Nanoscavenger based dispersion preconcentration; sub-micron particulate extractants for analyte collection and enrichment

Abstract

A new approach has been developed for the preconcentration of analytes from solution using nanoscavengers; monodisperse functionalised particles of sub-micron dimensions, that can be suspended as a quasi-stable sol in an aqueous solution, and quantitatively recovered with the analyte by conventional filtration. No external agitation of the sample is required as the particles move naturally through the sample by Brownian motion, convection and sedimentation. By careful choice and control of their particle size and surface chemistries, nanoscavengers can be designed to suit a number of different analytical problems. Surface modification of these nanometre-sized particles, through the grafting of complexing or partitioning functional groups, can produce nanoscavengers having affinities for specific analytes and operating through a wide range of mechanisms from covalent bonding to hydrophobic interaction. The approach is illustrated by the development of an extraction-based preconcentration of metals from solution that employs sub-micron Stöber-type silica spheres, the surfaces of which have been modified using chelating diamine and dithiocarbamate groups. The concept has potentially widespread applicability as it is neither limited to metal extractions, nor to the use of silica-based nanoscavengers. Minimal involvement of organic solvents make nanoscavengers a potentially environmentally benign (“green”) alternative to many conventional solvent extraction techniques.

---

Introduction

The concepts of extraction and preconcentration play important roles in many analytical procedures. Whilst primarily associated with increasing the concentration of an analyte, the approach brings with it a number of associated benefits, ranging from the removal of interfering species to a potentially beneficial change in the matrix composition.¹ Many approaches have been taken to achieve this analyte preconcentration, from the conventional solvent extraction systems that are common to many pesticide analyses, to the built-in preconcentration of metal analytes during the electrodeposition stage of anodic stripping voltammetry.

Solvent extraction and solid phase extraction are arguably the most commonly executed forms of preconcentration and for many years they have dominated approaches to the enrichment of pesticides, drugs and trace metals. There are, however, physical difficulties associated with the use of solvent extraction for the enrichment of large numbers of samples and/or enrichment from large sample volumes. Solvent extraction requires vigorous agitation to ensure complete partition of the analyte between two immiscible phases, and this can only be achieved by the application of significant human or mechanical effort. In addition, there are increasing environmental and cost pressures to replace, or at the very least reduce, the volumes of solvents employed in analytical procedures.

Modified silica has been widely and successfully used for the preconcentration of heavy metals, drugs and pesticides. Following initial investigations of the use of amine-modified Fiberglas² and Controlled Pore Glass³ for the collection of trace elements from solution, Leyden et al.^4–6 were amongst the first to modify silica gel to permit it to be employed as a chelating agent for the preconcentration of trace metals. This they achieved by grafting amine and dithiocarbamate functional groups onto the silica surface for the preconcentration of Zn, Cu and Ni; the final determination of the metals being carried out by X-ray fluorescence spectroscopy. Many other complexing agents have since been attached to silica surfaces, examples of which include 8-hydroxyquinoline,^7,8 3-mercaptopropylsilyl groups,^9,10–12N-propylsalicylaldimine,¹³ pyrogallol¹⁴ and resacetophenone.¹⁵

Over recent years there have been major developments in nanoscience and syntheses of particles with nanometre dimensions have been reported for many substances.^16–21 These particles fall within the colloidal range, exhibiting typical colloidal properties such as the long-term stability of their sols. Monodisperse populations of colloidal silica spheres can be produced by the hydrolysis of tetraalkoxysilanes in ammoniacal alcohol solution²² with the subsequent condensation of monomeric species^23–26 into high density spherical particles. With careful control of synthesis conditions the size of these particles can be controlled over a wide range (ca. 5 to 5000 nm).

There have been few applications to date of nanoparticulate materials being used as extractants. In biochemical applications cationic polystyrene nanoparticles have been employed for the solid phase extraction of oligonucleotides from plasma^27–29 and magnetic nanobeads have been developed for the isolation of DNA.³⁰ Nanoparticulate titania has been recently employed for the preconcentration of zinc and cadmium.³¹ In each of these cases the recovery of the particles has been by centrifugation. The authors can find no published examples of sub-micron chelating particles having been employed in the dispersion preconcentration of trace metals or of methods having been developed that permit the recovery of the sub-micron sized solid phase extractants from large water samples by simple filtration.

