Continuous-flow fractionation of trace metals in environmental solids using rotating coiled columns. Some kinetic aspects and applicability of three-step BCR leaching schemes

Abstract

The applicability of the three-step BCR leaching scheme to the continuous-flow fractionation of trace metals (TM) using rotating coiled columns (RCC) has been investigated taking soil and sediment reference samples (SRM-2710, CRM-601, BCR-701) as examples. A particulate sample was retained in the rotating column as the stationary phase under the action of centrifugal forces while different eluents, used according to the original and optimised BCR protocols, were continuously pumped through. The whole procedure required 3–4 h instead of at least 50 h needed for the traditional sequential extraction. It has been shown that in comparison with batch sequential extraction procedures (SEP), the recoveries of Cd, Zn, Cu, and Pb at the first stage (most mobile and potentially dangerous acid soluble forms) are somewhat higher, if a dynamic extraction in RCC is used. Nevertheless, the distribution patterns for TM in the first two leachable fractions (acid soluble and reducible forms) are similar in most cases. Since no heating is used in RCC, the recoveries of TM at the third stage (when hydrogen peroxide is applied to oxidize the organic matter) may be incomplete and matrix-dependent. The effect of eluent volume and flow rate on the recovery of TM in different forms has been investigated. It has been shown that the kinetics of heavy metal leaching vary significantly with samples. Hence, investigating the elution profiles can provide additional important information for risk assessment of TM mobilization.

---

Introduction

Risk assessment related to trace metals (TM) in environmental solids is of special importance for ecology, agriculture and environmental management. In the analysis of soils and sediments, batch sequential extraction procedures are traditionally used for the fractionation of TM according to their mobility and bioavailability.^1–4 The nominal “forms” determined by operational fractionation can help to estimate the amounts of TM in different reservoirs which could be mobilized under changes in chemical properties of soil.^4,5 However, the traditional sequential extraction procedures (SEP) are rather laborious and require at least a few days because in many cases the kinetics of TM recovery from solid samples can be slow. In addition, all SEP are based on a series of batch extraction experiments while naturally occurring processes are always dynamic.

An interesting continuous-flow extraction system for the fractionation of TM has been recently proposed.^6,7 The extraction was performed in a mixing chamber (10 ml volume) closed by a membrane filter while reactants were subsequently passed through. Apart from the simplicity, the system has many other advantages. Disadvantages of the system include the fact that the eluent in contact with the particulate matter is only partially renewed with time and that the flow rate is not stable throughout the experiment.

Other techniques employ microcolumns⁸ or microcartridges⁹ filled with dried solid samples. It becomes possible to perform the continuous-flow leaching with on-line determination of elements by inductively coupled plasma mass spectrometry. The proposed method looks very attractive, however, its sensitivity may be somewhat limited by the sample weight (10–80 mg). In contrast to micro techniques, very big columns have been used in hydrogeochemical studies.¹⁰ 10 kg of sample were treated to determine the differences between the binding forms of iron and trace metals in sediments with and without the addition of alkaline substances.¹¹

In previous work¹² we have proposed a new approach to performing an accelerated sequential extraction of TM from environmental solid samples. It has been shown that rotating coiled columns (RCC) earlier used mainly in counter-current chromatography can be successfully applied to the continuous-flow leaching of heavy metals from soils and sediments. A solid sample was retained in the rotating column as the stationary phase under the action of centrifugal forces while different eluents (used according to the Kersten–Foerstner¹³ or McLaren–Crawford¹⁴ protocols) were continuously pumped through. The developed procedure is time saving and required only 4–5 h, with complete automation being possible. Losses of solid sample were minimal. In most cases a dynamic extraction in RCC provided higher (compared with SEP) recoveries of readily bioavailable and leachable forms of Pb, Zn, and Cd. The Kersten–Foerstner and McLaren–Crawford leaching schemes have been correlated, the former has been found to be preferable.

In the present study the applicability of the three-step BCR scheme to the dynamic leaching of TM using RCC will be studied. The extractable contents of TM in reference materials CRM-601 and BCR-701 were certified especially for this scheme^15–20 adopted by the Standards, Measurements and Testing Program of the European Commission (formerly the BCR Program). Hence, investigating CRM-BCR samples is of special importance for evaluation of the proposed continuous-flow fractionation technique. In addition, some kinetic aspects of continuous-flow leaching in RCC will be considered.

