Metal sulfidenanoparticles synthesized via enzyme treatment of biopolymer stabilized nanosuspensions

Abstract

Nanoparticles of CuS, Cu_xS, Ag₂S and CdS were successfully prepared using a novel general and green synthetic process to give dextranbiopolymer stabilised metal sulfifde nanosuspensions. Following preparation, dextranase enzyme was used to remove the bulk of the bound dextran to give pure stable metal sulfidenanocrystals for application in for example aspects of medicine, photonics and solar cells. Particles of good homogeneity were obtained and the CuSnanoparticle size was controlled to 9–27 nm by adjusting the reaction conditions. Cu₂Snanoparticles were 14 nm, Ag₂Snanoparticles were 20–50 nm and CdS nanoparticles were 9 nm is size. The complexing mechanism of nanoparticlesulfides to dextrans was further studied using carboxylmethyl dextran as a complexing agent and crosslinked Sephadex (dextran) `beads as substrate. Particles were characterized by TEM, XRD, TGA, FT-IR and zeta-potential measurement, and their UV–vis spectroscopic absorption properties were determined. Stabilization of the sulfidenanoparticles with soluble hydroxylated biopolymers such as dextran is previously unreported and is here interpreted in terms of viscosity, pH of the system and weak polar S–H or S(metal)OH₂⁺ interactions with dextran depending on the material. Notably, the complexing mechanism appears to differ significantly from that taking place in known dextran–metal oxide systems. The process shown here has good potential for scale-up as a biosynthetic route for a range of functional sulfidenanoparticles.

---

1. Introduction

The fabrication of semiconductor nanostructures has been an important goal in recent years,¹ and interest and research into semiconductor chalcogenides, in particular copper sulfides and silver sulfide (Ag₂S) especially when formed as nanoparticles, has recently increased sharply.²Copper sulfides exist as a range of stable and metastable phases ranging between copper ‘poor’ CuS (covellite) to copper ‘rich’ Cu₂S (chalcocite) at room temperature, common Cu_xS phases are Cu_1.75S(anilite), Cu_1.8S(digenite) and Cu_1.94S (djurleite).³ Conductivity in copper sulfides arises due to copper vacancies and decreases from copper-poor to copper-rich Cu_xS.⁴Copper sulfides have an important application in photonics as a semiconductor material for new-generation solar cells,⁵ as a lithium battery cathode material,⁶ as a catalyst and as a polymer-coating material.⁷Covellite is also, at present, the only known superconducting naturally occurring mineral with a T_c of 1.63K.⁸Ag₂S has recently been recognised as semiconductor with unique properties such as enhanced and rapid optical nonlinearity and high luminescence, and has been applied in new optical and electronic devices.⁹CuS and Cu₂S are indirect semiconductors with bandgaps around 1.7 and 1.2 eV respectively. Cu_xS compounds have indirect bandgaps between 1.05 and 1.4 eV for values of x of 1.8 to 1.96.¹⁰Ag₂Snanoparticles can be both a direct and an indirect band semiconductor, with bandgaps E_g,dir of ∼1 and E_g,ind of ∼2.3 eV.¹¹ The bandgap varies depending on particle size with bulk Ag₂S having E_g,dir of 1.1 eV. CdSnanocrystals have previously been prepared by a great many routes and have a wide bandgap of E_g,dir 2.41 eV.¹² However there are few reports on the preparation of copper sulfide nanomaterials. Methods of synthesizing CuS include templating in liquid crystal phases,¹³ or spray pyrolysis.³ Copper-rich Cu_xS has been made using copper salts reacted under argon in a mixed solvent of a thiol and oleic acid,² or thermolysis of a copper–dodecanethiol complex under argon.¹⁴Cu_xS can also been made by the heating of CuS, resulting in the loss of sulfurous gases, however, large and aggregated crystals are obtained by this route.¹⁵ Furthermore, preparation of Ag₂S is complicated by the strong tendency of Ag₂S clusters to aggregate to form bulk precipitates.¹⁶ Reverse micelle techniques have been employed,¹⁷ along with silver thiolate polymers¹⁸ and thermal decomposition under argon of a silver(I) dialkyldithiophosphate.⁹

Here we report a green and general route applied to the preparation of photonic chalcogenides with CuS (covellite), Cu_∼1.9S, Ag₂S (acanthite) and CdS (hawleyite) nanoparticles as examples. To our knowledge, the use and interaction of dextran with metal sulfides has not been reported. Here we describe the use of the biopolymerdextran as a stablising agent for the preparation of the sulfidenanoparticles in the form of nanosuspensions, or as nanoparticle-coated Sephadex beads. This was followed by use of the enzyme dextranase for the liberation of the nanoparticles from their biopolymer coat leaving high-purity nanoparticles with only a trace level of dextran on the sulfide materials that prevented subsequent aggregation and growth. Dextran is a biocompatible polysaccharide composed of a majority of α1-6 and a minority of α1-3 or α1-4 glycosdic linked glucose units. Crosslinked dextran is available commercially as Sephadex beads of various types and used widely for size-exclusion chromatography.¹⁹

Dextran has previously been employed as a bound coating on magnetite and other metal oxides and forms an important class of materials for use in medicine as MRI contrast agents and in targeted drug delivery.²⁰ Previously, starch, polyurethane and polyacrylamide have been used as capping agents for preparation of Ag₂S and CdS nanoparticles.²¹Starch provided a green synthesis route to nanocomposite solids of 8 wt% Ag₂S, polymers dissolved in organic solvents or microwaved melts enabled polymer-coated nanoparticles to be synthesized. Previous polymer stabilisation routes to nanoparticles have a limitation in obtaining high-purity products and typically the nanoparticles can only be released from the polymer by calcination.

