COMPOSITION, STABILIZATION, AND LIGHT ABSORPTION OF FE(II)FE(III) HYDROXY- CARBONATE ('GREEN RUST')

Clay Minerals (1989) 24, 663-669 COMPOSITION, S T A B I L I Z A T I O N , A N D LIGHT A B S O R P T I O N OF FE(II)FE(III) H Y D R O X Y C A R B O N ...
Author: Nelson Bryan
8 downloads 0 Views 355KB Size
Clay Minerals (1989) 24, 663-669

COMPOSITION, S T A B I L I Z A T I O N , A N D LIGHT A B S O R P T I O N OF FE(II)FE(III) H Y D R O X Y C A R B O N A T E ( ' G R E E N RUST') H. C. B. H A N S E N Royal Veterinary & Agricultural Highschool, Chemistry Department, Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark (Received 12 December 1988; revised 13 February 1989)

ABSTRACT: The chemical composition of the pyroaurite-type compound Fe(II)Fe(III) hydroxy-carbonate('green rust') synthesised from freshly precipitated ferrihydrite and Fe(II) chloride solution at pH 7.0 ('induced hydrolysis') was determined. The compound was nearly stoichiometric, with the formula FeInFe[l(OH)12CO3, the Fe(II):Fe(III) ratio being independent of the Fe(II):Fe(III) ratio in the initial reaction mixture. It shows all the XRD peaks reported for this compound. Visible-nearIR showsa broad peak at 650 nm which is ascribable to intervalence charge transfer and therefore the absorbance maximum decreases with increasing degree of oxidation. Wetting the material with compoundscontaining hydroxylgroups (such as glycerol or glucose) retards the oxidation of the otherwise very oxidation-sensitive compound. Similar polar organic compoundsmay stabilize the Fe(II)Fe(III) hydroxy-carbonatein wet soils. The importance of the green rust phase as an intermediate in the formation and transformation reactions of oxides and oxyhydroxidesof iron in natural environments should not be ignored.

Green rust compounds are Fe(II)Fe(III) hydroxy-compounds containing a certain amount of a non-hydroxyl anion which is reported to be sulphate, bromide, fluoride (Bernal et al., 1959), chloride (Feitknecht & Keller, 1950), iodide (Vins et al., 1987). nitrate (Gancedo et al., 1983) or carbonate (Stampfl, 1969). Certain organic anions probably also fit into the structure (Feitknecht, 1953). The compounds, isostructural with the mineral pyroaurite (Allmann, 1970), are layered structures consisting of a positively charged brucite-like sheet and negatively charged interlayers of the anion and a variable amount of water (Allmann, 1968). The Fe(II): Fe(III) ratio of the synthetic green rust seems to be variable (Feitknecht & Keller, 1950). For synthetic pyroaurites produced by co-precipitation, the Mg :Fe(III) ratio has been found to depend on the Mg:Fe(III) ratio in the synthesis mixture (Hashi et al., 1983). The green rusts can be synthesised by oxidation of an Fe(II) solution (Feitknecht & Keller, 1950), by oxidation/anodic electrolysis of Fe (Butler & Benyon, 1967; Dasgupta & Mackay, 1959), or by 'induced hydrolysis' (Taylor & McKenzie, 1980). Stampfl (1969) has detected a carbonate green rust in inner corrosion films of iron waterpipes. The compound probably also exists in other iron-containing hydromorphic environments, but the very high sensitivity towards oxidation makes isolation and detection difficult. Taylor (1982) used high pressures of CO2 for storage and physical measurements. In this paper the chemical composition of the Fe(II)Fe(III) hydroxy-carbonate synthesised by 'induced hydrolysis' has been studied and a simple technique for protecting the material against oxidation is described. The visible-near I R absorption spectrum of the stabilized green rust was obtained. 9 1989 The Mineralogical Society

