WO2009080522A1

WO2009080522A1 - Synthesis of au, pd, pt or ag nano- or microcrystals via reduction of metal salts by cellulose in the ionic liquid 1-butyl-3-methyl imidazolium chloride

Info

Publication number: WO2009080522A1
Application number: PCT/EP2008/067216
Authority: WO
Inventors: Andreas Taubert; Zhonghao Li
Original assignee: Universität Potsdam
Priority date: 2007-12-19
Filing date: 2008-12-10
Publication date: 2009-07-02
Also published as: EP2222883A1; EP2222883B1

Abstract

The invention refers to a preparation method for nano- or microcrystals of Au, Pd, Pt or Ag comprising the steps of: a) preparing an ionic liquid mixture of cellulose and at least one metal salt selected form the group consisting of Au, Pd, Pt and Ag in l-butyl-3-methylimidazolium chloride; and b) (i) thermally inducing a reduction of the metal salt with cellulose by heating of the mixture to a temperature in the range of 50 to 250°C; or (ii) photoreduction of the metal salt by irradiation of the mixture with light having a wave length in the range of 200 to 800 nm.

Description

Synthesis of Au, Pd, Pt or Ag nano- or microcrystals via reduction of metal salts by cellulose in the ionic liquid l-butyl-3-methyl imidazolium chloride

Among others, the invention deals about a simple and partly sustainable approach towards nano- or microcrystals of Au, Pd, Pt and Ag with defined properties.

Technological Background and State of the Art

Ionic liquids (ILs) have successfully been used in organic, inorganic, and electrochemical synthesis of new or improved materials. In particular, inorganic materials chemistry in ionic liquids has recently attracted quite some attention. This is due to the fact that in ILs, some inorganic compounds can be prepared easily and occasionally with superior properties than via conventional pathways.

It has recently been shown that ILs with a long range order (ionic liquid crystals) can not only act as efficient solvents or templates for inorganic materials synthesis, but also as "all-in-one" solvent-template-reactants, so-called ionic liquid crystal precursors (ILCPs; see A. Taubert Angew. Chem. Int. Ed. 2004, 43, 5380). Zhu et al. have shown that the solvent-template-reactant principle is also applicable to ionic liquids without a long range order (see H. Zhu; J.-F. Huang; Z. Pan; S. Dai Chem. Mater. 2006, 18, 4473). They have termed their reactive ILs "ionic liquid precursors" (ILPs), in analogy to the ILCPs. Some IL(C)Ps have been studied in quite some detail, in particular the CuCl platelet formation from ascorbic acid-containing ILCPs (see e.g. A. Taubert; C. Palivan; O. Casse; F. Gozzo; B. Schmitt J. Phys. Chem. C 2007, 111, 4077). Few reports have shown that silver and gold can be grown from IL(C)Ps (see A. Taubert; I. Arbell; A. Mecke; P. Graf Gold Bulletin 2006, 39, 205; K.-S. Kim; S. Choi; J.-H. Cha; S.-H. Yeon; H. Lee J. Mater. Chem. 2006, 16, 1315; W. Dobbs; J.-M. Suisse; L. Douce; R. Welter Angew. Chem. Int. Ed. 2006, 45, 4179; CK. Lee; CS. Vasam; T.W. Huang; H.M.J. Wang; R.Y. Yang; CS. Lee; I.J.B. Lin Organometallics 2006, 25, 3768).

Gold particles are among the best-studied particles in modern materials science. This is due to the ease of their preparation, their high (chemical) stability, and the wide range of applications from catalysis to sensing and biological tagging. There are countless examples of spherical and near-spherical gold particles, including, for example, spherical particles with bimodal size distributions and porous self-assembled solid state structures. However, the number of publications on polyhedral gold particles and on gold plates is comparatively low. Gold plates have been fabricated via wet chemistry (X. Sun; S. Dong; E. Wang Angew. Chem. Int. Ed. 2004, 43, 6360), via biological methods (see S. S. Shankar; A. Rai; B. Ankamwar; A. Singh; A. Ahmad; M. Sastry Nature Mater. 2004, 3, 482), via a polyol process (see CC Li; WP. Cai; B.Q. Cao; F.Q. Sun; Y Li; CX. Kan; L.D. Zhang Adv. Fund. Mater. 2006, 16, 83), via a solid state reaction within an ILCP (A. Taubert; I. Arbell; A. Mecke; P. Graf Gold Bulletin 2006, 39, 205), and via microwave and thermal reactions in the ILs l-butyl-3 -methyl- imidazo Hum ([BMIM]) hexafluorophosphate and tetrafluoroborate ([BMIM]PF₆ and [BMIM]BF₄), respectively (see Z. Li; Z. Liu; J. Zhang; B. Han; J. Du; Y. Gao; T. Jiang J. Phys. Chem B 2005, 109, 14445).