To exploit the partitioning of analyte species onto nanometre-sized particles for the purposes of extraction and preconcentration, a fine balance has to be established between particle settling time and the ease with which the particle-associated analyte can be recovered from solution. Whilst centrifugation can be employed to recover extractant particles from solution, this is not a practical approach when employed in the analysis of large volume water samples. There is a small window of opportunity within which the size of a particle is sufficiently small for its settling time to be very long, but over which the particle is large enough to be readily removed from its suspension by simple membrane filtration. This critical size region occurs at approximately 250 nm.

This paper introduces a new concept for the extraction and preconcentration of materials from aqueous solution that is based on the dispersal in the sample of sub-micron chelating silicas (nanoscavengers). Extraction is performed as the particles move naturally through the sample as a result of Brownian motion, convection and sedimentation. Continuous physical agitation is not therefore required. The nature of these particles ensures that they do not settle during the equilibration phase and the particles can be recovered from the sample by filtration. The approach is illustrated by the development of a procedure for the preconcentration of metals from solution using surface-modified silicas, but the concept is equally applicable to many other sub-micron sized materials and analyte classes. The basis of the exemplar methods is a well-established solid phase extraction chemistry based on the grafting of chelating ethylenediamine functional groups onto a silica surface. These diamine reagents have a strong affinity for elements such as copper and their corresponding dithiocarbamates exhibit higher binding constants for the softer metals.

Experimental

Materials

Tetraethoxysilane (98%) and sodium diethyldithiocarbamate were obtained from Aldrich (Poole, UK). The N-[3-(trimethoxysilyl)propyl]ethylenediamine was Dow Corning Z-6094, supplied by Aldrich (Poole, UK). Hydrochloric acid (37% (w/v), Fluka, Gillingham, UK), ammonia solution (SG = 0.88), 2-propanol, ethanol, methanol and carbon disulfide were reagent grade. Toluene was reagent grade, dried by Dean–Stark distillation. Water (>14 MΩ cm) was purified by reverse osmosis followed by deionisation using an Elga Option 4 system. Cellulose nitrate membrane filters (47 mm diameter, pore sizes 0.1 and 0.2 µm) were supplied by Millipore (Billerica, USA). All apparatus was soaked in hydrochloric acid (10% v/v) and rinsed with deionised water before use. Tris buffer solution was prepared by dissolving 12.1 g of tris-hydroxymethylmethylamine in 1 L of water and adjusting the pH to 7.40 using concentrated hydrochloric acid. Stock dithiocarbamate solution was prepared by dissolving 100 mg of sodium diethyldithiocarbamate in deionised water to give a final volume of 100 mL. Standard copper solution containing 1000 µg Cu mL⁻¹ was prepared by dissolving 3.9292 g of copper(II) sulfate pentahydrate in deionised water to give 1 litre of solution.

Synthesis of 250 nanometre silica particles

The preparation of silica spheres by the hydrolysis of alkoxysilanes has been the subject of many previous investigations.^22,32–35 Bead size and the dispersity of the particle population depend strongly on the concentrations of alkoxysilane, ammonia and water. Temperature is also a controlling factor in the synthesis. A number of different approaches were investigated and it was found that the desired ca. 250 nm spherical silica particles could be reproducibly prepared by the following method.

250 mL of methanol and 250 mL ammonium hydroxide (SG = 0.88) were mixed in a 1000 mL conical flask for 5 minutes using a magnetic stirrer. 5 mL of tetraethoxysilane was slowly added, whilst continuing the gentle stirring. The reaction was left to proceed at ca. 20 °C for one hour. The product was then centrifuged for 60 minutes at 3000 rpm and the supernatant was removed. The silica was rinsed with methanol several times, discarding the solvent at each stage after centrifugation. The silica product was dried under vacuum.