Materials and methods

Samples and reagents

Montana Soil No. 1 (SRM-2710, NIST), reference samples of lake sediment (CRM-601, BCR-701), and real sludge samples taken after a flood in Germany (A-72651, A-72660) were under study. The total element contents in the samples CRM-601, A-72651, and A-72660 were determined by energy-dispersive X-ray fluorescence spectroscopy (XRF). As is shown (Table 1), most samples are characterised by elevated trace element concentrations. The sludge samples were passed through a 90 μm sieve. Since small sample particles mainly consisting of silt and clay may be more “mobile” under natural conditions, the sub-fractions of sludge less than 90 μm were chosen for leaching experiments (total concentrations of elements as XRF values are given in Table 1 for these sub-fractions).

Table 1 Total concentrations of some elements in the samples under investigation

Sample	Concentration/mg kg^−1
	Cd	Cr	Cu	Ni	Pb	Zn
a Certified values. b XRF values. c Indicative values aqua regia extractable.
Soil (SRM-2710)^a	21.8 ± 0.2	(39)	2950 ± 130	14.3 ± 1	5532 ± 80	6952 ± 91
Soil (SRM-2710)^b	22	37	2900	16	5630	6840
Sediment (CRM-601)^b	10.7	180	237	76.5	287	825
Sediment (CRM-601)^c	11.5 ± 1.9	112 ± 10	230 ± 15	78.8 ± 6.7	288 ± 52	833 ± 17
Sediment (BCR-701)^b	12.1	329	295	102	165	500
Sediment (BCR-701)^c	11.7 ± 0.1	272 ± 20	275 ± 13	103 ± 4.0	143 ± 6.0	454 ± 19
Sludge (A72660)^b	41	103	402	50	3529	2586
Sludge (A72651)^b	4	62	167	39	116	257

---

All chemicals used were analytical grade reagents.

Continuous-flow fractionation of TM using RCC

The continuous leaching of heavy metals was performed on a planetary centrifuge (Fig. 1) with a vertical one-layer coiled column drum developed at the Institute of Analytical Instrumentation (St. Petersburg, Russia) and fabricated at the Institute of Spectrochemistry and Applied Spectroscopy (Dortmund, Germany). The planetary centrifuge has a revolution radius R = 140 mm and a rotation radius r = 50 mm. The β value (β = r/R) is 0.36. The two axes of the instrument are parallel. The column was made from a PTFE tube with an inner diameter of 1.5 mm and the total inner capacity of the column was 20 ml (consequently, the tube length was about 10 m). The rotation and revolution speeds (ω) were 600 rpm. The mobile phase pumping rate (F) was 1.0 or 0.5 ml min⁻¹.

Fig. 1 Planetary centrifuge with a vertical column drum (also used in the dynamic extraction of PAH from soils²³).

Before commencing the leaching procedure, the spiral column was filled with water, after which the solid sample (about 0.5 g) was introduced into the column (in the stationary mode) as a suspension in 10 ml of 0.11 mol l⁻¹ acetic acid. Then, while the column was rotated, aqueous solutions of different reagents, used as the mobile phase, were continuously fed to the column inlet. The solid sample was retained inside the rotating column as the stationary phase by the action of centrifugal force throughout the experiment. The recovery of TM in different forms was achieved by successively changing the eluents used according to the original¹⁵ and optimised¹⁷ BCR protocols (Table 2). After each leaching step 10 ml of water were passed through the column. Fractions (10 ml each) of the mobile phase (effluent) were collected. The continuous extraction in RCC required 3–4 h. After this leaching procedure was finished, the residue of the solid sample was removed from the column for subsequent analysis.

Table 2 Original and optimised BCR leaching schemes: applied reactants and corresponding fractions

Step	Reagent	Fraction	Mobility
a Original scheme.^15 b Optimised scheme.^17
1 ^ab	0.11 mol l^−1 CH3COOH (16 h)	Acetic acid soluble	Mobile and easily mobilizable
2 ^a	0.1 mol l^−1 NH2OH·HCl	Reducible	Readily and poorly mobilizable
	+ HNO3 to adjust pH 2 (16 h)
2 ^b	0.5 mol l^−1 NH2OH·HCl	Reducible	Readily and poorly mobilizable
	+ HNO3 to adjust pH 1.5 (16 h)
3(1) ^ab	8.8 mol l^−1 H2O2 (2 × 1 h, 85 °C)
3(2) ^ab	1.0 mol l ^−1 CH3COONH4	Oxidizable	Poorly mobilizable
	+ HNO3 to adjust pH 2 (16 h)

