Improvements to peroxide oxidation methods for analysing sulfur in acid sulfate soils.
Acid sulfate soil (ASS) methods are used to analyse a diverse range of soils and materials--from sands, coffee rock, and peats, through to silts and heavy clays, as well as stockpiled dredge material and spoil. These materials have a wide range of chemical and physical properties, making their chemical analysis complicated and sometimes difficult. Part of this analytical difficulty stems from the range of oxidation states (-2 to +6) and the many chemical species of sulfur that are encountered in ASS.
Pyrite (Fe[S.sub.2]) is quantitatively the most abundant sulfur-containing compound in reduced sulfidic sediments (Metson et al. 1977; Bush and Sullivan 1997). Variable and sometimes significant quantities of iron `monosulfides', such as amorphous FeS, greigite ([Fe.sub.3][S.sub.4]), and mackinawite (Fe[S.sub.1-x]) can also occur (Bush and Sullivan 1997). Appreciable amounts of organically bound sulfur are also found in peaty ASS. In oxidised or partly oxidised ASS, sulfate is present in sparingly soluble salts such as gypsum and relatively insoluble minerals including jarosite [K[Fe.sub.3][(S[O.sub.4]).sub.2][(OH).sub.6]] and natrojarosite [Na[Fe.sub.3][(S[O.sub.4]).sub.2][(OH).sub.6]] (Fanning et al. 1993), or as the soluble ion adsorbed on soil clays and colloids. Saturated ASS, especially those influenced by seawater, often contain appreciable quantities of soluble sulfate.
Superimposed on the complex sulfur chemistry of ASS is the chemistry of acid generation and the by-products that are generated or solubilised by this acid (e.g. iron and aluminium species). In addition, these species' subsequent interaction with and neutralisation within the soil matrix further increases the complexity of ASS analysis (as illustrated schematically by Fig. 1).
[FIGURE 1 OMITTED]
The aim of laboratory analysis is to predict the net acid generation possible from the disturbance of ASS. Because of the complexity of ASS materials, more than one laboratory method will commonly be needed to dependably predict the environmental outcome.
Many chemical methods used on ASS (as well as on minespoils) have focussed on estimation of pyrite content and from this potential acid generation. Some of these methods determine pyrite content directly by measuring iron or sulfur species, then calculate the theoretical acidity possible from the stoichiometry of the complete pyrite oxidation reaction (Eqn 1):
(1) Fe[S.sub.2] + 15/4 [O.sub.2] + 7/2 [H.sub.2]O [right arrow] Fe[(OH).sub.3] + 2 S[O.sub.4.sup.2-] + 4 [H.sup.+]
The chromium reducible sulfur method is an approach that can accurately and specifically measure reduced inorganic sulfur in ASS, without apparent interference from the presence of organic sulfur compounds (Sullivan et al. 1999). As is the case with other solely sulfur-based methods, no account is taken of carbonates or any other acid neutralising components that may be present in the soil, nor is the contribution of preexisting acidity in the form of relatively insoluble sulfate minerals (such as jarosite of natrojarosite) determined.
Other methods use chemical means (e.g. hydrogen peroxide) to oxidise the iron sulfides (predominantly pyrite) present in ASS. Peroxide methods are based on the stoichiometry or the following pyrite oxidation equation (Evangelou and Zhang 1995; Eqn 2):
(2) Fe[S.sub.2] + 15/2 [H.sub.2][O.sub.2] [right arrow] Fe[(OH).sub.3] + 2 S[O.sub.4.sup.2-] + 4 [H.sup.+] + 4 [H.sub.2]O
Various peroxide-based methods have either followed the `sulfur trail' and measured the sulfate produced (e.g. the POSA method, Lin and Melville 1993), or followed the `acid trail' and titrated the acidity generated by the peroxide oxidation of pyrite (Konsten et al. 1988; Dent and Bowman 1996).
The POCAS (peroxide oxidation-combined acidity and sulfate) method (Ahern et al. 1996) combined measurements of both acid and sulfur trails (as well as calcium and magnesium), before and after peroxide oxidation. The POCAS method comprises 3 distinct parts:
(i) Titration (in 1 M KCl) of existing acidity (total actual acidity, TAA) and determination of the KCl-soluble sulfur ([S.sub.KCl]), calcium ([Ca.sub.KCl]), and magnesium ([Mg.sub.KCl]).
(ii) After oxidation with 30% w/v [H.sub.2][O.sub.2], titration (in 1 M KCl) of total potential acidity (TPA) and determination of the sulfur ([S.sub.p]), calcium ([Ca.sub.p]), and magnesium ([Mg.sub.p]) soluble in the peroxide digest solution.
(iii) The differences between respective analytes giving:
Total sulfidic acidity (TSA) = TPA - TAA
Peroxide oxidisable sulfur ([S.sub.POS]) = [S.sub.p] - [S.sub.KCl]
Reacted calcium ([Ca.sub.A]) = [Ca.sub.p] - [Ca.sub.KCl]
Reacted magnesium ([Mg.sub.A]) = [Mg.sub.p] - [Mg.sub.KCl]
As with the chromium method, POCAS does not account for jarosite sulfur in its determinations. The POCAS method (Ahern et al. 1996) has progressed as a result of continuing experimentation (Latham et al. 2000). A modified method (referred to as POCASm) was published in June 2000 (Ahern et al. 2000).
Recently, a paper by Ward et al. (2002) reported that both POCAS and POCASm methods did not reliably measure reduced inorganic sulfur in a study of 4 acid sulfate soils. This result has cast doubt on all peroxide methods. Although the POCASm method (Ahern et al. 2000) fared better than the original POCAS method in the Ward et al. (2002) investigation, the results still indicated a poor performance by POCASm. Atmospheric loss of sulfur and precipitation of jarosite were suggested as mechanisms for the poor performance of these peroxide methods.
Since publication of POCASm, continued development of the method has given iterative improvements, resulting in the method detailed here. After learning of the results reported in Ward et al. (2002), we were keen to determine whether this method would successfully analyse the 4 soils used in the Ward et al. study (soils which they reported to be subject to analytical interferences using the earlier peroxide methods). Unfortunately, those authors only had a very limited quantity of one of the soils remaining (Tuckean Swamp), which they willingly gave to us and we gratefully acknowledge.
This paper details modifications made to the POCASm method (Ahern et al. 2000) to ensure complete recovery of oxidisable sulfur from ASS and to minimise interference from jarosite precipitation. Oxidation conditions used in the improved suspension peroxide method are described and explained. Sulfur results (specifically the peroxide oxidisable sulfur or [S.sub.POS] values) are compared with the reduced inorganic sulfur ([S.sub.CR]) results obtained using the chromium method (Sullivan et al. 2000) for a range of soils (with [S.sub.CR] contents from 0.03 to 5.5 %S). A comprehensive sulfur budget is made for each soil, showing that no measurable loss of sulfur occurred under the peroxide digest conditions detailed here. The performance of the acidity trail of the improved method will be discussed in a subsequent paper.
Materials and methods
Soils used and their preparation
The location, description, Australian Soil Classification, and some chemical properties of the soils examined in this study are detailed in Table 1. Eight ASS were sampled from various locations in southeastern Qld and northern NSW. After collection, the samples were dried at 85[degrees]C for approximately 48 h in a high capacity fan-extraction oven and then ground in a Glenn Creston hammer mill, passing a 1-mm sieve. Subsequently, samples were ground for 5 min in a ceramic ring mill. When not in use, samples were kept at or below 5[degrees]C, inside 2 polyethylene bags (with most of the air excluded) inside a polyethylene jar.
