1932
GERALDs. GOLDEN .4ND HERBERT AT. CLlRK
spectra and tjhespectrum of hydrated ferric chloride indicate that water is not closely connected to the tetrachloroferrate anion; ie., the anion resembles the anhydrous salt much more than the hydrated ferric chloride. Acknowledgment.-The authors wish to express their gratitude to Dr. Norval J. Hawkins for his
Tol. 63
instructive discussions concerning magnetic resonance phenomena. The iron analyses were performed by Carolyn J. McCoy and David A. Del Grosso. Certain materials used in this research were made available a t Rensselaer Polytechnic Institute by the U. S. Atomic Energy Commission under Contract No. AT(30-1)-1663.
THE EXTRACTION OF FERRIC BROMIDE BY DIETHYL ETHER1 BYGERALD S. GOLDENAND HERBERT M. CLARK Department of Chemistry, Rensselaer Polytechnic Institute, Troy, New York Received December 8.9, 1960
The evtraction of Fe(II1) fron aqueous acidic bromide solutions into diethyl ether a t 24.9" was studied as a function of the concentration of hydrobromic acid, iron and salting agents. Included in the study was an investigation of the extraction of hydrobromic acid into diethyl ether. Iron extracts m a strong acid, HFeBra, which forms ion clusters in the organic phase a t high concentrations. The amount of water accompanying the extracting species into the ether decreases with an increase of the ionic strength of the aqueous phase. When the molar ratio of water to hydrogen ion in the ether phase reaches four, two ether phases form. With further increase in aqueous salt concentration, the ratio remains four in the lighter ether phase but continues to decrease approaching unity in the heavier phase.
Although many investigators have studied the solvent extraction of iron(JI1) as tetrachloroferric acid by various organic solvents, relatively have studied the extraction of tetrabromoferric acid. I n this paper the results of an investigation of the extraction of Fe(II1) from acidic bromide solution by diethyl ether are described. The investigation included a study of the extraction of hydrobromic acid since it not only co-extracts with tetrabromoferric acid, but also largely controls the relative volumes of the equilibrated phases Experimental Materials.-Reagent grade diethyl ether was treated to remove peroxides, dried over CaClz and distilled over calcium hydride. The fraction boiling a t 34.8 f 0.2" was collected. Reagent grade 48% hydrobromic acid was distilled in a light-protected Pyrex column. The constant boiling fraction wm colkcted and stored in low-actinic glass vessels. Ferric bromide stock solution (1 M ) was prepared by dissolving C.P.ferric oxide in an excess of concentrated hydrobromic acid and stored in low-actinic glassware. Analysis showed less than 0.05% Fe(I1). -411 other materials were reagent grade. Procedure.-Equal volumes of an aqueous solution of the desired romposition and of diethyl ether were mixed thoroughly in stoppered, low-actinic glass cylinders and allowed to equilibrate with intermittent shaking in a water-bath a t a temperature of 24.9 rt 0.1" for a t least 48 hours. This equilibration time was necessary in order to ensure that the Fmall amount of non-extractable Fe(I1) (approximately 0.1% of the total iron) formed in the ether phase migrated to the aqueous phase. Conductance measurements for ether extracts wwe made with a dipping-type cell having a cell constant of 0.100 cm.-l. Absorption spectra were obtained with either a Beckman Model B spectrophotometer or a Perkin-Elmer Spectracord, Model 4000. Methods of Analysis.-In the absence of iron, the HBr concentration in the aqueous phase wae determined by t i t r e tion with standard sodium hydroxide and that in the ether (1) Abstracted from a thesis presented by Gerald S. Golden to Rensselaer Polytecbnio Institute in partial fulfillment of the reqriirements of the Ph.D. degree. This work was supported by the U. S. Atomic Energy Commission, Contract No. AT(30-1)-1663. (2) I. Wada and R. Ishi, Sci. Papers Inst. Phys. Chem. Research (Tokyo), 24, 135 (1934). (3) W. A. E. McBryde and J. H. Yoe, Anal. Chem., 20, 1094 (1948). (4) R. Bock, H. Kusche and E. Book, Z . anal. Chem.. 136, 167 (1953). (5) H. G. Richter, S.M. Thesis. Mass. Inst. of Tech., 1950.