Preparation of a chelating diamine nanoscavenger

80 mL of toluene was transferred to a 250 mL round bottom flask fitted with a reflux condenser. 1.04 g of sub-micron silica (dried in an oven at 120 °C overnight) was dispersed in the toluene by magnetic stirring for 15 minutes, and then heated at 90 °C whilst still stirring. After the temperature had stabilized, 4 mL of N-[3-(trimethoxysilyl)propyl]ethylenediamine was added to the flask. The reaction was left to proceed for 4 hours under nitrogen and then allowed to cool. The supernatant was removed after centrifugation and the silica product was then rinsed thoroughly with toluene, and dried overnight under vacuum.

Preparation of a bis-dithiocarbamate nanoscavenger

To a 250 mL three necked round bottomed flask was added 50 mL of deionized water and 1 mL NaOH solution (0.1 M). 2.4 g of the diamine nanoscavenger was added to the reaction flask. Whilst sonicating, 4 mL of 2-propanol was added to the flask. After a further 10 minutes of sonication, 4 mL of carbon disulfide was added and the reaction was left to proceed under nitrogen at 20 °C for 30 min. The product was centrifuged for one hour at 3000 rpm and the white precipitate was rinsed four times with 2-propanol, until free of the CS₂ odour. The product was dried in a vacuum desiccator and then stored under nitrogen at 4 °C.

Particle size distribution

The silica powder was suspended in deionized water (pre-filtered using a 0.1 µm cellulose nitrate membrane filter) and ultrasonicated for 1 minute. The particle size distribution of the suspension was then measured using a COULTER N4 Plus instrument under the following conditions: angle selected: 90°, equilibrium time: 5 minutes, temperature 20 °C.

Surface property measurements

Nitrogen adsorption isotherms (77 K) were obtained with a Quantachrome NOVA automated gas sorption system, using version 1.11 NOVA for Windows software. All samples were pre-dried at 120 °C and evacuated in a desiccator. The multi-point BET equation was applied to estimate the surface area of the silica.

Scanning electron microscopy (SEM)

The sample was prepared by dispensing drops of an aqueous suspension of silica onto a glass plate. This was allowed to dry at room temperature and was then coated with a thin gold film. Scanning electron microscopy of the silica samples was carried out using a Philips XL-30 SEM instrument.

Particle settling rate

The settling time of a colloidal suspension of the nanoscavengers was assessed by following the light transmission at 600 nm over time. A 1 g L⁻¹ suspension of silica in water, held in a 1 cm pathlength quartz cuvette, was measured using a Perkin Elmer Lambda 3 spectrophotometer.

Atomic absorption spectroscopy

Copper, cadmium, lead and nickel measurements were carried out by flame atomic absorption spectroscopy using a Perkin Elmer 2380 instrument. The conditions employed were those recommended by the instrument manufacturer.

Copper capacity

The maximum copper loading capacity was measured by dispersion of a known weight of surface-modified silica in 10 mL of a 1000 µg Cu²⁺ mL⁻¹ solution. After 1 hour the suspension was centrifuged and the solid was rinsed 5 times with deionized water. Hydrochloric acid (3 mL, 37% w/v) was added to the tube to release the bound copper, the solution was diluted with deionized water and the concentration of copper in solution was measured by flame AAS.

Nanoscavenger based preconcentration using amine-modified silica

The potential application of modified sub-micron silica in the dispersion preconcentration of trace metals from aqueous solution was assessed by the extraction of copper using amine-modified silica. 1.20 g of diamine–silica (ca. 250 nm diameter) was sonicated in 100 mL of 0.1 M tris buffer (pH = 7.40). 20 mL of the suspension was transferred to a volumetric flask, which was then three quarters filled with deionised water. 1 mL of a stock copper solution, containing 10 µg copper per mL, was added to the flask. The flask was filled to the mark with deionised water, shaken and left for 2 hours. The suspension was then vacuum filtered using a 0.2 µm cellulose nitrate membrane filter. The filter was carefully transferred to a 20 mL glass vial, 5 mL of hydrochloric acid (1 M) was added and the vial was shaken for 30 minutes. The acidic solution was transferred into an appropriate volumetric flask and the membrane filter was rinsed with deionized water. These washings were then added to the first acid extract. The copper concentration in solution was measured by flame atomic absorption spectroscopy, normally after filtration through a Whatman GF/F filter. At the risk of aspirator blockage, once the particles have aggregated and settled, direct aspiration of the supernatant into the AAS instrument is possible.