Nicholas Ward kindly supplied the ninth soil (the Tuckean Swamp soil, sample G), which was also used in the investigation of Ward et al. (2002). Chemical analysis of the subsample that was provided showed that some oxidation of sulfides had occurred since Ward's original chromium analysis (of 4.38 %S) more than a year earlier. On the supplied sample, the mean of duplicate chromium reducible sulfur determinations was 2.85 %S (Environmental Analysis Laboratory, Southern Cross University) or 3.00 %S (the mean of 4 replicates in this study). The small quantity of the Tuckean Swamp soil supplied meant there was insufficient sample to perform the full range of analyses that were performed on the other 8 soils. To avoid sample loss this soil was not ring mill ground. Despite this limitation of quantity and its apparent oxidation with time, it was felt important to perform as many analyses as the sample quantity permitted on this soil from the Ward et al. (2002) investigation.
A summary of the terms and symbols used to describe analytical results is given in Appendix 1.
Improved peroxide method
Peroxide digest. Two grams of ring mill ground soil was accurately weighed into pre-weighed 250-mL tall-form beakers. Then 10 mL of AR grade 30% w/v [H.sub.2][O.sub.2] was added with swirling and the resulting reaction moderated where necessary by addition of deionised water. After 30 min, deionised water was added to make the total volume between 45 and 50 mL. Beakers were placed on a hotplate for a maximum of 30 min and maintained at 80-90[degrees]C. Samples were periodically swirled and deionised water added to maintain a constant volume and to wash soil residue from the sides of beakers. Those samples that had not reacted before being placed on the hotplate and had begun to do so vigorously while on the hotplate were removed (and had deionised water added to them if the reaction was overly vigorous), then replaced when the reaction had moderated. Samples that had `ceased reacting' while on the hotplate before the full 30 min had elapsed (i.e. samples where no effervescence was evident) were removed. Samples that had ceased effervescing prior to being placed on the hotplate (after an initial vigorous reaction) were removed from the hotplate after 5-10 min if no further reaction was apparent.
All samples were allowed to cool to room temperature and the pH of suspensions measured. A second 10-mL aliquot of [H.sub.2][O.sub.2] was added and after 10 min samples were returned to the hotplate for a maximum of 30 min (following the procedure outlined above). After removal from the hotplate and cooling, the pH of suspensions was measured. Those samples with pH <2 were repeated using 1 g of soil. It is suggested that those suspensions with pH >6.5 (and a likelihood of containing carbonates) should be digested for at least a further 2 cycles, as peroxide oxidation at near neutral to alkaline pH is less efficient (Zhang and Evangelou 1998). (At these pH values, extended digestion times are less liable to break down clays and jarosite is unlikely to form.) In this investigation, all samples had a pH <6.5 after 2 additions of peroxide so the further additions were unnecessary.
When peroxide digestion was complete, 1 mL of Cu[Cl.sub.2]*2[H.sub.2]O solution (400 mg/L in Cu) was added to suspensions (and blanks) to decompose any remaining peroxide. Beakers were returned to the hotplate until peroxide decomposition had ceased (or for maximum of 30 min) with volume maintained at about 45 mL. Duplicate blank and internal laboratory quality control samples were also run with each digestion batch. All other samples were run in quadruplicate.
Titration of peroxide suspension (TPA determination). After cooling, the beaker's contents were quantitatively transferred to titration vessels using 32 mL of 2.5 M KCl (with a final rinse of deionised water), giving approximately 80 mL, 1 M KCl suspensions (a final 1: 40 soil solution extraction ratio). Suspensions were titrated to pH 5.5 with standardised approximately 0.05 M or 0.25 M NaOH titrant (Mettler DL77 auto-titrator). The 0.25 U titrant was used for suspensions with an initial pH [less than or equal to]3, and 0.05 M NaOH used where the pH was >3. Then 2 mL of 30% w/v [H.sub.2][O.sub.2] (adjusted to pH 5.5) was added (to oxidise any ferrous ion) and the suspension re-titrated to pH 5.5 (if the pH had dropped below 5.5) and then to pH 6.5.
Determination of peroxide sulfur, calcium, and magnesium. Immediately after titration, contents of the titration vessels were quantitatively transferred (using deionised water) to a tared 600-mL beaker. While on a balance, suspensions were made up to a weight equivalent to 400 mL of 0.2 M KC1 plus soil, using deionised water. The suspensions were then filtered through Whatman #3 paper. The filtrate was analysed for [S.sub.p], [Ca.sub.p], and [Mg.sub.p] by ICP-AES (Thermo Jarrell Ash IRIS/AP HR Duo). The final volume/weight of the titrated suspension may be varied, depending on the technique/s used to determine these analytes. If sulfate specific techniques (such as turbidimetry or ion chromatography) are used, this may minimise any inflation of results that may result from ICP-AES determination of additional non-sulfate sulfur species (particularly those attributable to organic matter). Subsequent ion chromatography (IC) analysis (Dionex) for sulfate gave similar values to ICP, but replication was much poorer for the IC results compared with ICP.
1 M KCl extractable sulfur, cations, and TAA. To determine TAA and 1 M KCl extractable sulfur and cations, 2 g of soil was suspended in 80 mL of 1 M KCl (i.e. a 1:40 extraction ratio) and shaken for 4 h, before standing overnight and being briefly shaken the next morning to re-suspend the soil. The entire suspension was quantitatively transferred to a titration vessel and titrated to pH 5.5 and 6.5 endpoints, and finally to neutrality (on a Mettler DL77 auto-titrator). The titration results were used to determine TAA. Following titration, the contents of the vessels were quantitatively transferred to beakers with deionised water, made to a weight equivalent to 400 mL of 0.2 M KCl plus soil, then filtered through Whatman #3 paper. The filtrate was analysed for [S.sub.KCl], [Ca.sub.KCl], and [Mg.sub.KCl] by ICP-AES. Duplicate blank and internal laboratory quality control samples were also run with each batch. All other samples were run in quadruplicate with the exception of sample G (Tuckean Swamp), which was duplicated.
Chromium reducible sulfur
Chromium reducible sulfur ([S.sub.CR]) was determined for all soils (in quadruplicate) using the method of Sullivan et al. (2000). Sample weights were optimised to obtain the most accurate and reproducible values. To verify these results, samples were also analysed in duplicate by Environmental Analysis Laboratory, Southern Cross University, Lismore (referred to as Laboratory A). At the request of one reviewer, a subsample of soil H was sent to 2 additional independent laboratories for analysis (Australian Laboratory Services in Brisbane, referred to as Laboratory B, and Natural Resource Sciences Laboratories, Qld Government, referred to as Laboratory C).
Sulfur and cations soluble in 4 M HCl
Acid extractable [S.sub.HCl], [Ca.sub.HCl], [Mg.sub.HCl], [Na.sub.HCl], [Al.sub.HCl] and [Fe.sub.HCl] were determined (in quadruplicate) by shaking 2 g of soil in 80 mL of 4 M HCl for 16 h, filtering through Whatman #3 paper, followed by elemental analysis using ICP-AES (Thermo Jarrell Ash IRIS/AP HR Duo). The 4 M HCl extracts sulfate from relatively insoluble minerals such as jarosite, in addition to soluble, exchangeable, and gypsic sulfate. However 4 M HCl is unlikely to dissolve pyrite sulfur. The calculation [S.sub.HCl] - [S.sub.KCl] (or [S.sub.j]) provides an estimate of sulfur in insoluble minerals such as jarosite.