Here, synthesis of the metal sulfidenanoparticles was conducted by addition of Na₂S solution to a solution of dissolved metal nitrate and dextran, followed by mild heating to assist the bonding of dextran to the precipitated sulfidenanoparticle. This gave stable coloured suspensions of the sulfidenanoparticles. In order to prepare the copper-rich Cu_xS, Cu⁺ as CuCl and Na₂S in degassed distilled water were used as starting reagents. Since dextran is a natural substance, the enzyme dextranase was then used to remove the bulk of the biopolymer shell following the synthesis, this then liberated the surface to act effectively in photonic or other applications. Dextranase occurs in a number of fungi and is extracted and utilized by the sugar industry for the breakdown of unwanted dextrans in harvested sugar. Dextranase cleaves the glucosepolymer primarily into solubilized isomaltose and isomaltotriose by catalyzing the endohydrolysis of the α1-6 glycosidic linkages at any point in the polymer chain. Dextranase has been used for controlled drug release from core–shell polymer particles and for the characterization of dextran-coated polycaprolactonenanoparticles and more recently for the preparation of metal oxidenanoparticles.²² An important aspect and advantage of the methodology shown here is the potential to develop a green biosynthesis for production of a range of nanoparticles on a large scale. Dextranase enzyme bound to montmorillonite, calcium phosphate microparticles or acrylic beads can be used for nanoparticle liberation with a useful maltose byproduct, the enzyme-coated beads could then be easily extracted by filtration and re-used.²³ The dextranbiopolymer component can be readily generated from sucrose solution using dextran synthase enzyme directly or viabioreactors inoculated with Leuconostoc mesenteroides bacteria.²⁴ Also, it would be possible to control biopolymer/nanoparticle relative content by enzyme treatment time.

The results showed that dextran had two essential roles in the synthesis process, firstly in providing a moderately viscous environment for the initial precipitation of sulfide. The viscosity reduced diffusion of ions and inhibited aggregation of nanoparticles and growth of the embryonic nanoparticles. Secondly, subsequent mild heating increased binding of the dextran to the metal sulfide particles to give stable nanosuspensions of polymer-coated nanoparticles. The importance of dextranM_r was shown, as dextran of 500 000 M_r was necessary to form stable nanoparticles. Attempts to synthesize the sulfides with dextran at M_r of 6000 and 70 000 gave only or almost entirely highly aggregated bulk precipitations upon addition of S²⁻, with the exception of Ag₂S which formed stable and high-yield nanosuspensions with dextran of M_r 70 000.

Interestingly, in order to help elucidate the binding mechanism of biopolymer to sulfides, the stabilisation of copper sulfidenanoparticles with synthetically prepared fully carboxylated dextran, (where the dextran OH groups have been replaced by COOH) was studied. The capping effect was found to be very weak and nanoparticle yields very low. The pH of the system was 4.1 and the pK_a of dextran carboxyls is reported to be 6.1,²⁵ thus the carboxyl groups are largely protonated and interact only very weakly with the metal sulfide surface.

2. Results and discussion

Dextran (M_r 500 000) stabilized CuS (dex–CuS) was prepared with stoichiometric concentrations of Cu and S ions. Preparations were also conducted using progressively lower concentrations of S²⁻ to the point where dark-green coloured supernatant suspensions following centrifugation and SEC purification (Fig. 1A) were no longer obtained. With Cu⁺ as a starting reagent in the preparation of dextran-stabilized Cu_xS (dex–Cu_xS) brown coloured suspensions were formed (Fig. 1B). Grey/black suspensions were obtained with dextran-stabilized Ag₂S (dex–Ag₂S) preparations (Fig. 1C) and lemon-yellow coloured nanosuspensions were formed with dextran-stabilized CdS (dex–CdS) preparations (Fig. 1D). When Sephadex beads were used in place of dextran for CuS preparations, dark-green Sephadex–CuS beads were formed (Fig. S1 of the ESI† )

Fig. 1 Photograph of dextran-stabilized nanosuspensions of A) CuS; B) Cu_xS; C) Ag₂S; D) CdS.

2.1 Enzymatic liberation of nanoparticles

Treatment of nanosuspensions or Sephadex–CuS beads with dextranase enzyme resulted in a transformation into flocs of suspended particles in colourless solution after several hours treatment at 34 °C. Dex–CdS behaved somewhat differently to the other sulfides prepared, as floc formation occurred within 30 min. Furthermore, dextranase treatment of dex–CdS for 24 h (open to light) was observed to result in gradual dissolution of the suspended particles until a colourless solution was obtained. Control experiments of CdS nanoparticles stored in distilled water at pH 3.5 showed only very minimal dissolution compared to dextranase-treated dex–CdS systems which were at pH 6 gradually dropping to pH 4 due to dissolution of the CdS with liberation of acidic H₂S. This result is in accordance with literature reports on the significantly enhanced photodissolution of CdS in the presence of complexing agents.²⁶ With CdS, S⁰ is formed at the particle surface which readily reacts to give S–OH and S–H functional groups. Also CdS relative solubility (K_sp 10⁻²⁷) is significantly greater than the other sulfide samples which have K_sp values ranging from 10⁻³⁷ to 10⁻⁵¹. The results suggests that the combination of dextran as complexing agent and its binding to the surface S, followed by removal by the enzyme, results in a particularly efficient dissolution system for CdS. Control samples of all dextran–metal sulfidenanoparticle suspensions warmed under the same conditions without added enzyme showed no separation or any observable changes after 48 h. This showed that firstly separation from the dextran capping agent was not due to warming, and also it showed that the photocatalytic surface dissolution process only occurred significantly with dex–CdS systems.

