116 - 116 - 116 0.491 116 Mean = 116 m.k x lo5 1.647 128 1.830 136 1.764 132 1-66 123 1.572 145 - 122 1-81 112 - 137 - 137 - 129 - 131 Where no1 value is given for m h is regarded as constant and in thei lastl tjhreo elxperimelnts thei value of has been neglected, as not apprelciably aflecting tho total value of h. Also in the last three elxpelrimmts a was obtained by extrapolation and is there-fore only approximate. I n expelrhent XI eight values all between 300 and 312 were obtained for the constant k”xlO4 mentioned on p. 1327. In experiment XX similarly all ten values were between 107 and 112. The following Seveln experiments were in a solution molar with respect ta both phosphoric acid and potassium dihydrogen phos-phate as ‘‘ regulator ’’ (see p.1326). Formula (13) is used in each easel. Experiment SXITI is given in detail. Experinmat X X T I I . t . 0 2 6 10.2 15 20 25 33 45 a - s. 4.996 4.596 3,898 3.246 2.578 1.996 1.622 0.950 0.432 S. -0.400 1.098 1-750 2.418 3.000 3.474 4.046 4.564 1 - s. 10.50 10-10 9.40 8-75 8-08 7-50 7.03 6-40 5.04 k’ x 10’. -2 74 267 268 272 274 274 276 276 Xcan 272 3 o* 1332 MITCHELL STUDIES ON Expt. a. I. I;‘ x 104. XXIV. ......... 4- 970 5.224 271 XSV. 2.494 5250 254 XXXI. ......... 0.4992 1.050 364 XXXI1. 0.2496 1 -050 29 1 XXXIIJ. 0.4992 0.525 266 ......... ......... ......... ......... XXXIV. 0.4994 5.250 240 For XXV all eight values of liN x lo* (p.1327) are between 600 and 626. Experiment XXVI was exactly comparable with XXV except that i t contained potassium iodide in the proportion of 15-8 mole,-cules to elach molleculel 6f iodinei; XXV and all othes experiments except XII have tthe proportion 3.76. The1 constant 1;’ x 104 was 255; the1 agreement with 254 obtaineld in XXV shows that iodine funations as the 1,’ ion. Experimelnt XXVII atl l l * G o gave a constant k/ x lO4=54*8, and as it is exacltly parallel with XXV the temperaturecoefficient is 4-64 for 13.4 degrees. Logarithmic! proportion reduces this t o 3-14 for 10 degrees and 2.22 for 7 degrees. Steelel found 3.1 for 10 degreles for solutions without added acid or “ regulator.” Since the term involving r! the ratio beltween two rapid reaction veloci-ties (sea formula 12) accounts for morel than half the1 value1 of the constants it appears that this ratio is practically unaltered over the rangel of temperature employed.The agreement beitween the oonstants obtained in this seiries of “ regulator ” experiments shows clearly that the1 iodine must func-tion in the manner shown in the equations useld. Tho effect of iiegleoting itl is velry much more apparentl in this series than in the1 earlier experiments. In order t o detect possiblel oxidation to phosphoric acid the following procedure was adopted. If phosphorous acid only is producied each molecule o€ iodine reduced gives rise1 to 2 mole-cules of hydriodio acid and also changes 1 molecule of inonobasic hypophosphorous acid to 1 molecule1 of dibasic phosphorous acid, phenolphthalein being used as indicator ; whereas i f phosphoric acid is produceld 2 molecules of iodine give rise t o 4 molecules of hydriodic acid and still only raise1 the basicity by unity.There-fore in thel former case the1 increase of acidity should be 1.5 times the1 numbex of equivalents of iodine consumed and in the latter case the ratJo should be 1.25. I n every case tested the ratio has been very Close to 1.5 so that the production of phosphoric acid is very slight. Expt. I.-At tI4‘, 20 C.C. relquired 14.84 C.C. of NIlO-NaOH more than a t f,; iodine consumed=9*93 C.O. of iV/lO-Na,S,O,; r a h = 1 *495 HYPOPHOSPHOROUS ACID. PART 11. 1333 Expt. I1 -Increase of alkali titre a t t,,=14.63 C.C. ; iodine consume)d=9-75 C.G.; ratiol=l.501.Expt. V.-Inorease of alkali titre a t t,,,=19*58 c.c.; iodine consumed = 13.10 0.0. ; ratio= 1.495. For the samel purpose in the (‘ regulator ” series an elxperhent (XXVIII) was carried out exactly parallel with XXV except that hypophoisphorous acid was replacwd by the same concentration of phosphorous acid. I n three hours less than ons-half per cent. of the ioldine was reduced whereas in experiment XXV 50 per cent. was relducetd in twenty-two minutes. Experiment X was earrieid out’ with hypophosphoroius acid which was previoasly neutralised (to methyl-orange) and is comparable with I. I n five hours less than 5 per cent. of the iodine was re#ducmd; as the second hydrogen ion of the phosphorous acid, originally present to a slight exteat would be1 sufficient to1 start the reaction and so1 provide more hydrogen ions it< is probable, that reaction in neutral solution is etxtremdy slow and that1 the H,PO,’ ion does not relack except possibly in the presence of hydrogen ions in which case reaction is elvidently so slow as not qreatly to1 affect the1 constancy of values obtained on the1 assumption ;hat8 i t does not take place