664

H. C. B. Hansen

MATERIALS

AND METHODS

The Fe(II)Fe(IlI) hydroxy-carbonate was synthesised by a modification of the method described by Taylor (1985). Stock solutions and synthesis mixtures were prepared in 100 ml gas-proof cylindrical injection vials with rubber septums. A 0.2 u Fe(II) chloride stock solution was prepared by reacting a deoxygenated hydrochloric acid solution with an excess of Fe powder (Leussing & Kolthoff, 1953). Amorphous Fe(III) hydroxide was precipitated 1 h before by addition of 1 M sodium carbonate solution to a stirred 0.05 M Fe(III) nitrate solution with nitrogen bubbling through it. Weighed amounts of Fe(II) chloride solution and Fe(III) hydroxide suspension were injected in the synthesis vessel through 0.85 m m polythene tubing connected to hypodermic needles using nitrogen. Nitrogen had already been bubbled through a known quantity of water in the synthesis vessel for 1 h. Oxygen was removed from the nitrogen by bubbling through three gas washbottles in series containing Cr(II) solutions (Shriver, 1969), and from carbon dioxide by bubbling through 5 M sodium hydroxide. Finally the gas was forced through pure water and into the solutions/suspensions through 0.85 m m polythene tubing with a hypodermic needle. The flow rate was 30 ml/min. Radiometer Titralab equipment was run in the 'pH-stat' mode for the automatic addition of 1 M sodium carbonate into the reaction vessel. After the reaction had reached completion (no further base consumption), the suspension was stirred for 1 h with nitrogen bubbling and for a further I/2 h without nitrogen bubbling. A sample of the suspension was then withdrawn with a syringe through the septum for determination of the carbon dioxide content. Iron and carbon dioxide were also determined in the supernatant (after settling of the solids) from samples also withdrawn with a hypodermic syringe but through a 0-45/~m Millipore filter. Carbon dioxide was determined by absorption in base and back titration of the excess with hydrochloric acid. Iron was determined by the o-phenanthroline method using flow injection analysis (Mortatti et al., 1982). The uncertainty in the determination of iron is negligible compared to that of the carbon dioxide determination. The product was separated on an ice-cooled Bfichner fritted-glass filter (por. 4), washed with 25 ml of ice-cooled water and finally, while still on the filter, wetted with glycerol, glucose-saturated water or fatty acids. X-ray diffraction (XRD) was performed using a Philips diffractometer (Co-K~ radiation) with a smear of the product on a glass slide. A visible-near IR absorption spectrum was recorded on a Spectronic 1201 spectrophotometer using a transparent smear. RESULTS AND DISCUSSION Due to the transient nature of the green rust, only an indirect evaluation of its chemical composition was possible. It was assumed that all the added Fe(III) was contained in the green rust (the material dissolved readily in the 2 M hydrochloric acid forming a clear solution), that the compound contained carbonate only (no hydrogen carbonate), that no Fe(II) was oxidized to Fe(III), that the carbon dioxide-carbonate system was in equilibrium at the time of sampling and, finally, that the amounts of Fe(II) species and carbon dioxide adsorbed on to the hydroxy-carbonate were negligibly small. It was not possible to determine in this way whether the compound contained oxo groups as claimed by Misawa et al. (1974). No chloride or nitrate was found in the products. The contents of Fe(II), Fe(III), hydroxide and carbonate were determined in the following manner:

Composition, stabilization and light absorption of 'green rust"

665

Fe(III) = Fe(III) added as Fe(OH)3 Fe(II) = Fe(II) added - Fe(II) unreacted at termination of reaction CO32-

= Cr(suspension) - Cr(supernatant) where Cr = HCO~ + CO3~- + CO* and CO* = CO2 + H2CO3

OH-

= {Cr(total Na2CO3 addition) - Cr(suspension) + CO*(supernatant)} 2 + HCO~(supernatant)