Analysis of soil residue from peroxide digest
Any soil residue remaining in beakers was quantitatively transferred to the filter paper using deionised water and the paper washed with 2 x 20 mL aliquots of 1 M KCl, then rinsed 3 times with deionised water to remove any soluble or adsorbed sulfate. The soil residue was then analysed for various sulfur fractions as part of the complete sulfur accounting approach used in this study.
Residual [S.sub.HCl] When dry, filters papers (containing the soil residue) from 2 of the 4 replicates for each soil were placed in HDPE bottles along with 80 mL of 4 M HCl for overnight (16 h) extraction. These extracts were then filtered through Whatman #3 paper for ICP-AES analysis of S, Ca, Mg, Na, Fe, and Al (after appropriate dilutions where required). The 4 M HCl extract of the digested soil residue should contain insoluble sulfate minerals (e.g. jarosite) originally present in the soil plus any additional sulfates precipitated during peroxide digestion. (The beakers used in the original peroxide digest were treated with 4 M HCl to dissolve and account for any insoluble sulfate that may have been coating beaker walls, with the 4 M HCl solutions analysed by ICP-AES.) Begheijn et al. (1978) used 4 M HCl to extract jarosite sulfate, following extraction with 0.1 M 3Na. EDTA as part of their S speciation scheme for ASS analysis.
Residual [S.sub.CR]. The 4 M HCl extraction procedure is unable to determine whether there is unoxidised pyrite remaining in the soil residue. To test for the presence pyrite in soil residues following peroxide oxidation, the remaining 2 residues (from the 4 replicates) from each soil were analysed by the chromium reducible sulfur method (Sullivan et al. 2000). If the pyrite in the original sample had been fully oxidised by the peroxide digest then this result should be negligible.
Determination of total sulfur
Total sulfur was determined by 2 independent laboratories using different methods. One laboratory (Laboratory A) determined total sulfur by combustion furnace (Leco CNS 2000), while the other (Laboratory B) determined total sulfur after HF digest followed by ICP determination of sulfur.
Other analytical methods used
A number of pertinent soil properties were measured in addition to those determined in the improved peroxide method. Unreplicated determinations of soil p[H.sub.w] and electrical conductivity (EC) in a 1:5 soil water extract were made using Method 4Al in Rayment and Higginson (1992). Chloride was measured on the same extract by an automated colorimetric technique (as per Method 5A2, Rayment and Higginson 1992). Exchangeable bases ([Ca.sup.2+], [Mg.sup.2+], [Na.sup.+], and [K.sup.+]) in 1 M N[H.sub.4]Cl at pH 7.0 were measured in quadruplicate (using Method 15A1 in Rayment and Higginson 1992), without pre-treatment for soluble salts. Total C and organic C (after treatment with sulfurous acid to remove inorganic C) (after Nelson and Sommers 1982) were determined (in duplicate) with a Leco CNS 2000 combustion furnace operating at 1300[degrees]C.
Results and discussion
Comparison of peroxide oxidisable sulfur with chromium reducible sulfur
The chromium reducible sulfur ([S.sub.CR]) measurement from the chromium method comes by a chemical reduction process, while the peroxide oxidisable sulfur ([S.sub.POS]) value from the peroxide method comes by an oxidative technique. As these methods use analytically disparate approaches to determine an effectively equivalent quantity, a good agreement between the two should engender confidence in both methods. Our results show a highly significant ([F.sub.1,6] = 4311; P = 0.001; [R.sup.2] = 0.999) linear relationship between peroxide oxidisable sulfur ([S.sub.POS]) (i.e. Sp - [S.sub.KCl]) and chromium reducible sulfur ([S.sub.CR]) for 8 soils:
[S.sub.POS] = 0.967 x [S.sub.CR] + 0.051 (%S)
The slope was not significantly (P = 0.05) different from unity, and the intercept was not significantly (P = 0.05) different from zero. For comparison purposes, the ideal 1:1 relationship is shown along with the regression line in Fig. 2. Clearly the [S.sub.POS] results using the improved peroxide method were not significantly lower than the [S.sub.CR] results by the chromium method, as found by Ward et al. (2002) for previous peroxide methods.
[FIGURE 2 OMITTED]
For the only soil common to both studies (Tuckean Swamp, soil G), Ward et al. (2002) reported that the [S.sub.POS] result for the earlier peroxide methods ranged between 48% and 69% of the chromium reducible sulfur result (the worst result of the 4 soils they investigated). In our study, the [S.sub.POS] result for the Tuckean Swamp by the improved peroxide method was 98% of our chromium reducible sulfur result (or 103% of the chromium result determined by Environmental Analysis Laboratory, Lismore). This is good evidence that the improvements made have overcome problems of poor sulfur recovery associated with earlier peroxide methods.
For highly organic samples such as peats, peroxide oxidation of the organic sulfur fraction may result in a higher [S.sub.POS] value compared with [S.sub.CR]. This is the case for soil I, which is a highly organic (peaty) soil (total C = 10.7%). Its results were not included in the preceding regression, as sulfur from organic material is likely to have had inflated the [S.sub.POS] value. Also, as it had by far the highest [S.sub.POS] value ([S.sub.POS] = 6.1%), its inclusion would have unduly biased the regression (see Fig. 2). However, if soil I is included in the data set, the resultant regression ([S.sub.POS] = 1.064 x [S.sub.CR] - 0.062; [R.sup.2] = 0.992) is still highly significant ([F.sub.1,7] = 985; P = 0.001), the slope not significantly (P = 0.05) different from unity, and the intercept not significantly (P = 0.05) different from zero. When analysing highly organic samples (such as soil I), the chromium method is normally recommended, as it should provide a better measure of the reduced inorganic sulfur content (McElnea and Ahern 2000; Sullivan et al. 2000).
In summary, there is extremely good agreement between the sulfur trail from the improved method (i.e. [S.sub.POS]) and the chromium reducible sulfur ([S.sub.CR]) result, despite the two methods using contrasting analytical approaches.
Overall sulfur budgets
Various workers have reported sulfur loss during chemical analysis when conducting sulfur budgets for ASS (Konsten et al. 1988; Ward et al. 2002). For a method to be accepted as being robust, a sulfur budget should demonstrate minimal or no sulfur loss. In order to determine if sulfur loss did occur in the improved method, the sum of the sulfur measured in the peroxide digest solution ([S.sub.p]) and the sulfur remaining in the digested soil residue (res-[S.sub.HCl] and res-[S.sub.CR]) was compared with the total sulfur.
Reliability of total sulfur values
When preparing a sulfur budget there is a heavy dependence on the total sulfur value used. It is important to ensure that the total sulfur measurement is not biased by either the method or analytical laboratory. Soils were thoroughly mixed and subsamples sent to 2 independent laboratories for analysis by 2 different methods--HF digest and Leco Combustion Furnace (Table 2, Lines 1 and 2).
There was a highly significant ([F.sub.1,6] = 7451; P = 0.001; [R.sup.2]= 0.999) linear relationship between Leco furnace and HF digest total sulfur for 8 soils:
[S.sub.T] (Leco) = 1.024 x [S.sub.T] (HF digest) + 0.061 (%S)
The slope was not significantly different (P = 0.05) from unity, and the intercept was not significantly different (P = 0.05) from zero. This indicates that the 2 independent methods (and laboratories) yielded equivalent results, providing confidence in the total sulfur results used as the basis for the sulfur budget in this investigation.
Peroxide digest sulfur budget
The sulfur in the digested solution was measured (Table 2, Line 6; [S.sub.p]). The sulfur contained in this solution should include soluble and adsorbed sulfate, sulfides converted to sulfate by the peroxide, plus any organic sulfur released by the digestion of any organic matter originally present. Where there are no insoluble sulfates present (e.g. jarosite) and all sulfides have been fully oxidised, [S.sub.p], should approximate the total sulfur value.
