Microchim Acta (2008) 160: 351–356 DOI 10.1007/s00604-007-0798-8 Printed in The Netherlands

Original Paper Organometallic synthesis and electrophoretic characterization of high-quality ZnS:Mn==ZnS core==shell nanoparticles for bioanalytical applications Oliver Ehlert1 , Wendelin Bu¨cking1 , Ju¨rgen Riegler1, Alexey Merkulov1, Thomas Nann2 1 2

Freiburg Materials Research Centre (FMF), University of Freiburg, Freiburg, Germany School of Chemical Sciences and Pharmacy, University of East Anglia (UEA), Norwich, United Kingdom

Received 3 April 2007; Accepted 26 April 2007; Published online 9 July 2007 # Springer-Verlag 2007

Abstract. A novel and facile preparation method for colloidal ZnS nanoparticles doped with Mn2þ is introduced, using a simple one pot heating process followed by a capping procedure for saturation of the surface bound doping atoms to increase the nanoparticles’ stability and photoluminescence quantum yield. The particles were transferred into water with a standard ligand exchange method and investigated by means of laser Doppler electrophoresis, agarose gel electrophoresis, and isotachophoresis. Keywords: Zinc sulphide; manganese; non-toxic nanocrystals; electrophoresis; bovine serum albumin (BSA)

Semiconductor nanoparticles (NPs) such as CdSe=ZnS core=shell, CdSe=ZnS=SiO2 core=shell=shell or CdTe [1–4] have received much interest within the recent years – most likely due to their size dependent optical properties. A major drawback for the application of such NPs is the toxicity of Cd2þ for living cells. Lesstoxic particles like doped and colloidal ZnS or ZnSe NPs would be interesting alternatives for biological imaging or other electro-optical applications. Even though, the high energy excitation requirements of Correspondence: Alexey Merkulov, Freiburg Materials Research Centre (FMF), University of Freiburg, Stefan-Meier-Strasse 21, D-79104 Freiburg, Germany e-mail: [email protected]

such particles are a drawback for bioanalytical applications, the advantages may outweigh. Recently, several groups have invented synthesis methods for ZnSe:Mn2þ NPs using diethyl zinc (Et2Zn), dimethyl manganese (Me2Mn) and solution of Se in tri-n-octylphosphine (TOPSe) [5] or Zn- and Mn-stearate and TOPSe [6]. Due to the fact that these synthetic procedures are based on (multiple)-injection methods the problem of formation of non-doped nanoparticles still exists. For the synthesis of doped ZnS NPs precipitation methods in water or other suitable solvents with different sulfur sources were reported. For a review of this big scientific field see Ref. [7] and references therein. The biggest problem of those synthetic procedures is the co-precipitation of the doping-metals’ sulfide. Additionally, in most cases nanocrystalline powders and not colloidal solutions were obtained. Recently, we have reported a procedure for the synthesis of ZnS NPs (co)-doped with Cu and=or Pb in a homogenious diluted synthetic system, which were additionally capped by shell of ZnS utilising Et2Zn and (TMS)2S in order to increase the amount of corebound doping metal cations [8]. This synthetic procedure was used here for the synthesis of ZnS:Mn2þ = ZnS core-shell NPs. Different amounts of doping metal have been investigated to find the optimum Mn2þ -ion concentration. The particles have been characterized by absorbance and photoluminescence

