- Email: [email protected]
Kinetics of the Formation of Goethite in the Presence of Sulfates and Chlorides of Monovalent Cations
JOURNAL OF COLLOID AND INTERFACE SCIENCE ARTICLE NO.
190, 269–277 (1997)
CS974845
Kinetics of the Formation of Goethite in the Presence of Sulfates and Chlorides of Monovalent Cations Aida Frini and Mohamed El Maaoui 1 Laboratory of Industrial Inorganic Chemistry, Sciences University of Tunis, 1060 Tunis, Tunisia Received February 27, 1996; accepted February 25, 1997
The synthesis of goethite by oxidation of Fe2/ in presence of metallic iron was undertaken in an aqueous medium containing indifferent salts such as Na2SO4 , (NH4 )2SO4 , NaCl, and NH4Cl. Temperature and bubbling air rate were maintained, respectively, at 707C and 1 L/min. The influence of anions and cations on the kinetics of each step of the process has been followed distinctly, the iron dissolution rate has been determined by the variation of the medium acidity, and the precipitation of goethite has been determined by gravimetric measurements. With respect to Cl 0 , / the SO 20 4 anion decreases the rate of the two reactions. NH 4 acts as an inhibitor when it is present at low concentrations and as an accelerator for higher concentrations; the limit corresponding to the change of NH / 4 behavior depends on the nature of the counter ion. The reaction product is composed of pure goethite in the presence of sulfate salts, whereas a mixture of goethite and lepidocrocite, respectively, 60–70 and 40–30%, was observed in the presence of chloride salts. q 1997 Academic Press Key Words: goethite; iron dissolution; influence of anions; influence of cations; iron corrosion.
I. INTRODUCTION
Slow oxidation of ferrous solutions by air bubbling provides one or several of the following products: goethite ( a-FeOOH), lepidocrocite ( g-FeOOH), magnetite (Fe3O4 ), and hematite ( a-Fe2O3 ) (1–13). Ferrihydrite (Fe5HO8r4H2O) is a natural product (1). Rapid oxidation, using H2O2 , leads to the precipitation of feroxyhyte ( d-FeOOH) (14). Ferrous iron oxidation has been the main subject of many investigations. An extensive study which described the influence of temperature, pH, and Fe2/ concentration is due to Olowe et al. (14–17). No kinetic considerations were presented; however, the authors indicated the conditions which may yield goethite, lepidocrocite, or magnetite. They use the ratio R Å Fe2/ /OH 0 to define the experimental conditions; for example, at 607C, goethite is obtained when R Å 3 (16). At room temperature, Schwertmann and Cornell (1) indicated that goethite and lepidocrocite precipitate at pH 6–8, whereas magnetite is formed at pH 8–14. 1
To whom correspondence should be addressed.
A second method is widely used to synthesize iron oxyhydroxides; it consists of the hydrolysis of Fe3/ cations. This is achieved by pouring freshly prepared ferric solutions into concentrated NaOH or KOH solutions; the aging of the sols takes place at 60–807C for a period of time ranging from 1 to several days. Goethite precipitates if the pH of the final solution is above 12.8; for less basic media a mixture of goethite and hematite is observed (18–20). Matijevic et al. (21–24) have thoroughly studied the hydrolysis of ferric solutions in acidic medium; b-FeOOH and a-Fe2O3 are the end products. Livage et al. (25) gave theoretical considerations on the nucleation and germination processes of various species resulting from the hydrolysis of a variety of metallic cations. In this paper we have chosen a third synthetic method, controlled air oxidation of an iron powder suspension in aqueous medium; the presence of Fe2/ cations, even in low content, makes this method very similar to the oxidative method cited earlier. However, this procedure is interesting because of the inconveniences presented by the previous procedures. Thus, the hydrolysis of ferric solutions has limited applications because of problems with chemical products consumption and the maintenance of boiling solutions for a relatively long time. The oxidative method presents many problems concerning the constancy of the product characteristics, especially the yellow color, which is very sensitive to experimental conditions; this factor depends on the presence of small proportions of the other oxyhydroxides, on the crystal morphology, and on substitution of other cations such as Cr(III) for Fe(III) (1). One inconvenience of the oxidative method is the regulation of pH by continuous addition of NaOH or NH4OH solutions. The introduction of these alkalis produces unavoidably high local concentrations of alkalis where black Fe3O4 precipitates; in addition, solutions are more and more charged with Na2SO4 or (NH4 )2SO4 which have, as we will see later, an effect on the product characteristics. The presence of iron powder in the aqueous system keeps constant the experimental conditions such as pH, ferrous iron concentration, and salt content of the solution. For example, pH is stabilized by iron dissolution without introduction of alkalis. However, this method has the
269
AID
JCIS 4845
/
6g25$$$$61
06-03-97 12:18:06
0021-9797/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
coida
270
FRINI AND EL MAAOUI
major inconvenience of requiring the use of very pure iron powder; impurities such as carbon, which are always present, cause a change in the color of goethite and they must be kept as low as possible. This paper deals with the effect of NaCl, NH4Cl, Na2SO4 , and (NH4 )2SO4 on both the reaction kinetics and the nature of the final products. These salts are expected to be ‘‘indifferent salts’’ since they are absent from the reaction equations; in fact, many observations report that they really affect the system. For example, it is well known that the rust deposit changes chemically and morphologically when iron is submitted to different corrosive media (2–4, 6, 15, 26–28). To our knowledge, no references are cited in the literature regarding the study of the effect of indifferent salts where the process is considered as a whole. In contrast, data are available when we consider each of the successive following stages separately: iron dissolution, oxidation of the ferrous cations to give ferric species, formation of (Fe III-OH0) complexes, hydrolysis of the ferric cations, and, last, condensation of the monomeric species to form polymeric and crystalline products. For the first stage, we know that the dissolution of metallic iron in aqueous media is well explained by the oxidoreduction properties of the solution compared to the iron potential. Temperature, pH, and other oxidoreduction couples have well-defined effects. Among the chosen anions and cations, Cl 0 has been the most widely studied because of its negative effect on iron corrosion. Pourbaix (29) presents a comprehensive review where it appears that, unlike other anions which form a protective layer on iron, Cl 0 produces a permeable oxide layer which allows oxidation to continue; this anion produces generalized corrosion or pitting corrosion according to the pH, the iron potential, and the Cl 0 content. So chloride anion acts in a way dependent on the nature and the morphology of the reaction products. Studies dealing with the action of other anions on iron corrosion are numerous, especially for high complexing anions such as phosphate, chromate, and carbonate, which are often used as corrosion inhibitors (29). Similarly we note that ferrous chloride, used as raw material, seems to favor the precipitation of lepidocrocite; thus aqueous 0.05 M FeCl2 containing NaCl readily gives a brown precipitate of g-FeOOH at pH 6–7 (30). A well-known synthesis of g-FeOOH consists of adding pyridine to a FeCl2 solution and then oxidizing the yellow precipitate by air until its complete transformation into lepidocrocite (8). Schwertman and Cornell (1) use FeCl2 for goethite preparation at pH 7, but they add sodium bicarbonate to the solution and operate with a low rate of air bubbling (30–40 ml/min); the lepidocrocite is obtained using the same ferrous salt at pH 6.7–6.9, without bicarbonate and with a higher air flow (100–300 ml/min). Goethite was also obtained from FeCl2 by lowering the temperature to 77C and allowing oxidation by air contact without bubbling (31). FeSO4r7H2O is widely
AID
JCIS 4845
/
6g25$$$$62
06-03-97 12:18:06
TABLE 1 Granulometric Distribution of Iron Powder Fraction size (mm)
Weight %
õ63 63–90 90–125 125–180 ú180
1.3 16.5 32.4 17.2 32.6
used in industry as a raw material; it is an abundant byproduct of the titanium dioxide manufacture. Olowe et al. have used this salt (15); they obtained a mixture of gFeOOH and a-FeOOH at 257C in an acidic medium corresponding to R ú 0.5, where R is equal to Fe2/ /OH 0 . On the other hand, the influence of indifferent salts on the hydrolysis of ferric cations has been studied principally in alkaline media; Weiser and Milligan (32) have shown that the hydrolysis of sulfate, nitrate, bromide, oxalate, and acetate iron III leads to the formation of a-FeOOH, while the use of perchlorate gives either a- or g-FeOOH depending on the quantity of alkali added (33). It seems that there is no reference concerning the influence of the cation on the nature of the oxidation products, except one indicating that the addition of Cu 2/ in the aqueous medium favors hematite precipitation (27). All these considerations show that the influence of indifferent salts on goethite synthesis is a very interesting topic and hitherto not well studied. This paper constitutes a preliminary work on the subject. II. EXPERIMENTAL