For the purposes of accounting for all the sulfur present in the soils of this study, soil residues (from the 4 replicates of each soil) were retained and analysed. Two residues from each soil were tested for undigested pyrite using the chromium method (Table 2, Line 9; res-[S.sub.CR]). The insoluble sulfur content was determined on the remaining 2 residues by 4 M HCl extraction (Table 2, Line 10; res-[S.sub.HCl]). This extractant will not recover pyrite sulfur as evidenced by Rice et al. (1993) who could not detect sulfur after extraction of pyrite with 6 N HCl + Sn[Cl.sub.2].
Measurement of sulfur in soil residues served 2 useful functions. Chromium reducible sulfur analysis on the residue helped determine if the sample had been insufficiently digested (i.e. whether there was any unoxidised pyrite present). The HCl extraction of the soil residue determined whether there had been additional insoluble sulfur (e.g. jarosite) formed by the peroxide oxidation (i.e. whether the sample had been `over-digested').
The sum of the sulfur components in the filtered peroxide solution and digest residue (i.e. Table 2, Line 11; [S.sub.p], + res-[S.sub.CR] + res-[S.sub.HCl]) was either equal to or slightly greater than the total sulfur by either HF digest or Leco furnace methods (Fig. 3). A larger error would be expected with the addition of 3 sulfur values derived from the peroxide digest and residues, compared with a direct, single determination of total sulfur. This may explain why the sum of the sulfur components from the peroxide digest was slightly greater than the total sulfur on some soils. Nevertheless, there was good agreement between this sum of sulfur components and the total sulfur measurements.
[FIGURE 3 OMITTED]
Most importantly, for the 9 soils tested, there was no measurable loss of sulfur, demonstrating that the improved peroxide oxidation procedure does not result in the sulfur loss reported by Ward et al. (2002) for earlier peroxide methods. In their sulfur budgets, Ward et al. (2002) found that for 4 soils substantial losses of sulfur occurred during peroxide digestions using the original POCAS and modified POCASm methods. For the only soil common to both studies (Tuckean Swamp, soil G), they reported unaccounted losses of sulfur that ranged between 23 and 28% of the total sulfur result, which they stated was mostly likely due to atmospheric losses. In our study, the sum of sulfur components for the Tuckean Swamp soil was slightly greater than the total sulfur result.
In summary, the results from our sulfur budgets provide strong evidence that the improved peroxide procedure does not experience any sulfur loss, a problem that had been reported for earlier peroxide methods.
Efficiency of the improved peroxide method in oxidising sulfides
If oxidation of the sulfides has been efficient then little or no sulfide should be measured in the soil residue. With one exception, the sulfide content of the residues was <1% of the value found in the original soils (see Table 2, Line 9). Only the Tuckean Swamp sample (soil G) showed a small amount of unoxidised sulfide in the residue (with 92.4% of the original sulfide oxidised). Soil G was the only soil that had not been finely ground. Despite the incomplete digestion of this soil, [S.sub.CR] and [S.sub.POS] (3.00% and 2.93%, respectively) results were effectively equivalent, making little practical difference to the management of such a soil. Subsequent ring mill grinding of the remainder of this sample resulted in the complete oxidation sulfide after peroxide digestion.
Hence, the more moderate digestion conditions of the improved peroxide oxidation method are capable of efficiently oxidising sulfides in ASS, provided soils have been very finely ground.
Development of recent peroxide methods
Soil extraction ratio
The trend in the development of peroxide methods has been to an increase in the soil to solution ratio. The early POSA method (Lin and Melville 1993) used a 1:5 extraction ratio. With this low ratio, the solubility of gypsum (CaS[O.sub.4].2[H.sub.2]O) could easily be exceeded, lowering the sulfur result (McElnea and Baker 1998). This is especially the case for soils high in sulfides and CaC[O.sub.3], since the sulfuric acid produced by peroxide oxidation of pyrite can react with CaC[O.sub.3] to produce gypsum. In the improved method, the final extraction ratio is 1:40 (or 1:80 where 1 g of sample is used for high sulfur samples), making gypsum precipitation or insolubility much less likely. Although this wider extraction ratio may decrease precision for very low sulfur soils, it makes the analysis of high sulfur ASS much more reliable.
A number of changes to the peroxide digestion procedures used in earlier methods have been made. The duration of the digest has been substantially shortened. Samples are no longer left on the hotplate (or water bath) after the added peroxide has been consumed. As previously mentioned, the extraction ratio has been increased to 1:40. A clearer, more concise definition of the peroxide digest procedure has decreased the potential for operator bias in the method (e.g. the determination of when peroxide oxidation is complete), making the method more reproducible.
The less aggressive digestion procedures employed are designed to minimise the chance of jarosite precipitating during peroxide oxidation, which can result in lower [S.sub.POS] results. Jarosite precipitation is favoured at low pH and high [Fe.sup.3+], S[O.sub.4.sup.2-], and [K.sup.+] concentrations (Wang et al. 1988; Das et al. 1996). In contrast to the original POCAS method, 1 M KCl (whose presence can favour jarosite precipitation by the common ion effect) is not present during the digestion. The KCl is only added to cooled suspensions just prior to titration. Additionally, to avoid extremely low pH, samples whose suspension pH is <2 after the addition of 2 aliquots of peroxide are repeated using half the sample weight.
Jarosite precipitation is also favoured as temperature increases (Stahl et al. 1993). The Ward et al. (2002) study showed a trend toward increased formation of jarosite with length of digest time (in days), especially for hot digests. In the improved method, the duration of the peroxide digestion has been considerably shortened to further minimise the possibility of jarosite precipitation.
Many of the soils tested already contained measurable quantities of jarosite or `insoluble sulfur' (Table 2, Line 8) prior to oxidation. The sulfur in the Kerkin's Levee (soil B, a spoil sample typical of surface ASS material found at sites where there has been human disturbance) was principally in an insoluble form. However the insoluble sulfur content did not increase in the residue after peroxide oxidation (Table 2, Line 10). For many of the soils there was a slight but not appreciable increase in insoluble sulfates after peroxide treatment.
One sample in which appreciable insoluble sulfur did form during peroxide oxidation was soil H (res-[S.sub.HCl] = 0.46%). This soil was overwhelmingly the most rapidly and violently reacting sample when peroxide was added during the digestion procedure. The high temperatures associated with such a vigorous reaction may have contributed to the formation of insoluble sulfates. Despite the apparent formation of jarosite (or natrojarosite) in this high sulfur sample, the [S.sub.POS] result was only 2.5% lower (in relative terms) than the [S.sub.CR] result (3.19 %S v. 3.275 %S). The subsample analysed by laboratory A for chromium reducible sulfur gave a considerably lower result ([S.sub.CR] = 2.29%), illustrating that this soil can be a problematic sample to analyse by either peroxide or chromium methods (see later discussion).
Neither the peroxide nor the chromium method provides a measure of the insoluble (or jarositic) sulfur present in the original soil and hence of the latent acidity stored in some ASS (such as soil B). As was done in this investigation, an additional step to determine insoluble sulfur in the washed soil residue (after peroxide digestion) can be performed to obtain an estimate of jarosite content. Alternatively, a 4 M HCl extraction can be made ([S.sub.HCl]) on a fresh subsample, which in combination with the [S.sub.KCl] result can be used to estimate the maximum jarosite content. Soils such as B and H emphasise the difficulties inherent in analysis of ASS (by chromium or peroxide methods) and illustrate the need for a range of analytical methods to be employed in order to fully characterise some ASS.