---

Analysis of separated leachable fractions

The concentrations of elements in the separated fractions (10 ml each) were determined by ICP-AES and ICP-MS using the CIROS (Spectro A.I.) and ELAN 5000 (Perkin Elmer) instruments, respectively. The limits of determination for Cd, Cr, Cu, Ni, Pb, Zn were 0.04; 0.03; 0.05; 0.06; 0.06; 0.04 mg l⁻¹ (ICP-AES) and 0.1; 0.13; 0.18; 0.17; 0.08; 0.3 μg l⁻¹ (ICP-MS), respectively. The measurements were performed in undiluted and diluted samples because of the wide range of concentrations of the different analytes.

Analysis of residual fractions

The residue fixed on the membrane filter (0.45 μm; cellulose acetate, Sartorius) was partly dissolved employing a closed microwave digestion device (Multiwave, Perkin-Elmer) in aqua regia. Briefly, 1.2 ml of nitric acid (65% v/v) and 3.6 ml of hydrochloric acid (30% v/v), both suprapur quality (Merck), were added to the residue. The samples were heated to 220 °C within 20 min and held at this temperature for an additional 20 min. After cooling down, the resulting solution was separated by centrifugation. The supernatant was transferred to 25 ml polyethylene bottles and filled to volume with deionised water. The resulting solution was analysed using ICP-MS.

Results and discussion

Fractionation of TM in the soil sample SRM-2710 on the basis of the original BCR scheme

Since the soil SRM-2710 has already been studied by using SEP²¹ and continuous-flow extraction in a mixing chamber,⁶ it is of special interest to test this sample for the evaluation of the fractionation procedure in RCC. The experimental and published data are summarised in Table 3. As is seen, in comparison with batch methods, the continuous-flow extraction in RCC results in significantly higher recoveries of Cd, Cu, and Zn at step 1. The extractable contents of Pb are similar. As has been shown (Fig. 2), the kinetics of Pb leaching in RCC at step 1 are considerably slower than those for Cd, Cu, and Zn. This may explain the similar batch and dynamic data for Pb (contact times between the sample and extractant are 16 and 1 h, respectively). In general, despite the slow kinetics of TM recovery from solid samples, the use of multistage continuous extraction in RCC allowed us to reduce the contact time needed for the separation of exchangeable and acid soluble forms (fractions) of TM from 16 h down to 1 h. Taking into consideration that the flow rate is 1 ml min⁻¹ this corresponds to 60 ml of eluent pumped through the column. With regard to the continuous-flow extraction in the mixing chamber, data on Cd and Pb are not reported.⁶ The recoveries of Cu and Zn are somewhat higher in comparison with RCC, however, 150 ml of extractant were used for the recovery of each form in the mixing chamber. Hence, in RCC the extractability of Cu and Zn at step 1 may be increased by increasing the eluent volume.

Fig. 2 RCC: dynamics of heavy metals leaching from the soil sample SRM-2710 on the basis of the original BCR scheme. Eluent flow rate: 1 ml min⁻¹. Eluent composition: 1, 2, 3(1), 3(2)—see Table 2; 1′, 2′, 3′—water.

Table 3 Results of the fractionation of TM (mass fraction, mg kg⁻¹) in the Montana soil sample (SRM-2710) on the basis of the original BCR scheme. Comparison of SEP, continuous-flow extraction in a mixing chamber, and continuous-flow extraction in RCC