352

spectroscopy, X-ray diffraction, and TEM. Furthermore, the synthesised NPs were transferred into water by means of surface ligand exchange with different thiol-containing molecules, coated with bovine serum albumin protein (BSA), and characterized by laser Doppler electrophoresis (LDE), agarose gel electrophoresis (AGE), and isotachophoresis (ITP). The latter method is a capillary electrophoretic technique and allows to determine the electrophoretic mobility of nanocolloids without the influence of size neither due to the agarose gel network nor due to the influence of coagulated NPs in the LDE [9, 10]. Experimental Synthesis of core NPs All chemicals were of the highest purity grade available and used without further purification. MnCl2 (99.999%, Aldrich) was dissolved in small amounts of HPLC-grade N,N-Dimethylformamide (DMF, MultisolventTM, Scharlau Chemie S. A., Barcelona, Spain) to form an exactly concentrated solution. 300 mmol of anhydrous Znacetate (99.99%, Aldrich), 600mmol Oleic Acid (99%, Aldrich), and defined aliquots of the Mn2þ solution were dissolved in 20 mL Tri-noctylamine (TOA, 99%, Fluka) at 80
C, and the DMF was removed in vacuo. After the complete dissolving and degassing, 1.6 mmol of 1-Hexadecanethiol (97%, Merck) were added at 80
C with a syringe and the flask was heated to 300
C under nitrogen. The reaction time influenced the particle growth as depicted in Fig. 1. After the desired reaction time, the flask was cooled down to room temperature by means of compressed air and the particles were collected by addition of excess acetone (HPLC grade, ChromasolvTM, 99.9%, Aldrich) and centrifugation. The precipitated particles were washed several times with acetone and redispersed in chloroform or heptane. Shelling procedure For the shelling procedure the solvent of the purified particles was evaporated and they were dispersed in 2.5 mL Tri-n-octylphosphine

O. Ehlert et al. (TOP, 97%, Strem Chemicals) and 4 g tri-n-octylphosphineoxide (TOPO, 99%, Aldrich) at 80
C and degassed. 31 mL diethyl zinc (Aldrich) and 61 mL 1,1,1-3,3,3-hexamethyldisilathian (Aldrich) were dissolved in 2 mL TOP in a suitable nitrogen box and added to the flask drop wise at 130
C. The particles were stirred over night at 90
C. Afterwards 5 mL of 1-butanol were added and the flask was stirred for 4 more hours at 60
C. The nanoparticles were collected by addition of 10 mL of freshly distilled anhydrous methanol and centrifugation. Ligand exchange The stock solution of the synthesised ZnS:Mn2þ =ZnS NPs was prepared by dispersing them in 25 mL of CHCl3 (Aldrich). NPs maintained in an aliquot of 500 mL were precipitated by adding of 2 mL of acetone and suspended in 500 mL of DMSO (Aldrich) and centrifugation. After adding an excess of the ligand – 2-mercaptoethanol (ME, 20 mL, Fluka 99%), 3-mercaptopropionic acid (MPA, 20 mL, Fluka, 99%) or dithiothreitol (DTT, 10 mg, Pierce) – the sample was heated to 80
C for 3 h. The NPs were collected by addition of 1 mL Et2O, two drops of acetone and followed by centrifugation. The precipitate was dispersed in 500 mL tris-hydroxymethyl-aminoethane-glycine buffer (25 mM TRIS and 250 mM glycine, pH 8.3, both chemicals purchased by ROTH) for the gel electrophoretic experiments and the preparation of BSA-coated NPs. BSA-coated NPs The BSA-coated NPs were prepared by adding of 100 mL of 10 mg mL1 BSA (albumin, bovine serum, fraction V, approx. 99%, Sigma) solution to the 500 mL of ME-coated NPs and incubation at room temperature overnight. ITP measurements For the electrophoretic characterisation of the NPs an ItaChrom EA 101 (J&M Analytische Mess- und Regeltechnik, Aalen, Germany) with a two column setup was used. The injection volume was 30 mL. Thereby, for the pure ITP measurements the upper column was contacted with an electric source at a constant current of 250 mA and a voltage between 0.9 and 1.7 kV while for the lower column an electrical setup of 50 mA and a voltage between 1.3 and 8 kV was chosen. The analytes were detected by conductivity and UV absorption at 254 nm. The leading electrolyte was 10 mM HCl buffered with 60 mM TRIS to pH 9.1, the terminating electrolyte was 25 mM glycine. LDE measurements The LDE measurements were performed with a Nanosizer ZS (Malvern Instruments, Malvern, UK). The unique maintenance-free cells, developed for electrophoretic measurements in the Nanosizer ZS, were used. Samples were measured in TRIS–HCl buffer (see ITP measurements). Agarose gel