Concerning the supernatant, the following relationships were used: CO~- = Cr([H+]2/K*K2 + [H+]/K2 + 1)-1; HCO~ = Cr([H+]/K * + K2/[H +1 + 1)- t ; CO* = Cr(K* K2/[H+]z + K*/[H +] + 1)-1 where K* = (HCO~-)(H+)/{(H2CO3) + (CO2)} and K2 = (CO2-)(H+)/(HCO3-) The results of 5 syntheses performed at pH 7.0 are given in Table 1. The charge of the hydroxy-carbonate is assumed to be negligible compared with the calculated charge imbalance. Because the uncertainty in the determination of the hydroxyl content is at least twice the uncertainty in the determination of the carbonate content, the charge imbalance is apportioned 2/3 to hydroxide and 1/3 to carbonate. The following compositions were calculated: ~ :~ Synthesis no. 1 : F_IHFerl ~2 4.04,OH I, )12-31,x-~3)0-9 9 n :OH )13.0\'~-r ~ ::~ Synthesis no. 2: F -r2m F e 4.27~ n H )11.8(CO3)o.8. Synthesis no. 3: Fe 2IIl F e 3.72(0 :OH ~ ,~,-, Synthesis no. 4: ~-nlFe l c 2 n3.8ot )~.Tkk-~3)t.o. Synthesis no. 5: Fe~tIFe~t.12(OH)12.0(CO3)i.I.

TABLE 1. Chemical composition of green rusts from different experiments. Synthesis number

Fe(II) :Fe(III) in synthesis mixture Content in mmol of: Fe(III) Fe(II) OHCO32charges

1

2

3

3-9

5.4

5.5

1.02 2-07 6.66 0.53 -0,52

1.02 2,19 6-72 0.43 -0,13

1.15 2.13 6.78 0-48 -0.05

4

5

5-9

5.0

1-15 2.18 6.61 0.54 0,11

1-15 2.36 6.67 0.61 0-27

666

H. C. B. Hansen

1.46

/

1.$5 1 , 5 8

1.64

1.74

1.118

1,97

2.09

2.34

2,47

2,~7 2 . 7 2

J

FIG. 1. XRD traces of Fe(II)Fe(III) hydroxy-carbonate treated in different ways. A,B: Glycerol treated, freshly prepared and after storage for 35 days at 4~ respectively. C: Glucose treated, freshly prepared. D,E: Untreated, freshly prepared and after storage for 24 h at 4~ respectively, d-spacings in/~. TABLE 2. Comparison of XRD data for Fe(II)Fe(III) hydroxy-carbonate collected in this investigation, with those of McGiU et al. (1976). Glycerol-treated hydroxy-carbonate

McGill et aL (1976) d(A)

I/Io

d(A)

I/Io

003 006 101 012 104 015 107 018

7.504 3.755 2.718 2-666 2"462 2.343 2.086 1"964

100 25 3 20 l0 15 2 15

7.53 3"759 2.72 2"668 2.47 2"344 2"09 1.967

100 32 1 15 3 12 1 9

10,10 01,11 110 113 116

1.740 1-641 1"584 1.549 1-462

5 3 5 5 5

hkl

1"88

1

1.740 1"644 1"583 1.550 1.462

2 1 2 2 2

Composition, stabilization and light absorption of 'green rust'

667

I.O

0.9

(],El

13.7

0.6

0.5

0.4

0.3

0.2

0.1

i

0.0 ~

i

i

51~1 550

i ~

i

i

i

6~0

71~

750

1

i

B00,8180

i 900

9~1

FIG. 2. Visible-near IR absorption spect~m of a transparent smear of freshly prepared Fe(II)Fe(III) hydroxy-carbonate.

