PHOTOKINETICS OF SODIVM HYPOCHLORITE S9LUTIONS. 2565CCX XXI X.- Photokinetics of Sodium HypochloriteSolutions. Part 11.By LEO SPENCER.THE present paper is an account of the1 continuation of the investi-gation begun by W. C. McC Lewis on the photokinetics of sodiumhypochlorite solutions (T., 1912, 101, 2371), in which it wasshown that the main reaction involved in represented by theexpression NaClO -+ NaCl t- 0.For greater convenience and accuracy the method of estimatingthe hypochlorite was carried out, as follows: I n order t o diminishas far as possible any changes produced by the lowering of the levelof the liquid in the reaction space, the amount of liquid titratedwas limite'd t o 10 c.c., which made it difficult t o estimate thestrength of the solution by arsenious oxide and starch-iodide paper.To 10 C.C.of the hypochlorite solution was added 0.02N-arseniou2566 SPENCER : PHOTOKINETTCS OF SODIUMacid in very slight excess, then 0.5 C.C. of a 1 per cent. starchsolution, and 1 C.C. of a 10 per cent. solution of potassium iodide,the1 solution being afterwards tJtrated with O-OliV-iodide. Theamount of Cosine solution was generally less than 1 c.c.; thus errorsdue to evaporation of iodine were avoided, and a sharp end-pointobtained. With free alkali o r acid present, excess of sodium hydro-gen carbonate was added, i t having been noticed previously thatthe end-point was but little affected by thelatter or by sodium carbonate. Duplicate/6'titrations were done a t each determination,and the values obtained did not usuallydiffer more than 0*2-0*3 per cent.fromeach other.EXPERIMENTAL.As source of light the same uviol lampwas used throughout the work, and all theapparatus, except the outside vessel, was ofuviol glass (see figure). The lamp was en-closed in a uviol sheath, which fitted fairlyclosely t o it. This was held by small piecesof cork in a uviol cylinder (outside diameter,4.96 crn.), drawn out a t the bottom to anarrower tube, which served as an inlet fora continuous stream of water, which left bya sidetube a t the top, the object of thewater stream being t o keep the temperatureconstant. Around this was placed a seconduviol cylinder (outside diameter 7.03 cm.),leaving a space of 0.93 cm. between them forthe light-filter solution.On the outside of allwas the reaction vessel (internal diameter9.80 cm.), providing a thickness of 1.39 cm.of hypochlorite solution. The bottom of thisspace was 1.5 cm. above the level of the mercury in the lamp, andits height 30 cm. when full. The hypochlorite solution was stirredcontinuaIly by mechanical means. The solution was kept a t 1 5 O + l ounless otherwise stated.* The lamp was run off the 230-volt mainsin series with a self-induction and resistance; 3.6 amperes wereemployed as a rule, corresponding with a potential drop of 32.5volts across the lamp.* Thib approxiniate constancy in temperature which would be quite tooinaccurate for ordinary thermal reactions is in the present case sufficient owing tothe very small tempel-ature-coefficient (see later)HYPOCHLORITE SOLUTIONS.PART 11. 2567Prelimiizary Test.--The first point det,ermined was whether anyappreciable disturbaiice was prcduced by the level of the liquidfalling (due to withdrawals of solution for titration purposes)during the course of an experiment. The rate of decompositionwas measured, first with the reaction space full, and then only halffull. The result was as follows:Reaction space full.Time. Concentration. k .0 47.69 -0.8 43.63 0.04582-0 39-03 0.04353.0 35.39 0.0432AI \Reaction space half full.0 47.69 -1.0 42.94 0.04552.0 39.05 0.04343.0 35-33 . 0.0434,- h ,Time. Concentration. k.The time is measure,d in hours from the commencement of theexperiment.The concentration is that of the sodium hypochloriteexpressed in C.C. of 0*05N-arsenious acid. k is the unimolecularconstant 1 / t log,o a / a - x, where a is the initial concentration ofthe hypochlorite, and a - x is the concentration a t time t. Whenk is bracketed, thus [ k ] , it represents the unimolecular constantcalculat'ed for the interval just prezediiig the measurement, theinitial time and concentration being those of the previous reading,so [lc] for the interval between t, and t, is:a - .r21, - - t q a-x, - lo&lo---. 