Nanoscavenger based preconcentration using a bis-dithiocarbamate-modified silica

2.5 mL of a solution containing Cd²⁺, Cu²⁺, Pb²⁺ and Ni²⁺ (10 µg of each metal per mL) was transferred into 100, 250 and 500 mL volumetric flasks containing deionized water. The pH of each solution was adjusted to 7 using tris buffer. Dithiocarbamate-modified nanoscavenger (200 mg, pre-sonicated in 20 mL of buffer) was then dispersed in the metal solution. The flask was left for 2 hours before removing the solid material by filtration (0.2 µm cellulose membrane filter) and transferring the filter into a 20 mL vial. 10 mL of a 50 ∶ 50 mixture (by volume) of concentrated hydrochloric and nitric acids was added to the vial which was then shaken for 30 min on a rotating wheel. The acidic solution was transferred to a 10 mL volumetric flask and the flask was made up to the mark using deionized water, prior to analysis by flame atomic absorption spectroscopy.

Rate of uptake of copper by an amino-silica nanoscavenger

1 mL of 10 µg mL⁻¹ Cu(II) was added to a 100 mL flask, which was three quarters filled with deionised water. 10 mL of sub-micron amino-silica (10 mg silica mL⁻¹ in tris buffer, pH = 7.1) was dispersed in the solution and the flask was filled to the mark. Replicate solutions were left for different equilibrium times prior to vacuum-filtration through a 0.2 µm cellulose nitrate membrane filter. The filter was carefully transferred to a 20 mL glass vial and then 5 mL of 5 mol dm⁻³ HCl was added. The vial was shaken for 30 minutes and the acidic solution was then transferred to a 5 mL volumetric flask for analysis.

Colorimetric determination of copper as its dithiocarbamate complex

Copper was determined colorimetrically after solvent extraction of its diethyldithiocarbamate complex into dichloromethane. 100 mL of tap water was transferred into a 250 mL separating funnel and the pH was adjusted to 7.0 using a tri-ammonium citrate (10% w/v) solution. 1 mL of sodium diethyldithiocarbamate (1 mg mL⁻¹) was added to the funnel and the mixture was shaken for 5 minutes. 10 mL of dichloromethane was added. The funnel was shaken vigorously for 2 minutes, the organic layer was collected into a 10 mL volumetric flask and made up to the mark with dichloromethane. The absorbance was measured at 435 nm and the concentration was calculated based on calibration with standard solutions.

Results and discussion

Properties of the silica nanoscavenger particles

Crucial to this study was the production of monodisperse particles with a mean particle diameter that could be filtered from solution. It was also considered to be important that the particle size distribution was as narrow as possible with few small particles. The settling of some larger particles or aggregates is not a specific problem in the dispersion preconcentration approach with nanoscavengers, provided that the larger particles do not dominate the particle population. Smaller particles however are to be avoided as they will not be isolated by filtration and will result in reduced analyte recovery.

Refinement of the technique for the synthesis of spherical (Fig. 1a) sub-micron silicas resulted in monodisperse particle size populations having a mean particle size of ca. 250 nm. (Fig. 1b).

Fig. 1 Sub-micron silica particles (a) SEM image (5000× magnification) (b) particle size distribution.

After the introduction of diamine functionality onto the nanoparticulate silicas the mean particle size and monodispersity of the parent particles was essentially retained (Fig. 2).

Fig. 2 Diamine nanoscavenger (a) SEM image (12 500× magnification) (b) particle size distribution.

Subsequent conversion of the diamine nanoscavenger to its dithiocarbamate derivative had little impact on the particle size characteristics (Fig. 3).