Furthermore, it was noted that nanosuspensions of the dextran-capped CuS and Ag₂S particles prior to enzymatic treatment left in daylight for several weeks were unchanged. Aqueous brown Cu_xS nanosuspemsions were observed to undergo gradual oxidation into green coloured CuS after approximately 7 days, however it was found that this transformation was inhibited by the addition of a single drop of hydrazine solution.

Dex–CdS nanosuspensions showed the gradual dissolution process in daylight until no yellow coloration was observable after 4 days. Conversely dex–CdS nanosuspensions stored in dark conditions showed no dissolution, remaining lemon-yellow in colour. Presumably this difference is due to the relatively higher solubility and surface photoactivity of CdS in the presence of the dextran complexing agent.

CuS, Cu_xS and Ag₂S nanosuspensions were subjected to enzyme treatment for 24 h, CdS nanosuspensions had a reduced enzyme treatment of 5 h to minimize the dissolution effect. Upon light and brief centrifugation of the flocs, a coloured sediment and colourless supernatant liquid was obtained. The supernatant was found to consist of a mix of low-molecular-weight sugars.

2.1.1 TEM analysis. TEM of the thoroughly washed dark-green sediment from dex–CuS (1 : 1 molar ratio) preparations showed the presence of irregular nanocrystals with an average size of 27 nm (Fig. 2A). Upon decreasing S²⁻ a minimum Cu:S molar ratio of 1 : 0.625 that still resulted in the formation of green CuSnanoparticle suspensions was determined. Following the dextranase treatment the dark-green coloured particles were shown by TEM to be discrete and more monodisperse nanoparticles with an average size of 14 nm (Fig. 2B). Thus some degree of control over particle size was possible by reduction of the S²⁻ reagent concentration. Sephadex–CuS preparations yielded nanoparticles of average size 9.5 nm. The degree of homogeneity was less than that obtained with dextran solution systems however (Fig. S2 of the ESI† ). Copper-rich dex–Cu_xS preparations using CuCl as a starting reagent and a molar ratio of 2 : 1 Cu:S gave dark-brown coloured sediments following enzyme treatment, shown by TEM to be discrete nanoparticles with an average size of 14 nm (Fig. 2C). Although CuCl has low solubility in water, dextran is known to have strong complexion ability with copper ions,²⁷ which may enhance the salt solubility sufficiently for rapid precipitation of copper sulfide to occur. The presence of the weakly reducing long-chain dextran may also be beneficial in maintaining the Cu⁺oxidation state during the reaction. Ag₂S preparations using dextran at both M_r of 70 000 and 500 000 were shown to be more polydisperse with nanocrystals approximately 20–50 nm in size (Fig. 2D). CdS nanoparticles were shown to be roughly cubic crystals with an average size of 9 nm (Fig. 2E). Fig. 2F shows a TEM micrograph of an example green coloured dextran–CuS nanosuspension (see Fig. 1A) prior to any dextranase treatment. CuS crystals embedded within the dried dextran film are visible. (All TEM measurements were the average of 15 particle sizes).

Fig. 2 TEM micrographs of dextranase-treated and washed nanoparticles (A–E). A) CuS (Cu:S of 1 : 1 molar ratio), B) CuS (Cu:S of 1 : 0.625 molar ratio), inset shows a high-resolution image of a single nanoparticle with d-spacings of 3.3 Å (100); C) Cu_xS (Cu:S of 2 : 1 molar ratio), inset shows a high-resolution image of single nanoparticle with d-spacings of 1.68 Å; D) Ag₂S (Ag:S of 2 : 1 molar ratio); E) CdS (Cd:S of 1 : 1 molar ratio); F) Dextran–CuS nanosuspension (see Fig. 1A) prior to dextranase treatment showing crystals embedded in a dried dextran film (Cu:S of 1 : 0.625 molar ratio).

2.1.2 X-Ray analysis . X-Ray diffraction (XRD) analysis was conducted for all nanosuspensions following dextranase treatment and washing and drying of the coloured sediments (Fig. 3). All CuS preparations using Cu(NO₃)₂ as a starting reagent and following dextranase treatment gave broad reflections corresponding to hexagonal CuS (covellite) with d-spacings (Å) at 3.21 (101), 3.03 (102), 2.79 (103), 1.89 (110), 1.74 (108) and 1.56 (116) (JCPDS: 01-078-0876). Cu_xS prepared with CuCl in degassed water gave broad reflections with d-spacings (Å) at 3.19, 2.76 and 1.96, characteristic of a number of copper-rich sulfides of Cu_1.8–1.97S. Samples prepared as Ag₂S gave multiple reflections, all corresponding to monoclinic Ag₂S (acanthite) with major reflections with d-spacings (Å) at 3.074 (111), 2.83 (−112), 2.596 (−121) and 2.43 (112) (JCPDS: 00-014-0072). CdS nanoparticles gave broad reflections with d-spacings (Å) at 3.32 (111), 2.05 (220) and 1.76 (311) corresponding to cubic CdS (hawleyite) (JCPDS: 00-010-454). The broad reflections of the CuS, Cu_xS and CdS confirmed the nanoparticle nature of these products. In contrast, XRD of samples prior to enzymatic treatment gave only very weak reflections due to the high-polymer, low metal sulfide content.