phase by the Volhard method. The latter method was used also for the determination of bromide when iron was present. In the ether phase, where all the iron was present as Fe(III), iron analyses were performed on the same aliquot used for the determination of bromide. After removal of the AgBr precipitate, the solution was acidified with HC1, the iron reduced by Zimmermann-Reinhardt procedure and titrated with Ce(1V) to the ferroin end-point. In the aqueous phase, the total iron concentration was determined by precipitation with ammonia, then dissolution in HCl and titration with ceric sulfate after a stannous chloride reduction. Foy very low iron concentrations aluminum was used as a carrier. The concentration of Fe(I1) was found by adding an excess of AgNOa to an aliquot of the aqueous phase and titrating immediately with ceric sulfate. The aqueous Fe( 111) concentration was found by difference. In the ether phase the hydrogen ion concentration was calculated as the difference between the bromide ion concentration and three times the ferric ion concentration. Water in the ether phase was determined by the Karl Fischer method. In the presence of iron, the Laurenee modification of the latter method was used.
Results and Discussion Extraction of HBr.-The distribution of HBr is shown in Fig. 1. The data of Chalkely and Williams' a t 13' are included for comparison. Above 4 M initial aqueous HBr concentration the solubility of diethyl ether in the aqueous phase increases rapidly with resulting increase in the volume of the aqueous phase and decrease in the volume of the ether phase. Above 6.7 M HBr only one liquid phase remains after mixing. Over the two-liquid-phase range of HBr concentration the equilibrium concentration of HzO in the ether phase decreases with increasing initial HBr concentration from 0.444 41 for water-saturated ether to 0.332 M at 3.01 M HRr and to 0.227 M a t 6.17 M HBr. As the water in the aqueous phase becomes insufficient to fully solvate the proton, a t the higher concentrations of HBr, water is withdrawn from the ether phase, the solubility of ether in the aqueous phase increasps, and conditions become favorable for solvation of (6) A. H. Laurene, Anal. Chem.. 24, 1496 (1952). (7) D. E. Chalkely and R. J. P. WilliarnP, J . Chem. Soc., 1920
(1955).
Nov., 1961
E X T R A C T I O N OF FERRIC
BIZOMIDE BY DIETHYL ETHER 102
I
I
1933
1
I
10'
10'
=: 10-1
lo-:
2.0 4.0 6.0 Aqueous HBr concn., M. Fig. 1.-Extraction of HBr by diethyl ether: 0 , initial aqueous HBr, M ; 0, equilibrium aqueous HBr, M ; t, data of Challrely and Williams7 a t 13", equilibrium aqueous HBr, M. 0
lo-,
3 4 5 6 7 Aqueous bromide concn., AI. Fig. 3.--0 for Fe(II1) as a function of aqueous HBr concii.: 0 , initial total bromide concn.; 0, equilibrium totitl bromide concn. 2
2
2
4 6 8 10 12 Initial bromide concn., M . Fig. 2.-Kffect of salts on the extraction of HBr by diethyl ether. Salt8 added to 1.06 M HBr: 0, AIBra; 0,NH4Br; V, BaBr2; 0 , CaBrs; 0, LiBr; -0-, RIgBr2; KBr; X. XaBr; @, SrBrz.
i),
the proton by ether and formation of ion-paired or even molecular HBr.
4 6 Initial ionic strength. Fig. 4.-Effect of ionic strength on the extraction of Fe(II1) into diethyl ether from solutions 0.2 M in Fe(II1) and 1.1 M in HBr.
When salts are present in the aqueous phase, extraction increases as illustrated in Fig. 2 for solutions having an initial HBr concentration of 1.06 M . The over-all salt effect is in part a common-ion effect and in part a measure of the lowering of the activity of water in the aqueous
GERALD S. GOLDEN A N D HEHBERT 31.CLARK
1934
I
10-41 10-3
I
1
10-2 10-1 Aqueous phase Fe( 111),M . Fig. 5.--L)ependence of extraction of Fe( 111) on total iron concn.