Laboratory iron powder was provided by PROLABO; it had a minimum purity of 98% and a carbon content below 0.01%. The powder granulometric distribution is given in Table 1. A 250-ml round-bottomed flask is used; it is equipped with a condenser, a pH combined electrode, and a glass tube allowing the inlet of the compressed air. Air is preheated to avoid local cooling of the solution. The bubbling rate is maintained as constant as possible at about 0.9 to 1 L/min. Magnetic stirring is used. The use of glass vessels is usually not recommended for the precipitation of goethite by the hydrolysis method, where a basic medium and a long boiling period are needed; however, in our case the solutions are weakly acidic and so there is no possibility of contaminating them with silicate anions. One hundred milliliters of a given salt solution, such as Na2SO4 , is brought to 707C and then adjusted to pH 2.5 by means of the corresponding acid, such as H2SO4 ; air bubbling is started and the temperature is allowed to stabilize
coida
271
KINETICS OF GOETHITE FORMATION
again. One gram of iron (17.8 1 10 03 mol) and 0.05 g of a-FeOOH, used as seed crystals, are added at the same time. When FeSO4 is needed, solid FeSO4r7H2O is introduced at the same time as iron powder. After a given reaction time, the water suspension is filtered out and the remaining iron and a-FeOOH are separated magnetically. The precipitate is washed several times with distilled water and dried at 607C until constant weight is attained. The weight of unreacted iron and that of the aFeOOH formed may be used to calculate the soluble iron fraction, which is obtained by recording the balance between the initial mole number of iron and the sum of the mole numbers of the final solids. Precipitates are identified by X-ray diffraction using a RIGAKU apparatus; CuKa rays were used. The infrared absorption spectra are recorded using a Perkin–Elmer 783 spectrometer; they serve both to confirm the X-ray results and to evaluate the proportions of a-FeOOH and g-FeOOH. The standardization was accomplished using the spectra of mixtures containing known proportions of pure products. For a given spectrum, let S1 and S2 be the areas of the peaks at 888 and 1018 cm01 , which correspond, respectively, to the characteristic peaks of goethite and lepidocrocite; S1 and S2 are proportional to the quantities of these products. The coefficients of proportionality are the extinction coefficients e1 and e2 . If we consider m as the weight of the mixture introduced into the KBr pastille and C1 and C2 as the mass percentages of a-FeOOH and g-FeOOH, we have C1 / C2 Å 100 S1 Å e1mC1 S2 Å e2mC2 S1 /S2 Å ( e1 / e2 )C1 /C2 . So the standardization graph representing S1 /S2 versus C1 / C2 is a line whose slope is e1 / e2 . Experiments give e1 / e2 Å 0.6. C1 and C2 are known by means of their ratio and their sum. They can be calculated easily. III. THEORETICAL CONSIDERATIONS
In aerated weak acidic media, an aqueous system containing an electrically insulated iron may induce possible electrochemical equilibra: Water reduction, 2H / / 2e 0 ` H2 ,
[1]
whose electric potential depends on the pH according to the relation
AID
JCIS 4845
/
6g25$$$$62
06-03-97 12:18:06
E1 Å 0.000 0 0.059 pH 0 0.0295 log P(H2 ). Water oxidation, 2H2O ` O2 / 4H / 0 4e 0 , E2 Å /1.228 0 0.059 pH / 0.0147 log P(O2 ). [2] P(H2 ) and P(O2 ) are the partial pressures of H2 and O2 upon the aqueous solution. For neutral pure water and in the absence of air, log P(H2 ) and log P(O2 ) are evaluated, respectively, at 027.56 and 027.86 (29); under these conditions E1 Å E2 Å 0.40 V. Iron dissolution is due to the equilibrium Fe ` Fe2/ / 2e 0 E3 Å 00.440 / 0.0295 log(Fe2/ ).
[3]
Oxyhydroxide precipitation is usually written as (29) Fe2/ / 3H2O ` Fe(OH)3 / 3H / / e 0 . Free Fe3/ is not considered here because of the simultaneous presence of Fe and Fe2/ ; Fe(OH)3 can be considered as FeOOH / H2O, hence this reaction will be written as Fe2/ / 2H2O ` FeOOH / 3H / / e 0 E4 Å /1.507 0 0.1773 pH / 0.059 log(Fe2/ ). [4] Reaction [1] takes place only in reducing media when oxygen is absent; it becomes more important when the pH decreases. For aerated medium, equilibria [2, 3, 4] are involved; they take place in different points of the system. This is the well-known phenomenon of the Eavans cells where aerated zones and nonaerated zones adopt two distinct potentials. On the iron particle surface, situated in the aerated zones, we assist to the equilibrium [2]. Iron adopts a high value of potential, equal to E2 ; it is not corroded. The expression for E2 , given above, depends on the pH as well as on the oxygen pressure (the rate of air bubbling is involved here). In the nonaerated zones, iron particles are dissolved owing to reaction [3]. The iron potential is E3 ; it depends on the ferrous iron content of the solution. We see that the existence of these cells assumes the establishment of electric links between anodes and cathodes. For iron powder, this is possible when the electrochemical cell takes place on one grain or when the oxide layer is absent, porous, nonadherent, or continuously removed by agitation. For the next part of this work it is useful to write the reaction corresponding to the iron dissolution as a combination of [2] with [3] and the one which corresponds to the goethite precipitation as a combination of [2] with [4]:
coida
272
FRINI AND EL MAAOUI
2Fe / O2 / 4H / ` 2Fe2/ / 2H2O
[5]
4Fe2/ / O2 / 6H2O ` 4FeOOH / 8H / .
[6]
We note that iron dissolution is accompanied by an increase of pH; conversely, precipitation of goethite corresponds to a decrease of pH. The overall reaction allowing the transformation of metallic iron into goethite is 4Fe / 3O2 / 2H2O ` 4a-FeOOH.
[7]
which is the sum of reactions [5] and [6]; reaction [7] is independent of pH. If reactions [5] and [6] have the same rate, pH will remain constant; otherwise an increase of pH appears when the iron dissolution is more rapid than the goethite precipitation and vice versa. Consequently, the knowledge of pH Å f (t) curves will provide important information about the relative kinetics of the two reactions.
FIG. 2. IR spectra of different products: (a) 1 M NH4Cl; (b) 1 M NaCl; (c) 0.1 M NH4Cl; and (d) 0.1 M NaCl.
IV. RESULTS AND DISCUSSION
IV.1. Nature of the Reaction Products The nature and proportions of the formed products are known by IR absorption spectra and the standardization curve described under Experimental. Some examples of the spectra are given in Figs. 1 and 2. Data summarized in Table 2 indicate that anions have a great influence on the kind of the allotropic form which precipitates. In the presence of SO 20 4 , only goethite is observed. However, in Cl 0 medium the reaction product contains about 60 to 70% goethite and 40 to 30% lepidocrocite. These results are to be compared with those in the References. The introduction of Cl 0 shifts the oxidation of Fe2/ toward the formation of g-FeOOH (9–12); it shifts the
hydrolysis of Fe3/ toward the formation of b-FeOOH (8, 34, 35). Table 2 shows that cations have no influence on the kind of precipitate. The mechanism by which the anions move precipitation toward one or another product seems not to be elucidated yet. Different hypotheses can be proposed. Cl 0 , for example, favors g-FeOOH because of the existence of a similar compound containing Cl 0 ; Thus FeOCl has the same structure as g-FeOOH (36–38). Fe(III) oxyhydroxysulfate Fe8O8 (OH)6SO4 has a crystal structure similar to that of akaganeite (1). On an other hand, we know that goethite works as an anion exchanger in acidic media and a cation exchanger in basic solutions (20). The exchange capacity of goethite for the Cl 0 anion was studied by Paterson and Rahman (22). It is possible that anion or cation adsorption on the final products is responsible for the observed results, and that adsorption changes the kinetics of the different reacTABLE 2 Nature of the End Products of the Iron Oxidation as a Function of the Nature and the Concentration of the Indifferent Salt Anion
Salt
Concn mol/L
% a-FeOOH
% g-FeOOH
SO20 4
H2SO4 Na2SO4 Na2SO4 (NH4)2SO4 (NH4)2SO4 HCl NaCl NaCl NH4Cl NH4Cl
0.1 0.1 1.0 0.1 1.0 0.1 0.1 1.0 0.1 1.0
100.0 100.0 100.0 100.0 100.0 69.5 69.0 64.0 60.6 66.0
0.0 0.0 0.0 0.0 0.0 30.5 44.0 36.0 40.0 34.0
Cl0
FIG. 1. IR spectra of different products: (a) 1 M (NH4 )2SO4 ; (b) 1 M Na2SO4 ; (c) 0.1 M (NH4 )2SO4 ; and (d) 0.1 M Na2SO4 .
AID
JCIS 4845
/
6g25$$$$63
06-03-97 12:18:06
coida
273
KINETICS OF GOETHITE FORMATION
FIG. 4. Effect of initial FeII concentration on the curves pH Å f (t).
