The Thermodynamics of Aqueous Hydrofluoric Acid Solutions

The Thermodynamics of Aqueous Hydrofluoric Acid Solutions. Herman H. Broene, Thomas de Vries. J. Am. Chem. Soc. , 1947, 69 (7), pp 1644–1646. DOI: 1...
3 downloads 15 Views 382KB Size
1644

HERMAN H. BROENIF AND THOMAS DE VRIES

Vol. 69

chlorite. Withk five to fifteen minutes one mole of eyanopz efdmide is destroyed in this rapid reaction for each mole of hypochlorite present.

0jSwg.en chloride m c b with hypoeldorite stoi&rnetricdy cornto t f K oxidstim of the nitrogen it contains to gaseous &%en. Since the reaction is mml~td#ymote rapid thsn summary The kinetic8 of hydrolysis of cyanogen chloride hydrolysis under the 58me conditions, there must in alkaline, bicarbonate-buffered and phosphate- be a rapid direct reaction between cyanogen chfobuffered solutions -appear to be in satisfactory ride and fiypoclxlorite. This was shown to be even more rapid than subsequent disappearance of agreement with the following expression. "active chlorine." -d[ClCN]/dt = 6 X lO*[ClCN][OH-] 3 X 10-2 [ClCN][phosphate]

+

RECEIVED JUNE 20,1946

NOTREDAME,INDIANA

[ CONTB.XBUTION FROM THE DEPARTMENT OF CHEMZSTRY, PuaDUE UN-1

The Thermodynamics of Aqueous Hydrofluoric Acid sqlutions BY HERMANH. BROENE'AND THOMAS DE VRIES Two instances are recorded in scientific literature in which a hydrogen electrode has been used in hydrofluoric acid solutions. Wynne-Jones and Hudleston2 measured the e. m. f. a t 25' of the cell Ha, HF ( N ) , EC1 (sat.), Hg2C12, Hg, where N varied from 0.03 to 1. 5. Jahn-Held and ellhekg of the cell Pb ( 5 k amalm d the gam), PbF2(s),NF(l N ) , H1 a t 15,25 and 39'. Ivett and De Vries' have made a study.of the cell Na(amdgam), NaF(m), PbFZ(s), Pb(amalgam), and have determined the standard potential of the lead amalgam-lead fluoride electrode. In the present investigation use was made of this standard value to determine thermodynamic properties of hydrofluoiic acid solutions. The activity coefficients of hydrofluoric acid were calculated from measurements made on cells of the type: I Pb(5oJo amalgam), PbF2(s), HF(m), Ht(g); and the equilibrium constants from cells of the type: 11, Pb(5% amalgam), PbF&), HF(ml), Nfi(n2a), Hdg). Experimental M&&s.--Lead amalgam, lead fluoride and sodium fluoride were prepared in the manner described by Ivett and De Vriss.' Hydr~genfrom a commergisl cylinder was bubbled through a chromous (11) sulfate' or chromous chloride6 sofutMnto remove traces of oxygen,and through B solution of copper sulfate to remove any hydrogen sulfide which might have been formed in the first solution.' The hyd r m was bubbled twice through distilled water, and finally through a hydrofluoric acid solution of the same comehttation and temperature as that used in the cell. A few of the hydrofluoric acid solutions were prepared by diluting Baker and Reagent 48% hydroflu& acid with conductivlty water. Most solutions, howew, were made from a stock solution prepared by bubwlrg HatrrhafR ahydrous hydrogen fluoride into c&&dty water.

+

11) Abstract of the Ph. D. dhertation of H. H. Branc whose present oadnsa is Boreman E0d.k Ca., Rochester, N.Y. (2) Wyn-Joneurd Htldleston, J . C h . h.Iu, , 1081 (1924). (0) J.hn-€€dd rrtd JdlincL, 2. EJsklrockn., Q, 401 (1886). (4) I n r t r d Po Vriw, T a r r J o w a ~N, ~ ~asSl(1941). , (6) S t o r a d - , I d . &u. C k n . , A U d &., 8,188 (I=). (w. ST, 4M (leu). -.ad (a S m w m , J . N 4 &r. Slodadr, U, 46 ClOSn.