Fig. 3 Powder X-ray diffractograms of dextranase-treated and washed nanoparticles of CuS, Cu_xS, Ag₂S and CdS.

2.1.3 FT-IR analysis. FT-IR spectroscopy of pure dextran (M_r 500 000) and also CuS, Cu_xS, Ag₂S and CdS preparations following dextranase treatment and thorough washing are shown in Fig. 4. Native dextran itself showed characteristic bands around 1000–1100 cm⁻¹ and further bands at 1430, 1630, 2920 and 3400 cm⁻¹.²⁸ Dried dex–metal sulfide samples showed near identical bands to native dextran due to the large excess of bound dextran present which obscured any sulfur–metal absorbance bands (not shown). Conversely, dextranase-treated nanoparticle samples showed weak metal–sulfur bands present over 500–600 cm⁻¹ together with bands due to dextran at ca. 1000–1100 cm⁻¹ of relative size that corresponded to the nanoparticle dimension. Smaller particles showed a higher level of residual surface-bound dextran oligomer due to the higher surface-to-volume ratio. Thus prominent dextran bands were found with the smaller sized CuS (1 : 0.625 molar ratio) Cu_xS and CdS nanoparticles. Weaker bands from residual oligomer were obtained with the larger sized CuS (1 : 1 molar ratio) and Ag₂S samples.

Fig. 4 FT-IR spectra of pure dextran (M_r 500 000), other spectra are all dextranase-treated and washed nanoparticles of: CuS (1 : 0.625 molar ratio, 14 nm), CuS (1 : 1 molar ratio, 27 nm), Cu_xS, Ag₂S and CdS.

2.1.4 Thermogravimetric analysis . Thermogravimetric analysis (TGA) in air of dried dextran–metal sulfides showed significant onset of decomposition from around 300 °C due to decomposition of dextran into carbon with loss of water vapour and then removal of carbon as CO₂. Dex–CuS nanosuspensions (Fig. S1 of the ESI† ) then showed a further weight loss due to decomposition of CuS to CuO and copper sulfates that occurred up to around 430 °C. A residual mass of oxosulphates and CuO of 9.7 wt% was obtained. Cu₂S showed a similar profile to CuS but had the lowest residual mass of the materials prepared with a yield of copper oxysulfates of 4 wt%. This perhaps reflected the quite low solubility of the CuCl reagent in solution. TGA of dried dex–Ag₂S showed weight loss due to dextran removal and also decomposition of Ag₂S to give a residual mass of Ag of 10 wt% from around 650 °C. XRD measurement of a sample heated to 400 °C showed major reflections due to Ag d-spacings (Å) at 2.35 (111), 2.04 (200) and 1.44 (220), showing decomposition of the Ag₂S and carbothermal reduction to Ag had already largely taken place at this temperature (not shown). Upon heating in air therefore it is likely the acanthite converted to the higher temperature argentite phase around 180 °C followed by reduction of the Ag₂S,²⁹ with the loss of sulfurous gases to give Ag. Dex–CdS gave a similar profile to the dex–copper sulfides again showing biopolymer and sulfidedecomposition to give a residual mass of 6 wt% cadmium oxide and oxysulphates.

In contrast, TGA (with air flow) of dextranase-treated dex–CuS, dex–Cu_xS and dex–Ag₂S did not show significant weight loss on heating to 400 °C and had far greater residual masses confirming that the bulk of the dextran coating was removed by the enzymatic hydrolysis prior to the TGA analysis (Fig. S2 of the ESI† ). For the CuS sample, onset of conversion of the CuS into oxysulphates occurred around 300 °C with a weight gain to 108% before decomposition into CuO and mixed oxysulphates at around 620 °C. Cu_xS showed a weight loss of 2 wt% over the temperature range 100–400 °C which we attribute largely to decomposition of trace residual dextran, by 800 °C the sample weight had increased by 5 wt% overall due to conversion into oxosulphates with uptake of oxygen. Ag₂S samples showed a gradual increase in mass during heating, showing the absence of significant amounts of carbon from associated dextran meant that reduction to Ag was not promoted and conversion to oxysulfates occurred with uptake of oxygen.

2.2 UV–vis spectroscopic analysis

UV–vis spectroscopy of enzyme dex–CuS nanosuspensions showed the characteristic absorption in both the UV and near-IR regions of the spectrum (Fig. 5).³⁰UV spectra of enzyme-treated Sephadex–CuS were near-identical to dex–CdS (not shown). The Cu_xS phase showed absorption only in the blue–UV region which is typical for Cu_xS with x = 1.7–2.0. Ag₂S and dextran nanosuspensions show absorption in the blue–UV region with a quite broad absorption in the visible region due to overlap of absorption bands of differing energies from moderately polydisperse sized nanoparticles. CdS preparations show onset of absorption at 475 nm and an exciton peak at approximately 405 nm indicating the exciton Bohr radius (∼5 nm) is comparable to the radius of the prepared nanoparticles.³¹

Fig. 5 Plots of UV–vis absorbance of the dextranase-treated and washed nanoparticles of CuS (1 : 0.625 molar ratio), Ag₂S, Cu_xS and CdS.

