Iron(III) - American Chemical Society

(8) H. Thommen, bit. .... color of the oxidized solutions gradually intensified with time, and the cell ... between acetonitrile and water(8), the inc...
0 downloads 0 Views 470KB Size
drying were carefully scrapped off. Elution was repeated except that EtOH-benzene was used long enough to move the testosterone-azine derivative from the origin. A series of 4 samples containing benzoic acid, phenol, aniline, and benzamide, respectively, and p-nitrobenzaldehyde was prepared. Quantitative conversion of the aldehyde to the azine was found and none of the compounds interfered in the analysis. Products from the reaction of glucose and xylose with the hydrazone reagent were isolated. However, they were not identified and at this point it does not appear that the reagent is useful for the analysis of sugars. Because of the fluorescent nature of the azine product, the reaction is also a very useful qualitative reaction. Derivatives for 60 carbonyl compounds have already been reported (3)

while this paper reports data for an additional 40 compounds. Furthermore, the hydrazone reagent is not limited to the common organic aldehydes and ketones. In addition to the various types of carbonyl compounds studied in this report, naturally occurring keto carotenoids and carotenals have been isolated as azine-derivatives (8). RECEIVED for review August 1, 1969. Accepted October 13, 1969. Presented before the 17th Detroit Anachem Conference, Detroit, Michigan, September 1969. Financial Assistance from the National Science Foundation (GP 3957) is gratefully acknowledged.

(8) H. Thornmen, Znt. Z . Vitaminforsch.,37, 175 (1967); C.A., 67, 105323.

Iron(III)-(II) Couple in Acetonitrile Oxidation of Thiocyanate by lron(lll) Byron Kratochvil and Robert Long Department of Chemistry, University of Alberta, Edmonton, Alberta, Canuda The formal reduction potential of the iron(ll1)-(11) couple in acetonitrile is estimated to be 1.57 f 0.05 volts VI. a silver-0.01M silver nitrate in acetonitrile reference. Titrations of thiocyanate with iron(ll1) gave inflections at iron to thiocyanate ratios of 0.19 and 1.00, corresponding to formation of an iron(ll1)thiocyanate complex followed by oxidation of thiocyanate to thiocyanogen. The scope of titrations with iron(ll1) in acetonitrile is discussed.

INA NONAQUEOUS SOLVENT the effects of acid concentration, ionic strength, and complexation on a half-reaction may be much greater than in water, especially if ion association is promoted by a solvent of low dielectric constant. In addition, specific solvation of ions is likely, with the result that the potential difference between any two half-reactions generally varies appreciably upon going from one solvent to another. This difference in relative reduction potentials can be analytically useful, as has been shown for the copper(II)-(I) couple in acetonitrile. In this case the stabilization of copper(1) relative to copper(I1) allows use of copper(I1) as an analytical oxidant and as a reagent for investigating the oxidation of organic compounds (1-5). Polarographic data on reactions of metal ions in acetonitrile indicate that iron(II1) should also be an effective oxidant (6). We report here a study of the iron(III)-(II) couple in acetonitrile, along with a potentiometric and spectrophotometric investigation of the reactions of iron(II1) and iron(I1) with thiocyanate. A brief survey of the analytical scope of titrations with iron(II1) in acetonitrile is included. (1) B. Kratochvil, D. A. Zatko, and R. Markuszewski, ANAL. CHEM., 38, 770 (1966). (2) B. Kratochvil and D. A. Zatko, ibid., 40,442 (1968). (3) D. A. Zatko and B. Kratochvil, ibid., p 2120. (4) P. Quirk and B. Kratochvil, 157th National Meeting, American

Chemical Society, Minneapolis, Minn., April 1969. (5) H. C. Mruthyunjaya and A. R. Vasudeva Murthy, ANAL. CHEM., 41, 186 (1969). (6) I. M. Kolthoff and J. F. Coetzee, J . Amer. Chem. SOC.,79, 1852(1957).