Fig. 3 Dithiocarbamate nanoscavenger (a) SEM image (25 000× magnification) (b) particle size distribution.

Adsorption isotherm characteristics. Nitrogen adsorption isotherms indicated that the amino-nanoparticles had a multipoint BET surface area of 14.5 m² g⁻¹, a total pore volume of 0.0301 cm³ g⁻¹ and an average pore diameter of 8.2 Å. These isotherms were both type II with a small hysteresis between the adsorption and desorption branches indicating some mesoporosity. The indicated surface area is close to the theoretical surface area of spherical, non-porous particles having a diameter of 250 nm and a density³⁶ of 1.8 g cm⁻³ (13.3 m² g⁻¹). The results are also not dissimilar to those obtained by Dekany et al.³⁷ when studying slightly larger silica particles. The 77 K nitrogen adsorption isotherm, analysed using the BET equation, is a widely accepted general approach to the measurement of surface area, but it is now well established that, with these Stöber-type silicas, gas adsorption studies frequently do not accurately reflect the surface area calculated from silanol concentrations, due to the inaccessibility of its pores to gases.^38,39 Indeed, surface area results from 77 K BET nitrogen adsorption surface area studies have been shown not to be consistent with several other measures of surface area and the BET results may underestimate the surface area by a factor of 30 times.^37,40 If it is the case that the silica particles are swelling and opening up in aqueous solution, it is to be expected that the surface area and consequent analyte capacity of the sub-micron silica beads will be significantly larger than those indicated from the gas adsorption results.

Settling time. The stability of the particle dispersion is of particular importance in the development of the preconcentration approach. Light transmission at 600 nm was therefore employed to assess the rate of settling of the particles. Under low ionic strength conditions the dispersion was found to be stable for several hours. A small proportion of the particles settled during the early stage of the experiment, presumably as a consequence of the settling of larger particle clusters that had not been completely broken up by the ultrasonic agitation and due to the presence of a small number of larger particles (Fig. 4A). The settling of a small proportion of the silica as large particles should be of little significance as any analyte collected by the particles will be included in the final determination step.

Fig. 4 Settling of the amine nanoscavenger from its suspension in water (curve A) and in 1 M NaCl solution (curve B).

The ionic strength of the sample solution has the potential to destabilise the sol through the compression of the double layer causing premature deposition of particle aggregates. This effect was therefore assessed by measuring the transmission change occurring with nanoscavengers that were suspended in aqueous 1 M sodium chloride solution (cf ionic strength of seawater is 0.7 M). Major destabilization of the sol did not occur immediately and over a period of more than 30 minutes there was no significant deposition evident (Fig. 4B) from the transmission measurements. Once started, particle sedimentation continued relatively slowly over a further 2 hour time period. In other experiments this 30 minute nanoscavenger suspension period was found to be sufficient to ensure essentially complete metal uptake from solution. The scavenging would not however be expected to stop during the subsequent slow sedimentation phase.

Copper capacity. The capacity of the amine-modified silica, measured by saturation of reactive sites with a concentrated copper(II) solution, was found to be ca. 2.1 mg Cu g⁻¹.

Rate of copper(II) uptake by an amine nanoscavenger

A number of experiments were performed to assess the time that would be required to achieve quantitative collection of a metal from aqueous solution by a nanoscavenger dispersion. A diamine-based nanoscavenger was dispersed in a spiked sample solution for time periods ranging from 1 minute to two hours. Analysis of the copper recovered from the nanoscavenger showed that maximum, and essentially quantitative, recovery of the copper could be achieved within a 15 minute equilibration period (Table 1).

Table 1 Rate of copper collection by a diamine nanoscavenger

Time/min	Average of recovery (%) ± SD n = 3
1	94.1 ± 1.3
5	94.4 ± 0.8
15	98.4 ± 0.5
30	98.4 ± 0.5
60	98.2 ± 0.7
120	98.1 ± 0.5

---

Preconcentration of copper(II) using diamine nanoscavengers

The preconcentration system was based on the dispersion of diamine nanoscavengers in solution to collect copper(II). Quantitative scavenging of copper was achieved for a number of different preconcentration factors (Table 2).