Particle size and grinding
One of the aims of this investigation was to achieve complete peroxide oxidation of pyrite, without having to resort to prolonged, high temperature digestion conditions that favour jarosite precipitation. Fine grinding (by a ceramic ring mill) was performed in order to decrease soil particle size, maximise surface area of pyrite, release pyrite from within fine root remains and organic matter, and reveal fresh crystallite surfaces where they may have originally been protected by surface coatings. A finely ground sample has the added advantage of facilitating better sample mixing and improving homogeneity. This permits a smaller subsample to be used during analysis, while still being representative of the entire sample. Additionally, a smaller sample requires lesser quantities of chemical reagents resulting in a cost saving for laboratories.
Konsten et al. (1988) in their original soil `TPA' method (using wet, `as-received' samples) concluded that not all pyrite was oxidised by hydrogen peroxide. They found this was particularly the case for soils rich in organic matter (citing chemical and microscopic evidence and intimating that `stable' pyrite may be enclosed within the organic matter and protected from oxidation). Such evidence supports the contention that drying and pulverisation of the sample prior to analysis needs to be performed if quantitative pyrite oxidation is to be achieved.
Nicholson et al. (1988) state that the rate at which pyrite oxidises is a linear function of surface area. Hence, the greater the exposed pyrite surface area (resulting from the fine grinding) of the material to be oxidised, the faster pyrite reacts and the shorter the duration of the digest necessary.
In this investigation the peroxide digestion procedure was tested using finely ground samples, but it cannot be assumed that the shortened peroxide digest procedure will necessarily perform as well on more coarsely ground samples. This procedure is not necessarily transferable to pyritic acid mine samples. `Mineral' pyrite (from metamorphic or igneous material where crystal size is typically macroscopic) responds differently to oxidation in terms of effective reaction rates, compared with the sedimentary pyrite of ASS, because of the latter's framboidal texture, micro-crystallinity, and greater specific surface area (Caruccio 1975).
To some extent, ASS peroxide methods have been influenced by mine spoil analysis methodology and have been judged by their ability to quantitatively recover pure, finely ground mineral pyrite added to samples. We are of the opinion that the use of complete oxidation and recovery of added mineral pyrite as the gauge to determine the vigour and extent needed for peroxide oxidation procedures is inappropriate for ASS. The use of this gauge has undoubtedly biased method development and resulted in unnecessarily prolonged oxidation procedures that can lead to precipitation of insoluble sulfate salts (such as jarosite). In this investigation, added mineral pyrite (<63 gm) was not quantitatively recovered by the shorter and milder digestion conditions. However, the digestion conditions used efficiently oxidised and recovered sedimentary pyrite in the ASS studied, as shown by chromium reducible sulfur analysis on the soil residues.
Effect of measuring sulfur after titration
Earlier POCAS methods filtered the soil suspension after digestion and separate aliquots were taken for titration and sulfur determination. The aliquot taken for sulfur analysis often had a low pH and a high concentration of soluble iron. An important alteration made in the improved method involved the titration of soil suspension to pH 6.5 and subsequent filtering before analysis for sulfur. The most important effect of this change was to increase the pH of the analysed solution, largely removing any soluble iron present prior to sulfur determination. This may be advantageous for sulfur determination by ICP-AES where the presence of large concentrations of iron in solution has the potential to cause interference. The decrease in the concentration of soluble iron may also be beneficial in the determination of sulfur by techniques other than ICP-AES.
General comments on methods
Precautions with the peroxide digestion
While the oxidation procedure has been detailed in the methods section and its implications discussed in previous sections, it is desirable to alert potential users to some precautions that should be taken to minimise the chance of poor sulfur recovery. Some soils (especially those high in pyrite) can react extremely violently and almost uncontrollably, particularly following the first addition of peroxide. Such `volcanic' reactions should be avoided at all costs--it wastes peroxide and the high temperatures reached during these violent reactions favour jarosite formation. The potential for loss of sulfur from spitting of solution is also greatly enhanced with such a reaction. The addition of deionised water (via a wash bottle) at the first sign of a vigorous reaction will moderate the subsequent reaction, making it easier to control. If the reaction does becomes uncontrollable, the sample should be repeated with greater care and/or with a lesser sample weight.
After the initial reaction period and before being placed on the hotplate (or water bath), sample volumes are made up to approximately 45 mL. It is important to maintain this volume by regular addition of deionised water and also to periodically swirl the sample container to prevent the soil from settling on (and sticking to) the bottom of the beaker while it is on the hotplate. Rinsing the sides of beakers with small squirts of deionised water also serves to dissolve any salts that may have accumulated there.
Precautions with the chromium reducible sulfur method
While our investigation focussed on the peroxide digest, a few points in relation to the chromium method are worth noting. Care needs to be taken using the chromium method when analysing high sulfur samples. This investigation suggests a slight weight bias associated with the method when analysing high sulfur ASS.
For all 9 soils, a plot of the mean [S.sub.CR] measured by independent laboratory A versus the mean [S.sub.CR] measured in our investigation is presented (Fig. 4). There was a highly significant (P = 0.001) linear relationship ([R.sub.2] = 0.999) for soils A to G, demonstrating very good inter-laboratory agreement for these soils (with [S.sub.CR] of up to 3%) using this method. However, laboratory A had a much lower mean [S.sub.CR] result for sample H compared with our result. At the request of one referee, subsamples of soil H were sent to 2 additional laboratories (B and C) for chromium reducible sulfur analysis to try to resolve the discrepancy between our data and laboratory A on this soil. The means of duplicate determinations by Laboratory B were 2.82 %S for 0.5 g of soil and 3.00 %S for 0.2 g, of soil, while for Laboratory C the means were 3.24 %S and 3.28 %S for 0.5 g and 0.2 g, respectively. Our mean [S.sub.CR] result on sample H was 3.275 %S after optimising the sample weight (0.3 g), which was essentially equal to laboratory C and slightly higher than laboratory B (see Fig. 4). The range of mean results for this `problematic' sample illustrates that the chromium reducible sulfur method, like many other soil methods, can be prone to considerable variation on some soils.
[FIGURE 4 OMITTED]
For soil I (which had the highest SCR result), laboratory A had a slightly lower mean SCR (see Fig. 4) compared with our mean result using an optimised weight. Only when the sample weight was decreased to 0.1 g did we find that replicates were consistent and recovery maximised for this soil. Lowering the sample weight to between 0.2 and 0.3 g still gave variable and lower values compared with the optimised weight. It should be noted that the published method (Sullivan et al. 2000) only recommends dropping sample weight to about 0.5 g for soils with [S.sub.CR] > 1%.
A potential error in the chromium reducible sulfur method is likely to result from the failure of the trapping solution to capture all gaseous [H.sub.2]S being released by soils during the digest. While the [H.sub.2]S trapping capacity of the collection solution should not be exceeded on the basis of reduced inorganic sulfur content in soil I, the instantaneous rate of [H.sub.2]S generation might exceed the rate at which the solution can trap [H.sub.2]S (Sullivan et al. 2000), yielding a low result. This indicates that the chromium reducible sulfur method could underestimate reduced inorganic sulfur in routine laboratory use, particularly on reactive, high sulfur samples. Despite some evidence for weight bias and variability on very high sulfide content samples, the chromium method, when optimised, is a reliable measure of the reduced inorganic sulfur content in acid sulfate soils (Sullivan et al. 2000) and has an advantage over peroxide methods for measuring reduced inorganic sulfur contents of highly organic soils and those soils with very low sulfide contents.