EXPERIMEhTAL

Reagents and Apparatus. Acetonitrile was purified by an adaptation of the method of O'Donnell, Ayres, and Mann (7). About 2500-ml batches of technical-grade acetonitrile (Matheson, Coleman and Bell, Norwood, Ohio) were distilled rapidly from 30 g of potassium permanganate and 30 g of sodium carbonate. The distillate was acidified with several drops of concentrated sulfuric acid to remove ammonia formed in the previous step, vacuum distilled rapidly, and then distilled again at high reflux in a 30-plate column at 10 ml per hr. The first liter and the last 200 ml were discarded. Solvent purified in this way showed no absorption in the ultraviolet down to 200 mp and no oxidizable impurities could be detected polarographically at 2.0 V us. a silver-0.1M silver nitrate in acetonitrile reference. Karl Fischer titrations showed the water content to be in the range of 1 to 2 x 1 0 - 4 ~ . Where this amount interfered, acetonitrile prepared as outlined above was stirred with calcium hydride and filtered in a drybox immediately before use; in this way a water concentration less than 5 X 10-5M was achieved. This level corresponds to one drop of Karl Fischer reagent per 50-ml sample. The end point was followed electrometrically with a Metrohm Model 436E automatic titrator. Dry acetonitrile avidly scavenges water from its surroundings ; accordingly the water content began to increase immediately on removal of the calcium hydride, even in a closed container in a drybox. Therefore the water content of solutions prepared from very dry solvent was always considerably higher than the minimum level. Hydrated iron(I1) perchlorate was prepared from perchloric acid and electrolytic iron. Hydrated iron(II1) perchlorate (nonyellow) was used as received (G. F. Smith Chemical Co., Columbus, Ohio, and Alfa Inorganics, Beverly, Mass,). Anhydrous silver perchlorate was prepared by reacting silver carbonate with a slight excess of perchloric acid, evaporating, filtering, and drying under vacuum at about 60 "C. Anhydrous iron(I1) perchlorate, Fe(C10& 6CH3CN, was prepared by azeotropic distillation of acetonitrile solutions (7) J. F. O'Donnell, I. T. Ayres, and C. K. Mann, ANAL.CHEM., 37, 1161 (1965).

ANALYTICAL CHEMISTRY, VOL. 42, NO. 1, JANUARY 1970

43

r'