	Procedure	Step 1	Step 2	Step 3	Residue	Sum^a	Recovery^e
a Sum of residual and three leachable forms. b Step 2 at 65 °C, step 3 at 85 °C. c Step 2 at 65 °C, step 3 at 55 °C, at each step 150 ml of extractant were pumped for recovery of elements, flow rate is not stable (3–5 ml min^−1). d All steps at room temperature, at each step 60 ml of extractant (eluent) were pumped for recovery of elements, flow rate is 1 ml min^−1. e Recovery based on the total concentration.
Cd	SEP (Ho et al.^21)	11.7 ± 0.4	5.1 ± 0.2	1.3 ± 0.2	<1	18.2 ± 0.5	83%
	RCC	15.2 ± 0.3	2.4 ± 0.1	<1	3.5 ± 0.5	21.1 ± 0.6	97%
Cu	SEP (Ho et al.^21)	991 ± 54	850 ± 22	388 ± 23	427 ± 97	2660 ± 130	90%
	SEP^b (Shiowatana et al.^6)	850 ± 160	800 ± 160	740 ± 380	440 ± 100	2830 ± 230	96%
	Mixing chamber^c (Shiowatana et al.^6)	1580 ± 180	980 ± 100	130 ± 20	250 ± 80	2940 ± 110	100%
	RCC^d	1323 ± 23	980 ± 18	288 ± 12	389 ± 70	2980 ± 80	101%
Pb	SEP (Ho et al.^21)	773 ± 79	324 ± 25	2540 ± 130	950 ± 170	4950 ± 170	83%
	RCC^d	778 ± 18	3950 ± 90	321 ± 11	430 ± 90	5480 ± 90	99%
Zn	SEP (Ho et al.^21)	1211 ± 61	1120 ± 34	519 ± 42	3610 ± 350	6460 ± 350	93%
	SEP^b (Shiowatana et al.^6)	990 ± 160	2100 ± 50	520 ± 80	2980 ± 220	6550 ± 290	94%
	Mixing chamber^c (Shiowatana et al.^6)	1630 ± 60	2390 ± 260	1380 ± 320	2440 ± 420	7840 ± 300	113%
	RCC^d	1555 ± 56	767 ± 17	267 ± 12	3700 ± 150	6290 ± 150	90%

---

The data on the reducible forms of TM obtained by different methods can only be easily correlated for Cu. For Cd, the lower recovery using RCC in comparison with SEP can be explained by a “pre-extraction” of this element at step 1. The sums of the first two leachable fractions obtained by both methods are similar. For Pb and Zn, the data are poorly comparable. The recovery of Pb in reducible form is dramatically lower in the case of SEP, though both procedures were performed at room temperature for step 2. This cannot be related simply to the difference in batch and multistage continuous extraction. Another assumption may be made. The extraction of Pb by hydroxylamine is known to be very pH sensitive.¹⁷ The continuously renewed eluent in RCC provides a constant pH value (at the inlet of the column) for the recovery of TM at step 2 whereas under batch conditions the pH value of the reaction mixture may change throughout the extraction decreasing the extractability of Pb. For Zn, if the batch and continuous extraction are performed at room temperature, the overall recoveries in the first two leachable fractions are similar. Increasing the temperature at step 2 leads to significantly higher recoveries of Zn both under batch and continuous-flow conditions (mixing chamber). Since at step 2 amorphous and crystalline Fe oxides are supposed to be dissolved, this could be related to peculiarities of Fe-mineralogy and the type of Fe–Zn binding in the sample under study.

The difference in TM recoveries at step 3 (oxidizable form) can be observed first of all as a consequence of different “pre-extraction” at steps 1 and 2. Besides, no heating was used in the rotating column though a high temperature (85 °C) is normally required for oxidation of the organic matter in soil by hydrogen peroxide. Without heating, the dissolution of trace metals in oxidizable form may be dependent on the nature and type of the sample under study.

In general, it can be concluded that SEP under batch conditions may lead to the underestimation of extractable TM at step 1. Since metals in acid soluble form are considered to represent an easily mobilizable/bioavailable pool of TM,¹ potential danger of contamination may also be underestimated.

If steps 1 and 2 are performed at room temperature, the overall recoveries of TM, except Pb, in the first two leachable fractions are similar for SEP and continuous-flow extraction. In order to evaluate, in detail, the peculiarities of batch and continuous leaching, other samples should be studied.

Fractionation of TM in the sediment sample CRM-601 on the basis of the original and modified BCR schemes

The data on the fractionation of TM in the sample CRM-601 obtained by batch SEP¹⁵ and continuous-flow fractionation in RCC are presented in Table 4 and Fig. 3. As shown in Fig. 3, the extractions in RCC and SEP (according to the BCR original protocol¹⁵) result in different patterns of element recoveries. The recoveries of Cd, Cu, Pb, and Zn into the first fraction (most bioavailable forms of TM) are considerably higher if a dynamic multistage extraction in RCC is used. The recovery of Cu is four times higher than the certified value obtained by batch SEP (38.7 and 10.2 mg kg⁻¹, respectively). In step 2 the recoveries of Cd and Zn in RCC are somewhat lower apparently due to “pre-extraction” at the previous stage. The sums of the first two leachable fractions for these elements are comparable. For Cu and Pb in reducible form, the data obtained by the batch SEP are dramatically lower. As has been discussed above, this may be strongly related to the pH sensitivity of TM extraction by hydroxylamine. The recoveries of all TM under investigation in oxidizable form (step 3) are somewhat lower for RCC. This difference may be observed either due to some “pre-extraction” of TM in the preceding two steps, or due to partial oxidation at room temperature.