Bandgap energies and transition types were derived by using the following equation that describes the relationship for near-edge absorption:

α = [k(hν − E_g)^n/2]/hν

α is a constant that depends on transition probability, ν is the frequency, h is the Planck's constant, k equals a constant while n can be 1 or 4. The bandgap, E_g, can be obtained from a straight line plot of (αhν)^2/n as a function of hν. Extrapolation of the line to the baseline, where the value of (αhν)^2/n is zero, will give E_g. If a graph with a portion of straight line is obtained for n = 1, it indicates a direct electron transition between the semiconductor states, whereas an indirect transition is indicated if a straight-line graph is obtained for n = 4.

Direct transition plots for Ag₂S gave straight-line portions that give a bandgap of E_g,dir 2.4 eV consistent with that reported.⁸ Plots of CdS samples extrapolate to zero (hν) at E_g,_dir 2.62 eV (∼475 nm) and show a blue shift of 40 nm relative to bulk CdS with a bandgap of 2.41 eV (515 nm), this gives a particle size of approximately 10 nm by calculation.²⁰ For Cu_xSnanoparticles plots of (αhν)² gave an E_g,dir of 1.8 eV quite close to the 1.7 eV reported for this direct transition (Fig. 6A). A plot of (αhν)^1/2 against hν for absorption of CuS and Cu_xSnanoparticles gave straight-line portions that could be extrapolated to zero hν to give an indirect (E_g,ind) bandgap of 1.60 eV for the CuS which closely matches the 1.55 eV reported for the transition of this material (Fig. 6B).¹⁰ For Cu_xSnanoparticles, extrapolation gave an E_g,ind bandgap of 1.5 eV which is close to the value of 1.4 eV reported for bulk Cu_1.96S (djurleite) (Fig. 6B).¹⁰ This suggest the Cu_xSnanoparticles obtained in this synthesis most closely match this polymorph of Cu_xS. The indirect plot for Ag₂S could be extrapolated to give an E_g,ind of 1.1 eV (Fig. 6B).

Fig. 6 Plots of (αhν)² (direct transitions) as a function of photon energy (hν) for prepared Cu_xS, Ag₂S and CdS nanoparticles. B. Plots of (αhν)^1/2 (indirect transitions) as a function of photon energy (hν) for CuS, Cu_xS and Ag₂Snanoparticles.

Overall it was observed that, except for Ag₂S, bandgaps increased relative to bulk samples due to quantum size effects of the prepared nanoparticles. Also, the observed increase over bulk bandgaps corresponded to the relative sizes of nanoparticles obtained, with a marginal increase obtained with CuSnanoparticles and a more significant increase obtained with the smaller CdS nanoparticles. Sizes of prepared nanoparticles as measured by TEM after dextranase treatment and washing, the calculated surface areas and zeta-potentials are shown in Table 1.

Table 1 Table of prepared nanoparticles following enzymatic treatment showing average sizes, calculated surface areas and zeta-potential

Sample	Cation:anion molar ratio	Size/nm	Surface area/m^2g^−1	Zeta-potential/mV
CuS	1 : 1	27	47	−24.8
CuS	1 : 0.625	14	91	−27.4
CuS (onto Sephadex bead)	1 : 1	9.5	130	−23.2
CuxS	2 : 1	14	∼75	−8.49
Ag2S	2 : 1	20–50	∼25	+0.43
CdS	1 : 1	9	137	−8.08

---

2.2.1 Zeta-potential analysis . Zeta-potential measurements of dex–metal sulfide nanosuspensions all gave slightly negative surface charge indicating the surface charge was dominated by the bound dextran. Following dextranase treatment and washing of the sediment, zeta-potential measurements were in accordance with values reported for pure materials.³²CuS suspensions at both molar ratios indicated a moderately negative surface charge of approximately −26 mV for 27 nm sized and for 14 nm sized CuS. This moderate negative charge was reflected in the stability of these suspensions which showed little or no aggregation upon standing for several weeks. The copper rich Cu_xS and CdS were measured as having significantly less negative charge of with zeta-potential of approximately −8.5 mV. For Cu_xS this reflects the higher positively charged copper content of this material’s surface. Upon standing, nanosuspensions of Cu_xS and CdS showed a tendency to flocculate, however, upon very brief sonication good redispersion occurred and TEM showed that discrete nanoparticles of unchanged dimension were present. It is likely that the small quantity of remnant surface-bound dextran oligomer prevents aggregation and growth of the nanocrystals (Table 1).

3. Conclusions

Overall we have demonstrated a simple and green methodology for the preparation of copper sulfide in copper ‘poor’ and ‘rich’ forms, silver sulfide and also cadmium sulfide nanoparticles as stable nanosuspensions, which can have medical application for fluorescence imaging of biological samples.³⁰Dextranbiopolymer of M_r 500 000 was required to prevent aggregation of the precipitated sulfides during preparation. However, XRD, FT-IR and TGA data show that biopolymer-coated nanoparticles could subsequently be readily enzymatically liberated from the bulk of their biopolymer shells giving near-pure nanoparticles that were still stabilized by a remnant dextran oligomer coat.