4 h

/SILVER COUNTER ELECTRODE

ll

PLATINUM GAUZEWORKING ELECTRODE PLATINUMINDICATING ELECTRODE SILICONE -I

B

800

"MAGNETIC

z

600-

2

5 t-

Z 400 -

-SILVER REFERENCE ELECTRODE -POROUS GLASS PLUG

Fe(III) /SCN-

of hydrated iron(I1) perchlorate under reduced pressure in the presence of Linde 3A molecular sieves, followed by removal of the excess solvent under vacuum at room temperature. The dry product is stable to 125 "C. Above this temperature slow decomposition occurs. Acetonitrile solutions of the iron(I1) salt are stable to at least 50 "C. (Note: Safety precautions must always be exercised when heating perchlorate salts in the presence of organic materials.) Analysis of the salt by titration with aqueous potassium dichromate solution gave a purity of 100.3%. Attempts to prepare anhydrous iron(II1) perchlorate, Fe(ClO& .6CHsCN, by metathesis and dehydration were unsuccessful. In each case the solutions turned red-brown, indicating the formation of iron(II1)-hydroxy species despite precautions to keep water contamination to a minimum. The formal potential of the anhydrous iron(II1)-(11) couple therefore was estimated by electrolytic oxidation of 50 of an anhydrous iron(I1) perchlorate solution in a glass H-cell (Figure 1) to give a 1 :1 mixture of the two oxidation states. An anion-exchange membrane in the perchlorate form was used to separate the electrolysis compartments because glass frits and porous Vycor were too permeable to cation transfer under the conditions used. The membrane (AMF A-lOCEC, donated by American Machine and Foundry Co., Springdale, Conn.), along with two silicone-rubber gaskets, was placed between the cell flanges and the unit fastened together with a horseshoe clamp. The IR drop across the membrane was low at the current levels used (6 to 10 mA), and cation diffusion was minimal; in one test no detectable concentration of silver ion diffused into the working electrode compartment in 5 hr. The counter-electrode compartment contained a 0.1 M acetonitrile solution of anhydrous silver perchlorate. Solution volumes of about 25 ml were used on each side of the membrane. All compounds studied, with the exception of the phenothiazines, N,N'-diphenyl-p-phenylenediamine (DPPD), and tetrachlorohydroquinone, were recrystallized once from acetonitrile. Phenothiazine and tetrachlorohydroquinone were vacuum sublimed, The phenothiazine drugs were converted from the hydrochloride or carboxylate salts to the corresponding perchlorates by precipitation from aqueous solution through addition of a concentrated solution of sodium perchlorate, followed by filtration and drying under vacuum. This conversion was made to provide adequate solubility in acetonitrile and to remove anions that might complex with iron(II1). DPPD was recrystallized four times from benzene with Norite A decolorizing charcoal.

RATIO

Figure 2. Titration of KSCN with 0.1M hydrated Fe(C10& solution

STIRRING BAR

EXCHANGE MEMBRANE

Figure 1. Electrolysis and coulometry cell

44

-

2

\ U L T R A FINE FRIT

'\MAGNETIC \ANION STIRRING BAR

9

Measurement of Iron(lII)-(II) Potential in Acetonitrile. The reduction potential of the anhydrous iron(III)-(II) couple was measured in a glass H-cell containing a platinumfoil indicating electrode and a 1.00 X 10-2M silver nitrate in acetonitrile-silver foil reference electrode (hereafter called the silver reference electrode) separated by a salt bridge of 0.1714 tetraethylammonium perchlorate in acetonitrile. In a typical procedure, an approximately 0.1M solution of iron(I1) perchlorate was electrochemically oxidized in a dry box as described previously to give about equimolar concentrations of iron(II1) and iron(II), then diluted to 4 X lOW3M and transferred to the H-cell. The cell was stoppered, removed from 0.05 "C the drybox, and placed in a water bath at 25.00 for measurement. Potentials were measured with a Leeds and Northrup Type K4 potentiometer. Hydrolysis of the iron(II1) by traces of water absorbed by the system caused the potential to fall at a rate of several millivolts a minute. The value reported here was obtained by extrapolation of the potentials to zero time and is considered accurate to +0.05 V. Potentiometric Titrations. Most of the potentiometric titrations were run under nitrogen in a Teflon-stoppered titration vessel described previously ( I ) . Titrant delivery and potential recording were done with a Metrohm Model 436E automatic titrator. Titration times ranged from 20 to 200 min, depending on the rate of reaction.

*

RESULTS AND DISCUSSION Anhydrous I r o n ( I I I ~ ~ I 1 Couple ) in Acetonitrile. The formal reduction potential of the iron(II1)-(11) couple in acetonitrile (2 X 10-3M in each), obtained coulometrically by 50% oxidation of solutions of iron(I1) perchlorate, was 1.57 + 0.05 V L;S. the silver reference electrode. The yellow color of the oxidized solutions gradually intensified with time, and the cell potential drifted downward. The color is attributed to the formation of iron(II1)-hydroxy complexes from traces of water in the solutions. Although this contamination prevented measurement of an accurate thermodynamic potential, the high formal potential indicates iron (111) to be a powerful oxidant in acetonitrile. By use of the rubidium(I)-(O) couple as a reference for compa:ing potentials between acetonitrile and water (8), the increase in the ironcouple reduction potential in acetonitrile over that in water was estimated to be 1.3 V. This increase may be partly attributed to increased stabilization of iron(II), since the crystalfield splitting energy for iron(I1) in acetonitrile, measured (8) J. F. Coetzee and J. J. Campion, J. Amer. Ckem. Soc., 89,

ANALYTICAL CHEMISTRY, VOL. 42, NO. 1, JANUARY 1970

2513 (1967).

