Spectrophotometric Investigation of the Interaction between Iron (II

Spectrophotometry of Iron(II) and Iron(III) Interaction in HC1. 5557. [Contribution No. 1490 from the. Gates and. Crellin. Laboratories of Chemistry, ...
0 downloads 0 Views 436KB Size
Dec., 1950

SPECTROPHOTOMETRY O F IRON(I1) AND

[CONTRIBUTION No. 1490 FROM

THE

I R O N ( I I 1 ) INTERACTION IN HC1

5557

GATESAND CRELLINLABORATORIES OF CHEMISTRY, CALIFORNIAINSTITUTE OF TECHNOLOGY]

Spectrophotometric Investigation of the Interaction between Iron(I1) and Iron(II1) in Hydrochloric Acid Solutions1 BY HARDENMCCONNELL AND NORMAN DAVIDSON~ This communication describes a study of “optical interaction absorption” in hydrochloric acid solutions containing iron in two oxidation states, iron(I1) and iron(II1). That is, a 5-12 F hydrochloric acid solution containing both iron(I1) and iron(II1) exhibits a total light absorption in the 550-800 mp wave length range which is markedly greater than the sum of the light absorptions of (a) complexes containing only iron(I1) and (b) complexes containing only iron(III), present in the solution. Investigations in this Laboratory of the interaction absorption exhibited by hydrochloric acid solutions of antimony(II1) and (V),3 by hydrochloric acid solutions of tin(I1) and (IV), and by mixed hydrochloric acid-perchloric acid solutions of copper(1) and (11)4have already been reported?

Experimental Hydrochloric acid solutions containing iron(II1) were prepared by slowly dissolving anhydrous ferric chloride in hydrochloric acid of the desired formality. It was assumed that the change of the hydrochloric acid formality was negligible. The iron(II1) concentration was determined by reduction t o the ferrous state with amalgamated zinc and titration with standard potassium permanganate solution, using the Zimmerman-Reinhardt procedure. Hydrochloric acid solutions containing both iron(I1) and (111) were prepared by washing crystals of FeClz.4H20 with hydrochloric acid and then dissolving the crystals in hydrochloric acid solutions containing iron(II1). The iron (11) concentration was determined with standard potassium Permanganate solution and the total iron concentration was determined as above. Solutions containing manganese(I1) and iron(II1) were prepared by dissolving weighed samples of crystalline MnClz.4Hz0in 12 F hydrochloric acid solutions containing iron(II1). The crystals of MnClr4Ht0 were analyzed by the method of Lingane and Karplus6 to insure that the water of hydration corresponded to this formula. Weighed crystalline samples of MgC12.6HzO were dried a t 85“ (dec. temp. 117’) for a period of thirty minutes. Zinc chloride solutions were prepared from fused zinc chloride. Hydrochloric acid solutions containing only iron(I1) were prepared by dissolving powdered iron metal in hydrochloric acid of known formality, and then filtering the solution through a fine sintered glass funnel. The operations of dissolving the iron, filtering the solution and measuring volumetric samples were all performed in the same closed air-free apparatus, which was completely flushed with carbon dioxide. Titrations of iron(I1) were performed in a carbon dioxide atmosphere. Hydrochloric acid concentrations (1) Presented in part at the San Francisco Meeting of the Amencan Chemical Society, April 1, 1949. (2) To whom inquiries concerning this article should be addressed. (3) J. E. Whitney and N. Davidson, THISJOURNAL, 69, 2076 (1947); 71, 3809 (1949). (4) H. McConnell and N. Davidson, ibid., 74, 8168 (1950). (5) Dr. Harvey Diehl and collaborators at Iowa State College, Ames, have independently studied the iron(II), (111)system (private communication). (6) J. Lingone nnd U,Puplur, I r d E*#. Chrm., AN&. E d . . SO, 191 (1940).

were determined by titration with standard sodium hydroxide solution to the methyl orange end-point Absorption spectra were measured with the model DU Beckman spectrophotometer. All the absorption data except as given in Figs. 3, 4 and 7 were obtained with one centimeter light paths. The absorption data of Fig. 3 were determined with 0.10, 0.03 and 0.01 cm. light paths obtained with calibrated quartz spacers placed in a quartz cell of 1.00 * 0.002 cm. light path. Some of the data of Figs. 4 and 7 were determined with a 10.0 cm. light path.