6.0
r
Yl HEAVY ETHER PHASE
u
E \
5.0
l
--
I
L/J
3.0 2.0
O
,
4.0 6.0 8.0 10.0 Initial aqueous bromide concn., M. Fig. 6.-Variation of Br/Fe(III) in the ether phase with initial total bromide concentration: 0 , HBr; 0, HBr, 1.1M plus metal bromides.
phase. A general treatment of the salt effect has been given by Diamond.* Extraction of Iron(III).-The variation of D, the distribution ratio, for Fe(II1) with HBr concentration for extraction from aqueous phase initially 0.2 M in Fe(II1) as bromide is represented in Fig. 3. D is plotted us. both the total initial and the total equilibrium aqueous bromide concentration. Volume changes are the same as those for the extraction of HBr alone a t corresponding acid concentrations. D becomes essentially constant when the equilibrium aqueous HBr concentration also becomes constant. If the aqueous bromide concentration is increased by adding a bromide salt a t constant initial FeBr3 and HBr concentrations (0.2 and 1.1 M , respectively), D increases reaching values of about 50 for 2.2 M alkaline earth bromides or 3.8 M lithium bromide. The extraction of Fe(II1) increases with increasing salt concentration because (1) with increasing bromide concentration there is greater conversion (8) R. M. Diamond, J P h y s . Chern., 65, 669 (1959).
1'01.
65
of Fe(II1) to the extractable HFeBr4 and (2) with increasing cation concentration there is a lowering of the water activity in the aqueous phase with resulting greater etheration and extraction of the proton. There is also greater conversion of various hydrated Fe(II1) cations having a coordination number of six to FeBrc- having a coordination number of four. These common-ion and the diverse-ion salt effects are illustrated in Fig. 4. Extraction increases with increasing concentration of Fe(II1) as shown in Fig. 5, which is based on data obtained by varying the initial aqueous iron concentration from 0.02 to 0.5 M while holding the equilibrium aqueous concentrations of acid and bromide constant a t 2.3 and 2.9 44, respectively, using HBr and NaBr. This behavior and the variation of the electrical conductance of extracts when diluted with ether are consistent with polymerization or ion-clustering in the ether phase and resemble those of the strong acid tetrachloroferric acid in isopropyl ether.g The conductance per equivalent of iron as tetrabromoferrate ion decreases from 29.7 to a minimum of 4.29 and then increases to 18.1for ethereal iron concentrations of 0.306, 0.00245 and 0.000098 M , respectively. Further evidence of clustering may be obtained by examining the concentration ratio of bromide to iron in the ether phase. In Fig. 6 the ratio, corrected for the bromide expected in the absence of iron, is plotted as a function of the initial aqueous bromide concentration corresponding to 0.2 JI FeBr3and variable HBr concentration for one curve and to 0.2 JI FeBrs, 1.1 AI HRr and variable bromide salt concentration for the other. The initial rise from three to four could result from a slight extraction of FeBra a t low aqueous bromide concentration. Values above four could be an indication of the extraction of higher anionic complexes, e.q., FeBrs=, except that spectral measurements in the region 320-520 mF show identical spectra for ether extracts over the entire range of aqueous bromide Concentration studied. Values above four are attributed to increased solubility of hydrobromic acid in the ether phase resulting from the formation of mixedion clusters, Le., Br-H+FeBr4- and H+FeBr4-H+, in the ether phase at high aqueous bromide concentrations. There is, then, increased solubility of hydrobromic acid beyond that already ascribed to ion pair formation in the absence of iron. The extent of hydration of the proton in ether extracts of tetrabromoferric acid varies with the conditions of extraction. Thus, the concentration ratio H 2 0 / H + decreases from 8.45 to 5.99 for an increase of initial HBr concentration from 2.70 to 5.40 X. Khen LiBr is added to the system, initially 0.2 M in FeBra and 1.131 in HBr, the ratio decreases to four when the LiT concentration is about 8 Jl. At higher aqueous Li' concentrations two ether phases form. For the heavy ether phase in which mixed ion clustering favors an increase in the iron concentration with increasing aqueous lithium concentration, the ratio (9) D. E. Campbell. H. >I Clark and W. H. Bauer, zbzd. 62, 506 (1958).