1.2

I.

V

z a

9 0,8U 0.4 -

0 300

WAVELENGTH (millimicrons)

400

3500 uu

600 ouu

WAVELENGTH (millimicrons)

Figure 3. Spectra taken at numbered points in Figure 2 Left. To first equivalence point Right. From first to second equivalence point spectrophotometrically by Hathaway and Holah, is some 600 cm-1 greater in acetonitrile than in water (9). Destabilization of iron(II1) also may contribute to the decrease, however, for the lower dielectric constant of acetonitrile would not be expected to stabilize species of high charge. Hydrated Iron(IIIHI1) System. An estimate of the potential of the hydrated iron(II1)-(11) couple was obtained by measurements 100% past the equivalence point in iron(II1) titrations of iodide, hydroquinone, and thiourea. The formal reduction potential in the presence of several moles of water per mole of iron was found to be 1.1 V us. the silver reference electrode. This compares to values measured polarographically at a rotating platinum microelectrode of 0.8 V for iron(II1) perchlorate reduction and 1.3 V for iron(I1) oxidation (6). When the concentration of hydrolyzed iron(II1) species in solution was minimized by keeping the water concentration as low as possible or by the addition of small amounts of perchloric acid to repress hydrolysis, plots of log [Fe(III)]/ [Fe(II)] us. potential gave straight lines with slopes in the range of 63 to 68 mV. Thus the system neared reversibility as the concentration of iron(II1)-hydroxy species was reduced. Thiocyanate Titrations with Iron(II1). Titrations of potassium thiocyanate with hydrated iron(II1) perchlorate gave plots with two well-defined breaks at iron(II1) to thiocyanate ratios of 0.19 and 1.00 (Figure 2). The first titration inflection is attributed to the formation of a mixture of Fe(SCN)e-a and Fe(SCN)60H-8 species. Curves 1 to 6 in Figure 3 illustrate ultraviolet and visible spectra of samples taken at the corresponding numbered points on the potentiometric titration curve of Figure 2. The absence of absorption maxima for the iron(I1)-thiocyanate complexes (log Kl = 5.5, log KZ = 3.7) at 299 and 289 mp in curves 1 to 3 indicates that these species were absent and therefore no oxidation had taken place. Acetonitrile solutions of iron(II1) perchlorate alone gave maxima at 218, 252,278, and 366 mp, while iron(I1) perchlorate produced one broad absorption at 205 to 215 mp. Maxima for thiocyanate occurred at 198 and 234 mp, while thiocyanogen did not absorb in this region. The absorptions at 250,315, and 419-540 mp in Figure 3 are attributed to Fe(SCN),a-2 species, where x is estimated to be 4 to 6. As the iron(I1I) to thiocyanate ratio was increased above 0.19, iron(II1) began to be reduced, as shown by the appear(9) B. J. Hathaway and D. G. Holah, J. Chem. SOC.,1964,2408.