.

Results and Discussion Figure 1 illustrates the effects of added ferrous chloride, manganous chloride, zinc chloride and magnesium chloride on the absorption spectrum (550-725 mp) of a solution of ferric chloride in 12 F hydrochloric acid. All of the above divalent halides are essentially colorless in this wave length range. The data of this figure clearly indicate the marked and specific effect of iron(I1) on the absorption spectrum of solutions of iron (111).

0.6

0.5

d 0.4

0.3

550

650 700 mp. Fig. 1.-Effects of several salts on the absorption spectrum of iron(II1) in 12 F hydrochloric acid: curve I, 0.88 F FeClr, no added salt; curve 11,0.88 F FeC18,0.13 FMgClz; 8 , O . S FFeCb, 0.13 FMnClg; Q O . 8 8 FFeCk, 0.13 F ZnClr; curve 111, 0.88 F FeCb, 0.13 F FeCh. 000

By analogy with other investigationrr of interaction ab~0fption,(14 it i5 anticipated that the total optical dcndty, D(X, GII, CIIX), of ZL hydro-

(11) and magnesium(I1) used for Fig. 1. The agreement between the two sets of the data supports equation (2). A test of this relation D(h, CII, CIII) = (ClI - CD)eII “r (CIII - c D ) t l I I f (CII over a wider range of iron concentrations was CD)(CIII - C D ) ~ (1) precluded by the rather small values of k and by The three terms on the right hand side of (1) the limited solubility of ferrous chloride in conrepresent the optical densities of iron(I1) com- centrated hydrochloric acid solutions of ferric plexes, iron(II1) complexes and “interaction chloride. There is agreement to within about complexes,’’ each interaction complex containing 10% for the values of k obtained using the one atom of iron(III), one atom of iron(II), and a magnesium and the manganese corrections. number of coordinating chloride ligands. In An attempt was made to determine the intereq. (l),CII and 6111are the formal concentrations action absorption a t shorter wave lengths, 450of iron(I1) and (111), CD is the concentration 550 my by employing short light paths. As of the interaction dimer, $11 and 21x1 are the shown in Fig. 3 the optical density of interaction formal extinction coefficients of iron(I1) and iron- absorption is small compared to the large optical (111). The interaction constant, k , is a function densities due to the iron(III)-chloro complexes in of the formation constants and extinction co- this wave length range. Furthermore, the differefficients of the various interaction complexes. ence between the optical densities of an iron(II1)By analogy, with previous studies, it is further magnesium(I1) mixture and an iron(II1)-mananticipated that the interaction dimer is strongly ganese(I1) mixture is of the same order of magnicolored and is present a t sufficiently low concen- tude as the difference between the optical densitration so that equation ( 2 ) holds. ties of an iron(II1)-iron(I1) mixture and an D(X, CII, 6111) = crrtrr ciircili C I I C I I I ~ (2) iron(II1)-manganese(I1) mixture. The values To evaluate the interaction function, k , we as- of the interaction absorption function, k , are sume that there is a small non-specific effect of fer- therefore quite different depending on whether rous chloride on the crrIZIII term of ( 2 ) which may mixtures of iron(II1) with manganese(I1) or be estimated from the effect of manganous chlo- with magnesium(I1) are used to determine the iron(II1) absorption; it is not even possible ride or magnesium chloride on the c r 1 1 h term. Figure 2 exhibits the values of the interaction to say with certainty whether the interaction function, k, calculated from the data of Fig. 1 absorption in this range is greater or less than in using equation ( 2 ) and also values of k calculated the 5FjO-700 mp range. from a series of measurements on 12 F hydrochloric acid solutions containing one-half the concentrations of iron(III), iron(II), manganesechloric acid solution containing iron(I1) and (111) can be expressed by the equation

+

+

100

80

d

60

12.0 Qi 8.0

10

4.0

0.0

1.0 20

500

600

700

800

410

m. Fig. 2.-Dependance of k on wave length and hydrochloric acid concentration: I, 12 F HC1 solutions. Points 0 and 9 of this figure calculated from the data of Fig. 1, using magnesium and manganese corrections, respectively ; points 0 and 8 , solutions of Fig. 1, diluted by a factor of two,magnesium and manganese corrections, respectively. II, 5.0 PHC1 solutions, data of Fig. 5 (ordinates, k X 10).