Fig. 3 Fractionation of trace metals (recoveries, % of total content) in the sediment sample CRM-601 on the basis of original BCR scheme. Continuous-flow extraction in RCC and batch extraction SEP. Recoveries based on indicative values for aqua regia extraction.

Table 4 Data on the extractable contents of TM (mass fraction, mg kg⁻¹) in the reference sediment sample CRM-601. Comparison of the batch extraction¹⁷ and dynamic fractionation in RCC on the basis of the original and modified BCR schemes (for RCC n = 3 and α = 0.05)

	Procedure	Cd	Cr	Cu	Ni	Pb	Zn
CRM-601: grain size < 90 µm a Original scheme.b Optimised scheme.c n. d. = not detectable.
Step 1^a	SEP	4.46 ± 0.63	0.35 ± 0.09	10.2 ± 0.8	8.22 ± 0.83	2.07 ± 0.49	259 ± 13
	RCC	7.76 ± 0.47	n.d.^c	38.7 ± 0.6	6.70 ± 1.12	10.5 ± 1.1	333 ± 4
Step 2^a	SEP	3.09 ± 0.88	1.42 ± 0.79	7.87 ± 5.14	5.55 ± 1.42	37.3 ± 17.7	175 ± 15
	RCC	2.21 ± 0.65	12.9 ± 3.6	104 ± 18	8.34 ± 0.94	134 ± 12	122 ± 21
Step 3^a	SEP	2.01 ± 1.22	20.1 ± 2.0	116 ± 9	6.75 ± 0.82	108 ± 18	124 ± 17
	RCC	0.83 ± 0.18	7.83 ± 1.16	92.8 ± 5.3	5.14 ± 0.67	71.3 ± 6.4	76.5 ± 8.2
Step 1^b	SEP	4.45 ± 0.67	0.35 ± 0.08	10.5 ± 0.8	7.82 ± 0.84	2.28 ± 1.17	261 ± 13
	RCC	7.81 ± 0.45	n.d.^c	38.8 ± 0.7	6.70 ± 1.12	10.7 ± 0.9	334 ± 3
Step 2^b	SEP	3.95 ± 0.53	10.6 ± 0.9	72.8 ± 4.8	10.6 ± 0.3	205 ± 11	266 ± 17
	RCC	1.67 ± 0.38	8.19 ± 1.13	85.1 ± 6.7	7.83 ± 2.1	187 ± 5	133 ± 26
Step 3^b	SEP	1.91 ± 1.43	14.4 ± 2.6	78.6 ± 8.9	6.04 ± 1.25	19.7 ± 5.8	106 ± 11
	RCC	0.74 ± 0.29	5.58 ± 1.76	58.6 ± 3.5	4.19 ± 0.96	23.7 ± 4.4	78.0 ± 8.3

---

Use of the modified BCR scheme allowed us to eliminate the main contradictions between the data obtained by batch and continuous-flow extraction. As is seen from Table 4 and Fig. 4, at step 2 (reducible form) the recoveries of TM obtained by the use of SEP and dynamic extraction in RCC are principally comparable especially taking into consideration some “pre-extraction” of elements in RCC in the preceding step (step 1 is the same for original and modified protocols). The sum of the first two leachable fractions are only considerably lower for Cu in the case of SEP. In contrast, this value is somewhat higher for Ni and Zn. At step 3 the recoveries of elements obtained by the use of SEP and dynamic extraction in RCC are principally comparable (except for Cr) though the oxidation of organic matter in RCC at room temperature may be incomplete.

Fig. 4 Fractionation of trace metals (recoveries, % of total content) in the sediment sample CRM-601 on the basis of the optimised BCR scheme. Continuous-flow extraction in RCC and batch extraction SEP. Recoveries based on indicative values for aqua regia extraction.