ance of iron(I1)-thiocyanate absorption bands (Figure 3, curves 4 to 6). The end products of the reaction were iron(I1) and thiocyanogen. Thiocyanogen was identified by evaporation of a solution at the equivalence point to dryness and extraction of the thiocyanogen with carbon disulfide. The solid was kept below 0 "C during these steps. Removal of the carbon disulfide by evaporation gave free thiocyanogen, whose behavior paralleled that of the product produced by bromine oxidation of silver thiocyanate. At room temperature both products polymerize to dark red parathiocyanogen, (SCN),. Anal. Calcd for (SCN),: C, 20.6; N, 24.1. Found for the iron(I1) oxidation product: C, 21.7; N, 23.3. Hydrated Iron(II1) Perchlorate as an Analytical Titrant. Since anhydrous solutions of iron(II1) perchlorate in acetonitrile could not be prepared conveniently, titrations with solutions of commercial hydrated iron(II1) perchlorate were investigated. A solution of approximately 0.1M iron(II1) perchlorate was prepared from commercial material that had been stored over magnesium perchlorate. Its stability was checked by titrations with potassium iodide, ferrocene, and tetrachlorohydroquinone. Although purified dry acetonitrile was used to prepare the original solution, no special precautions were taken to avoid contamination by water after preparation. The molarity decreased about a per cent a day. Stability is appreciably improved (to about 1 part per thousand decrease in molarity per day) if precautions are taken to avoid water uptake. In any event, for accurate work the titrant should be standardized the day it is used. The decrease in molarity is due to slow hydrolysis of the iron(II1) rather than reduction to iron(II), because the addition of a small amount of perchloric acid to an aged solution raises the molarity to near the original value. Additional evidence for hydrolysis was obtained from ultraviolet spectra of hydrated iron(II1) perchlorate in acetonitrile solutions, which show a strong absorption maximum at 259 mp. This maximum is similar to one found at about 350 mp in methanol-water solutions, which has been assigned to the FeOH2+species (IO). This absorption accounts for the yellow-to-red color of hydrated iron(II1) perchlorate solutions. A number of organic compounds were titrated in an investigation of the scope of oxidation by iron(II1) in acetonitrile. Table I lists those that gave well-defined titration (10) G. Popa, C. LUM, and E. Iosif, Z . Phys. Chem., 222, 49 (1963).

ANALYTICAL CHEMISTRY, VOL. 42, NO. 1, JANUARY 1970

45

Table I. ComDounds Oxidized bv Hvdrated Iron(II1) Perchlorate in Acetonit& Potential Compound break, mVa Stoichiometry Benzidine 150 2 Carbazole 170 2 Dimethoxybenzidine 100, 100 0.5,2 Diphenylbenzidine 50,350 0.5,2 900 Diphenylguanidine 0.25 N,N'-diphenyl-p-phenylenediamine 120,250 1.0,2.0 Ferrocene 700 1.00 Hydroquinone 400 2.0 600 2-Mercaptobenzothiazole 1 2-Naphthalenethiol 550 1 Phenothiazine 150,175 1.0,2.0 N-Phenylcarbazole 50 2 N-Phenylphenothiazine 300,50 1.O, 2.5 Potassium thiocyanate 300,300 0.19,l.O Promazine perchlorate 150,100 1.o, 2.1 Tetrabutylammonium iodide 300,500 0.66,l.O Tetrachlorohydroquinone 250 2.0 Tetramethylbenzidine 250 2 Tetramethylthiourea 700 1 2 Tetraphenylpyrrole 200 600 1 Thioacetamide Thioacetanilide 120 2 1 Thiocarbanilide 600 Thiourea 500 1 0 Defined as difference in potential at 90% and 110 of end point.

breaks. In general, compounds oxidized by copper(I1) were also oxidized by iron(III), as well as many more for which copper is too weak. Thus, although carbazole was not attacked by copper(II), a well defined stoichiometric reaction with iron(II1) was obtained. Ferrocene, which is oxidized stoichiometrically by copper to the ferricinium ion, was further oxidized to colorless products by iron(II1). This additional oxidation can be avoided by lowering the iron(II1) potential through the addition of a small amount of water to the ferrocene solution before titration. Tetrachlorohydroquinone showed a single two-electron oxidation.

OH

OH

0

The reaction appears analogous to that of copper with hydroquinone. N,N'-diphenyl-p-phenylenediamine (DPPD), which is insoluble in water, gave two well-defined, reversible, oneelectron breaks similar to those reported in 90% acetic acid (11). The reactions are: ~

\

/

C6H5

\ N-C6Hb-N-H H

/

'

-le-

+

f

C&I6=N