-4.0

480

520

560

IriG.

Fig. 3.-Optical

interaction in the 460-560

rnp

range:

T, optical density, D (referred to a 1.00 cm. light path), of 0.88 F FeC4 in 12 F HC1 and containing 0.13 F MgCL. 11, 111, values of k calculated using magnesium(I1) and manganese(I1) corrections, respectively. Concentrations identical with those of Fig. 1. (Curves are dashed in regions of large probahlr error

Dec., 1950

SPECTROPHOTOMETRY OF I R O N ( I 1 ) AND

IRON(II1) INTERACTION

IN

HC1

5559

Figure 4 gives the optical densities, D(I1, 111), and CIIZII, measured with the Beckman spectrophotometer from 720 to 1,000 mp. As shown in Fig. 2, the interaction absorption is observed to decrease continuously with increasing wave length. At 920 mp, the optical interaction absorption is equal to zero within the experimental error of 5%. At wave lengths greater than 920 mp, k was found to be zero within the experimental error; however, the high intensity of scattered light in the spectrophotometer a t these wave lengths makes this conclusion uncertain. CIIIZIII

6

I

I

1 520

600

w.

680

760

Fig. 5.-Optical densities of 5 F hydrochloric acid solutions containing I, 0.68 F iron(II1). 0.71 F iron(I1); 11, 0.68 F iron(II), 0.71 F magnesium(I1). (The light absorption by iron(II), not shown in the figure, is important a t wave lengths above 625 r n w and has been taken into account in calculating k for these wave lengths.)

. .

* -. .-.

0.1

It is known that iron(II1) forms chloro complexes in hydrochloric acid solutions. Rabinowitch and Stockmayera have determined the formation constants and the absorption spectra of the complexes, FeClf-', for i = 1,2,3 in dilute hydrochloric acid solutions. Metzler and Myers9 have suggested that FeC14- is the predominant

840, 920 1000 mp. Fig. 4.-Optical densities of 12 F hydrochloric solutions containing I, 0.88 F iron(II1) and 0.13 F iron(I1); 11, 0.88 F iron(II1) and 0.13 F magnesium(I1); 111, 0.13 F iron(I1).

760

Figure 5 gives the data used in the determination of optical interaction absorption in 5 F hydrochloric acid solutions. The optical interaction absorption in 5 F hydrochloric acid solutions is not sufficiently intense to allow an experimental verification of eq. (2). Assuming eq. (2) to hold for these solutions, values of k have been calculated from the data of Fig. 5 and are plotted in Fig. 2. It may be seen that k decreases (by a factor of approximately 20) when the hydrochloric acid concentration is decreased from 12 to 5 F. Figure 6 shows the absorption spectra of solu350 450 550 tions containing Fe(C104)2 and Fe(C104)3 in 4 F perchloric acid.? In these solutions the mp. values of k (eq. (5)) are calculated to be 0 * 0.05 Fig. 6 . 4 p t i c a l densities of 4.0 F perchloric acid conand we may infer that the interaction absorption taining: I, 1.08 F Fe(ClO&, 11, 0.54 F Fe(ClO&, 0.54 F between Fe(H20)e++and Fe(HzO)e++f is much Fe(C104)2; 111, 1.08 F Fe(ClO4)n. less than that between the chloro complexes of (8) E. Rabinowitch and W.Stockmnyer, TmS JOWRNAL, 64, 336 these ions. (1942). (7) The data of Fig. 6 were taken by Mr. Wendell Miller.