Fig. 1. Absorbance spectra taken after different reaction times, starting at 30 min in 30 min intervals

Agarose gel experiments were performed in TRIS-glycine buffer by constant voltage of 100V at room temperature utilizing agarose purchased by ROTH (agarose HEEO, ultra-quality). Gels were prepared using running buffer without dilution, sample buffer (5 stock solution, 50% glycerine in TRIS-glycine buffer) was used for loading (sample volume 15 mL).

353

Organometallic synthesis and electrophoretic characterization

Results and discussion Synthesis and characterisation of NPs NPs were synthesized according to the description in the experimental section, whereas the preparation was based on a previously published procedure reported for Cu and=or Pb (co-)doped ZnS nanocrystals [8]. We did not use an injection method at high temperatures, but a homogeniously diluted solution, which was heated to the desired reaction conditions. This procedure guarantees the homogeneous formation of the host lattice and the doping atoms within the reaction and prevents the nucleation of undoped ZnS or the formation of MnS during an injection process, which is the main disadvantage of the injection methods at high temperatures [11]. As it follows from absorbance spectra (Fig. 1) the ZnS particle growth can be easily adjusted by the reaction time. The particle size can be calculated according to the procedure published by Tiemann et al. [12]. To prevent quantum confined properties the particle growth was stopped after 150 min when the exciton radius of ZnS of 2.3 nm [13] should be exceeded. Figure 2 shows the photoluminescence spectra excited at 3.9 eV (320 nm) for NPs after 150 min reaction time. Both, the interstitial sulfur luminescence around 2.9 eV (420 nm) [14] and the bulk PL of undoped

Fig. 2. PL spectra taken after 150 min reaction time using different percentages of Mn2þ , excited at 3.9 eV (320 nm). The bulk ZnS PL at 3.5 eV (350 nm) can be seen in the 2% spectrum. The 2.19 eV peak (580 nm) ist increasing with increasing Mn2þ content

ZnS at 3.5 eV (350 nm), are decreased with increasing amount of manganese under these conditions. This indicates the increasing Mn2þ ion incorporation inside of the NPs or on the surface, which leads to the typical Mn2þ luminescence at about 2.19 eV (580 nm). Figure 3 shows on the left side a TEM micrograph of the core NPs with a mean particle diameter of about 5 nm (relative dispersion of approx. 7%). The X-ray

Fig. 3. Left: TEM picture of the core NPs. Right: XRD pattern of the NPs. The cubic sphalerite phase is indexed

354

O. Ehlert et al.

clear and transparent solution under daylight and an orange luminescent solution under UV excitation. Phase transfer and electrochemical characterization

Fig. 4. PL spectra before (dot line) and after (solid line) the shelling procedure as described in the experimental section

diffraction pattern of the NPs on the right side shows the cubic sphalerite phase and the highly crystalline nature of the nanocrystals prepared in this work. From this XRD data an average particle size of 4.8 nm is calculated using Scherrer’s equation. No co-nucleated MnS is observed in the XRD. This means that after 150 min the exciton radius of ZnS of 2.3 nm [13] is definitely exceeded and no quantum confinement effects can be detected anymore. Figure 4 displays the PL spectra of the core and core=shell NPs. The spectra provide evidences that the shelling procedure was successful, because the spectrum of the ‘‘shelled’’ NPs was comparable with the bulk one (data not shown), and the PL peak is much narrower after the procedure. This indicates the covering of surface bound Mn2þ ions by additional ZnS to increase the amount of core-bound Mn2þ and to decrease the number of surface bound doping atoms, which compete with the regular luminescence mechanism by non-radiative decay channels. This effect can also be explained by the results of Beermann et al. [15] who synthesized their ZnS:Mn2þ NPs by an aqueous precipitation method. They prepared ZnS NPs with only surface-bound Mn2þ and observed only low luminescence values compared to their core-doped particles, which also showed the typical PL for ZnS: Mn2þ as depicted in Fig. 2. This confirms that with our organometallic synthesis procedure core- and surface-doped NPs are yielded, while surface-bound Mn2þ ions turn into core-bound ions after the shelling procedure. The asprepared core=shell NPs exhibit the colloidal nature: a