Table 2 Preconcentration of copper(II) by nanoscavenger extraction

Sample volume/mL	Mass of Cu(II)/µg	Preconcentration factor	%Recovery ± % rsd (n = 4)
100	10	10	105 ± 4
500	10	50	105 ± 4
1000	10	100	96.6 ± 5

---

The preconcentration factors quoted in Table 2 are based on the collection of copper from volumes of sample up to one litre, with an analyte release into 10 mL of acid. Small volumes of nanoscavenger are employed in such experiments and it should therefore be possible to further enhance the preconcentration factor to 1000× by reducing the acid digest volume to 1 mL.

Release of metals from dithiocarbamate nanoscavengers

The softer trace metals form strongly bound complexes with dithiocarbamate functional groups and the release of metals from dithiocarbamate nanoscavengers can require more rigorous conditions than the corresponding amine. Indeed, release of the metal through destruction of the dithiocarbamate is often the most reliable approach with such complexation systems. Metal release from the dithiocarbamate nanoscavengers was therefore carried out using 10 mL of a 50 ∶ 50 (by volume) mixture of concentrated hydrochloric and nitric acids. Under such conditions Cu²⁺, Cd²⁺, Pb²⁺ and Ni²⁺ were rapidly decoupled from the nanoscavenger giving high recoveries within 30 minutes (Table 3).

Table 3 Release of metal ions from a dithiocarbamate nanoscavenger

Treatment time/min	%Recovery
	Cd(II)	Cu(II)	Pb(II)	Ni(II)
10	89 ± 2.6	92 ± 1.5	97 ± 6.6	76 ± 2.4
30	104 ± 3.9	97 ± 2.3	97 ± 1.2	90 ± 3.4
120	105 ± 0.7	99 ± 2.2	105 ± 2.8	92 ± 1.6
240	104 ± 1.8	97 ± 4.2	106 ± 2.0	92 ± 2.2

---

Preconcentration of trace metals using a bis-dithiocarbamate nanoscavenger

The effectiveness of the dispersion preconcentration approach with dithiocarbamate nanoscavengers was assessed for Cu²⁺, Cd²⁺, Pb²⁺ and Ni²⁺. At the highest preconcentration factor, the surrogate sample concentration of 50 µg Cu L⁻¹ (50 ppb) was increased to 2.5 µg mL⁻¹ (2.5 ppm) with essentially quantitative recovery of all metals (Table 4).

Table 4 Preconcentration of metal ions with the dithiocarbamate nanoscavenger

Preconcentration factor	%Recovery
	Cd	Cu	Pb	Ni
10	102 ± 6.4	98.8 ± 1.5	93.1 ± 1.6	94.5 ± 1.8
25	100 ± 0.7	96.7 ± 1.6	92.1 ± 2.3	92.5 ± 2
50	99.3 ± 1	96.6 ± 1.1	91 ± 1.7	91.8 ± 1.7

---

A comparison of the nanoscavenger extraction method and a solvent-based extraction approach

In order to assess the performance of the nanoscavenger approach in the determination of copper, replicate analyses of a water sample were undertaken using both the amino-silica nanoscavenger dispersion preconcentration procedure and a well-established colorimetric procedure employing solvent extraction of the copper as a dithiocarbamate complex. The two approaches produced comparable results for the concentration of copper in tap water (Table 5).

Table 5 Analysis of a water sample by nanoscavenger and chelation–solvent extraction approaches

Method	Average of concentration/µg L^−1 (n = 5), X ± %rsd
Solvent extraction/colorimetry	88.2 ± 7.2
Nanoscavenger/AAS	84.4 ± 2.3

---

A t-test, at the 95% confidence level, showed there to be no significant difference between the results obtained by the two methods.