1I n ordinary circumstances the fall in level did not amourit t omore than a quarter of the length of the reaction space, so com-paratively little error could have been introduced from thiscause.Effect of Change of Concentration of the Hypochlorif e Solution.I n the previous paper it was found that for the first six hours orso the decomposition was fairly well represented by a unimolecularformula, it being noted, however, that there was a slight tendencyfor the constant t o rise towards the end of that period.The iiewmethod of analysis allowed the reaction to be followed over amuch greater range of concenhation, and as will be seen from thefollowing figures the valu? began t o rise very rapidly as lowconcentrationr were reached.Time in Concentrationhours in C.C. offrom 0-02N-beginning. As*O,. 12. P I0.0 30.77 - -3-0 21.93 0.0490 0.04905.6 16-94 0.0505 0.05228.5 10.57 0.0546 0.062711-5 6.04 0.0612 0.081015-5 1.91 0.0780 0.125019.5 0.30 0- 1027 0.201022.0 0.04 0.1401 0.3502568Hypochlorite alone 0-08N.\ Time.Concentration. k.0.0 44.71 -1.0 40.84 0.03932.3 35.66 0.04214.0 30.34 0.0421cSPENCER : PHOTOKIKETICS OF SOD1U"Mchloride 0.34N.A >Time. Concentration. k.0.0 44.21 -1.0 40.66 0.03612-0 37-39 0.03643.5 32.69 0.0375Abseiice of Ch Pmicnl A ~ t o c a ~ f n l y t i c Effects.Other products possibly formed in small amounts in the decom-position might have had a tendency t o accelerate the decompositionas they accumulated. I n order to put this t o the test, two solutionsof the same concentration with respect t o the hypochlorite wereprepared, (a) one by photochemical decomposition of a more con-centrate'd solution, and ( b ) one by simply diluting with distilledwater.If a catalyst were produced in the decomposition itself, theni t ought, t o have been present, in ( a ) and absent in ( b ) .Prepared by decomposition.Conientration. Wl '21.5 0.05910.6 0.0846-0 0.120Prepared by dilution.7 Wl A Concentration.22.68 0.05711.43 0.0825.47 0.130The equality in the value of the two rates indicated the absenceof autocatalysis, which could not therefore account for the rise invalue of the constant. As chemical changes in the reacting systemitself appear to be excluded as the cause of the disturbance, i t isnecessary t o ascribe i t t o alteration in the optical absorption whichdiffers for various parts of the spectrum. This will be consideredlaterHYPOCHLORITE SOLUTIONS.PART 11. 2569Effect of IIiin.2itc iski?ig t h e Ali:dt C'otrtetit.In the previous paper (Lewis, Zoc. c i t . ) it was found that theneutralisation of the small amount of alkali present in the solutionwas followed by an imrease in the velocity of decomposition.Further experiments have shown that the addition of a smallamount of hydrochloric acid increased the velocity, but that i t soonbegan appoently to fall as more was added. The strength of thesodium hypochlorite solution was usually 0.075N ; the amounts ofsodium carKonate and sodiu-m hydroxide present a t the ssme timewere 0.0089il' and 0.0053.N respectively. The following table con-tains the mean velocity constant li corresponding with a series ofdeterminations over a time of exposiire varying from two t o threehours, varying amounts of hydrochloric acid being added a t thecommencement .Number of experiment 1 2 3 4 5 6Concentration of hydro-chloric acid added,lents per litre of thehypochlorite solutionexpressed in equiva- 0 0.0091 0.0171 0.0253 0.0379 0.0495 1 E ........................... 