The synthesised ZnS:Mn2þ =ZnS NPs can be used for biological labelling after a phase transfer into water like e.g. CdSe=ZnS NPs, since they are completely dispersible in non-polar solvents and possess comparable or the same surface ligands like their CdSe counterparts. In order to show this similarity, the phase transfer of the synthesized ZnS:Mn2þ =ZnS NPs into water was successfully accomplished using three SH-group containing ligands: 2-mercaptoethanol (ME), dithiothreitol (DTT), and 3-mercaptopropionic acid (MPA) as it is known from the literature [16, 17] and is circumstantially described in the experimental section. It was also found that the ‘‘non-shelled’’ Mn2þ doped NPs lose their luminescent properties after application of the same procedure (data are not shown). This observation confirms that some Mn2þ ions are concentrated on the surface of NPs after the first preparation step and are simply washed from the surface during the phase transfer step in the absence of a protecting ZnS layer. The LDE measurements of the electrophoretic mobility were performed in TRIS–HCl buffer (pH 9.1), because the same buffer was also used for the ITP experiments. According to LDE (Fig. 5) the ME- and MPA-coated ZnS:Mn2þ =ZnS NPs exhibit almost the same electrophoretic properties as the CdSe=ZnS NPs of the comparable size with the same ligands [18]. In contrast to this the DTT-coated ZnS:Mn2þ =ZnS NPs reveal a very broad distribution with a maximum at the lower absolute value (see Table 1). The detection in LDE could be influenced by coagulated particles due to the Rayleigh scattering. Therefore, only a little amount of bigger particle clusters would falsify the result of the measurement.

Fig. 5. Electrophoretic mobility of ZnS:Mn2þ =ZnS NPs coated with ME (1), MPA (2) and DTT (3) measured by LDE

Organometallic synthesis and electrophoretic characterization

355

Table 1. Electrophoretic mobility of ZnS:Mn2þ =ZnS NPs with different coating ligands measured by LDE (peak maximum) and ITP Electrophoretic mobility (m  109 m2 V1 s1 )

ZnS:Mn2þ =ZnS=ME ZnS:Mn2þ =ZnS=DTT ZnS:Mn2þ =ZnS=MPA

LDE

ITP

22 12 24

28 23 26

Fig. 7. 1% agarose gel image on an UV-table (TRIS-glycine buffer, 100V). Samples in vials from left to right (1) and (2): ME-coated CdSe=ZnS; (3) and (4) non-transferred ZnS:Mn2þ =ZnS directly from the synthesis; (5) and (6) ME-coated ZnS:Mn2þ =ZnS, the image reveals that the NPs are coagulated to small extent forming trail with lower mobility

Fig. 6. Conductivity (upper line) and UV-absorption at 254 nm (lower line) of an isotachophoretic measurement of ZnS:Mn2þ = ZnS=DTT NPs. Leading electrolyte (LE): 10 mM HCl, buffered with 60 mM TRIS to pH 9.1, terminating electrolyte (TE): 25 mM glycine. Current: 50 mA