The dynamics of heavy metal leaching from the sediment sample CRM-601 on the basis of the original and optimised BCR schemes is shown in Fig. 5. At step 1 (the same for both schemes) a retarded leaching of Cu is observed. Nevertheless, as has been noted above, the recovery of Cu at this stage within one hour is four times higher than the certified value obtained by batch SEP in 16 h. The elution curves obtained at step 2 clearly illustrate the difference between the original and improved BCR leaching schemes. Increasing the concentration of hydroxylamine hydrochloride and the acidity of the leaching solution leads to more effective and faster extraction of trace metals. The variation in dynamics of element leaching is especially evident for Cu and Pb. In the case of the modified extractant, larger amounts of Cu and Pb are recovered in 30–40 ml of the effluent while employing the originally proposed extractant requires at least 60 ml of the mobile phase to be pumped through the column. If the traditional batch procedure is used, even 16 h of contact time between the sample and leaching solution might be insufficient for complete extraction, since the recovery is dependent on shaking conditions, etc. Hence, the correlation of elution profiles obtained at stage 2 may help to explain why the concentrations of some metals (including Cu) could not be certified according to the original BCR procedure due to the high variability between results obtained by different laboratories.^15–17 The variations in the recoveries of elements at step 3 are evidently observed as a consequence of employing different extractants at step 2.

Fig. 5 RCC: dynamics of heavy metal leaching from the sediment sample CRM-601 on the basis of the original (above) and optimised (below) BCR schemes. Eluent flow rate: 1 ml min⁻¹. Eluent composition: 1, 2, 3(1), 3(2)—see Table 2; 1′, 2′, 3′—water.

Hence, the use of the modified BCR scheme is preferable both for batch and continuous-flow procedures. On one hand, the modified scheme enabled us to markedly improve the reproducibility of SEP results obtained by different laboratories.²⁰ On the other hand, it allowed us to eliminate the main contradictions between batch and dynamic data.

Fractionation of TM in the sediment sample BCR-701 on the basis of the modified BCR scheme

In order to optimise the continuous-flow fractionation procedure, the effect of the eluent flow rate on the dynamic recovery of TM has been studied using the reference sediment sample BCR-701. The sequence of extracting reagents was applied according to the modified BCR scheme.¹⁷ The time of contact between the sample and leaching solution was kept constant (60 min at each step). Consequently, the volumes of eluent pumped at each step were 60 and 30 ml (for the flow rates 1.0 and 0.5 ml min⁻¹, respectively). The results obtained are shown in Fig. 6. As is seen, both continuous-flow leaching experiments result in quite similar patterns of element recoveries, except for Cu. The recovery of Cu at step 1 is somewhat higher, if a flow rate equal to 1.0 ml min⁻¹ is used. This can be explained by the peculiarities of extraction at step 1. In comparison with other TM under investigation, the extraction of Cu by acetic acid is characterised by slower kinetics (Fig. 7). Hence, a higher flow rate and/or a larger eluent volume passed through the column may affect the extraction equilibrium that results in a somewhat higher recovery of Cu.

Fig. 6 Continuous-flow fractionation of trace metals (recoveries, % of total content) in the sediment sample BCR-701 on the basis of the optimised BCR scheme. Effect of the flow rate: 1.0 ml min⁻¹ and 0.5 ml min⁻¹. Recoveries based on indicative values for aqua regia extraction.

Fig. 7 Dynamics of trace metals leaching from the sediment sample BCR-701 on the basis of the optimised BCR scheme. Eluent flow rate: 1 ml min⁻¹. Eluent composition: 1, 2, 3(1), 3(2) – see Table 2; 1′, 2′, 3′ – water.

In general, the data obtained at the two different flow rates are very similar. Since the lower flow rate (0.5 ml min⁻¹) can provide more stable retention of particulate matter in the rotating column, it is reasonable to choose this value for further leaching experiments. It should be noted that the stable retention is especially important at step 3 employing hydrogen peroxide when the oxidation of organic matter may be accompanied by a violent reaction and gas release. Besides, the lower flow rate allows us to reduce the volume of effluent collected at each step. This may be useful if the sample under study is characterised by low contents of extractable TM.

The certified SEP data on the extractable contents of TM in the sample BCR-701 and results of continuous-flow fractionation are presented in Table 5. For the dynamic extraction in RCC, the eluents, used according to the modified BCR scheme, were pumped at a rate of 0.5 ml min⁻¹. The contact time between the sample and leaching solution was 60 min at each step. It should be noted that the reproducibility of data obtained by the use of RCC is very high. Although only three parallel experiments were performed, in all cases the confidence intervals (α = 0.05) are within 4% of the corresponding mean values (Table 5).