QUASTITATIVE DIFFERENTIAL THERMAL AXALYSIS
Xov., 1961
continues to decrease and approaches unity. For the light phase, however, the ratio remains essentially constant a t four and the iron concentration decreases. Thus, there is no single value for the number of water molecules associated with the extracting species for all conditions of extraction.
1935
The number varies with the aqueous acid or salt concentration, i.e., with the activity of water in the aqueous phase. Similarly, the formation of two ether phases is governed by the activity of water in the system as controlled by non-extractable salts in the aqueous phase.
QUANTITATIVE DIFFERENTIAL THERMAL AXALYSIS BY CONTROLLED HEATING RATES’ BY EDWARD STURM Department of Geology, Texas Technological College, Lubbock, Texas Received January 26, 1961
The determination of quantities of heat, associated with thermal changes occurring in a sample when investigated by the method of differential thermal analysis, requires difficult evaluations of experimental variables such as the thermal conductivity of the sample and the heat leakage through the thermocouple wires. The method suggested here is based on the experimental determination of the over-all thermal conductivity of the total sample. The over-all thermal conductivity is computed from data obtained when heating the sample a t a rate resulting in a constant thermal gradient. The conductivity thus found is incorporated in a constant of proportionality which relates the heat of the reaction to the area under the differential thermal curve.
Introduction The aini of quantitative differential thermal analysis is the evaluation of heat changes taking place in a substance while it undergoes an exothermic or endothermic reaction. I n the conventional method, the sample substance and a thermally inert substance are heated at a constant rate. By means of a differential thermocouple, the e.m.f.’s developed by the two junctions embedded in the sample and inert substances can be continuously compared. The differential e.m.f., which is proportional to the differential temperature, is plotted against the temperature prevailing in the center of the inert reference substance by means of :m X-Y recorder or two strip-chart recorders. Thermal changes occurring in the sample are recorded as deviations from the base line (Fig. 1). The area under the curve is approximately proportional to the amount of heat liberated or absorbed during the thermal r e a ~ t i o n . ~ -The ~ numerical relationship between the quantity of heat involved and the area under the curve may be expressed as Q
=
Several workers6J made use of the divergence theorem (Green theorem) to obtain an expression for the total heat. Boersma6 derived a general expression (2) for the total heat
where Q’ = quantity of heat per unit volume ( ~ a l . / c m . ~ ) V = volume of sample (cm.3) S = surfwe of sample (cm.z) k, = thermal conductivity of sample (cal./sec. cm. deg.)
For a sample holder of cylindrical shape, whose height (h) equals or exceeds twice its radius ( T ) , Boersma6 derived the relationship
where r
=
radius of the sample holder (cm.)
Although the lower limit of integration should be r = radius of the thermocouple bead, no significant error is introduced by making this limit equal zero. Further, since Q = Q’V, and V = nr2h
$Lf2edt
where Q = quantity of heat (cal.) = differential temperature (deg.) tl, t2 = initial and final time of the reaction (sec.) $ = proportionality constant to be evaluated experinientally (cal./sec. deg.)
e
To evaluate the results of differential thermal analysis quantitatively, one must find the constant of proportionality for the given experimental conditions. (1) This work was supported by a research grant from Texas Technological College, Lubbock, Texas. (2) F. C. Kracek. J . Phys. Chem., 34, 225 (1930). (3) L. G. Berg, Compt. rend. Acad. Sci. U.R.S.S., 49, 648 (1945). (4) S. Speil, et al., U. S. Bureau of Mines Tech. Paper 664, 1945. (5) M .J. Vold, Anal. Chem., 21, 683 (1949).
In practice, the conductivity, k,, is the over-all conductivity of the total sample. This bulk conductivity, henceforth referred to as the effective conductivity, ke, is a function of not only the material under test but also of the effects of the thermocouple mires and the sample holder. The Effective Conductivity In the conventional experimental arrangement, the thermocouple beads are embedded in the center of the sample and inert substances contained in (6) S. L. Boersma, J . Am. Ceram. SOC.,38, 281 (1955). Note: the symbols used here are different from those used by Boersma. (7) C. M. A. deBruijn and W. Van der Marel, Geol. en Mignbouzu, 16, 69 (1954).