The synthesised Fe(II)Fe(III) hydroxy-carbonate seems to be rather uniform in composition, with an Fe(II):Fe(III) ratio in the range 1.86-2-14. The stoichiometry is close to that of Fe[UFe~(OH)l 2CO3 determined by Stampfl (1969) on natural material, but diverging from that of pyroaurite, M(III)2M(II)6(OH)16CO3. On the basis of the XRD data of McGill et al. (1976), Brindley & Bish (1976) have predicted an Fe(II):Fe(III) ratio of 2:1 for the hydroxy-carbonate which has been confirmed by M6ssbauer measurements by Murad & Taylor (1984) on freshly prepared compounds. The same ratio has also been found by Tamaura et al. (1984) and Kanzaki & Katsura (1986) for the green rust II. Treatment of the green rust crystals with glycerol inhibits the oxidative breakdown of the pyroaurite structure since after storage at 4~ for I month, only a thin yellow brown skin had developed on the blue-green material. XRD data confirmed this structure-stabilizing effect of glycerol. No changes in d-spacings or relative intensities were observed after exposing the hydroxy-carbonate to the atmosphere for 35 days at 4~ (Fig. 1A and B). On the contrary, Fig. 1D and 1E show that the untreated material rapidly oxidizes. Glucose-saturated water has an effect similar to glycerol (Fig. 1C), whereas fatty acids such as 9-octadecenoic acid(c/s) or 9,12-octadecadienoic acid(cis,cis) have no significant effect. In Table 2, the d-spacings and relative intensities of the glycerol-treated hydroxycarbonate are compared with the XRD data reported by McGiU et aL (1976) indexed in the hexagonal system. The relatively high 00! intensities of the glycerol-treated material are probably caused by a preferred orientation of the crystals produced during preparation of the smear. The corresponding d-spacings indicate that neither glycerol nor glucose penetrate the interlayer region of the pyroaurite structure. The oxidation-retarding effect of the compounds containing hydroxyl groups is not only

668

H. C. B. Hansen

caused by a slow velocity of diffusion of oxygen towards, or carbon dioxide out of, the hydroxy-carbonate in the viscous liquid films surrounding the crystals. The glycerol/glucose may chemisorb on to the hydroxy-carbonate, their hydroxy groups having a high affinity for the hydroxy-carbonate surface. Some of these hydroxy groups are easily oxidized to aldehyde or carboxylic acid groups, further lowering the probability of oxygen reaching and oxidizing the structural Fe(II). Fig. 2 shows the visible-near I R absorption spectrum of the hydroxy-carbonate (unpolarized light). A broad peak at 650 nm was ascribed to intervalence {Fe(II)-Fe(III)} charge transfer (Bums, 1980). It has also been reported by Misawa et al. (1973) for green rust II. With increasing oxidation the peak gradually disappears, since after oxidation for 1 day the peak height above background had dropped to 45% of the original value, and after 5 days the peak had vanished. CONCLUSIONS Fe(II)Fe(III) hydroxy-carbonate prepared by induced hydrolysis at pH 7.0 seems to be a stoichiometric compound with the approximate composition FeI2"Fen(OH)12CO3 in which the Fe(II):Fe(III) ratio is not dependent on the overall Fe(II):Fe(III) ratio in the synthesis mixture. The hydroxy-carbonate has a broad adsorption peak at 650 nm, the absorbance of which decreases upon oxidation, and therefore can be used to follow the oxidation kinetics. Wetting the blue-green material with compounds containing hydroxyl groups (such as glycerol or glucose) protects it against rapid oxidation and this effect should be useful during sampling in the field and also during characterization. The interaction between the green rust and the glycerol/glucose may be worth studying, because a similar effect of organic compounds on green rust materials in soil seems possible. The implication of this stabilization effect may be crucial to understanding the genesis of oxides and oxyhydroxides of iron in hydromorphic soils, because green rust phases are known from laboratory experiments to be important intermediates in the formation and transformation reactions of iron oxides. REFERENCES ALLI~L~NNR. (1968) The crystal structure of pyroaurite. Acta Cryst. B24, 972-977. ALl,MANNR. (1970) Doppelschichtstrukturen mit brucitahnlichen Schichtionen [Me(II)l_xMe(III),(OH)2]x+. Chbnia 24, 99-108. BEltNALJ.D., Dg~3ulrrAD.R. & MACKAYA.L. (1959) The oxides and hydroxidesof iron and their structural inter-relationships. Clay Miner. Bull. 4, 15-30. BItlNDLEYG.W. & BIstl D.L. (1976) Green rust: a pyroaurite type structure. Nature 263, 353. BUXNSR.G. (1980) Mixed-valenceminerals of iron and titanium: Correlations of structural, M6ssbauer and electronic spectral data. Pp. 295-336, in Mixed-Valence Compounds (D. B. Brown, editor). Proc. NATO Adv. Study Inst. Oxford.