For 9 actual and potential ASS, with a wide range of properties and sulfur levels, there was good agreement between the peroxide oxidisable sulfur ([S.sub.POS]) result by the improved peroxide method and the reduced inorganic sulfur ([S.sub.CR]) value by the chromium method. Complete oxidation of sulfides in soils was achieved by the peroxide oxidation procedure employed, provided samples were finely ground.
Sulfur budgets for soils analysed using the improved method confirmed that there was no volatile sulfur loss occurring as found by others using earlier peroxide-based methods. Additionally, changes made to the digestion conditions seem to have largely overcome the problem of jarosite precipitation experienced with earlier peroxide methods. Highly sulfidic and organic soils were shown to be the most difficult to analyse routinely using either peroxide or chromium methods.
Finally, it must be emphasised that ASS can be inherently complex and all analytical methods may have some shortcomings that limit their abilities to accurately predict the likely environmental risk posed by some ASS. The data presented restore confidence in peroxide oxidation as an effective approach for the analysis of ASS, though it is important that the comparison between the improved peroxide method and the chromium method be extended to more soils. Peroxide-based methods, combined with the chromium reducible sulfur method and the 4 M HCl extraction, form a sound platform for understanding and managing acid sulfate soils.
Appendix 1. Terms and symbols Terms Units ANZECC POCAS POCASm POSA ANC equivalent % CaC[O.sub.3] p[H.sub.KCl] -- p[H.sub.OX] -- TAA mol [H.sup.+]/kg (or mol [H.sup.+]/t) TA[A.sub.Ba] mol [H.sup.+]/kg (or mol [H.sup.+]/t) TPA mol [H.sup.+]/kg (or mol [H.sup.+]/t) TSA mol [H.sup.+]/kg (or mol [H.sup.+]/t) [S.sub.CR] %S [S.sub.KCl] %S [S.sub.HCl] %S [S.sub.HCl]-[S.sub.KCl] %S (or [S.sub.J]) [S.sub.P] %S [S.sub.POS] %S [S.sub.T] %S S-TSA equivalent %S res-[S.sub.CR] %S res-[S.sub.HCl] %S [Ca.sub.A] %Ca [Ca.sub.KCl] %Ca [Ca.sub.HCl] %Ca [Ca.sub.P] %Ca [Mg.sub.A] %Mg [Mg.sub.KCl] %Mg [Mg.sub.HCl] %Mg [Mg.sub.P] %Mg Terms Expansion ANZECC Australian and New Zealand Environmental Conservation Council POCAS Peroxide oxidation-combined acidity and sulfate method (Ahern et al. 1998) POCASm Modified peroxide oxidation-combined acidity and sulfate method (Ahern et al. 2000) POSA Peroxide oxidisable sulphidic acidity method (Lin and Melville 1993) ANC Acid neutralising capacity p[H.sub.KCl] Potassium chloride pH p[H.sub.OX] Peroxide oxidised pH TAA Titratable actual acidity (formerly Total actual acidity) TA[A.sub.Ba] Titratable actual acidity (in Ba[Cl.sub.2]) TPA Titratable peroxide acidity (formerly Total potential acidity) TSA Titratable sulfidic acidity (formerly Total sulfidic acidity [S.sub.CR] Chromium reducible sulfur [S.sub.KCl] KCl sulfur [S.sub.HCl] HCl sulfur [S.sub.HCl]-[S.sub.KCl] Insoluble (jarositic) sulfur (or [S.sub.J]) [S.sub.P] Peroxide sulfur [S.sub.POS] Peroxide oxidisable sulfur [S.sub.T] Total sulfur S-TSA res-[S.sub.CR] Residual chromium sulfur res-[S.sub.HCl] Residual insoluble (or [S.sub.JR]) (jarositic) sulfur [Ca.sub.A] `Reacted' calcium [Ca.sub.KCl] KCl calcium [Ca.sub.HCl] HCl calcium [Ca.sub.P] Peroxide calcium [Mg.sub.A] `Reacted' magnesium [Mg.sub.KCl] KCl magnesium [Mg.sub.HCl] HCl magnesium [Mg.sub.P] Peroxide magnesium Terms Description ANZECC Australian and New Zealand Environmental Conservation Council POCAS Peroxide oxidation-combined acidity and sulfate method (Ahern et al. 1998) POCASm Modified peroxide oxidation-combined acidity and sulfate method (Ahern et al. 2000) POSA Peroxide oxidisable sulphidic acidity method (Lin and Melville 1993) ANC Measurement of a soil's ability to neutralise or buffer added acid p[H.sub.KCl] pH of 1:40 1 M KCl suspension (prior to TA[A.sub.KCl] titration) p[H.sub.OX] pH of suspension after peroxide digestion, prior to TPA titration (1:40 1 M KCl) TAA Acidity titration with standardised NaOH of soil suspension or extract with unbuffered salt solution (e.g. 1 M KCl, 0.5 M Ba[Cl.sub.2]). TAA method indicated with subscripts showing extracting salt and titration pH (e.g. TA[A.sub.KCl6.5]) TA[A.sub.Ba] TAA in 0.5 M Ba[Cl.sub.2] 1:5 suspension, titrated to pH 5.5 TPA Acidity titration with standardised NaOH on 1:40 1 M KCl soil suspension after peroxide digest (to pH 5.5 or 6.5; specified with subscript, e.g. digest (to TP[A.sub.6.5]) TSA Net acid generated after peroxide (TPA - TA[A.sub.KCl]; titrated to the same pH, specified with subscript, e.g. TS[A.sub.6.5]) [S.sub.CR] Sulfide measured by iodometric titration after acidic chromous chloride reduction [S.sub.KCl] S in 1 M KCl (after TAA titration) [S.sub.HCl] S after 1:40 16 h 4 M HCl extract [S.sub.HCl]-[S.sub.KCl] Net acid soluble S ([S.sub.HCl] - (or [S.sub.J]) [S.sub.KCl]) [S.sub.P] S after peroxide digest and TPA titration [S.sub.POS] S oxidised by peroxide digestion ([S.sub.P] - [S.sub.KCl]) [S.sub.T] Total sulfur S-TSA TSA converted to equivalent %S (assuming 4 mol [H.sup.+] equiv. to 2 mol S) for comparison with [S.sub.POS] or [S.sub.CR] res-[S.sub.CR] Sulfide measured by iodometric titration after acidic chromous chloride reduction of washed peroxide digested soil residue res-[S.sub.HCl] S in 1:40 16 h 4 M HCl extract of washed (or [S.sub.JR]) peroxide digested soil residue [Ca.sub.A] Ca solubilised by acid generated from peroxide digest ([Ca.sub.P] - [Ca.sub.KCl]) [Ca.sub.KCl] Ca in 1 M KCl (after TAA titration) [Ca.sub.HCl] Ca after 1:40 16 h 4 M HCl extract [Ca.sub.P] Ca after peroxide digest and TPA titration [Mg.sub.A] Mg solubilised by acid generated from peroxide digest ([Mg.sub.P] - [Mg.sub.KCl]) [Mg.sub.KCl] Mg in 1 M KCl (after TAA titration) [Mg.sub.HCl] Mg after 1:40 16 h 4 M HCl extract [Mg.sub.P] Mg after peroxide digest and TPA titration Table 1. Soil information and properties Cation data are mean values of four replicates. Carbon data are mean values of two replicates, pH/EC/[Cl.sup.