Several factors are responsible for the dimensions of the nanoparticles obtained; the pH of the system, the relative solubilities of the precipitated sulfides, the relative reagent concentration, and the particular affinity of the metal sulfide surface for dextran. Dextran-to-metal-ion interaction has been previously reported for metal oxide systems.²¹ Here the systems are strongly basic and COO⁻ bonds strongly with the metal ion surface. At higher dextranM_r hydrogen bonding between the metal ion and dextranhydroxyls also becomes important.²²

For copper ions, hydrogen bonding to dextranhydroxyls has been shown to only begin occurring at pH 7 or higher.²⁷ For the systems prepared here, the pH of the reaction mixtures varied from pH 3–6 (due to the acidic nitrate salts), in this pH range dextranhydroxyls and terminal carboxyls will be protonated (the pK_a of dextranhydroxyls is ∼11).³³ Also, surface S is reported to exist as SH altering to include some S(metal)OH⁺ functionalities nearer to pH 6, due to the dissociative adsorption of water.³⁴

Except for Ag₂S, stabilisation of the nanoparticles during preparation was found to only become possible with dextran at a high M_r of 500 000, which suggests that in these systems complexation of the dextranbiopolymer to the sulfides was occurring by weak electrostatic interactions between surface S groups and dextranhydroxyls. The preparation steps and suggested metal sulfidebiopolymer interactions are represented in Scheme 1.

Scheme 1 Diagrammatic representation of stabilised metal sulfidenanoparticle preparation steps. Top: metal ions complexed by dextranhydroxyls. Middle: suggested metal sulfide complexation by hydroxyls of high M_r or crosslinked polymervia weak polar interactions. Bottom: metal sulfides released from the bulk of the biopolymer shells by enzymatic treatment. A low level of remnant oligomer inhibits subsequent aggregation and growth of the nanocrystal products.

Dextran polymer changes from an open-linear to tightly coiled structure in solution as its M_r increases from a few thousand to tens and hundreds of thousands.²⁰ The results suggest that highly coiled dextran, tightly enwrapping the sulfidenanoparticles was necessary for the weak S–H bonding to be sufficient to stabilize the nanoparticles. This is supported by related experiments in which fully carboxylated dextran gave only weak stabilisation of the copper sulfides. Thus it is likely that binding to the terminal carboxyl group of dextran chains of high-M_rdextran (500 000) made only a very minor contribution to complexion of nanocrystals for copper and cadmium systems. Crosslinked dextran as Sephadex beads was also demonstrated to stabilise CuSnanoparticles. Here the open meshwork of interconnected polymer chains at the bead surface presents a similar environment to the coiled high-M_rpolymer, and is able to stabilise nanoparticles almost as effectively as uncrosslinked dextran.

Ag₂S demonstrated a somewhat different complexing behaviour to the other materials prepared. It is possible that a silver sulfide–dextran complex of SAg₂OH₂⁺ at the pH 5 of this system exists, as although relatively larger and polydisperse particles were formed in this case, lower dextranM_r stablisation was possible. Furthermore, zeta-potential measurements also suggested a weakly positively charged surface for dextranase de-shelled Ag₂Snanoparticles.

For Cu_xS, resulting particle size may be complicated by the lower solubility of the CuCl reagent leading to smaller sized nanoparticles. For CdS, dextran in combination with dextranase enzyme appears to act as an enhanced dissolution system, here the smaller size of nanoparticles obtained may be partly the result of the continuous gradual dissolution of the capped nanoparticles. Zeta-potential measurements showed that dextran-capped nanoparticles all carried slightly negative surface charge at pH 6.3. Following dextranase treatment zeta-potential values were significantly altered and reflected the true surface charges of the native nanoparticles. TGA measurements also showed that prior to enzyme treatment the dried composites were ∼80–95 wt% biopolymer which was reduced to a few weight percent polymer by subsequent enzyme treatment. The materials prepared have application in photonics as materials for medical fluorescence imaging, nanoparticlebioassays , fourth generation nanoparticle and conductive polymer-based solar cells, and for use in conductive coatings. Further studies on developing the process into a large-scale general and green synthetic route to a range to useful nanoparticles are in progress.

4. Experimental

Dextran was purchased from Fluka Chemical Co and all other chemicals were purchased from Sigma-Aldrich Chemical Co. Dextranase (1,6-α-D-glucanohydrolase) was purchased as Penicillium sp. partially purified lyophilized powder. All chemicals were used as supplied.

4.1 Preparation of CuS–dextran nanosuspensions (dex–CuS)

Dextran-stabilised CuS was prepared by a combined dextrancarboxylation and precipitation reaction by modification of a published method for preparation of dextran-stabilised metal oxides.²² 2.5 g of Cu(NO₃)·2.5H₂O (0.01 M) and 8 g of dextran (M_r 500 000) were dissolved in 75 mL of deionised water with stirring. Then 2.4 g of Na₂S·9H₂O (0.01 M) was dissolved in 25 mL of distilled water and rapidly added with vigorous stirring. The black-brown mixture was stirred for 25 min, followed by heating in a standard 800 W microwave to 80 °C. The mixture was then cooled in an ice bath for 10 min followed by centrifugation at 5300 rpm for 30 min. The pH of the mixture was 3.0. The green coloured supernatant liquid was then purified by size exclusion chromatography (SEC). A Sephadex G-50 column was used with deionised water as an eluent. The initial 50 mL of coloured fraction was collected; the pH of the eluted solution was 6.5. The above process was then repeated using a lower reagent amount of 1.5 g Na₂S·9H₂O (0.00625 M). At lower concentrations of this reagent a green supernatant was not formed. (Microwave heating was employed to minimize heating time in order to eliminate the possibility of charring of the dextran reagent).