(w

Anal lis of Hydrofiudc Acid.-A measured - o a t of hydrof&oric acid was plwed in an Erlenmeyer &t& and phenolphthalein added. Sodium hydroxiae was run in from a buret until the acid was almost The mixture was heated to boiling and tlze tirah-2 while hot. The sodium hydroxide &tibn iern$iLiing io the buret was used to titrate a sample ctf standprd hydrochloric acid in the same manner. The resufts of t h r e or more analyses were averaged. All dtmietric ware was carafully calibrated and a carreetion made to each titration forearbon dioxide dissolved in the acids and wash water. While the titration in bare Pyrex involves the formation of some fluosilicicacid, this wm not interfere if the solution is kept above 8oo.* The molality of each solution was calculated from the normality using the density data of Winteler.@ In order to prepare mixed solutions of hydrofluoric acid and sodium fluoride, the acid solutian was first prepared and standardized. To a known weight of this solution, a weighed amount of dry sodium fluoride was added and the molality calculated. Apparatus and Procedure.-The apparatus used was of conventional design. The anode andathode were, however, in two separate compartments connected by a bridge containing the electrolyte. This was necessary to prevent poisoning of the platinized platinum electrode by lead ions. The electrodes were sealed in place with paraffin wax to make the cell air tight. Hydrogen escaped through a trap containing distilled water. The cell was placed in aowater thermostat which held the temperature to *0.02 Hydrogen was allowed to ow slowly past the platinized platinum electrode, and after three to five hours the e. m. f . became cotlstant. When changed to another temperature, readings became constant after about an hour and a half. Measurements were made on a Rubicon, type B, potentiometer, using a carefully calibrated standard cell. A fresh platinized platinum electrode was used for each cell assembled. Platinization was done in two ways: one using chloroplatinic acid only, and one using acid containing a trace of lead acetate. Both methods gave satisfactory dectrbdes which agreed very closely in e. m. f. readiip. Platinization, however, took longer without lead acetate, and it gave a dull gray deposit instead of a black one. The cell apparatus was constructed of Pyrex glass. The interior was coated with a special lacquer" and then with ceresin wax. Before a satisfactory coating of !this lcind was developed, considerable trouble was experienced.

.

THERMODYNAMICS OF AQUBOUS HYDRUFLUORIC ACIDSOLUTIONS

July, 1947

Apparently a piat&ed plathum electrode is very sensitive to pqsomy by substances formed in the attack of hydrofluopc .aud on g h . l%hw g resulted in a potentid whch j h p e d sewed rmlfivolts as mch bubble of hydrogen passsd over the deetrode. The combination d a lacquer awl wax costing, however, usually dorded adequate protection. Measurements were made on approximately 80 cells of Types I and 11. The experimental results are given in Tables I and 11. All readings were corrected to 760 mm. partial pressure of hydrogen.

+

where k = g(m q - r ) f . The activity coeffiaent,f , r e p r a t s activity divided by ionac molality. It was assumed to be the same for all ions present. The Eo value of Ivett and De Vriee' was used in the above equation. The two equilibrium constants may be obtained in the forms k

K1

15O, v.

m

.m57

.00295 .OOO986

25', v.

0.15497 .12636 .09710 .0677 .0406 .0163 -.0061

(m,

Kz (m1

350, v.

0.15458 .12491 .09493 .0649 .0379 .0133 -.0087

">

0.15415 .12381 .09297 .0625 .0355 .0107 -.0113

TABLEI1 FORCE OF CELL8 Pb(Hg), PbFds), ffF(m), N a N m ) , Hnk) mr

1&O,

For each acid-salt solution we obtain two equations like (1) and (2), but three unknowns: R,, Kp and p, The results from two or more solutions may be combined to find K I and Kz. This was done by a graphical method. Various.values of p were inserted in equations (1) and (2), and the carrespondiag values of K1 and K2 determined. These were plotted against each other as shown in Fig. 1 for the calculations a t 25'. We thus obtain 15

35', v.