(9) D. E. Metzler and R. J. Myers, ibid., 71,3776 (1950)~

complex species of iron(II1) in 12 F hydrochloric acid.

tion that interaction absorption due to hydroxyl bridging may occur in mixed solutions of Fe(C10& and Fe(ClO,), which are less acid than the ones studied here.l1 The increase in the interaction constant, k, with increasing hydrochloric acid concentration that has been observed for all the cases of interaction absorption studied in this Laboratory, viz., Sb(II1, V), Sn(I1, IV), Cu(1, 11) and Fe(11,111) may be due to one or both of the following factors. Due to both mass action and activity effects, the concentrations of the interaction dimers may be increased by increasing the hydrochloric acid concentration. Furthermore it is to be expected that of the various interaction Complexes formed, those having a larger number of chloride ligands will be the more strongly colored in the long wave length part of the absorption spectrum. This is similar to the shift of the "electron transfer" spectra of complexes containing a single cation to longer wave lengths 700 800 900 .is the number of chloride ligands is increased.s*12 w. Acknowledgments.-This research has been Fig. 7.--Dependence of the formal extinction coefficient supported by the Office of Naval Research. We of iron(II), i I 1 , on hydrochloric acid concentration: I are indebted to Professor Dan H. Campbell for 4.1 F HCI, 0.31 F FeClr; 11, 5.0 F HCl. 1.42 F FeCle; many kindnesses in connection with the use of 111, 10.4 FHCI,0.29 FFeC12; IV,12 F HC1,O.OS FFeClp. the spectrophotometer. It is a special pleasure Figure 7 shows that the formal extinction co- to acknowledge that Mr. Rollie Myers, formerly efficient of iron(II), VII, decreases as the hydro- of this Laboratory, made the original observation chloric acid concentration is increased. It is of the occurrence of interaction absorption for likely that this is due to the formation of iron(I1)- mixtures of iron(I1) and (111) in concentrated chforo complexes. (It is highly improbable hydrochloric acid solutions. that there is sufficient Fe(0H)f a t the acidities Summary used to account for the light absorption by the The light absorption in the 450-900 mp wave solutions.) Thus it is seen that a strongly colored length range by mixed solutions of iron(I1) interaction dimer can be formed in solutions containing chloro complexes of both iron(I1) and iron(II1) in hydrochloric acid is interpreted and iron(II1). The species Fe(HzO)s+++ and as evidence for the formation of unstable but Fe(H20I6++do not interact. Mr. J. A. Ibers, strongly absorbing interaction complexes, each in this Laboratory, has observed that there is no interaction complex containing one atom of ironinteraction absorption in mixed solutions of Fe- (11), one atom of iron(II1) and a number of (CN)F and Fe(CN)p. There appears to be coordinating chloride ligands. The light abinteraction absorption in the solid ferroferri- sorption by interaction complexes decreases with cyanides, in FeaOs, in freshly precipitated mix- decreasing hydrochloric acid concentration and tures of Fe(0H)z and Fe(OH)a, and in numerous there is no interaction absorption by solutions minerals containing iron(I1) and (111), e. g., containing Fe(H20)e++, Fe(H20)s+++ and no biotite and tourmaline.'O It is possible that chloride ion. Absorption spectra of iron(I1) in solutions of direct chemical bridging of a ligand between the two iron atoms is important for the occurrence varying hydrochloric acid concentration observed of interaction absorption. Thus the bridged in the 700-900 mp wave length range are used to structure, Fe-N :Fe, occurs in the ferri-ferro- show the presence of iron(II)-chloro complexes. cyanides, and Fea04 has 0" bridges between the PASADENA, MARCH10, 1950 CALIFORNIA RECEIVED two kinds of iron atoms. The interaction dimer ( 1 1 ) The one known case where there is not a direct bridge of the in the hydrochloric acid solutions may contain type discussed above is for the intensely black salts of the type a halogen bridge between the iron(I1) and the CstSbCle. in which each antimony ion is surrounded by an octahedron chlorides and each chloride is co6rdinated to only one antimony iron(lI1) atom. This theory leads to the sugges- of ion (L. A. Jensen, 2. a m # . Chcm., am, 817 (1944)). (10)L. Pauling, Chrm. Rag. News, 96. 2970 (1947).

(12) E. Rabinowitch, Rev Mod. Phys., 14, 112 (1942).