0-0418 0.0510 0.0487 0.043'7 0-0304 0.0250It would appear from the above that the maximum rate isreached when the solution is about neutral. The fact that thegreatest normality of the acid added did not exceed 0.05N was dueto the large amount of chlorine formed, which rendered the titra-tion values unsteady. The first action of the hydrochloric acid isnaturally t o neutralise the sodium hydroxide present in the solutionas ordinarily supplied for bleaching purposes.Owing to the weakcharacter of hypochlorous acid the sodium salt will be hydrolysedand the degree of hydrolysis will be much increased by the addi-tion of the hydrochloric acid. Apparently, theref ore, t'he undisso-ciated hypochlorous acid molecule is less photosensitive than theion. Continued addition of hydrochloric acid introduces complica-tions owing to the reaction between hydrochloric and hypochlorousacids.EffPct of I t i c r m s i t q the AAll;ulL Content.As free alkali was present in the original solution i t was desirableto find its action on the decomposition. It was found, however,that sodium hydroxide, even in strength several times that of thehypochlorite, had only a small effect on the reaction.The effect,such as it is, represents a diminution in velocity as the alkdiincreases2570 SPENCER : PHOTOKINETICS OF SODIUMHypochlorite 0.076N.Time. Concentration. k .0.0 47.20 -0.5 45.09 0.03961.0 42.98 0.04061.75 39.92 0.04153.25 34-57 0.0416h ,- -.Hypochlorite O.O76N+sodium hydroxide 0-143N.Time. Concentration. IC.0.0 38.261.2 34-31 0-03941.7 32.79 0.03982.0 31.84 0.03993.5 27-58 0-0408-I n a parallel experiment the value of the constant was alteredfrom 0.0416 to 0*0400 by the addition of 0*358N-sodium hydroxide.Y'emperature-coe ficierat of the Reaction.The temperatuse-coefficient of the reaction was determined byI. 11. 111. Mean.k at 10" . .... . . 0.0369 0-0360 0.0373 0.0367k at 24' ....... 0.0388 0.0402 0.0397 0-0396observations* of the rates a t loo and 2 4 O .The mean values give a temperature-coefficient of 1.06 for anThis is in agrwment with many of the results interval of loo.obtaiuesd in other photochemical reactions.The Effects of Differelit Spectral Regions.I n the foregoing experiments the light from the lamp was useddirectly.For quantitative examination, however, it is more impor-tant to investigate the effect of separate regions. The action ofapproximately monochromatic light was studied by employingWinther's light filters (Zeitsch. Elektrochem., 1913, 19, 389), whichtransmit a known amount of light of a particular wave-length fromthe uviol lamp whilst ahorbing all other lines. Assuming Beer'slaw t o hold for the filters, the concentrations of the constituentswere so altered that the same amount of light passed through thefilter in the present apparatus as passed through the thicknessused by Winther.I n order t o effect this, the concentration of tlhefilter had to be altered in the inverse ratio of t.he two thicknesses.Since the thickness of the Alter layer in the present experimentsdid not differ by a largo amount from that of Wintlier, this applica-tion of Beer's law was considered to be justifiable. Winther'smeasurements were of considerable importance for the presentinvestigation, sinco they constitute the only quantitative work onthe spectrum of the uviol lamp available a t the present time.Winther's method is, however, not one of extreme precision.As a check, two experiments were carried out with water onlyin the filter space, one a t the beginning of the series, the other* The temperatnres of these measurements were kept constant to 0.5"HYPOCHLORITE SOLUTIONS.PART 11. 257 1a t the end. The respective values were k=0*0411 and 0.0423,mean 0.0417. The regions examined corresponded with the lines578, 436, 405, 366, and 313pp. The results are as follows:A 578. Plotnikov's filter (isolating the yellow region). Filter trans-mits 80 per cent. of the line ~ 5 7 8 (Plotnikov).Time in Concentration in C.C. ofhours. As,O, solution. k.0.0 35.06 Zero (that is, no appreci-4.2 35.17 able change in con-centration).A 436. Wint'her's filter, transmitting 38 per cent. of line A 436.Time.0.01.31.83.74.75.6Concentration.36.4036.0135.8535.2434-9634.73k.0-003613663803720.00365-r\ 405.Winther's filter, transmitting 34 per cent. of line A 405.Time.0.01-72.32.8Concentration.37.8137.2137.0226-85k.