Table 5 Data on the extractable contents of TM (mass fraction, mg kg⁻¹) in the reference sediment sample BCR-701. Comparison of the batch extraction (certified data) and dynamic fractionation in RCC on the basis of the modified BCR scheme (for RCC n = 3 and α = 0.05)

		Cd	Cr	Cu	Ni	Pb	Zn
a Data determined by ICP-AES due to interferences in ICP-MS.
Step 1	Certified	7.34 ± 0.35	2.26 ± 0.16	49.3 ± 1.7	15.4 ± 0.9	3.18 ± 0.21	205 ± 6
	RCC	8.24 ± 0.10	1.94 ± 0.08	58.2 ± 0.8	(13.3)^a	3.67 ± 0.10	201 ± 3
Step 2	Certified	3.77 ± 0.28	45.7 ± 2.0	124 ± 3	26.6 ± 1.3	126 ± 3	114 ± 5
	RCC	2.65 ± 0.10	28.7 ± 0.6	119 ± 2	(15.9)^a	117 ± 2	55.6 ± 0.9
Step 3	Certified	0.27 ± 0.06	143 ± 7	55.2 ± 4.0	15.3 ± 0.9	9.3 ± 2.0	45.7 ± 4.0
	RCC	0.30 ± 0.02	36.0 ± 0.4	33.5 ± 0.2	(7.7)^a	8.1 ± 0.2	15.2 ± 0.3

---

At step 1 continuous-flow fractionation and SEP result in quite comparable values of element recoveries, except for Cu. The recovery of Cu, characterised by slow extraction kinetics, is higher in the case of RCC due to the multistage continuous extraction process. At step 2 the use of RCC leads to the lower recoveries of Cr, Ni, and Zn. The difference for Zn is very significant. As can be seen from the elution curves (Fig. 7), the low extraction of Zn in RCC at step 2 cannot be attributed to the extraction kinetics. Hence, further study is needed to explain the data obtained.

As was discussed above, the oxidation of organic matter by hydrogen peroxide at room temperature may be incomplete and matrix-dependent even if a multistage extraction in RCC is used. Nevertheless, the data on Cd and Pb in oxidizable forms are similar for RCC and SEP.

Hence, despite some contradictions between batch and dynamic data, the continuous-flow procedure developed can be successfully applied to the fast and efficient leaching of TM in most bioavailable forms (acetic acid soluble and reducible).

In addition, since the kinetics of heavy metal leaching varies with samples the extractable species of metals may be classified as labile or moderately-labile. For example, acid soluble forms of Zn and Cu in the sample BCR-701 can be considered as labile and moderately-labile, respectively. A similar approach has already been applied to the leaching of TM forms under batch extraction conditions.²²

Analysis of real sludge samples

The developed fractionation procedure was used for the risk assessment related to the flood in Germany in summer 2002. Two sludge samples (A72660 and A72651) taken respectively in the Freiberg region (Mulde River) and the Heidenau region (Elbe River) were under investigation.

The results on the continuous-flow fractionation of TM in the sludge samples are presented in Table 6. The distribution patterns of elements between leachable forms vary from sample to sample. It should be noted that the sludge A72660 is characterised by high concentrations of easily mobilizable Cd, Pb, and Zn extracted at step 1.

Table 6 Data on the continuous-flow fractionation of TM (mass fraction, mg kg⁻¹) in the real sludge samples taken after the inundation in Germany, summer 2002

		Cd	Cu	Pb	Zn
a Bold type indicates highly elevated concentrations. b Data may be underestimated as no heating was used in RCC.
A72660	Step 1	28.0 ± 1.1^a	31.2 ± 1.6	592 ± 12^a	800 ± 18^a
	Step 2	8.36 ± 0.37	228 ± 6^a	2200 ± 27^a	547 ± 31^a
	Step 3	(7.72 ± 0.45)^b	(158 ± 12)^b	(224 ± 8)^b	(573 ± 46)^b
A72651	Step 1	1.12 ± 0.11	20.8 ± 1.4	1.92 ± 0.12	44.3 ± 2.7
	Step 2	0.53 ± 0.08	96.3 ± 3.5^a	88.4 ± 5.7^a	32.8 ± 3.1
	Step 3	(0.12 ± 0.04)^b	(24.8 ± 2.9)^b	(8.35 ± 0.64)^b	(18.4 ± 0.9)^b

---

Conclusions

Taking soil, sediment and sludge samples as examples, the applicability of the three-step BCR leaching schemes to the fast continuous-flow fractionation of TM using RCC has been studied. The original and modified BCR schemes have been tested, the latter has been found to be preferable because it enables elimination of the main contradictions between data obtained by batch and continuous-flow extraction.