26O, v.

v. *

(2)

- rn - 2p + ")("> nf2 nf

kECTROMOTIVE

mi

(1)

- m2 - 2p + PfZ ma+q-$ k

TABLEI ELECTROMOTIVE FORCE OF CELLS Pb(Hg), PbFds), HF(m), H h )

0.985 .3174 .lo06 .(BO47

1645

0.0821 0.1000 0.09112 0.08848 0.08538 .08088 .07783 .07367 ,0921 .3015 . 0 1 W . W 2 8 .04393 .03944 ,03480 .012!39 .13388 .03891 .03431 .Om1

Calculations Equilibrium Cmstants of Hydrofft~oricAcid.It seems to be generally agreed" that aqueous hydrofluoric acid contains H +, F-, H F and HF2-. From the results in Table I1 (cells containing sodium fluoride), it is possible to determine the equilibrium constants K1 and KZfor the reactions 0 2 4 6 8 (1) H F = H+ F- and (2) H F 4- F- = HF2-. Ka. It was assumed that in any solution of this series Fig. l.-Graphical solution at 25". the solubility of lead fluoride was negligible, and that the mean activity coefficient of sodium fluo- one line for each of the four solutions studied. ride in the hydrofluoric acid solutions was the Tkae lines are seen to intersect in a point within same as that found by Ivett and De Vries4 for the limits of experimental error. The values of Kl solutions of sodium fluoride only. Both of these and KZ corresponding to the intersections are assumptions were later found to be justifiable. TABLEI11 For example, in a solution of 0.0921 m H F and EQUILYBRIUM CONSTANTS OF HYDROFLUORIC ACID 0.3015 m NaF, the concentrations of the H+, F15' 25' 350 and HF2- ions are 1.21 X 0.2559 and 0.0457 K~ x 104 7.93 (*o.io) 6.71 5.64 m,making the ionic strength 0.3016 (HF concenKa 3.94 (t0.20) 3.86 4.32 tration is 0.0463). The equations were set up as TABLEI V

+

HF ml--g-r HF ml-q-r

+

= H+ P

+

F-

m+g-r = HFa-

Fmrfq-r

r

whereml = total molality of HF, m = total molality of NaF, q = acbal molality of H+ ion, r = actual molality of H R - ion. The e. m. f. of the cell will be given by E = E" 2.303 RT/nF log k,

+

(11) Anthooy and Hudleston, J . Chsm. Soc., 117, 1122 (1826); Dovie a d Hudleston, ibid., 11&,260 (1924); Pick, "Nanst's FeW&ri?t," W .Ktrppp, Halle, 1912, p. 360.

EQWXLIBRIUM &NSTANTS K~ x

104

Ki

OF

HYDROFLUORIC A m

AT

25"

RefeZWCC

Anthony and Hadleaton" Aumetasb 4.7 Davies and Hudleston" 5.6 Pick" Roth, et a1.0 3.98 This work Ref. 11. Aumeras, Comfit. m d . , 184,1659 @f@7). ORoth, Pahlke, Bertram and Borger, 2. E k k r o c h . , 43, 3 m (1937).

7.4 1.67 7.4 7.2 5.5 6.71

4.7

.. ..

HERMAN H. BROENEA m THOMAS DE VRIES

1646

Vol. 69

taken to be the actual values. The graphs a t 15 The activities called for in equation (7) were and 35’ were Similar to the one shown for 25’. obtained by solving the equations 1 to 6 in the Results are listed in Table 111. They compare preceding section. Equation (7) then led to the favorably with values found by other workers, values for EL given in Table V. The values a t 25’ Table IV. calculated from ionic activities by means of the Ionic Activities.-The activities of the various Henderson equationla are shown for comparison. ions in any hydrofluoric acid soluton may be de- The agreement is good considering that both methtermined from the data in Table 111 by means of ods are but approximations. the equations (LHt(1.F-

=

aHFs-

= K f i F - UHF

aHt

=

?? =I UHF

(3) (4) (5) (6)

&UHF aF-

f

+

aH+/f

UHF*-

+

aHFr-/f

TABLEV LIQUIDJUNCTION POTENTIAL

15O, v.

m

35’, v.

2 5 O , v.

25O, v,, Calcd. with Henderson eq.