ITP measurements with all described ligands were performed to examine the electrophoretic properties of the NPs in more detail. For example, Fig. 6 shows an isotachopherogram for the ME ligand, in which the NPs were detectable either by conductivity or UV detection. The conductivity of the zones represents linearly the electrophoretic mobility between leader (Cl ), terminator (NH2CH2COO , glycine) and hydrogen carbonate (HCO3  ) anions, the latter being almost inescapably present in the basic conditions. The changeable conductivity of the zone containing NPs (Fig. 6) indicates a distribution in the electrophoretic mobility either due to the ligands, size dispersity or due to coagulated NPs. For the evaluation of the conductivity, the values at the upper end of the zone were used, representing the particles with the lowest electrophoretic mobility. There is a good agreement between the results of LDE and ITP measurements as displayed in Table 1 with exception of DTT-coated NPs. These NPs exhibit a very broad distribution of electrophoretic mobility according to LDE measurements that could be a reason for the observed disagreement. Above all, the relative

sequence in electrophoretic mobility of the NPs with different ligands was observed in LDE and ITP and different electrophoretic methods gave comparable results with high reproducibility. The migration mobility of water dispersable ZnS:Mn2þ =ZnS NPs compared to non-transferred NPs and ME-coated CdSe=ZnS NPs in agarose gel is demonstrated in Fig. 7. The experiment shows that ZnS:Mn2þ =ZnS NPs possess almost the same migration mobility as ME-coated CdSe=ZnS NPs of the comparable size, but are coagulated to small extent forming trail with lower migration ability, while non-transferred ZnS:Mn2þ =ZnS NPs ‘‘stick’’ on the vial’s edge. Unfortunately, the colloidal stability of the described NPs could be still not enough for applications in water due to the degradation and dissociation processes. Due to this fact we pursue a plan to stabilise NPs with ZnS surface modification using cheap and easily available Bovine Serum Albumin (BSA) protein, because this method works perfectly for CdSe= ZnS NPs [19]. As revealed by the LDE and ITP experiments, ME-coated ZnS:Mn2þ =ZnS NPs were the most suitable for further experiments and were used for preparation of BSA-coated NPs by simple addition of BSA to ZnS:Mn2þ =ZnS NPs dispersed in TRIS-glycine buffer. The obtained BSA-coated ZnS:Mn2þ =ZnS NPs were characterized by gel elec-

356

Organometallic synthesis and electrophoretic characterization and Research (BMBF) within the project 13N8644 for financial support. The help of Dr. Martin Ade, Institute for Inorganic and Analytical Chemistry, for the XRD measurements, and Dr. Ralf Thomann for the TEM pictures, Institute for Macromolecular Chemistry, both University of Freiburg, Germany, are gratefully acknowledged.

References

Fig. 8. 4% agarose gel image on an UV-table (TRIS-glycine buffer, 100V). Samples in vials from left to right (1) BSA (B); (2) and (3) ME-coated ZnS:Mn2þ =ZnS (C); (4) BSA-coated ZnS:Mn2þ = ZnS (A) and BSA excess (B). Dark spots in vials (2) and (3) are the loading dye – Coomassie Blue. Coomassie Blue forms a weak fluorescent complex with BSA, which can be recognized under UV-light in vials (1) and (4)

trophoresis (Fig. 8). This experiment confirmed that BSA adheres to the ZnS surface and reproduces the migration mobility sequence obtained for CdSe=ZnS NPs [19]: ME-coated ZnS:Mn2þ =ZnS>BSA>BSAcoated ZnS:Mn2þ =ZnS. Conclusion In conclusion we reported a new and facile organometallic preparation method for colloidal ZnS nanoparticles doped with Mn2þ . The particles display the typical orange Mn2þ -luminescence around 580 nm and possess cubic sphalerite crystal structure. The additional shelling with ZnS improves the luminescence stability and is required for the maintaining of the luminescent properties after phase transfer into aqueous media. The ZnS:Mn2þ =ZnS NPs were transferred into water using thiol-containing ligands exhibiting the similarity in the surface properties between ZnS:Mn2þ =ZnS and CdSe=ZnS NPs. The biological capability of the fluorescent ZnS:Mn2þ =ZnS NPs was demonstrated by preparation of BSA-coated NPs. Future work includes the isolation of purified BSAcoated ZnS:Mn2þ =ZnS NPs after gel electrophoresis and the preparation of non-toxic conjugates for luminescent cell imaging within this system. Acknowledgements. We thank German Research Foundation (DFG, Grant No. NA 373=4-1) and German Federal Ministry of Education