FIG. 3. Effect of initial pH on goethite reaction synthesis.
tions involved in the product formation; these changes certainly have an effect. In the Introduction, we have seen that Schwertmann and Cornell (1) have proposed two similar operating methods to obtain a-FeOOH or g-FeOOH. The two methods seem to differ only in the rate of air bubbling. For the synthesis of goethite the rate of bubbling is 30–40 mL/min and the reaction is carried for 48 h; for lepidocrocite the rate is 100–300 mL/min and the oxidation is complete after 1 to 3 h. We will see in Section IV.3 that the kinetic of formation of goethite is faster in the presence of Cl 0 than in the presence of SO 20 4 ; this probably explains the presence of lepidocrocite in the former case.
thus, for pH0 § 4, we have observed that the reaction product is black, it remains stuck to the magnet, and the X-ray diffraction spectrum presents the peaks of magnetite, Fe3O4 . We note that the literature situates magnetite precipitation at higher values of pH (pH ú 8) (see Ref. 15); in our case, precipitation of Fe3O4 at pH 4 is probably due to the high temperature used. IV.2.2. Influence of initial Fe2/ concentration on pH Å f(t) curves. Figure 4 contains different graphs, pH Å f (t), corresponding to various initial Fe2/ concentrations. We note than the iron dissolution rate is lowered when Fe2/ content increases; this is due to the iron potential which becomes higher, Fe ` Fe2/ / 2e 0 E Å E0 / 0.03 log[Fe2/ ].
IV.2. Kinetics of the Iron Dissolution IV.2.1. Influence of the initial pH. Figure 3 shows the influence of the initial pH on the curves pH Å f (t). At the beginning of the experiment, where only iron dissolution is present, pH increases rapidly owing to reaction [5]; when the pH becomes sufficiently high, goethite precipitation takes place suddenly, the pH remains constant, and equilibrium is reached between reactions [5] and [6]. Sometimes we notice a decrease of the pH immediately after the sudden precipitation. This is characteristic of a supersaturation phenomenon, which is not reliable; two experiments, carried out under the same conditions, may or not present a maximum pH. Thus we have chosen to present in Fig. 3 only the curves having no maximum. We observe that when the pH of the initial solutions decreases, the time to reach precipitation increases and the equilibrium pH decreases. The increased time is evident since a greater quantity of protons must be consumed to reach the pH precipitation. The decrease of the equilibrium pH can be explained by the greater quantity of Fe2/ liberated by reaction [5] before precipitation. This will be confirmed in the following section in which different initial values of Fe2/ are used. For the next part of this work, kinetic curves are drawn with pH0 2.5. Higher initial pH risks giving impure products;
AID
JCIS 4845
/
6g25$$$$64
06-03-97 12:18:06
The asymptotic pH continuously decreases while Fe2/ increases; this can be explained by equilibrium [6]. As a matter of fact, an increase of Fe2/ concentration shifts the system to the right, releasing more H / ions. For the next part of this work, we have chosen to use an initial Fe2/ corresponding to the addition of 5 g/L of FeSO4r7H2O. This permits one to control the pH of the asymptotic branch, which must remain below pH 4 in order to yield pure goethite (see Section IV.1.1). IV.2.3. Influence of sulfate and chloride anions on the kinetics of iron dissolution. Figures 5 and 6 show the varia-
FIG. 5. Effect of NaCl content on the curves pH Å f (t).
coida
274
FRINI AND EL MAAOUI
FIG. 6. Effect of Na2SO4 content on the curves pH Å f (t).
tion of pH versus time in the presence of increasing NaCl and Na2SO4 content. Very little effect is observed in the presence of Cl 0 . Conversely, SO 20 4 has a marked effect; Fig. 6 shows that the higher the sulfate concentration, the greater the time to reach precipitation. The anion effect, observed here, concerns reaction [5], which is responsible for the pH increase. As reaction [5] is the sum of [2] and [3] we can assume that only reaction [3] is affected by the presence of anions. Anions adsorb on the metallic iron surface, which is positively charged by the ferrous ions being liberated. SO 20 forms a compact layer 4 limiting the diffusion of O2 to the iron, whereas Cl 0 , which has a lower charge, gives a much more permeable layer. IV.2.4. Influence of the nature of cations. The effect of the NH 4/ cation is studied by progressive replacement of Na / by NH 4/ keeping the anion content constant. The chosen mixtures of NaCl and NH4Cl, for example, go from pure NaCl to pure NH4Cl solutions. Figure 7 to 10 show the experimental results corresponding to the conditions mentioned in Table 3. Obviously, the effect of Na / may be neglected; only NH 4/ is responsible for the observed changes. We notice that the ammonium cation has a double effect: / 4
—Low NH content causes an inhibition effect: in the presence of 0.1 M Cl 0 (see Fig. 7) the kinetics of iron
FIG. 7. Effect of 0.1 M (Na, NH4 )Cl on the curves pH Å f (t).
AID
JCIS 4845
/
6g25$$$$64
06-03-97 12:18:06
FIG. 8. Effect of cations 1 M (Na, NH4 )Cl on the curves pH Å f (t).
dissolution decreases until NH 4/ concentration equals 0.05 M; in presence of 0.1 M SO 20 (Fig. 9) the decrease is 4 observed up to 0.09 M. The curve corresponding to 0.1 M Na2SO4 , which seems to be abnormal, has been repeated about 10 times; the same results have been obtained. —High NH 4/ content produces an accelerating effect (See Figs. 8 and 10). Different hypotheses can be proposed to explain these observations. NH 4/ probably acts in the two following ways: (a) Adsorption of NH3 on the slightly charged iron surface: Fe / xNH 4/ ` Fe(NH3 )x / xH / .
[8]
(b) Formation of ferrous iron complexes: Fe2/ / nNH 4/ ` [Fe(NH3 )n ] 2/ / nH / .
[9]
Reaction [8] can be achieved even when small NH 4/ quantities are used, just to form an NH3 layer on the iron surface. Reaction [9] necessitates more ammonium salt, especially if n is high. We note that these reactions contribute to the proton liberation, so when the pH Å f (t) curves rise slowly,
FIG. 9. Effect of cations 0.1 M (Na, NH4 )2SO4 on the curves pH Å f (t).
coida
275
KINETICS OF GOETHITE FORMATION
FIG. 11. Variation of the iron oxidation kinetic as function of the concentration and the cation nature in the presence of SO 20 4 . FIG. 10. Effect of cations 1 M (Na, NH4 )2SO4 on the curves pH Å f (t).
this could be attributed either to a lower iron dissolution rate or to H / release. When the ammonium salt is present at a low concentration level, we can assume that only the adsorption effect takes place, so the increase of pH Å f (t) curves appears to be slow, due either to lower iron dissolution rate or to release of H / . However, for high ammonium content, the Fe2/ complexes are predominant, and curves appear to increase rapidly although equilibrium [9] is accompanied by H / liberation. The rate of iron dissolution is enhanced because of the formation of ferrous complexes which contributes to shift equilibrium [3] to the right. IV.3. Kinetics of the Goethite Formation The rate of goethite formation is determined by direct measurement of the weight of the dry precipitate versus time (see Figs. 11 and 12). Variations of the weight of unreacted iron are given in Figs. 13–15; those relative to the variations of the dissolved iron are shown in Figs. 16 and 17. We see that the concentration of dissolved iron, which is Fe2/ in its free or complexed form, remains constant; this means that there is no Fe2/ accumulation in solution, the rate of iron dissolution being equal to the rate of goethite precipitation. This result was confirmed earlier by the asymptotic pH, which becomes constant after goethite precipitation.
We note that some differences may be observed between the whole kinetics, represented by the variations of the weight of goethite versus time, and the kinetics given by the graphs pH Å f (t). The curves pH Å f (t), especially in the period corresponding to the rapid growth of pH, were established for naked iron; they give only the influence of the salts on reaction [5]. Changes occur at the interface between metallic iron and the aqueous solution during the reaction. After a certain time the goethite forms an adherent layer on iron surface which causes a change in the kinetics of reaction [5] and consequently affects the rate of the whole process. The effect of an indifferent salt on the whole goethite formation process is actually its effect on the slowest stage between the two following reactions: 2Fe / O2 / 4H / ` 2Fe2/ / 2H2O 2/
4Fe
[5] /
/ O2 / 6H2O ` 4FeOOH / 8H .