Conclusions

In pure water the settling of nanoparticles is largely governed by a slow Stoke's law terminal settling velocity (for 250 nm particles this is 3.06 × 10⁻⁸ m s⁻¹) and Brownian motion. The stability of such sols, their Brownian displacement and movement in convection currents provides an effective means of maintaining a solid phase extractant in suspension without the need for additional agitation. Particles having mean particle sizes of ca. 250 nm can be suspended within the sample solution for several hours, during which time they can selectively scavenge dissolved species from solution. This specific particle size is well suited to the rapid recovery of the nanoscavengers after the prescribed equilibration time by conventional membrane filtration. Careful control of the size characteristics of the nanoscavengers is necessary to ensure that the particle population neither includes particles below the filter cut-off limit, nor contains significant quantities of larger particles which sediment rapidly from their suspension.

Sub-micron sized chemically modified silicas have been shown to be suitable replacements for organic solvents and reactive complexants in the extraction and preconcentration of trace metals from aqueous solution; the approach having been demonstrated by the preconcentration of copper from solution using a diamine-based chelating nanoscavenger and the application of the procedure to a number of metals using a dithiocarbamate-modified silica. Quantitative recovery of the metals was achieved using nanoscavengers to collect them from solution for determination by atomic absorption spectroscopy. Results obtained by this new approach were statistically indistinguishable from those obtained by a conventional solvent extraction of the copper as a dithiocarbamate complex. No organic solvents were employed in the application of the nanoscavengers and only small quantities of largely environmentally benign reagents were used in their synthesis. The quantity of nanoscavenger employed in each analysis also reduces the quantities of waste involved.

The physical advantages of the nanoscavenger approach over conventional solvent extraction are not to be underestimated. In routine use large numbers of samples can be rapidly treated with the nanoscavenger. This can be carried out at the sampling site, stabilizing the analyte during transport and pre-analysis storage. This leads to the extraction being carried out, without further intervention, during the sampling excursion. Filtration manifolds are readily developed to permit the simultaneous recovery of the nanoscavenger from many samples and this filtration can be carried out in the field to minimize transport weight, whilst protecting the analyte against contamination or loss by immobilizing it on the solid nanoscavenger.

There are many more potential applications of nanoscavengers that remain to be developed. This paper has introduced the concept using the examples of simple chelating nanoscavengers having an affinity for trace metals. Besides the many alternative modifications that can be carried out to broaden the range of applicability for the collection of metals from solution, the approach can also be applied to the extraction and preconcentration of trace organics. These approaches are currently under investigation.

Acknowledgements

The authors would like to gratefully acknowledge the support provided for this work by the Ministries of Health and of Higher Education of the Kingdom of Saudi Arabia.