1. Darbandi M, Thomann R, Nann T (2005) Hollow silica nanospheres: In situ, semi-in situ, and two-step synthesis. Chem Mater 17: 5720 2. Gerion D, Pinaud F, Williams S, Parak W, Zanchet D, Weiss S, Alivisatos A P (2001) Synthesis and properties of biocompatible water-soluble silica-coated CdSe=ZnS semiconductor quantum dots. J Phys Chem B 105: 8861 3. Nann T, Mulvaney P (2004) Single quantum dots in spherical silica particles. Angew Chem Int Ed 43: 5393 4. Kumar S, Nann T (2003) Hexagonal CdTe nanoparticles of various morphologies. Chem Commun: 2478 5. Zu L, Norris D J, Kennedy T A, Erwin S C, Efros A L (2006) Impact of ripening on manganese-doped ZnSe nanocrystals. Nano Lett 6: 334 6. Pradhan N, Battaglia D M, Liu Y, Peng X (2007) Efficient, stable, small, and water-soluble doped ZnSe nanocrystal emitters as non-cadmium biomedical labels. Nano Lett 7: 312 7. Hu H, Zhang W (2005) Synthesis and properties of transition metals and rare-earth metals doped ZnS nanoparticles. Opt Mater 28: 536 8. Ehlert O, Osvet A, Batentschuk M, Winnacker A, Nann T (2006) Synthesis and spectroscopic investigations of Cu- and Pb-doped colloidal ZnS nanocrystals. J Phys Chem B 110: 23175 9. Bu¨cking W, Nann T (2006) Electrophoretic analysis of gold nanoparticles: size-dependent electrophoretic mobility of nanoparticles. IEE Proc Nanobiotechnol 153: 47 10. Bu¨cking W, Ehlert O, Riegler J, Rettenmaier C, Merkoulov A, Nann T (2007) Isotachophoretic measurements of luminescent semiconductor nanocrystals. Biophys Rev Lett 2(1): 99 11. Pradhan N, Gorskeey D, Thessing J, Peng X (2005) An alternative of CdSe nanocrystal emitters: pure and tunable impurity emissions in ZnSe nanocrystals. J Am Chem Soc 127: 17586 12. Tiemann M, Weiss O, Hartikainen J, Marlow F, Linden M (2005) Early stages of ZnS nanoparticle growth studied by insitu stopped-flow UV absorption spectroscopy. Chemphyschem 6: 2113 13. Yoffe A D (2002) Semiconductor quantum dots and related systems: electronic, optical, luminescence and related properties of low dimensional systems. Adv Phys 51: 799 14. Denzler D, Olschewski M, Sattler K (1998) Luminescence studies of localized gap states in colloidal ZnS nanocrystals. J Appl Phys 84: 2841 15. Gonzales Beermann P A, McGarvey B R, Skadtchenko B O, Muralidharan S, Sung R C W (2006) Cationic substitution sites in Mn2þ -doped ZnS nanoparticles. J Nanoparticle Res 8: 235 16. Chan W C W, Nie S (1998) Quantum dot bioconjugates for ultrasensitive nonisotopic detection. Science 281: 2016 17. Riegler J, Ehlert O, Nann T (2006) A facile method for coding and labelling assays on polystyrene beads with differently colored luminescent nanocrystals. Anal Bioanal Chem 284: 645 18. Bu¨cking W, Riegler J, Ziegler J, Hins J, Merkoulov A, Nann T (2007) Electrophoresis of CdSe=ZnS nanoparticles with different coating ligands (manuscript in preparation) 19. Bu¨cking W, Merkoulov A, Nann T (2007) Preparation and characterization of colloidal BSA-coated CdSe=ZnS nanoparticles (manuscript in preparation)