[6]
The rate of reaction [5] can be known from the rise of pH at the start of the reaction (before the goethite precipitation); since the pH remains constant after precipitation, it is difficult to say which of [5] or [6] is the limiting step. It is expected that the horizontal curve will not remain so forever;
TABLE 3 Experimental Conditions for Figs. 7 to 10 Figure
AID
System
7 8 9 10
(Na, (Na, (Na, (Na,
NH4)Cl NH4)Cl NH4)2SO4 NH4)2SO4
JCIS 4845
/
6g25$$$$64
Concentration (mol/L) 0.1 1 0.1 1
06-03-97 12:18:06
FIG. 12. Variation of the iron oxidation kinetic as function of the concentration and the cation nature in the presence of Cl 0 .
coida
276
FRINI AND EL MAAOUI
FIG. 13. Variation of the iron dissolution kinetic as function of the concentration and the cation nature in the presence of SO 20 4 .
for long reaction times, metallic iron will diminish and its surface will be progressively coated by the oxide layer, so the dissolution kinetic will decrease. Conversely, the precipitation process, which is enhanced by the presence of goethite crystals, will become rapid, so a slow decrease of the asymptotic pH will be observed. Many experiments confirm this, which is why pH Å f (t) curves were presented here up to 60 min and not further. Figures 11 and 12 show clearly that for a sulfate medium, the kinetics of formation of goethite decreases when the salt concentration increases. Conversely, chloride salt concentration seems to have no effect. The chloride behavior can be explained by the high permeability of the rust deposit obtained in the presence of this anion; this fact was already signaled by Pourbaix (29). The results presented in Section IV.1 explain this permeability; thus, the reaction product in the presence of Cl 0 is a mixture of 60% goethite, having the shape of needles, and 40% lepidocrocite in hexagonal plates. The protective layer is then formed with heterogeneous crystals; it is a porous barrier easily crossed by the corrosive reagents. The kinetic seems to be independent of Cl 0 since it is the nature of the products which is important. For sulfate salts, the decrease in the kinetic rate when the sulfate anion content increases is identical to the decrease in the rate of reaction [5] (see pH Å f (t) curves in Fig. 6); reaction [5] is probably the limiting step. We have assigned
FIG. 14. Variation of the iron dissolution kinetic as function of the anion nature at a constant salt concentration, 0.1 M.
AID
JCIS 4845
/
6g25$$$$64
06-03-97 12:18:06
FIG. 15. Variation of iron dissolution kinetic as function of the anion nature at a constant salt concentration, 1 M.
this decrease to the adsorption of sulfate anions on metallic iron. It is worthy of note that the increased sulfate content also involves changes in the protective layer morphology. Thus, Domingo et al. (36), who studied the crystal growth of various oxyhydroxides, explain that adsorption of cations, for example, when an excess of Fe2/ is present, enhances the particles surface charges and diminishes their aggregation ability; therefore, small crystals are formed. Anions, when they adsorb on the goethite, give lower surface charges and thus larger crystal sizes. We note that Cl 0 also adsorbs readily on goethite (36), but its effect is less important because of its lower charge. V. CONCLUSION
The present work has shown that the synthesis of goethite by oxidation of the system (Fe, Fe2/ ) in weak acidic medium is very sensitive to the presence of indifferent salts such as Na2SO4 , (NH4 )2SO4 , NaCl, and NH4Cl. The reaction products are composed of pure goethite in the presence of SO 20 4 and of a mixture of goethite and lepidocrocite, respectively, 60–70% and 40–30%, in the presence of Cl 0 . The ammonium cation has no influence on the nature of the reaction outcome but it influences the reaction kinetics. An inhibition effect is observed for low ammonium concentrations, whereas an acceleration effect occurs for high con-
FIG. 16. Variation of aqueous Fe2/ content as function of the concentration and the cation nature in the presence of SO 20 4 .
coida
KINETICS OF GOETHITE FORMATION
FIG. 17. Variation of aqueous Fe2/ content as function of the concentration and the cation nature in the presence of Cl 0 .
centration; this is due to adsorption phenomena in the first case and to formations of ferrous complexes in the later. The kinetics of the iron dissolution step determined by the pH Å f (t) curves, and that of the whole process, determined by weight measurements, are sometimes affected differently by the anions and the cations studied. This is explained by a change in the morphology of the goethite protective layer. REFERENCES 1. Schwertmann, U., and Cornell, R. M., ‘‘Iron Oxides in the Laboratory: Preparation and Characterization.’’ VCH, Weinheim/New York/Basel/ Cambridge, 1991. 2. Tuula, L., and Markku, L., J. Thermochim. Acta 77, 177 (1984). 3. Getskin, L. S., Matveev, A. F., Bubnov, A. A., and Piskunov, V. M., USSR Patent 779374, November 15, 1980. 4. Solcova, A., Subrt, J., Tlaskai, J., and Zapletal, V., Chem. Prun. 8, 34 (1984). 5. Fan, Q., Guan, Z., et al., CN 1,056,688 (Cl.C01G49/06), 04 Dec. (1991) Appl. 91,106,200, 06 Jul. (1991). 6. Rezaie-Serej, S., and Cook, D. C., J. Hyperfine Interact. 41(1–4), 701 (1988). 7. Okazaki, K., Japanese Patent 62 36, 026 [87 36, 026] (Cl. CO1 G49/ 02), 17 Feb. (1987), Appl. 85/174, 379, 4pp, 09 Aug. (1985). 8. Bouth, H. S., Audrieth, L. F., Bailar, J. C., Fernelius, W. C., Johnson, W. C., and Kirk, R. E., ‘‘Inorganic Syntheses,’’ Vol. 1, p. 184. McGraw–Hill, New York/London, 1939. 9. Kaji, N., Japanese Patent 02,293,329 [90,293,329], 4 Dec. (1990), App.89/112,43, 01 May (1989).
AID
JCIS 4845
/
6g25$$$$65
06-03-97 12:18:06
277
10. Matsui, Y., Kamisaka, T., Goto, T., and Okasaki, K., Japanese Patent 62 65,937 [87 65 937] 25 March (1987), Appl. 85/203,922, 14 Sep. (1985). 11. Koike, N., Matsui, Y., Tsugaru, T., and Horiguchi, K., Japanese Patent 62 65,934 [87,65,934], 25 March (1987), Appl. 85/202,267, 12 Sep. (1985). 12. Kamisaka, T., Kaji, N., and Okasaki, K., Japanese Patent 62,167,222 [87,167,222], 23 Jul. (1987), Appl.86/7,310, 17 Jan. (1986). 13. Horst, B., Herivert, B., and Franz, H., British Patent Appl. 2,023,559,03, Jan. (1980), Ger. Appl. 2,826,941, 20 Jun. (1978). 14. Olowe, A. A., Refait, PH., and Genin, J. M. R., J. Corrosion Sci. 32, 1003 (1991). 15. Olowe, A. A., and Genin, J. M. R., J. Corrosion Sci. 32, 965 (1991). 16. Olowe, A. A., Pauron, B., and Genin, J. M. R., J. Corrosion Sci. 32, 985 (1991). 17. Olowe, A. A., and Genin, J. M. R., J. Corrosion Sci. 32, 1021 (1991). 18. Atkinson, R. J., Posner, A. M., and Quirk, J. P., J. Inorg. Nucl. Chem. 30, 2371 (1968). 19. Atkinson, R. J., Posner, A. M., and Quirk, J. P., J. Phys. Chem. 71, 550 (1967). 20. Paterson, R., and Rahman, H., J. Colloid Interface Sci. 98, 494 (1984). 21. Matijevic, E., and Cimas, S., J. Colloid Polym. Sci. 265, 155 (1987). 22. Paterson, R., and Rahman, H., J. Colloid Interface Sci. 94, 60 (1983). 23. Matijevic, E., J. Pure Appl. Chem. 50, 1193 (1978). 24. Flynn, C. M., J. Chem. Rev. 84, 31 (1984). 25. Livage, J., Henry, M., and Sanchez, C., Progr. Solid. State Chem. 4, 18 (1988). 26. Kroeckert, B., and Buxbaum, G. (Bayer A-G), European Patent, Appl. EP 406,633, 09 Jan. (1991), DE Appl. 3,922,106,05, Jul. (1989). 27. Cornell, R. M., and Giovanoli, R., Polyhedron 7, 5, 385 (1988). 28. E. Matijevic, E., Annu. Rev. Mater. Sci. 15, 483 (1985). 29. Pourbaix, M., ‘‘Lec¸ons en corrosion e´le´ctrochimique.’’ Centre Belge d’Etude de la Corrosion Cebelcor, 1975. 30. Lewis, D. G., and Farmer, V. C., J. Clay Minerals 21, 93 (1986). 31. Sato, T., Okamoto, S., and Hashimoto, K., Yogyo Kyokaishi 94(12), 1201 (1986). 32. Weiser, H. B., and Milligan, W. O., J. Phys. Chem. 39, 25 (1935). 33. Knight, R. J., and Silva, R. N., J. Inorg. Nucl. Chem. 36, 591 (1974). 34. Tatsuo, I., Kazuhiko, K., Kyoji, O., Hiroshi, M., and Toshio, K., Japanese Patent, 03, 109, 216, [91, 109, 216], 09 May (1991); App. 89 244,985,22, Sep. (1989). 35. Matijevic, E., J. Colloid Interface Sci. 58, 374 (1977). 36. Domingo, C., Rodriguez-Clemente, R., and Blesa, M., J. Colloid Interface Sci. 165, 244 (1994). 37. Ralph, W. G., Wyckoff ‘‘Crystal Structures,’’ Vol. I. Wiley–Interscience, New York/London/Sydney, 1965. 38. Wells, A. F., ‘‘Structural Inorganic Chemistry.’’ Clarendon, Oxford 1975.