-] data are single determinations Name A B C Byron Bay Kerkin's Levee Oyster Cove Sand jarosite Lock site Location Belongil, NSW Pimpama, Qld Hope Island, Qld Texture Sand Sandy loam Loamy sand Depth 1-1.1 m Surface 1.9-2.0 m (disturbed) ASC (A) Oxyaquic Spolic Sulfidic Hydrosol Anthroposol Oxyaquic Hydrosol Chemical properties E[C.sub.w] 1:5 (mS/cm) 0.22 2.24 2.04 p[H.sub.w] 1:5 4.13 3.36 3.98 [Cl.sup.-] (mg/kg) 98 744 1044 Total C (Leco) (%) 0.15 3.71 0.73 OC (Leco) (%) 0.15 3.70 0.73 Ca ([cmol.sub.c]/kg) 0.11 2.15 2.48 Mg ([cmol.sub.c]/kg) 0.34 4.37 6.75 Na ([cmol.sub.c]/kg) 0.35 3.21 2.83 K ([cmol.sub.c]/kg) 0.02 1.06 0.58 Name D E F Dux Creek Carrara Mischke's Bribie Is Speedway farm site 2074 Location Bribie Carrara, Qld Rocky Point, Island, Qld Qld Texture Silty loam Silty clay loam Silty light clay Depth 0.2-0.5 m 1.5 m 1.8-2.0 m ASC (A) Intertidal Sulfidic Sulfuric Sulfidic Oxyaquic Redoxic Hydrosol Hydrosol Hydrosol Chemical properties E[C.sub.w] 1:5 (mS/cm) 8.37 3.41 2.72 p[H.sub.w] 1:5 4.03 4.58 4.48 [Cl.sup.-] (mg/kg) 10289 129 1202 Total C (Leco) (%) 1.85 2.33 1.22 OC (Leco) (%) 1.83 2.11 1.21 Ca ([cmol.sub.c]/kg) 2.94 24.72 7.53 Mg ([cmol.sub.c]/kg) 10.53 14.25 15.44 Na ([cmol.sub.c]/kg) 25.68 2.72 9.26 K ([cmol.sub.c]/kg) 0.78 0.98 2.64 Name G H I Tuckean Swamp Redlands EAS Maroochydore site 144 SEA site 2295 Location Tuckean Swamp, Native Dog Eudlo Creek, NSW Creek, Qld Qld Texture Light clay Fibric silty Peat light clay Depth 1.0-1.4 m 1.3-1.5 m 3.5-3.7 m ASC (A) -- Peaty Oxyaquic Peaty Hydrosol Oxyaquic Hydrosol Chemical properties E[C.sub.w] 1:5 (mS/cm) n.d. 6.10 4.16 p[H.sub.w] 1:5 n.d. 3.98 3.17 [Cl.sup.-] (mg/kg) n.d. 6191 39 Total C (Leco) (%) n.d. 4.20 10.71 OC (Leco) (%) n.d. 4.21 10.71 Ca ([cmol.sub.c]/kg) n.d. 6.49 28.44 Mg ([cmol.sub.c]/kg) n.d. 21.95 22.78 Na ([cmol.sub.c]/kg) n.d. 34.90 2.71 K ([cmol.sub.c]/kg) n.d. 4.29 1.79 (A) Isbell (1996). n.d., Not determined due to insufficient sample. Table 2. Measured or calculated sulfur results for the soils tested using the improved peroxide digest Data in Lines 3-8 are means (n = 4); data in parentheses are standard deviations of the mean. Data in Lines 1 and 2 are single determinations. Data in Lines 8 and 9 are means of duplicate determinations A B C Byron Bay Kerkin's Oyster Cove Sand Levee jarosite Lock site Measurements on the soil (%S) Line 1. Total S HF Digest 0.079 1.72 0.905 ([S.sub.T]) Line 2. Total S Leco 0.056 1.725 0.895 ([S.sub.T]) Line 3. Chromium reducible S 0.032 0.145 0.520 ([S.sub.CR]) (0.0005) (0.004) (0.01) Line 4. 4 M HCl extractable S 0.050 1.506 0.363 ([S.sub.HCl]) (0.001) (0.003) (0.004) Line 5. 1 M KCl extractable S 0.017 0.315 0.184 ([S.sub.KCl]) (0.0005) (0.001) (0.001) Line 6. Peroxide S 0.064 0.546 0.694 ([S.sub.P]) (0.001) (0.003) (0.009) Line 7. Peroxide oxidisable S 0.047 0.231 0.510 ([S.sub.POS] = [S.sub.P] - [S.sub.KCl]) Line 8. `Insoluble' S 0.033 1.191 0.179 ([S.sub.J] = [S.sub.HCl] - [S.sub.KCl]) Measurements on digested residue (%S) Line 9. Residual [S.sub.CR] 0.001 0.002 0.002 (res-[S.sub.CR]) Line 10. Residual [S.sub.HCl] 0.013 1.180 0.256 (res-[S.sub.HCl] or [S.sub.JR]) Summation (%S) Line 11. [S.sub.P] + res 0.078 1.729 0.952 [S.sub.HCl] + res [S.sub.CR] D E F Dux Creek Carrara Mischke's Speedway farm site 2074 Measurements on the soil (%S) Line 1. Total S HF Digest 1.27 2.37 3.23 ([S.sub.T]) Line 2. Total S Leco 1.173 2.289 3.192 ([S.sub.T]) Line 3. Chromium reducible S 0.856 1.816 2.854 ([S.sub.CR]) (0.009) (0.005) (0.015) Line 4. 4 M HCl extractable S 0.354 0.520 0.255 ([S.sub.HCl]) (0.005) (0.007) (0.002) Line 5. 1 M KCl extractable S 0.243 0.521 0.262 ([S.sub.KCl]) (0.001) (0.004) (0.003) Line 6. Peroxide S 1.142 2.368 3.096 ([S.sub.P]) (0.008) (0.024) (0.049) Line 7. Peroxide oxidisable S 0.900 1.847 2.834 ([S.sub.POS] = [S.sub.P] - [S.sub.KCl]) Line 8. `Insoluble' S 0.111 0 0 ([S.sub.J] = [S.sub.HCl] - [S.sub.KCl]) Measurements on digested residue (%S) Line 9. Residual [S.sub.CR] 0.003 0.002 0 (res-[S.sub.CR]) Line 10. Residual [S.sub.HCl] 0.173 0.137 0.138 (res-[S.sub.HCl] or [S.sub.JR]) Summation (%S) Line 11. [S.sub.P] + res 1.318 2.508 3.235 [S.sub.HCl] + res [S.sub.CR] G H I Tuckean Redlands Maroochydore Swamp EAS site 144 SEA site 2295 Measurements on the soil (%S) Line 1. Total S HF Digest n.d. 4.07 7.29 ([S.sub.T]) Line 2. Total S Leco 4.817 4.226 7.384 ([S.sub.T]) Line 3. Chromium reducible S 2.996 3.275 5.52 ([S.sub.CR]) (0.037) (0.035) (0.13) Line 4. 4 M HCl extractable S 1.741 0.574 1.415 ([S.sub.HCl]) (0.007) (0.002) (0.002) Line 5. 1 M KCl extractable S 1.312 0.609 1.318 ([S.sub.KCl]) (0.012) (0.003) (0.001) Line 6. Peroxide S 4.244 3.801 7.443 ([S.sub.P]) (0.023) (0.037) (0.116) Line 7. Peroxide oxidisable S 2.932 3.192 6.126 ([S.sub.POS] = [S.sub.P] - [S.sub.KCl]) Line 8. `Insoluble' S 0.429 0 0.097 ([S.sub.J] = [S.sub.HCl] - [S.sub.KCl]) Measurements on digested residue (%S) Line 9. Residual [S.sub.CR] 0.229 (A) 0.005 0.030 (res-[S.sub.CR]) Line 10. Residual [S.sub.HCl] 0.583 0.455 0.262 (res-[S.sub.HCl] or [S.sub.JR]) Summation (%S) Line 11. [S.sub.P] + res 5.055 4.262 7.735 [S.sub.HCl] + res [S.sub.CR] (A) Sample digested and re-analysed after ring-mill grinding gave duplicate mean of 0.008 %S. n.d., Not determined due to insufficient sample.