It has been shown that in most cases the dynamic leaching results in higher recoveries of extractable TM (compared to the batch SEP) at step 1 (the same for the original and modified protocols). Metals in acetic acid soluble form are considered to represent an easily mobilizable/bioavailable pool of TM, consequently, this step is of particular importance. Since leaching kinetics vary with samples, metals in acid soluble form may be characterised as labile and moderately-labile.

In order to optimise the operational conditions for continuous-flow fractionation, the effect of eluent volume and flow rate on the recovery of TM in different forms was investigated. The procedure developed was used for the risk assessment related to TM in sludge transported during the last inundation in Germany.

Other environmental solids of different type and origin should be further investigated to evaluate general regularities of continuous-flow fractionation of TM on the basis of the modified BCR scheme. Besides, TM can be pre-extracted employing water as eluent to recover elements in water-soluble form.

Acknowledgements

We are indebted to Ines Volkmann (UFZ) for technical assistance.

This work was supported by the German Service of Academic Exchange and the Russian Foundation of Basic Research (grant 04-03-32837).

References

1. U. Foerstner, Int. J. Environ. Anal. Chem., 1993, 51, 5.
2. C. Gleyzes, S. Tellier and M. Astruc, Trends Anal. Chem., 2002, 21, 451.
3. A. V. Filgueiras, I. Lavilla and C. Bendicho, J. Environ. Monit., 2002, 4, 823.
4. C. M. Davidson, A. L. Duncan, D. Littlejohn, A. M. Ure and L. M. Garden, Anal. Chim. Acta, 1998, 363, 45.
5. L. J. Ehlers and R. G. Luthy, Environ. Sci. Technol., 2003, 37, 295A.
6. J. Shiowatana, N. Tantidanai, S. Nookabkaew and D. Nacapricha, J. Environ. Qual., 2001, 30, 1195.
7. R. Chomchoei, J. Shiowatana and P. Pongsakul, Anal. Chim. Acta, 2002, 472, 147.
8. D. Beauchemin, K. Kyser and D. Chipley, Anal. Chem., 2002, 74, 3924.
9. M. Jimoh, W. Frenzel, V. Müller, H. Stephanowitz and E. Hoffmann, Anal. Chem., 2004, 76, 1197.
10. F. Wisotzky and N. Cremer, Environ. Geol., 2003, 44, 799.
11. F. Wisotzky and N. Cremer, Environ. Geol., 2003, 44, 805.
12. P. S. Fedotov, A. G. Zavarzina, B. Ya. Spivakov, R. Wennrich, J. Mattusch, K. de P.C. Titze and V. V. Demin, J. Environ. Monit., 2002, 4, 318.
13. M. Kersten and U. Foerstner, Water Sci. Technol., 1986, 18, 121.
14. R. G. McLaren and D. Crawford, J. Soil Sci., 1973, 24, 172.
15. Ph. Quevauviller, G. Rauret, J. F. Lopez-Sanchez, R. Rubio, A. Ure and H. Muntau, Sci. Total Environ., 1997, 205, 223.
16. J. F. Lopez-Sanchez, A. Sahuquillo, H. D. Fiedler, R. Rubio, G. Rauret, H. Muntau and Ph. Quevauviller, Analyst, 1998, 123, 1675.
17. G. Rauret, J. F. Lopez-Sanchez, A. Sahuquillo, R. Rubio, C. Davidson, A. Ure and Ph. Quevauviller, J. Environ. Monit., 1999, 1, 57.
18. G. Rauret, J. F. Lopez-Sanchez, A. Sahuquillo, E. Barahona, M. Lachica, A. Ure, C. Davidson, A. Gomez, D. Lück, J. Bacon, M. Yli-Halla, H. Muntau and Ph. Quevauviller, J. Environ. Monit., 2000, 2, 228.
19. M. Pueyo, G. Rauret, D. Lück, M. Yli-Halla, H. Muntau, Ph. Quevauviller and J. F. Lopez-Sanchez, J. Environ. Monit., 2001, 3, 243.
20. G. Rauret and J. F. Lopez-Sanchez, Int. J. Environ. Anal. Chem., 2001, 79, 81.
21. M. D. Ho and G. J. Evans, Anal. Commun., 1997, 34, 363.
22. M. J. Gismera, J. Lacal, P. da Silva, R. Garcia, M. T. Sevilla and J. R. Procopio, Environ. Pollut., 2004, 127, 175.
23. P. S. Fedotov, C. Bauer, P. Popp and R. Wennrich, J. Chromatogr., A, 2004, 1023, 305.

---