coida
190, 269–277 (1997)
CS974845
Kinetics of the Formation of Goethite in the Presence of Sulfates and Chlorides of Monovalent Cations Aida Frini and Mohamed El Maaoui 1 Laboratory of Industrial Inorganic Chemistry, Sciences University of Tunis, 1060 Tunis, Tunisia Received February 27, 1996; accepted February 25, 1997
The synthesis of goethite by oxidation of Fe2/ in presence of metallic iron was undertaken in an aqueous medium containing indifferent salts such as Na2SO4 , (NH4 )2SO4 , NaCl, and NH4Cl. Temperature and bubbling air rate were maintained, respectively, at 707C and 1 L/min. The influence of anions and cations on the kinetics of each step of the process has been followed distinctly, the iron dissolution rate has been determined by the variation of the medium acidity, and the precipitation of goethite has been determined by gravimetric measurements. With respect to Cl 0 , / the SO 20 4 anion decreases the rate of the two reactions. NH 4 acts as an inhibitor when it is present at low concentrations and as an accelerator for higher concentrations; the limit corresponding to the change of NH / 4 behavior depends on the nature of the counter ion. The reaction product is composed of pure goethite in the presence of sulfate salts, whereas a mixture of goethite and lepidocrocite, respectively, 60–70 and 40–30%, was observed in the presence of chloride salts. q 1997 Academic Press Key Words: goethite; iron dissolution; influence of anions; influence of cations; iron corrosion.
I. INTRODUCTION
Slow oxidation of ferrous solutions by air bubbling provides one or several of the following products: goethite ( a-FeOOH), lepidocrocite ( g-FeOOH), magnetite (Fe3O4 ), and hematite ( a-Fe2O3 ) (1–13). Ferrihydrite (Fe5HO8r4H2O) is a natural product (1). Rapid oxidation, using H2O2 , leads to the precipitation of feroxyhyte ( d-FeOOH) (14). Ferrous iron oxidation has been the main subject of many investigations. An extensive study which described the influence of temperature, pH, and Fe2/ concentration is due to Olowe et al. (14–17). No kinetic considerations were presented; however, the authors indicated the conditions which may yield goethite, lepidocrocite, or magnetite. They use the ratio R Å Fe2/ /OH 0 to define the experimental conditions; for example, at 607C, goethite is obtained when R Å 3 (16). At room temperature, Schwertmann and Cornell (1) indicated that goethite and lepidocrocite precipitate at pH 6–8, whereas magnetite is formed at pH 8–14. 1
To whom correspondence should be addressed.
A second method is widely used to synthesize iron oxyhydroxides; it consists of the hydrolysis of Fe3/ cations. This is achieved by pouring freshly prepared ferric solutions into concentrated NaOH or KOH solutions; the aging of the sols takes place at 60–807C for a period of time ranging from 1 to several days. Goethite precipitates if the pH of the final solution is above 12.8; for less basic media a mixture of goethite and hematite is observed (18–20). Matijevic et al. (21–24) have thoroughly studied the hydrolysis of ferric solutions in acidic medium; b-FeOOH and a-Fe2O3 are the end products. Livage et al. (25) gave theoretical considerations on the nucleation and germination processes of various species resulting from the hydrolysis of a variety of metallic cations. In this paper we have chosen a third synthetic method, controlled air oxidation of an iron powder suspension in aqueous medium; the presence of Fe2/ cations, even in low content, makes this method very similar to the oxidative method cited earlier. However, this procedure is interesting because of the inconveniences presented by the previous procedures. Thus, the hydrolysis of ferric solutions has limited applications because of problems with chemical products consumption and the maintenance of boiling solutions for a relatively long time. The oxidative method presents many problems concerning the constancy of the product characteristics, especially the yellow color, which is very sensitive to experimental conditions; this factor depends on the presence of small proportions of the other oxyhydroxides, on the crystal morphology, and on substitution of other cations such as Cr(III) for Fe(III) (1). One inconvenience of the oxidative method is the regulation of pH by continuous addition of NaOH or NH4OH solutions. The introduction of these alkalis produces unavoidably high local concentrations of alkalis where black Fe3O4 precipitates; in addition, solutions are more and more charged with Na2SO4 or (NH4 )2SO4 which have, as we will see later, an effect on the product characteristics. The presence of iron powder in the aqueous system keeps constant the experimental conditions such as pH, ferrous iron concentration, and salt content of the solution. For example, pH is stabilized by iron dissolution without introduction of alkalis. However, this method has the
269
AID
JCIS 4845
/
6g25$$$$61
06-03-97 12:18:06
0021-9797/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
coida
270
FRINI AND EL MAAOUI
major inconvenience of requiring the use of very pure iron powder; impurities such as carbon, which are always present, cause a change in the color of goethite and they must be kept as low as possible. This paper deals with the effect of NaCl, NH4Cl, Na2SO4 , and (NH4 )2SO4 on both the reaction kinetics and the nature of the final products. These salts are expected to be ‘‘indifferent salts’’ since they are absent from the reaction equations; in fact, many observations report that they really affect the system. For example, it is well known that the rust deposit changes chemically and morphologically when iron is submitted to different corrosive media (2–4, 6, 15, 26–28). To our knowledge, no references are cited in the literature regarding the study of the effect of indifferent salts where the process is considered as a whole. In contrast, data are available when we consider each of the successive following stages separately: iron dissolution, oxidation of the ferrous cations to give ferric species, formation of (Fe III-OH0) complexes, hydrolysis of the ferric cations, and, last, condensation of the monomeric species to form polymeric and crystalline products. For the first stage, we know that the dissolution of metallic iron in aqueous media is well explained by the oxidoreduction properties of the solution compared to the iron potential. Temperature, pH, and other oxidoreduction couples have well-defined effects. Among the chosen anions and cations, Cl 0 has been the most widely studied because of its negative effect on iron corrosion. Pourbaix (29) presents a comprehensive review where it appears that, unlike other anions which form a protective layer on iron, Cl 0 produces a permeable oxide layer which allows oxidation to continue; this anion produces generalized corrosion or pitting corrosion according to the pH, the iron potential, and the Cl 0 content. So chloride anion acts in a way dependent on the nature and the morphology of the reaction products. Studies dealing with the action of other anions on iron corrosion are numerous, especially for high complexing anions such as phosphate, chromate, and carbonate, which are often used as corrosion inhibitors (29). Similarly we note that ferrous chloride, used as raw material, seems to favor the precipitation of lepidocrocite; thus aqueous 0.05 M FeCl2 containing NaCl readily gives a brown precipitate of g-FeOOH at pH 6–7 (30). A well-known synthesis of g-FeOOH consists of adding pyridine to a FeCl2 solution and then oxidizing the yellow precipitate by air until its complete transformation into lepidocrocite (8). Schwertman and Cornell (1) use FeCl2 for goethite preparation at pH 7, but they add sodium bicarbonate to the solution and operate with a low rate of air bubbling (30–40 ml/min); the lepidocrocite is obtained using the same ferrous salt at pH 6.7–6.9, without bicarbonate and with a higher air flow (100–300 ml/min). Goethite was also obtained from FeCl2 by lowering the temperature to 77C and allowing oxidation by air contact without bubbling (31). FeSO4r7H2O is widely
AID
JCIS 4845
/
6g25$$$$62
06-03-97 12:18:06
TABLE 1 Granulometric Distribution of Iron Powder Fraction size (mm)
Weight %
õ63 63–90 90–125 125–180 ú180
1.3 16.5 32.4 17.2 32.6
used in industry as a raw material; it is an abundant byproduct of the titanium dioxide manufacture. Olowe et al. have used this salt (15); they obtained a mixture of gFeOOH and a-FeOOH at 257C in an acidic medium corresponding to R ú 0.5, where R is equal to Fe2/ /OH 0 . On the other hand, the influence of indifferent salts on the hydrolysis of ferric cations has been studied principally in alkaline media; Weiser and Milligan (32) have shown that the hydrolysis of sulfate, nitrate, bromide, oxalate, and acetate iron III leads to the formation of a-FeOOH, while the use of perchlorate gives either a- or g-FeOOH depending on the quantity of alkali added (33). It seems that there is no reference concerning the influence of the cation on the nature of the oxidation products, except one indicating that the addition of Cu 2/ in the aqueous medium favors hematite precipitation (27). All these considerations show that the influence of indifferent salts on goethite synthesis is a very interesting topic and hitherto not well studied. This paper constitutes a preliminary work on the subject. II. EXPERIMENTAL