BtrrL~l G. & BENYONJ.G. (1967) The corrosion of mild steel in boiling salt solutions. Corros. Sci. 7, 385-404. DASOUFrAD.R. & MACKAYA.L. (1959)fl-ferricoxyhydroxideand green rust. J. Phy. Soc. Japan 14, 932-935. FEFrKNEOrr W. (1953) Die festen Hydroxysalzezweiwertiger Metalle. Fortschr. Chem. Forsch. 2, 670-757. FErrKNEOtr W. & KBLLERG. (1950) Uber die Dunkelgriinen Hydroxyverbindungendes Eisens. Z. Anorg. Chem. 262, 61-68. GANCEDOJ.R., MARTINEZM.L. & OTONJ.M. (1983) Formation of green rust in NH4NO3 solutions. Anal. Quim. 79, 470-472.

Composition, stabilization and light absorption o f 'green rust"

669

HASHIK., KIKKAWAS. & KOtZUMIM. (1983) Preparation and properties of pyroaurite-like hydroxy minerals. Clay Clay Miner. 31, 152-154. KANZAKIT. & KATSt~O~T. (1986) Mrssbauer Spectra at 77 K of products formed during transformation of Fe(OH)2 to Fe304 in aqueous oxidation. J. Chem. Soc. Dalton Trans. 1243-1246. LEUSSINOD.L. & KOLTHOFFI.M. (1953) The solubility product of ferrous hydroxide and the ionization of the aquo-ferrous ion. J. Am. Chem. Soc. 75, 2476-2479. MCGXLLJ.R., MCENANEYB. & S~rrH D.C. (1976) Crystal structure of green rust formed by corrosion of cast iron. Nature 259, 200-201. MISAWA T., HgSH1MOrO K. & SI-nMODAII~S. (1973) Formation of Fe(II)~ - Fe(III)l intermediate green complex on oxidation of ferrous ion in neutral and slightly alkaline sulphate solutions. J. Inorg. Nucl. Chem. 35, 4167-4174. MISAWA T., H.~SHIMOTOK. & SH1MODAIIL~S. (1974) The mechanism of formation of iron oxide and oxyhydroxides in aqueous solutions at room temperature. Cortes. Sci. 14, 131-149. MORTATTI J., KRUO F.J., PESSENDA L.C.R., ZAOOTTO E.A.G. & STOROAMtD JOItOENSEN S. (1982) Determination of iron in natural waters and plant material with 1,10-phenanthroline by flow injection analysis. Analyst 107, 659-663. MURADE. & TAYLOgR.M. (1984) The Mrssbauer spectra of hydroxycarbonate green rusts. Clay Miner. 19, 77-83. SHRIVERD.F. (1969) The Manipulation of Air-Sensitive Compounds, p. 196. McGraw-Hill, New York. STAMPFL P.P. (1969) Ein basisches eisen-II-III-Karbonat in Rest. Corros. Sci. 9, 185-187. T ~ U r , A Y., Yosmv^ T. & KATSV-SAT. (1984) The synthesis of green rust II (Fe~a-Fe~) and its spontaneous transformation into Fe304. Bull. Chem. Soc. Japan 57, 2411-2416. TAVLOg R.M. (1982) Stabilization of colour and structure in the pyroaurite compounds Fe(II)Fe(III)AI(III) hydroxycarbonates. Clay Miner. 17, 369-372. TAVLOg R.M. (1985) A rapid method for the formation of Fe(II)Fe(III) hydroxycarbonate. Clay Miner. 20, 147-151. TAYLORR.M. & McKENZlE R.M. (1980) The influence of aluminum on iron oxides. VI. The formation of Fe(II)-AI(III) hydroxy.chlorides, -sulfates, and -carbonates as new members of the pyroaurite group and their significance in soils. Clay Clay Miner. 20, 179-187. VINS L, SUngT J., ZAPLETALV. & HANOt~SEKF. (1987) Preparation and properties of green rust type substances. Collect. Czech. Chem. Comm. 52, 93-102.

Suggest Documents