-0.004120.004000.00400r\ 366. Winther's filter, trmsmit.ting 27 per cent. of line A 366.Time.0.01.63.86.35.9Concentration.37-0936.4335-5435.0534.80k.0,004824874620-00469-r\ 313. Winther's filter, transmitting 30 per cent,. of line A 313.Time.0.01.32.34.34.8Concentration.36.4636.2336.0535.0635.60k.0.002082132240-002 17-I n general, thel substitution or" approximately monochromaticlight in place of the entire spectrum yielded much more consistentvalues f o r k.Duplicate experiments were carried out, and agreedwell with those recorded. Wean values of both sets are employedin the summarising table2572 SPENCER : PHOTOKINETICS OF SODIUMThe values of the concentratioiis in tlie above tables are themean of three t?itratioW. The results are suiiiiuarisecl in the accom-panying table :I. 11. Ill. IV. v. VI. VII.Per cent. Percent. Per cent.A436 pp 38 0,00368 8.8 23.2 1.0 1.0A 405 34 0.00405 9.7 28.5 0.29 4.3A 366 27 0,00475 11.4 41.7 0.25 7.1A 313 30 0.00212 5.1 17.0 0.045 16.3Complete spectrum 100 0.0417 100 --Total 110.4The wave-length is given in column I. Column I1 containsWinther's values for the percentage of the line which passes throughthe light filtez.I n column I11 are the values of the unimolecularconstants directly observed with the light-filter in position. Thelast value in this column is the one observed with water only inthe filter space. Column IV contains the values of the velocityconstants of column I T 1 expressed as a percentage of the velocitywhen water only is preselnt in the fihr layer. Column V containstlie calculated amount of decomposition, that is, the velocity whichwould occur if the filter transmitted 100 per cent. of the line andkept back all the others completely. The rate is expressed as apercentage of the rate observed with water alone in the light-filterspace. The relative intensities of the lines from the uviol lamp aregiven in column VI, being taken from the tables given by Winther.Column V11 contains the relative rates of decomposition that wouldbe produced if e'ach line were of the same intensity and separatedfrom the rest..These figures show that the fastest decomposition is effected bythe wave-length ~ 3 6 6 , at the intensity at which it is omitted fronithe uviol source. This statement is true not only when the resultsare calculated (column V) on the basis of 100 per cent.of the linepassing through-as actually occurs, of course, when no filter isinterposed-but also when no such calculation is made, the directeffect through the filter bekg measured (column IV). It shouldbe noted, however, that A366 is not the most intense line of theuviol spectrum (compare column VI).The data in column V I Ishow that; if one calculates the velocity constants on the basis ofone and the same intensity for all the lines of the spectrum, oneobtains the result that the shorter the wave-length t'he greater theefficiency, that is, ~ 3 1 3 is more effective than h 366. This is inagreement with the concept of the energy quantum (hv) and withth3 photochemical-equivalent law of Einstein, for the shorter thewave-length the greater the frequency, and hence the greater thesize of a single quantum, This uniformity in light, intensity isHYPOCHLOHITE SOLUTIONS. PART 11. 25’73however, not the actual state of things in the experiment itself.I n the actual case there is a well-marked maximum efficiency a tA 366, as shown by the figures in column V, due to the simultaneousoperation of two effects, namely, the large intensity of this line inthe lamp and the large absorbability of this region by the solution(the head of the absorption band of sodium hypochlorite lies in theregion 200pp).The shortest wave transmitted by the uviol sheathis h 290 pp (compare Lewis, loc. c i t . ) . As the wave-length increasesbeyond ~ 3 6 6 the efficiency diminishes, so that on reaching h578(a yellow line) and using Plotnikov’s filter no chemical effect isobserved at all.Returning t o the data of column V i t will be noticed that thesum of the effects of the various regions apparently exceeds by asmall amount the total value of the lamp as a whole. The mostreasonable explanation of this is that the filters not being perfecta certain amount of superposition and repetition of certain parts ofthe spectrum occurs.The result as it stands favours the view thatthe effects of different portions of the spectrum are simply additive,a conclusion already come to by Luther and Forbes ( J . Amer.Chem. SOC., 1909, 31, 770).Owing to the absence of data on the absorption-coefficient ofsodium hypochlorite solution for