References

1. A. G. Howard and P. J. Statham, Inorganic Trace Analysis: Philosophy and Practice, John Wiley & Sons, Chichester, 1993.
2. D. M. Hercules, L. E. Cox, S. Onisick, G. D. Nichols and J. C. Carver, Anal. Chem., 1973, 45, 1973–1975.
3. K. F. Sugawara, H. H. Weetall and G. D. Schucker, Anal. Chem., 1974, 46, 489–492.
4. D. E. Leyden and G. H. Luttrell, Anal. Chem., 1975, 47, 1612–1617.
5. D. E. Leyden, G. H. Lutrell, W. K. Nonidez and D. B. Werho, Anal. Chem., 1976, 48, 67–70.
6. D. E. Leyden, M. L. Steele, B. B. Jablonski and R. B. Somoano, Anal. Chim. Acta, 1978, 100, 545–554.
7. K. Akatsuka, J. W. McLaren and S. S. Berman, Bunseki Kagaku, 1993, 42, 423–428.
8. B. K. Esser, A. Volpe, J. M. Kenneally and D. K. Smith, Anal. Chem., 1994, 66, 1736–1742.
9. A. G. Howard, M. Volkan and O. Y. Ataman, Analyst, 1987, 112, 159–162.
10. M. Volkan, O. Y. Ataman and A. G. Howard, Analyst, 1987, 112, 1409–1412.
11. F. Sahin, M. Volkan, A. G. Howard and O. Y. Ataman, Talanta, 2003, 60, 1003–1009.
12. G. Gokturk, M. Delzendeh and M. Volkan, Spectrochim. Acta, Part B, 2000, 55, 1063–1071.
13. K. S. Abou-El-Sherbini, I. M. M. Kenawy, M. A. Hamed, R. M. Issa and R. Elmorsi, Talanta, 2002, 58, 289–300.
14. N. V. Deorkar and L. L. Tavlarides, Hydrometallurgy, 1997, 46, 121–135.
15. A. Goswami and A. K. Singh, Anal. Chim. Acta, 2002, 454, 229–240.
16. R. Alexandrescu, F. Dumitrache, I. Morjan, I. Sandu, M. Savoiu, I. Voicu, C. Fleaca and R. Piticesu, Nanotechnology, 2004, 15, 537–545.
17. W. J. Stark, A. Baiker and S. E. Pratsinis, Part. Part. Syst. Charact., 2002, 19, 306–311.
18. R. P. Das and S. Anand, Indian J. Phys. A, 2004, 78, 165–174.
19. K. Okuyama and I. W. Lenggoro, Chem Eng. Sci., 2003, 58, 537–547.
20. E. Hosono, S. Fujihara, T. Kimura and H. Imai, J. Sol-Gel Sci. Technol., 2004, 29, 71–79.
21. P. Tartaj and L. C. De Jonghe, Journal of Materials Chemistry, 2000, 10, 2786–2790.
22. W. Stober, A. Fink and E. Bohn, J. Colloid Interface Sci., 1968, 26, 62–69.
23. A. Vanblaaderen, J. Vangeest and A. Vrij, J. Colloid Interface Sci., 1992, 154, 481–501.
24. T. Matsoukas and E. Gulari, J. Colloid Interface Sci., 1988, 124, 252–261.
25. C. F. Zukoski, J. L. Look and G. H. Bogush, in The Colloid Chemistry of Silica, ed. H. F. Bergna, Oxford University Press, Oxford, 1994, vol. 234, pp. 451–465.
26. C. F. Zukoski, G. H. Bogush and J. L. Look, Abstr. Pap. Am. Chem. Soc., 1990, 200, 261-COLL.
27. H. Fritz, M. Maier and E. Bayer, J. Colloid Interface Sci., 1997, 195, 272–288.
28. M. Maier, H. Fritz, M. Gerster, J. Schewitz and E. Bayer, Anal. Chem., 1998, 70, 2197–2204.
29. M. Gerster, J. Schewitz, H. Fritz, M. Maier and E. Bayer, Anal. Biochem., 1998, 262, 177–184.
30. X. Xie, X. Zhang, H. Zhang, D. P. Chen and W. Y. Fei, J. Magnet. Magnet. Mater., 2004, 277, 16–23.
31. P. Liang, T. Q. Shi and J. Li, Int. J. Environ. Anal. Chem., 2004, 84, 315–321.
32. A. Arkhireeva and J. N. Hay, J. Mater. Chem., 2003, 13, 3122–3127.
33. S. Sadasivan, A. K. Dubey, Y. Z. Li and D. H. Rasmussen, J. Sol-Gel Sci. Technol., 1998, 12, 5–14.
34. I. Nebot-Diaz, E. Cordoncillo, M. Irun, P. Escribano and J. B. Carda, Euro Ceramics VII, Pt 1-3, Key Eng. Mater., 2002, 206-2, 1203–1206.
35. S. L. Chen, P. Dong, G. H. Yang and J. J. Yang, Ind. Eng. Chem. Res., 1996, 35, 4487–4493.
36. G. H. Bogush, M. A. Tracy and C. F. Zukoski, J. Non-Cryst. Solids, 1988, 104, 95–106.
37. I. Dekany, J. Nemeth, M. Szekeres and R. Schoonheydt, Colloid Polym. Sci., 2003, 282, 1–6.
38. A. Burneau and B. Humbert, Colloids Surf., A, 1993, 75, 111–121.
39. J. P. Gallas, J. C. Lavalley, A. Burneau and O. Barres, Langmuir, 1991, 7, 1235–1240.
40. M. Szekeres, J. Toth and I. Dekany, Langmuir, 2002, 18, 2678–2685.

---