0.985 +0.0013 +0.0011 +0.0014 -0.0004 where a = activity, m = total molality of HF, f = ,3174 .0004 activity coefficient of H+, HF2- or F- ion, i. e., .0000 - ,0002 - .0006 .lo06 , - ,0014 - .0020 - ,0030 - ,0019 activity divided by ionic molality. These four equations were solved for each hydrofluoric acid .03047 - ,0049 - .0066 - .0085 - .OM4 solution used in cells of Type I. The value of f i n .00957 - ,0111 - .0131 - .0152 - .0113 .00295 - 0162 - . O M - .0207 - .0143 equation (6) was approximated until a value was found which satisfied both this equation and the .000986 - 0200 - .0216 - .0237 - .0157 equation logf = -A’dP/(l fi), where A’ is Activity Coefficients.-The activity coefficients the Debye-Hiickel constant and p is the ionic of hydrofluoric acid in aqueous solution were strength. As has been mentioned, the cells were assem- calculated from the ionic activities of equations bled in two parts, connected by a bridge contain- (3) to (6) with the equation y = (aHfaF-)”*/m. ing the electrolyte. It was assumed that at the The values determined in this way were plotted hydrogen electrode the activities could be ob- on a large graph from which the values listed tained by solving equations (3) to (6). At the in Table VI were taken. It must be noted that lead amalgam-lead fluoride electrode, however, these are stoichiometric activity coefficients with the solution was assumed to be saturated with reference to H+ and F- ions; i. e., the molality m lead fluoride; hence one more activity enters; is the total molality of the hydrofluoric acid, not that of lead ion, and one more equation, the activ- corrected for incomplete dissociation. ity product of lead fluoride. The value of the latTABLEVI ter quantity was calculated from sohlbility measACTIVITY COEFFICIENTS OF HF urements made by Kohlrausch, by Dundon, and 25 ’ 350 15’ m by Carter.12 The Debye-Huckel limiting law was 0.024 I . 022 1.00 0.026 used to estimate the activity coefficient of lead .031 .028 .034 0.50 fluoride in water. The value of the activity prod.044 .042 ,048 .30 uct was taken to be 3.66 X lo4 a t 15’, 4.37 X .077 .070 .083 .lo lo4 at 25’ and 5.17-X lo4 a t 35’. When consid,106 .097 .05 ,115 eringlead fluoride, equation (5) should be modified .136 .125 .03 .148 to

+

+

UH+

+

%pbt+ f+/f+t

= aF-

+

UHF*-

(5’)

where the activity coefiicient of the lead ions can be given by log f++ = - A I 4 4 and for the univalent ions by logf+ = -A ‘ 4 . Liquid Junction Potential.-The Nernst equation for cell I may be written &bi.

= E’sleotrode#

+ RT In

aH$afc

+ EL

(7)

.01 .005

.003 .OOl

,241 .320 .395 .573

.224 .300 ,371 .544

.208

.280 .347 .515

Summary From potentials of the cells Pb(Hg), PbFds), HF(md, NaWmz), H&)

it has been possible to calculate the equilibrium constants at 15,25 and 35’ for the reactions

The subscript x refers to concentration a t the hydrogen electrode (no lead fluoride present); the HF = H f + Fsubscript y to concentration a t the lead amalgamHF + F- = HFzlead fluoride electrode. The term EL represents the potential due to the liquid junction between Stoichiometric activity coefficients €or hydrothe two hydrofluoric acid solutions, one saturated fluoric acid were determined from measurements with lead fluoride and one free from it. The term made on cells of the type is the sum of standard electrode potenPb(Hg), PbFz(s), HNm), Hzk) ti& of the cell, using data of Ivett and De V r i e ~ . ~ (19) &hlrprueh, 2. physik. Chem., 64, 129 (1908); Dwdon. THIS JOIJRWAL, U ,a068 (1928); Carter,I n d . Eag. C h m . , M, 1195 (1928).

LAFAYETTE, INDIANA

RECEIVED FEBRUARY 7,1947

(18) Hendmwn, 2.physik. Chcm., 8S, 118 (IWn); W, 526 (1808).