the various lines it is not possiblea t present to compare fully the effect of each line, and thus testBruner’s hypothesis (Zeitsch. E’lelctrochem., 1913, 19, 555), t o bementioned in the next paragraph.Further Discussion of Results.If a substance in solution obeys Beer’s law, and if I, denotesthe intensity of incident light of wave-length A, c the1 concentra-tion, h the thickness of the layer, and m is a constant, namely, theabsorption-coefficient depending only on the substance and wave-length, then the amount of light absorbed will be given by theexpression I(,A( 1 - e-mciL).I n the paper of Luther and Forbes,referred to, the view has been put forward that the amount ofchemical action is proportional la the amount of light absorbed bythe solution, and is therefore given by the expressionwhere kA is a constant for each wavelength of the incident light.Bruner’s hypothesis is equivalent t o attributing the same valueto k~ whatever the wavelength.Froin the above “ab,sorption” equation it follows that the effectof dilution on the rate of decompositdon will be different accordingto the amount of absorption that the line undergoes.A line thatis almost completely absorbed by a given thickness of, say, a centi-SkAEhlOA(1 - e-m 1,VOL. cv. 8 2574 SPENCER : PHOTOKlNETICS OF SODIUMnormal solution will naturally be still more completely absorbedby a decinormal solution. It follows that practically the sameamount of optical abs,orption occurs in the two cases, t h a t is, it ispractically complete. If photochemical action depends on theamount of light absorbed, we would expect the same absoluteamount of chemical decomposition in the two cases. This is borneout by the following experiments, using the line ~ 3 1 3 . The meanrate of decomposition of a solution initially 0.075iV was foundt o be 0.185 C.C. per hom (reckoned in terms of the arsenious acidsolution).A solution initially 0.013N was likewise found to have amean decomposition value of 0.20 C.C. per hour, a quantity verynearly the same as the previous, although the absolute concentra-tion in the former case was approximately six times as great asin the latter.Strictly speaking, there should be no unimolecular const’antobtainable ir, this case, since the law is not d x / d t = k ( n - z ) , butcl.n./dt = li. For purposes of comparison, however, the values overshort experiments werel used for thel calculation of a constant,although i t will be clear fioni the two sets of values that as thedecomposition progressed the unimolecular ‘‘ constant ” reallyrose . -%Expt. ((0. Initial concentration ofsodium hypochlorite = 0 0769( approx .) .Tim;inhours.0.01.32.34.34.8Concentration u&-in C.C. molecular.of As,O,. k.36.46 -36.23 0.0020836.05 21335.66 22435.60 0-00217Expt. ( I , ) . Initial concentration ofsodium hypochlorite =O.O13,V(approx.)Time Concentration. Uni-r A ,in in C.C. molecular.0.0 6-67 -1.0 6.46 0.01391.5 6.37 1333.5 5.97 0.0138hours. of As,O,. E .(The li of Expt. ( b ) indicates the value tlo which the b ofExpt. ( a ) has risen when the decomposition has proceeded t o aConsiderable extent.)On the other hand, if the amount of the line absorbed is small,then doubling the concentration will be accompanied by a doublingof the amount of iight absorbed, and therefore a doubling of theactual quantity of substance decomposed.I n this case the rate ofdecompositicn is proportional to the concentration of the solutdon,which is expressed by the equation f o r a unimolecular reactiondx/dt==k(a-z). These conditions were found t o be satisfied by* The comparison mentioned refers to the efiects produced by the other lines iuall cases on solutions initially of approximately the same composition, namely,0-075NHYPOCHLORITE SOLUTIOSS. PART 11. 2575the line ~ 4 0 5 (which is only slightly absorbed), as the followingresults show :Expt . Ic). Initial concentration ofsodium hypochlorite = 0.076N(approx. 1Tim;in,hours. Titre. k.0.0 37.81 -1.7 37-21 0.0041 22.3 37.02 4002.8 36.85 0.00400Expt. (,d). Initial concentration ofsodium hypochlorite=O-Ol3.Vapprox.)