Laboratory iron powder was provided by PROLABO; it had a minimum purity of 98% and a carbon content below 0.01%. The powder granulometric distribution is given in Table 1. A 250-ml round-bottomed flask is used; it is equipped with a condenser, a pH combined electrode, and a glass tube allowing the inlet of the compressed air. Air is preheated to avoid local cooling of the solution. The bubbling rate is maintained as constant as possible at about 0.9 to 1 L/min. Magnetic stirring is used. The use of glass vessels is usually not recommended for the precipitation of goethite by the hydrolysis method, where a basic medium and a long boiling period are needed; however, in our case the solutions are weakly acidic and so there is no possibility of contaminating them with silicate anions. One hundred milliliters of a given salt solution, such as Na2SO4 , is brought to 707C and then adjusted to pH 2.5 by means of the corresponding acid, such as H2SO4 ; air bubbling is started and the temperature is allowed to stabilize
coida
271
KINETICS OF GOETHITE FORMATION
again. One gram of iron (17.8 1 10 03 mol) and 0.05 g of a-FeOOH, used as seed crystals, are added at the same time. When FeSO4 is needed, solid FeSO4r7H2O is introduced at the same time as iron powder. After a given reaction time, the water suspension is filtered out and the remaining iron and a-FeOOH are separated magnetically. The precipitate is washed several times with distilled water and dried at 607C until constant weight is attained. The weight of unreacted iron and that of the aFeOOH formed may be used to calculate the soluble iron fraction, which is obtained by recording the balance between the initial mole number of iron and the sum of the mole numbers of the final solids. Precipitates are identified by X-ray diffraction using a RIGAKU apparatus; CuKa rays were used. The infrared absorption spectra are recorded using a Perkin–Elmer 783 spectrometer; they serve both to confirm the X-ray results and to evaluate the proportions of a-FeOOH and g-FeOOH. The standardization was accomplished using the spectra of mixtures containing known proportions of pure products. For a given spectrum, let S1 and S2 be the areas of the peaks at 888 and 1018 cm01 , which correspond, respectively, to the characteristic peaks of goethite and lepidocrocite; S1 and S2 are proportional to the quantities of these products. The coefficients of proportionality are the extinction coefficients e1 and e2 . If we consider m as the weight of the mixture introduced into the KBr pastille and C1 and C2 as the mass percentages of a-FeOOH and g-FeOOH, we have C1 / C2 Å 100 S1 Å e1mC1 S2 Å e2mC2 S1 /S2 Å ( e1 / e2 )C1 /C2 . So the standardization graph representing S1 /S2 versus C1 / C2 is a line whose slope is e1 / e2 . Experiments give e1 / e2 Å 0.6. C1 and C2 are known by means of their ratio and their sum. They can be calculated easily. III. THEORETICAL CONSIDERATIONS
In aerated weak acidic media, an aqueous system containing an electrically insulated iron may induce possible electrochemical equilibra: Water reduction, 2H / / 2e 0 ` H2 ,
[1]
whose electric potential depends on the pH according to the relation
AID
JCIS 4845
/
6g25$$$$62
06-03-97 12:18:06
E1 Å 0.000 0 0.059 pH 0 0.0295 log P(H2 ). Water oxidation, 2H2O ` O2 / 4H / 0 4e 0 , E2 Å /1.228 0 0.059 pH / 0.0147 log P(O2 ). [2] P(H2 ) and P(O2 ) are the partial pressures of H2 and O2 upon the aqueous solution. For neutral pure water and in the absence of air, log P(H2 ) and log P(O2 ) are evaluated, respectively, at 027.56 and 027.86 (29); under these conditions E1 Å E2 Å 0.40 V. Iron dissolution is due to the equilibrium Fe ` Fe2/ / 2e 0 E3 Å 00.440 / 0.0295 log(Fe2/ ).
[3]
Oxyhydroxide precipitation is usually written as (29) Fe2/ / 3H2O ` Fe(OH)3 / 3H / / e 0 . Free Fe3/ is not considered here because of the simultaneous presence of Fe and Fe2/ ; Fe(OH)3 can be considered as FeOOH / H2O, hence this reaction will be written as Fe2/ / 2H2O ` FeOOH / 3H / / e 0 E4 Å /1.507 0 0.1773 pH / 0.059 log(Fe2/ ). [4] Reaction [1] takes place only in reducing media when oxygen is absent; it becomes more important when the pH decreases. For aerated medium, equilibria [2, 3, 4] are involved; they take place in different points of the system. This is the well-known phenomenon of the Eavans cells where aerated zones and nonaerated zones adopt two distinct potentials. On the iron particle surface, situated in the aerated zones, we assist to the equilibrium [2]. Iron adopts a high value of potential, equal to E2 ; it is not corroded. The expression for E2 , given above, depends on the pH as well as on the oxygen pressure (the rate of air bubbling is involved here). In the nonaerated zones, iron particles are dissolved owing to reaction [3]. The iron potential is E3 ; it depends on the ferrous iron content of the solution. We see that the existence of these cells assumes the establishment of electric links between anodes and cathodes. For iron powder, this is possible when the electrochemical cell takes place on one grain or when the oxide layer is absent, porous, nonadherent, or continuously removed by agitation. For the next part of this work it is useful to write the reaction corresponding to the iron dissolution as a combination of [2] with [3] and the one which corresponds to the goethite precipitation as a combination of [2] with [4]:
coida
272
FRINI AND EL MAAOUI
2Fe / O2 / 4H / ` 2Fe2/ / 2H2O
[5]
4Fe2/ / O2 / 6H2O ` 4FeOOH / 8H / .
[6]
We note that iron dissolution is accompanied by an increase of pH; conversely, precipitation of goethite corresponds to a decrease of pH. The overall reaction allowing the transformation of metallic iron into goethite is 4Fe / 3O2 / 2H2O ` 4a-FeOOH.
[7]
which is the sum of reactions [5] and [6]; reaction [7] is independent of pH. If reactions [5] and [6] have the same rate, pH will remain constant; otherwise an increase of pH appears when the iron dissolution is more rapid than the goethite precipitation and vice versa. Consequently, the knowledge of pH Å f (t) curves will provide important information about the relative kinetics of the two reactions.
FIG. 2. IR spectra of different products: (a) 1 M NH4Cl; (b) 1 M NaCl; (c) 0.1 M NH4Cl; and (d) 0.1 M NaCl.
IV. RESULTS AND DISCUSSION
IV.1. Nature of the Reaction Products The nature and proportions of the formed products are known by IR absorption spectra and the standardization curve described under Experimental. Some examples of the spectra are given in Figs. 1 and 2. Data summarized in Table 2 indicate that anions have a great influence on the kind of the allotropic form which precipitates. In the presence of SO 20 4 , only goethite is observed. However, in Cl 0 medium the reaction product contains about 60 to 70% goethite and 40 to 30% lepidocrocite. These results are to be compared with those in the References. The introduction of Cl 0 shifts the oxidation of Fe2/ toward the formation of g-FeOOH (9–12); it shifts the
hydrolysis of Fe3/ toward the formation of b-FeOOH (8, 34, 35). Table 2 shows that cations have no influence on the kind of precipitate. The mechanism by which the anions move precipitation toward one or another product seems not to be elucidated yet. Different hypotheses can be proposed. Cl 0 , for example, favors g-FeOOH because of the existence of a similar compound containing Cl 0 ; Thus FeOCl has the same structure as g-FeOOH (36–38). Fe(III) oxyhydroxysulfate Fe8O8 (OH)6SO4 has a crystal structure similar to that of akaganeite (1). On an other hand, we know that goethite works as an anion exchanger in acidic media and a cation exchanger in basic solutions (20). The exchange capacity of goethite for the Cl 0 anion was studied by Paterson and Rahman (22). It is possible that anion or cation adsorption on the final products is responsible for the observed results, and that adsorption changes the kinetics of the different reacTABLE 2 Nature of the End Products of the Iron Oxidation as a Function of the Nature and the Concentration of the Indifferent Salt Anion
Salt
Concn mol/L
% a-FeOOH
% g-FeOOH
SO20 4
H2SO4 Na2SO4 Na2SO4 (NH4)2SO4 (NH4)2SO4 HCl NaCl NaCl NH4Cl NH4Cl
0.1 0.1 1.0 0.1 1.0 0.1 0.1 1.0 0.1 1.0
100.0 100.0 100.0 100.0 100.0 69.5 69.0 64.0 60.6 66.0
0.0 0.0 0.0 0.0 0.0 30.5 44.0 36.0 40.0 34.0
Cl0
FIG. 1. IR spectra of different products: (a) 1 M (NH4 )2SO4 ; (b) 1 M Na2SO4 ; (c) 0.1 M (NH4 )2SO4 ; and (d) 0.1 M Na2SO4 .
AID
JCIS 4845
/
6g25$$$$63
06-03-97 12:18:06
coida
273
KINETICS OF GOETHITE FORMATION
FIG. 4. Effect of initial FeII concentration on the curves pH Å f (t).