4.2 Preparation of CuS using Sephadex bead substrates (Sephadex–CuS)

CuS nanosuspensions were prepared as in section 4.1 using a 1 : 1 Cu:S molar ratio with the following modifications: 8 g of Sephadex G-50 (fine) beads were hydrated with 170 mL of water for use as a biopolymer substrate. After addition of reagents, stirring, microwave treatment and cooling, the flask was sonicated for 30 s. The contents was then filtered using a 90 micron mesh with thorough washing with distilled water. The retained material consisted purely of dark-green coloured Sephadex beads.

4.3 Preparation of Cu_xS–dextran nanosuspensions (dex–Cu_xS)

The above process was repeated with following modifications, firstly, nitrogen degassed water was used throughout the synthesis and 1.5 g of Na₂S·9H₂O (0.0625 M) was added to 1.25 g (0.0126 M) of CuCl to give a brown coloured solution. Following the centrifugation a single drop of hydrazine was added to the brown coloured supernatant liquid to inhibit oxidation into CuS. The pH of the mixture was 6, following SEC, the brown coloured eluent was pH 5.96.

4.4 Preparation of Ag₂S–dextran nanosuspensions (dex–Ag₂S)

The dex–CuS preparation method was used (using both dextran of M_r 70 000 or 500 000) with 1.5 g Na₂S·9H₂O (0.00625 M) added to 2.13 g AgNO₃ (0.0125 M) that formed a black coloured solution. The pH of the mixture was 5.1. Following SEC the grey coloured eluent was pH 5.6.

4.5 Preparation of CdS–dextran nanosuspensions (dex–CdS)

The dex–CuS preparation method was used with modifications. 1.5 g Na₂S·9H₂O (0.0625 M) was added to 1.93 g of Cd(NO₃)₂·4H₂O (0.0625 M) to form an orange coloured mixture, the mixture was then warmed to 60 °C by microwave heating. Note: Due to dissolution of CdS by enzyme treatment, filtration of the nanosuspension through the SEC column was found to be unnecessary. The pH was adjusted to 6 with dilute acetic acid solution and the suspension treated directly with enzyme.

4.6 Dextranase enzymatic de-shelling of dextran-stabilised sulphidenanoparticles

25 mL of the eluted sulfide nanosuspension, (or unfiltered dex–CdS nanosuspension) was warmed in a water bath to 34 °C. 250 units of freeze-dried dextranasePenicillium sp. was then added and the mixture held at 34 °C with gentle stirring for 24 h for dex–CuS, dex–Cu_xS and dex–Ag₂S nanosuspensions. For Sephadex–CuS 0.4 g of hydrated beads were subjected to enzyme treatment. Dex–CdS nanosuspensions were observed to sediment within 30 min. Treatment was limited to 5 h as longer dextranase treatment resulted in gradual dissolution of the CdS particles. After the respective time periods the mixtures were centrifuged and the supernatant removed. A coloured sedimented pellet was obtained in each case that was washed and centrifuged several times in deionised water.

Instruments. TEM was conducted for samples prepared on carbon-coated cooper grids using JEOL 1200EX and JEOL 2010FX microscopes with attached digital cameras. Thorough washing of enzyme-treated samples was necessary to remove residual supernatant maltoses and dextran oligomers that would otherwise coat onto nanoparticles upon drying of grids. Particle sizes were measured as the average of 10 crystals. Powder X-ray diffraction was recorded on a Bruker D8 Diffractometer. UV–vis absorbances were recorded on a Perkin-Elmer Lambda 25 spectrometer. The zeta-potentials of dextran-stabilized nanosuspensions and dextranase-treated suspensions, adjusted to pH 6.3 with dilute acetic acid or ammonia, were recorded on a Malvern Instruments Zetasizer Nano 2.

Acknowledgements

We thank the EPSRC (UK) ARF (Grant No. EP C544803) (DW) for financial support of this study and Lawrence C. Thomson for technical assistance.