Funding support from the Natural Heritage Trust is gratefully acknowledged. The first author is thankful for financial support provided by an Australian Postgraduate Award scholarship. The authors would also like to acknowledge the assistance and facilities provided by the Analytical Centre at the Natural Resource Sciences Laboratories (Qld Government) at Indooroopilly. In particular the assistance given by Niki Latham, Tony Kelly, Ian Grant, Kate Dolan, David Lyons, and Dennis Baker was greatly appreciated. Steve Dobos made valuable comments and suggestions regarding the manuscript. We are grateful to Nicholas Ward and Leigh Sullivan for providing us with the Tuckean Swamp soil and also for making available to us an advance copy of their recent paper on peroxide oxidation methods.
Ahern CR, McElnea AE, Hicks W, Baker DE (1996) A combined routine peroxide method for analysing acid sulfate soils. In `Proceedings 2nd National Conference of Acid Sulfate Soils'. pp. 82-87. (Robert J Smith and Associates and ASSMAC: Australia)
Ahern CR, McElnea AE, Latham NP, Denny SL (2000) A modified acid sulfate soil method for comparing net acid generation and potential sulfide oxidation--POCASm. In `Acid sulfate soils: Environmental issues, assessment and management, Technical Papers'. (Eds CR Ahern, KM Hey, KM Watling, VJ Eldershaw) pp. 21/1-21/12. (Department of Natural Resources: Brisbane)
Begheijn LTh, van Breemen N, Velthorst EJ (1978) Analysis of sulfur compounds in acid sulphate soils and other recent marine soils. Communications in Soil Science and Plant Analysis 9, 873-882.
Bush RT, Sullivan LA (1997) Morphology and behaviour of greigite from a Holocene sediment in eastern Australia. Australian Journal of Soil Research 35, 853-861.
Caruccio FT (1975) Estimating the acid potential of coal mine refuse. In `The ecology of resource degradation and renewal'. (Eds MJ Chadwick, GT Goodman) pp. 197-205. (Blackwell Scientific Publications: London)
Das GK, Acharya S, Anand S, Das RP (1996) Jarosites: A review. Mineral Processing and Extractive Metallurgy Review 16, 185-210.
Dent D, Bowman G (1996) Quick, quantitative assessment of the acid sulphate hazard. CSIRO Division of Soils, Divisional Report No. 128, Canberra.
Evangelou VP, Zhang YL (1995) A review: pyrite oxidation mechanisms and acid mine drainage prevention. Critical Reviews in Environmental Science and Technology 25, 141-199.
Fanning DS, Rabenhorst MC, Bigham JM (1993) Colors of acid sulfate soils. In `Soil color'. SSSA special publication No. 31. (Eds JM Bigham, EJ Ciolkosz) pp. 91-108. (Soil Science Society of America, Inc.: Madison, WI)
Isbell RF (1996) `The Australian soil classification.' (CSIRO Publishing: Melbourne)
Konsten CJM, Brinkman R, Andriesse W (1988) A field laboratory method to determine total potential and actual acidity in acid sulphate soils. In `Selected papers of the Dakar Symposium on Acid Sulphate Soils'. Dakar, Senegal, January 1986. ILRI Publication 44. (Ed. H Dost) pp. 106-134. (ILRI: Wageningen, The Netherlands)
Latham NP, Denny SL, McElnea AE, Ahern CR (2000) Experimental evidence for modifications to the peroxide digest of POCAS. In `Acid sulfate soils: Environmental issues, assessment and management'. Technical Papers. (Eds CR Ahern, KM Hey, KM Watling, VJ Eldershaw) pp. 22/1-22/9. (Department of Natural Resources: Brisbane)
Lin C, Melville MD (1993) Control of soil acidification by fluvial sedimentation in an estuarine floodplain, eastern Australia. Sedimentary Geology 85, 271-284.
McElnea AE, Ahern CR (2000) Issues associated with laboratory methods for assessing acid sulfate soil risk--particularly hydrogen peroxide-based methods. In `Proceedings of Workshop on Remediation and Assessment of Broadacre Acid Sulfate Soils'. (Ed. P Slavich) pp. 1-15. (ASSMAC, Southern Cross University: NSW)
McElnea AE, Baker DE (1998) Solubility of gypsum in aqueous 1.0 M KC1 at various pH values--implications for acid sulfate soil (ASS) analysis. In `Proceedings of the ASSSI National Soils Conference'. Brisbane, Australia. (Ed. P Mulvey) pp. 540-543. (Australian Society of Soil Science Inc.: Sydney)
Metson AJ, Gibson E J, Cox JE, Gibbs DB (1977) The problem of acid sulphate soils, with examples from North Auckland, New Zealand. New Zealand Journal of Science 20, 371-394.
Nelson DW, Sommers LE (1982) Total carbon, organic carbon, and organic matter. In `Methods of soil analysis Part 2: Chemical and microbiological properties'. 2nd edn (Ed. AL Page) pp. 539-579. (American Society of Agronomy Inc. and Soil Science Society of America Inc.: Madison, WI)
Nicholson RV, Gillham RW, Reardon EJ (1988) Pyrite oxidation in carbonate-buffered solution: 1. Experimental kinetics. Geochimica et Cosmochimica Acta 52, 1077-1085.
Rayment GE, Higginson FR (1992) `Australian laboratory handbook of soil and water chemical methods.' (Inkata Press: Melbourne)
Rice CA, Tuttle ML, Reynolds RL (1993) The analysis of forms of sulfur in ancient sediments and sedimentary rocks: comments and cautions. Chemical Geology 107, 83-95.
Stahl RS, Fanning DS, James BR (1993) Goethite and jarosite precipitation from ferrous sulfate solutions. Soil Science Society of America Journal 57, 280-282.
Sullivan LA, Bush RT, McConchie D, Lancaster G, Clark MW, Lin C, Saenger P (2000) Chromium reducible sulfur--Method 22B. In `Acid sulfate soils: Environmental issues, assessment and management'. Technical Papers. (Eds CR Ahern, KM Hey, KM Watling, VJ Eldershaw) pp. 23/1-23/5. (Department of Natural Resources: Brisbane)
Sullivan LA, Bush RT, McConchie D, Lancaster G, Haskins PG, Clark MW (1999) Comparison of peroxide oxidisable sulfur and chromium reducible sulfur methods for determination of reduced inorganic sulfur in soil. Australian Journal of Soil Research 37, 255-265.
Wang MK, Houng KH, Lin YW, Lim KK (1988) Effect of pH, potassium, sulfate and iron concentration on the formation of jarosite and goethite. Journal of the Chinese Agricultural Chemical Society 26, 117-129.
Ward NJ, Sullivan LA, Bush RT and Lin C (2002) Assessment of peroxide oxidation for acid sulfate soil analysis. 1. Reduced inorganic sulfur. Australian Journal of Soil Research 40, 433-442.
Zhang YL, Evangelou VP (1998) Formation of ferric hydroxide-silica coatings on pyrite and its oxidation behavior. Soil Science 163, 53-62.
Manuscript received 13 November 2001, accepted 19 April 2002
Angus E. McElnea (AB), Col R. Ahern (A), and Neal W. Menzies (B)
(A) Queensland Department of Natural Resources and Mines (QASSIT), Indooroopilly, Qld 4068, Australia.
(B) School of Land and Food Sciences, University of Queensland, St Lucia, Qld 4072, Australia.