FIG. 3. Effect of initial pH on goethite reaction synthesis.
tions involved in the product formation; these changes certainly have an effect. In the Introduction, we have seen that Schwertmann and Cornell (1) have proposed two similar operating methods to obtain a-FeOOH or g-FeOOH. The two methods seem to differ only in the rate of air bubbling. For the synthesis of goethite the rate of bubbling is 30–40 mL/min and the reaction is carried for 48 h; for lepidocrocite the rate is 100–300 mL/min and the oxidation is complete after 1 to 3 h. We will see in Section IV.3 that the kinetic of formation of goethite is faster in the presence of Cl 0 than in the presence of SO 20 4 ; this probably explains the presence of lepidocrocite in the former case.
thus, for pH0 § 4, we have observed that the reaction product is black, it remains stuck to the magnet, and the X-ray diffraction spectrum presents the peaks of magnetite, Fe3O4 . We note that the literature situates magnetite precipitation at higher values of pH (pH ú 8) (see Ref. 15); in our case, precipitation of Fe3O4 at pH 4 is probably due to the high temperature used. IV.2.2. Influence of initial Fe2/ concentration on pH Å f(t) curves. Figure 4 contains different graphs, pH Å f (t), corresponding to various initial Fe2/ concentrations. We note than the iron dissolution rate is lowered when Fe2/ content increases; this is due to the iron potential which becomes higher, Fe ` Fe2/ / 2e 0 E Å E0 / 0.03 log[Fe2/ ].
IV.2. Kinetics of the Iron Dissolution IV.2.1. Influence of the initial pH. Figure 3 shows the influence of the initial pH on the curves pH Å f (t). At the beginning of the experiment, where only iron dissolution is present, pH increases rapidly owing to reaction [5]; when the pH becomes sufficiently high, goethite precipitation takes place suddenly, the pH remains constant, and equilibrium is reached between reactions [5] and [6]. Sometimes we notice a decrease of the pH immediately after the sudden precipitation. This is characteristic of a supersaturation phenomenon, which is not reliable; two experiments, carried out under the same conditions, may or not present a maximum pH. Thus we have chosen to present in Fig. 3 only the curves having no maximum. We observe that when the pH of the initial solutions decreases, the time to reach precipitation increases and the equilibrium pH decreases. The increased time is evident since a greater quantity of protons must be consumed to reach the pH precipitation. The decrease of the equilibrium pH can be explained by the greater quantity of Fe2/ liberated by reaction [5] before precipitation. This will be confirmed in the following section in which different initial values of Fe2/ are used. For the next part of this work, kinetic curves are drawn with pH0 2.5. Higher initial pH risks giving impure products;
AID
JCIS 4845
/
6g25$$$$64
06-03-97 12:18:06
The asymptotic pH continuously decreases while Fe2/ increases; this can be explained by equilibrium [6]. As a matter of fact, an increase of Fe2/ concentration shifts the system to the right, releasing more H / ions. For the next part of this work, we have chosen to use an initial Fe2/ corresponding to the addition of 5 g/L of FeSO4r7H2O. This permits one to control the pH of the asymptotic branch, which must remain below pH 4 in order to yield pure goethite (see Section IV.1.1). IV.2.3. Influence of sulfate and chloride anions on the kinetics of iron dissolution. Figures 5 and 6 show the varia-
FIG. 5. Effect of NaCl content on the curves pH Å f (t).
coida
274
FRINI AND EL MAAOUI
FIG. 6. Effect of Na2SO4 content on the curves pH Å f (t).
tion of pH versus time in the presence of increasing NaCl and Na2SO4 content. Very little effect is observed in the presence of Cl 0 . Conversely, SO 20 4 has a marked effect; Fig. 6 shows that the higher the sulfate concentration, the greater the time to reach precipitation. The anion effect, observed here, concerns reaction [5], which is responsible for the pH increase. As reaction [5] is the sum of [2] and [3] we can assume that only reaction [3] is affected by the presence of anions. Anions adsorb on the metallic iron surface, which is positively charged by the ferrous ions being liberated. SO 20 forms a compact layer 4 limiting the diffusion of O2 to the iron, whereas Cl 0 , which has a lower charge, gives a much more permeable layer. IV.2.4. Influence of the nature of cations. The effect of the NH 4/ cation is studied by progressive replacement of Na / by NH 4/ keeping the anion content constant. The chosen mixtures of NaCl and NH4Cl, for example, go from pure NaCl to pure NH4Cl solutions. Figure 7 to 10 show the experimental results corresponding to the conditions mentioned in Table 3. Obviously, the effect of Na / may be neglected; only NH 4/ is responsible for the observed changes. We notice that the ammonium cation has a double effect: / 4
—Low NH content causes an inhibition effect: in the presence of 0.1 M Cl 0 (see Fig. 7) the kinetics of iron
FIG. 7. Effect of 0.1 M (Na, NH4 )Cl on the curves pH Å f (t).
AID
JCIS 4845
/
6g25$$$$64
06-03-97 12:18:06
FIG. 8. Effect of cations 1 M (Na, NH4 )Cl on the curves pH Å f (t).
dissolution decreases until NH 4/ concentration equals 0.05 M; in presence of 0.1 M SO 20 (Fig. 9) the decrease is 4 observed up to 0.09 M. The curve corresponding to 0.1 M Na2SO4 , which seems to be abnormal, has been repeated about 10 times; the same results have been obtained. —High NH 4/ content produces an accelerating effect (See Figs. 8 and 10). Different hypotheses can be proposed to explain these observations. NH 4/ probably acts in the two following ways: (a) Adsorption of NH3 on the slightly charged iron surface: Fe / xNH 4/ ` Fe(NH3 )x / xH / .
[8]
(b) Formation of ferrous iron complexes: Fe2/ / nNH 4/ ` [Fe(NH3 )n ] 2/ / nH / .
[9]
Reaction [8] can be achieved even when small NH 4/ quantities are used, just to form an NH3 layer on the iron surface. Reaction [9] necessitates more ammonium salt, especially if n is high. We note that these reactions contribute to the proton liberation, so when the pH Å f (t) curves rise slowly,
FIG. 9. Effect of cations 0.1 M (Na, NH4 )2SO4 on the curves pH Å f (t).
coida
275
KINETICS OF GOETHITE FORMATION
FIG. 11. Variation of the iron oxidation kinetic as function of the concentration and the cation nature in the presence of SO 20 4 . FIG. 10. Effect of cations 1 M (Na, NH4 )2SO4 on the curves pH Å f (t).
this could be attributed either to a lower iron dissolution rate or to H / release. When the ammonium salt is present at a low concentration level, we can assume that only the adsorption effect takes place, so the increase of pH Å f (t) curves appears to be slow, due either to lower iron dissolution rate or to release of H / . However, for high ammonium content, the Fe2/ complexes are predominant, and curves appear to increase rapidly although equilibrium [9] is accompanied by H / liberation. The rate of iron dissolution is enhanced because of the formation of ferrous complexes which contributes to shift equilibrium [3] to the right. IV.3. Kinetics of the Goethite Formation The rate of goethite formation is determined by direct measurement of the weight of the dry precipitate versus time (see Figs. 11 and 12). Variations of the weight of unreacted iron are given in Figs. 13–15; those relative to the variations of the dissolved iron are shown in Figs. 16 and 17. We see that the concentration of dissolved iron, which is Fe2/ in its free or complexed form, remains constant; this means that there is no Fe2/ accumulation in solution, the rate of iron dissolution being equal to the rate of goethite precipitation. This result was confirmed earlier by the asymptotic pH, which becomes constant after goethite precipitation.
We note that some differences may be observed between the whole kinetics, represented by the variations of the weight of goethite versus time, and the kinetics given by the graphs pH Å f (t). The curves pH Å f (t), especially in the period corresponding to the rapid growth of pH, were established for naked iron; they give only the influence of the salts on reaction [5]. Changes occur at the interface between metallic iron and the aqueous solution during the reaction. After a certain time the goethite forms an adherent layer on iron surface which causes a change in the kinetics of reaction [5] and consequently affects the rate of the whole process. The effect of an indifferent salt on the whole goethite formation process is actually its effect on the slowest stage between the two following reactions: 2Fe / O2 / 4H / ` 2Fe2/ / 2H2O 2/
4Fe
[5] /
/ O2 / 6H2O ` 4FeOOH / 8H .