References

1. W.-T. Yao and S.-H. Yu, Adv. Func. Mat., 2008, 18, 1.
2. Yu. Wu, C. Wadia, W. Ma, B. Sadtier and A. P. Alivisatos, Nano Lett., 2008, 8, 2551.
3. L. Isac, A. Duta, A. Kriza, S. Manolache and M. Nanu, Thin Solid Films, 2007, 515, 5755; I. Grozdanov, C. K. Barlingay and S. K. Dey, Integr. Ferroelectr., 1995, 6, 205.
4. I. Grozdanov and M. Nadjoski, J. Solid State Chem., 1995, 114, 469; K. S. Leschkies, R. Divakar, J. Basu, E. Enache-Pommer, J. E. Boercker, C. B. Carter, U. R. Kortshagen, D. J. Norris and E. S. Aydil, Nano Lett., 2007, 7, 1793.
5. B. Tian, X. Zheng, T. J. Kempa, Y. Fang, N. Yu, G. Yu, J. Huang and C. M. Lieber, Nature, 2007, 449, 885; W. J. E. Beek, M. M. Wienk and R. A. J. Janssen, Adv. Mater., 2004, 16, 1009; J. S. Chung and H. J. Sohn, J. Power Sources, 2002, 108, 226.
6. S. Y. Kuchmii, A. V. Korzhak, A. E. Raevskaya and A. I. Kryukov, Theor. Exp. Chem., 2001, 37(1), 36.
7. M. H. Kunita, E. M. Girotto, E. Radovanovic, M. C. Gonçalves, O. P. Ferreira, E. C. Muniz and A. F. Rubira, Appl. Surf. Sci., 2002, 202, 223.
8. F. Di Benedetto, M. Borgheresi, A. Caneschi, G. Chastanet, C. Cipriani, D. Gatteschi, G. Giovanni Pratesi, M. Romanelli and R. Sessoli, Eur. J. Mineral., 2006, 18, 283.
9. R. Chen, N. T. Nuhfer, L. Moussa, H. R. Morris and P. M. Whitmore, Gas. Nanotech, 2008, 19, 1; X. B. Wang, W. Liu, J. Hao, X. Fu and B. Xu, Chem. Lett., 2005, 34, 1664.
10. M. T. S. Nair, L. Guerrero and P. K. Nair, Sci. Technol, 1998, 13, 1164.
11. A. Nuñez Rodríguez, M. T. S. Nair and P. K. Nair, Semicond. Sci. Technol., 2005, 20, 576.
12. M. Salavati-Niasari, M. R. Loghman-Estarki and F. Davar, Chem. Eng. J., 2008, 145, 346; C. Li, B. Yang, Y. Y. Yang and Y. T. Qian, J. Cryst. Growth, 2006, 291, 45; P. Sreekumari Nair, N. Revaprasadu, T. Radhakrishnana and Kolawolea, J. Mater. Chem., 2001, 11, 1555.
13. P. S. Khiew, S. Radiman, N. M. Huang and M. Soot Ahamd, J. Cryst. Growth, 2004, 268, 227.
14. T. Kuzuya, S. Yamamuro, T. Hihara and K. Sumiyama, Chem. Lett., 2004, 33, 352.
15. C. M. Simonescu, V. S. Teodorescu, O. Carp, L. Patron and C. Capatina, J. Therm. Anal. Calorim., 2007, 88, 71.
16. F. Gao, Q. Lu and D. Zhao, Nano Lett., 2003, 3, 85; D. Brühwiler, S. Leiggener, S. Glaus and G. Calzaferri, J. Phys. Chem. B, 2002, 106, 3770.
17. M. P. Pileni, L. Motte, F. Billoudet, J. Mahrt and F. Willig, Mater. Lett., 1997, 31, 255.
18. G. T. Schaaff and A. J. Rodinon, J. Phys. Chem. B, 2003, 107, 10416.
19. J. Porath and P. Flodin, Nature, 1959, 183, 1657.
20. T. Kawaguchi, T. Hanaichi, M. Hasegawa and S. Maruno, J. Mater. Sci.: Mater. Med., 2001, 12, 121.
21. Q. Wei, S. Z. Kang and J. Mu, Colloids Surf., A, 2004, 247, 125; B. K. Božanić, V. Djoković, J. Blanuša, P. S. Nair, M. K. Georges and T. Radhakrishnan, Eur. Phys. J. E, 2007, 22, 51; J. F. Zhu, Y. J. Zhu, M. G. Ma, L. X. Yang and L. Gao, J. Phys. Chem., 2007, 111, 3920; S. H. Liu, X. F. Qian, J. Yin, L. Hong, X. L. Wang and Z. K. Zhu, J. Solid State Chem., 2002, 168, 259.
22. C. Lemarchand, R. Gref, S. Lesieur, H. Hommel, B. Vacher, A. Besheer, K. Maeder and P. Couvreur, J. Controlled Release, 2005, 108, 97; Y.-Y. Kim, K. Hore, S. R. Hall and D. Walsh, Small, 2009, 5, 913.
23. F. A. Erhardt and H.-J. Jordening, Immobilization J. Biotech., 2007, 131, 440; Y. Aslan and A. Tanriseven, Biochem. Eng. J., 2007, 34, 8.
24. S. A. U. Qader, A. Abid Azhar, L. Iqbal, A. Aman and E. Shireen, Turk. J. Biochem., 2005, 31, 21.
25. R. Zhang, M. Tang, A. Bowyer, R. Eisenthal and J. Hubble, Biomaterials, 2005, 26, 4677.
26. A. P. Davis, Y. H. Hsieh and C. P. Huang, Chemosphere, 1995, 31, 3093.
27. E. Norkus, J. Vaiciuniene, T. Vuorinen and M. Heikkilä, Carbohydr. Polym., 2002, 50, 159.
28. A. N. J. Heyn, Biopolymers, 1974, 13, 475.
29. S. Ding and W. T. Petuskey, Solid State Ionics, 1992, 58, 123.
30. P. Zhang and L. Gao, J. Mater. Chem., 2003, 13, 2007.
31. D. R. Larson, W. R. Zipfel, R. M. Williams, S. W. Clark, M. P. Bruchez, F. W. Wise and W. W. Webb, Science, 2003, 300, 1434.
32. J. C. Liu and C. P. Huang, Langmuir, 1992, 8, 1851.
33. C. Larsen, Int. J. Pharm., 1989, 52, 55.
34. L. Ronngren, S. Sjoberg, Z. Sun, F. Willis and P. W. Schindler, J. Colloid Interface Sci., 1991, 145, 396.

---

Footnote

† Electronic supplementary information (ESI) available: Figs. S1–S4. See DOI: 10.1039/b9nr00194h

---