./- A\ Timeinhours.Titre. k.0.0 6.71 -1.5 6.63 0.003473.4 6.51 3824.2 6.46 3945.1 6.39 0.004 12Although the constant in experiment (d) is not' particularly goodi t is not very different in value from that of experiment ( c ) , thatis, the constant is independent of the absolute concentration. 111experiment ( c ) t h e change in the titration vaIue is of the order0.35 C.C. per hour as a mean, whilst in (d), the weaker solution, therate is approximately 0.063 C.C. per hour. A comparison withexperiments ( a ) and ( b ) illustrates the difference in behaviour inthe two cases. These results are in agreement with Luther's" absorption " thetory of the velocity of photochemical reactions.The gradu'al rise of the unimolecular constant with time observe(*when water only occupied the filter space is now explicable.Thelight' from the lamp can be considered as made up of tlwo sets oflines, the' short wave-lengths which are weak in intensity but arestrongly absorbed by the given thickness of solution, and the longwave-lengths of higher intensity, but weakly absorbed by the samethickness of solution. The effect of the long waves is greatest in astrong solution, but rapidly falls off (compare the line ~ 4 0 5 above).On the other hand, the effect, of the short waves is fairly constantthroughout. With both together the long waves are predominanta t the outset, and their action is fairly well represented by theunimolecular formula d z / d t = Ic(a - x).As the solution becomesmore dilute, however, the short wave-lengths become more impor-tant, with the consequence that the value of the unimolecular" constant " begins t o rise sinc3 their effect' approximates t o thecase of the line A 313, where d x l d t = Ic rather than dxldt =7i(a - x)expresses the reaction.A'ote O I L the Iiifluence of Light on, the Bleaching of Lii~enby Sodium Hypochlorite.Two bands of opaque paper welre fastened 3 cm. apart round theuviol lamp, which was enclosed in a uviol glass sheath. This wasimmersed in a jar of the hypochlorite solution, and a piece ofunbleached linen was placed loosely around the lamp. After a2576 PHOTOKINETICS OF SODIUM HYPOCHLORITE SOLUTIONSexposure of one to two hours, the portions of the linen exposedto the rays from thel lamp were1 found to’ be slightly more bleachedthan the parts in the shadow.The light had therefore slightlyaccelerated the bleaching action.Summary.(1) The further investigation of the1 kinetics of the reactionNaClO+NaCl+O, using the mercury uviol lamp, has beencarried out.(2) By employing Winther’s and Plot’nikov’s filters variousspectral regions have been examined and the relative photoclzemicalefficiencies of these regions determined.(3) The results are in favour of Luther’s absorption theory ofphotochemical action. The general reaction equation may bewrittenFor monochromatic light when the absorption-coefficient m is smallthe above equation reduces to dx/dt =constant x (a - x), that is,apparently unimoleculm ; when the absorption-coefficient m is largethe equation reduces t o d x / d t =constant, that is, a “zero niole-cular ” equation. These: relations have been verified in the caseof the’ lines I 405 and A 313 respectively.(4) The temperat’ure-coefficient has been measured, and like themajority of photo-reactions is small.(5) When the photo-effects are reduced to one and the sameintensity for the various lines i t is shown that the shorter thewave-length the more effective is the decomposition, that is, thegreater the velocity constant. Under actual conditions, however,the intensity of the lines differs with the result that 1366 is moreeffective than ~ 3 1 3 . The latter is the shortest strong line measur-able with the uviol lamp. The head of the absorption band ofsodium hypochlorite lies beyond this. Shorter lines will thereforebe investigated with the help of quartz apparatus.(6) It has been found that the presence of light from the uviollamp slightly accelerates the bleaching of linen by sodium hypo-chlorite.I n conclusion,. the author wishes t o express his thanks t oProfessor W. C. McC. Lewis for suggesting and supervising theresearch.‘THE MUSPRATT LIRORATORY OF PHYSICAL A S D ~ 1 , 1 2 C . ~ ~ O - C t I l ~ ~ I I S ’ I . R Y ,THE UNWEKSI I‘Y OF LIVEP.POUL