[6]
The rate of reaction [5] can be known from the rise of pH at the start of the reaction (before the goethite precipitation); since the pH remains constant after precipitation, it is difficult to say which of [5] or [6] is the limiting step. It is expected that the horizontal curve will not remain so forever;
TABLE 3 Experimental Conditions for Figs. 7 to 10 Figure
AID
System
7 8 9 10
(Na, (Na, (Na, (Na,
NH4)Cl NH4)Cl NH4)2SO4 NH4)2SO4
JCIS 4845
/
6g25$$$$64
Concentration (mol/L) 0.1 1 0.1 1
06-03-97 12:18:06
FIG. 12. Variation of the iron oxidation kinetic as function of the concentration and the cation nature in the presence of Cl 0 .
coida
276
FRINI AND EL MAAOUI
FIG. 13. Variation of the iron dissolution kinetic as function of the concentration and the cation nature in the presence of SO 20 4 .
for long reaction times, metallic iron will diminish and its surface will be progressively coated by the oxide layer, so the dissolution kinetic will decrease. Conversely, the precipitation process, which is enhanced by the presence of goethite crystals, will become rapid, so a slow decrease of the asymptotic pH will be observed. Many experiments confirm this, which is why pH Å f (t) curves were presented here up to 60 min and not further. Figures 11 and 12 show clearly that for a sulfate medium, the kinetics of formation of goethite decreases when the salt concentration increases. Conversely, chloride salt concentration seems to have no effect. The chloride behavior can be explained by the high permeability of the rust deposit obtained in the presence of this anion; this fact was already signaled by Pourbaix (29). The results presented in Section IV.1 explain this permeability; thus, the reaction product in the presence of Cl 0 is a mixture of 60% goethite, having the shape of needles, and 40% lepidocrocite in hexagonal plates. The protective layer is then formed with heterogeneous crystals; it is a porous barrier easily crossed by the corrosive reagents. The kinetic seems to be independent of Cl 0 since it is the nature of the products which is important. For sulfate salts, the decrease in the kinetic rate when the sulfate anion content increases is identical to the decrease in the rate of reaction [5] (see pH Å f (t) curves in Fig. 6); reaction [5] is probably the limiting step. We have assigned
FIG. 14. Variation of the iron dissolution kinetic as function of the anion nature at a constant salt concentration, 0.1 M.
AID
JCIS 4845
/
6g25$$$$64
06-03-97 12:18:06
FIG. 15. Variation of iron dissolution kinetic as function of the anion nature at a constant salt concentration, 1 M.
this decrease to the adsorption of sulfate anions on metallic iron. It is worthy of note that the increased sulfate content also involves changes in the protective layer morphology. Thus, Domingo et al. (36), who studied the crystal growth of various oxyhydroxides, explain that adsorption of cations, for example, when an excess of Fe2/ is present, enhances the particles surface charges and diminishes their aggregation ability; therefore, small crystals are formed. Anions, when they adsorb on the goethite, give lower surface charges and thus larger crystal sizes. We note that Cl 0 also adsorbs readily on goethite (36), but its effect is less important because of its lower charge. V. CONCLUSION
The present work has shown that the synthesis of goethite by oxidation of the system (Fe, Fe2/ ) in weak acidic medium is very sensitive to the presence of indifferent salts such as Na2SO4 , (NH4 )2SO4 , NaCl, and NH4Cl. The reaction products are composed of pure goethite in the presence of SO 20 4 and of a mixture of goethite and lepidocrocite, respectively, 60–70% and 40–30%, in the presence of Cl 0 . The ammonium cation has no influence on the nature of the reaction outcome but it influences the reaction kinetics. An inhibition effect is observed for low ammonium concentrations, whereas an acceleration effect occurs for high con-
FIG. 16. Variation of aqueous Fe2/ content as function of the concentration and the cation nature in the presence of SO 20 4 .
coida
KINETICS OF GOETHITE FORMATION
FIG. 17. Variation of aqueous Fe2/ content as function of the concentration and the cation nature in the presence of Cl 0 .
centration; this is due to adsorption phenomena in the first case and to formations of ferrous complexes in the later. The kinetics of the iron dissolution step determined by the pH Å f (t) curves, and that of the whole process, determined by weight measurements, are sometimes affected differently by the anions and the cations studied. This is explained by a change in the morphology of the goethite protective layer. REFERENCES 1. Schwertmann, U., and Cornell, R. M., ‘‘Iron Oxides in the Laboratory: Preparation and Characterization.’’ VCH, Weinheim/New York/Basel/ Cambridge, 1991. 2. Tuula, L., and Markku, L., J. Thermochim. Acta 77, 177 (1984). 3. Getskin, L. S., Matveev, A. F., Bubnov, A. A., and Piskunov, V. M., USSR Patent 779374, November 15, 1980. 4. Solcova, A., Subrt, J., Tlaskai, J., and Zapletal, V., Chem. Prun. 8, 34 (1984). 5. Fan, Q., Guan, Z., et al., CN 1,056,688 (Cl.C01G49/06), 04 Dec. (1991) Appl. 91,106,200, 06 Jul. (1991). 6. Rezaie-Serej, S., and Cook, D. C., J. Hyperfine Interact. 41(1–4), 701 (1988). 7. Okazaki, K., Japanese Patent 62 36, 026 [87 36, 026] (Cl. CO1 G49/ 02), 17 Feb. (1987), Appl. 85/174, 379, 4pp, 09 Aug. (1985). 8. Bouth, H. S., Audrieth, L. F., Bailar, J. C., Fernelius, W. C., Johnson, W. C., and Kirk, R. E., ‘‘Inorganic Syntheses,’’ Vol. 1, p. 184. McGraw–Hill, New York/London, 1939. 9. Kaji, N., Japanese Patent 02,293,329 [90,293,329], 4 Dec. (1990), App.89/112,43, 01 May (1989).
AID
JCIS 4845
/
6g25$$$$65
06-03-97 12:18:06
277
10. Matsui, Y., Kamisaka, T., Goto, T., and Okasaki, K., Japanese Patent 62 65,937 [87 65 937] 25 March (1987), Appl. 85/203,922, 14 Sep. (1985). 11. Koike, N., Matsui, Y., Tsugaru, T., and Horiguchi, K., Japanese Patent 62 65,934 [87,65,934], 25 March (1987), Appl. 85/202,267, 12 Sep. (1985). 12. Kamisaka, T., Kaji, N., and Okasaki, K., Japanese Patent 62,167,222 [87,167,222], 23 Jul. (1987), Appl.86/7,310, 17 Jan. (1986). 13. Horst, B., Herivert, B., and Franz, H., British Patent Appl. 2,023,559,03, Jan. (1980), Ger. Appl. 2,826,941, 20 Jun. (1978). 14. Olowe, A. A., Refait, PH., and Genin, J. M. R., J. Corrosion Sci. 32, 1003 (1991). 15. Olowe, A. A., and Genin, J. M. R., J. Corrosion Sci. 32, 965 (1991). 16. Olowe, A. A., Pauron, B., and Genin, J. M. R., J. Corrosion Sci. 32, 985 (1991). 17. Olowe, A. A., and Genin, J. M. R., J. Corrosion Sci. 32, 1021 (1991). 18. Atkinson, R. J., Posner, A. M., and Quirk, J. P., J. Inorg. Nucl. Chem. 30, 2371 (1968). 19. Atkinson, R. J., Posner, A. M., and Quirk, J. P., J. Phys. Chem. 71, 550 (1967). 20. Paterson, R., and Rahman, H., J. Colloid Interface Sci. 98, 494 (1984). 21. Matijevic, E., and Cimas, S., J. Colloid Polym. Sci. 265, 155 (1987). 22. Paterson, R., and Rahman, H., J. Colloid Interface Sci. 94, 60 (1983). 23. Matijevic, E., J. Pure Appl. Chem. 50, 1193 (1978). 24. Flynn, C. M., J. Chem. Rev. 84, 31 (1984). 25. Livage, J., Henry, M., and Sanchez, C., Progr. Solid. State Chem. 4, 18 (1988). 26. Kroeckert, B., and Buxbaum, G. (Bayer A-G), European Patent, Appl. EP 406,633, 09 Jan. (1991), DE Appl. 3,922,106,05, Jul. (1989). 27. Cornell, R. M., and Giovanoli, R., Polyhedron 7, 5, 385 (1988). 28. E. Matijevic, E., Annu. Rev. Mater. Sci. 15, 483 (1985). 29. Pourbaix, M., ‘‘Lec¸ons en corrosion e´le´ctrochimique.’’ Centre Belge d’Etude de la Corrosion Cebelcor, 1975. 30. Lewis, D. G., and Farmer, V. C., J. Clay Minerals 21, 93 (1986). 31. Sato, T., Okamoto, S., and Hashimoto, K., Yogyo Kyokaishi 94(12), 1201 (1986). 32. Weiser, H. B., and Milligan, W. O., J. Phys. Chem. 39, 25 (1935). 33. Knight, R. J., and Silva, R. N., J. Inorg. Nucl. Chem. 36, 591 (1974). 34. Tatsuo, I., Kazuhiko, K., Kyoji, O., Hiroshi, M., and Toshio, K., Japanese Patent, 03, 109, 216, [91, 109, 216], 09 May (1991); App. 89 244,985,22, Sep. (1989). 35. Matijevic, E., J. Colloid Interface Sci. 58, 374 (1977). 36. Domingo, C., Rodriguez-Clemente, R., and Blesa, M., J. Colloid Interface Sci. 165, 244 (1994). 37. Ralph, W. G., Wyckoff ‘‘Crystal Structures,’’ Vol. I. Wiley–Interscience, New York/London/Sydney, 1965. 38. Wells, A. F., ‘‘Structural Inorganic Chemistry.’’ Clarendon, Oxford 1975.
coida