at all.The following experiments of Steele’s are quoted after con-version to1 the units used herein tol show how the irregularities oherveld by him are largely accounted for in the light of the preselnt work. The first two columns are derived from his data; the remaindelr are obtained as in the author’s rwults. The constants obtained by each method are shown those of Steele being designated by Ks.The letter S after the Roman numerals indicates his experiments. Tha value 119 obtainetd f o r lc x 105 in X(S) is in agreement with the author’s results and shows less falling off with time than does K,. Moreover for the thrw experiments the resulte are far more concosdant than those obtained on Steele’s hypothesis. The fact that Steele used a very pure specimen of hypophosphorous acid may account f o r the slightly lower values in his elxperiments. Expt.. a. E. a. H p = h. k x 10’. Ks x 10’. X(S) 2-00 20.00 0.555 11.10 119 94 XI(S) 0.500 5.00 0-734 3.67 92 45 XII(S) 0.200 2-00 0.841 1.68 83 24.5 Altqhough the agreement is not so good as in tdhe other results, it is evidefntly far beltter than in the results obtained by Steelel’s melthod so that t h s present hypothesis which attributes a definite part to t’he iodine! molecules is apparently more justified by reeults than that’ which ignofree the ioldine and attributee the great decreas 1334 HYPOPROSPHOROUS ACID.PART 11. in the1 conshnt t'ol a delarelase,d a,ctivity of hypopho'sphorous aaid a.t low concelntrat,iolns. This is further suppolrteld by a recoasidera,tion of S,t,eele's experi-meat 11 for which he1 co1ul.d nolt olbtain a. constIantl witholut assuming the degree olf ioiiisaflio.n t,o b,e 0-20 inst8eamd of the a.otua3 0.775 whioh was aga.in supposeld to be! due tlo dscrelased activity. Aclcolrding to the! premnt melthold one o'bt'aJns modera,te constmanay. Experim,ent I I ( S ) nt lao. t. 0 390 570 1440 1710 1830 2805 3240 a - 9.s. 1 - s. H'"'. 4.50 - 4.00 -4.20 0.30 3.70 0.60 3.98 0.52 3.48 1-02 2.42 2-08 1.92 4-00 2.04 2.46 1-54 4.68 1.84 2.66 1.34 5-04 0.80 3.70 0.30 6.96 O.G8 3.82 0.18 7.24 a. 0.775 0-750 0.735 0.660 0-643 0.636 0.597 0.592 H'p. h. 3.11 3.11 3.00 3-60 2.94 3-9G 2.64 6-64 2.56 7.24 2-54 7.58 2-38 '3.34 2.36 9.GO m = 1.675. 1 -aH. k x 109. - -0.235 (25.3) 0.245 (28.1) 0.280 38.3 0.290 37.7 0.295 39.2 0.340 41-5 0.355 39.8 Mean = 39.3 I f the temperature-aolesfficient for sewn degrees is t'aken ass 2.22 (see p. 1332) this gives the low value of 87 for 25O but that the elxperiments correispolnd in geaeral is sewn from the fo'llowing summary of expelrirnenk for which the dat'a can be utilieed: Experiment.III(S). IV(S). V(S). VIII(S). XIII(S). XV(S). XXI(S) 11 x lo6 108 115 119 105 107 113 119 K S S lop 355 400 425 430 425 505 555 Steele's .Es x 104 give'n folr cornpasison clearly shows tqhO ina,delqualcy od his hypolt,hesis. The1 me,thod of trelat.tmentj nolw presenteld obviously p l a m t,he qu,es,t,ion o'n a molrel systma.t,ic basis a,lthoagh tlhe slight variation of tlhel " const,a.nt< " fro'm one elxperiment tol anothe,r is not a,ocount;eid for ; but tlhel sevelral small a.ppro.xima't,ions maBdet t'ogethe'r wit,h the unclert<aint!y in the ohoice; olf the value] of the constant' T th0 large influelncsel in wrtaJn ca,ses olf a. differelnc,el I - a whiclh is somet8imes very mall and t*het adtelration in t'he pro'portioln of the iodine whioh is preselnt as the! I,' ion may easily a,coount forr the variaikm of 10 per cent'. from the ineta,n valuel 128 x 10-5. It is. also1 probable t,hhat the H,PO,' ions reactl slowly and a,ccount pastly for the va,ria,tiolns8 since t'hel three higheet constant.8 are! given by t'hel three great eat' con (3en tqr ahi on s of h y p c~p h osp h or ou s' acid . Su!mmuay. (1) The1 previous a;tltemptt to explain the reladim had suggwted an hypothesis involving tthe preselncs of free ioai,u aharges. By takin PRIEDEL AND CRBFTS' REACTION. PART I. 1335 into consideration certain factors which had beea ignored this is shown t'ol be unnelaewary and most discrepancries disappelar. (2) The:se fa,ctors were (i) the influence 04 the ioldinel concen-taation oln the me:asurab,le vellocity t,his influesnoe being slight. at moldemte dilut,ion but rellatlively lasge akt grea'telr dilution and (ii) hhe eE& of hydrolchlolric and hydriodic acids on the ionisation of hypophosphorous a,oid. (3) The following hypotheisis is a,dvanmd ts elxplain all the anoma.lies f omelrly notleld. The hypophosphorous a,cid molecules a,rt?l in equi1ib'riu.m with a very small proipolrtion of an '' active" form (say H,P03) which rea,cts rapidly wit,h the iodine, The restmatioln olf the eiquilibrium thus displaaed is accellerat3ed hy hydrorgen ions a.nd forms the' me.a.surable! react.ion, H,PO,+ I,' -+ €I,PO,+ 2I-I'+ 31'. H3P02+ H,O - H,PO,. UNIVERSITY OF LONDON, SOUTH KICNSINCTON S.W. [Received Xeptembw 13th 1920.