Truncated gold tetrahedral, cubes, and icosahedra have been prepared by a polyol process. Gold octahedra were synthesized by thermal decomposition of HAuCl₄ in block copolymer micelles. Decahedra were synthesized by ultrasound- induced reduction Of HAuCl₄ on pre-synthesized gold seeds with poly(vinylpyrrolidone) (PVP) as a stabilizing polymer.

Among others, the shapes, sizes, size distributions, and therefore the physical properties of gold particles strongly depend on the reducing agent used in particle synthesis. Common reducing agents are, for example, NaBH₄, PVP, glycol, and ascorbic acid. Carbohydrates can also act as reducing agents, but their limited solubility in water or organic solvents prevented their use in the past. Summary of Invention

The present invention shows that solutions of cellulose and metal salts can be transformed into metal particles with a tunable structure. The IL can be recycled

(although in the current case, this will be a major challenge due to the presence of many small organic residues and some remaining metal salt in the IL), cellulose is a renewable raw material, the reaction temperatures are usually below 220 ⁰C, and the only side products are the oxidation products of cellulose and the metal salt. The present invention therefore introduces a cheap, simple, and at least partly sustainable process towards metallic microparticles (including also nanostructures of the metals). The unique solubility of cellulose in ionic liquids simplifies processing and chemical transformation of this otherwise hard-to -process biological material.

According to the present invention, there is provided a method of preparing nano- or microcrystals of Au, Pd, Pt and Ag (especially gold microcrystals) comprising the steps of: a) preparing an ionic liquid mixture of cellulose and at least one metal salt selected form the group consisting of Au, Pd, Pt and Ag in l-butyl-3-methylimidazolium chloride; and b) (i) thermally inducing a reduction of the metal salt with cellulose by heating of the mixture to a temperature in the range of 50 to 250⁰C; or (ii) photoreduction of the metal salt by irradiation of the mixture with light having a wave length in the range of 250 to 800 nm.

Preferably, the metal salt is selected from the group consisting of M(NOs)_x, MCl_x, MBr_x, MI_x, M(OAc)_x, M(TfO)_x, M(acac)_x and HAuCl₄ * 3 H₂O, wherein M represents Au, Pd, Pt or Ag and x is an integer from 1 to 4. Most preferred, the metal salt is HAuCl₄ * 3 H₂O.

Preferably, the cellulose is present in the mixture in an equimolar amount to or in a molar excess to the metal salt. In particular, a molar ratio of cellulose to metal salt may be from 1 :1 to 20:1. According to another preferred embodiment of the invention, the metal salt is an Au salt and step b) (i) of thermally inducing the reduction is performed at a temperature in the range of 180 to 220⁰C for preparing gold microcrystals having a plate-like shape and a plate thickness in the range of 700 to 1.000 nm.

Another aspect of the invention is directed to nano- or microcrystals of Au, Pd, Pt or Ag, especially gold microcrystals, prepared by the above described method.

Further, there is provided a cellulose product containing nano- or microcrystals of Au, Pd, Pt or Ag, especially gold microcrystals, prepared by the above described method. The formation of Au, Pd, Pt or Ag particles in the ionic liquid l-butyl-3 -methyl imidazolium chloride in the presence of cellulose has been studied. In this reaction, cellulose is the reducing agent for the Au(III)-Au(O) reduction. At the same time, cellulose is a template for the formation of the metal particles. The metal particle morphologies and sizes mainly depend on the reaction temperature or irradiation parameters. Morphology and size are less affected by the cellulose or metal precursor salt concentration. The change in the particle shapes can be assigned to the role of the cellulose as a template in conjunction with an effect provided by the ionic liquid.

Overall, the approach reported here presents a sustainable approach towards nano- and microparticles of Au, Pd, Pt or Ag.

The unique ability of ionic liquids to dissolve cellulose can be exploited for the fabrication of for example gold microparticles via the thermally induced reduction of an Au(III) salt by cellulose or the photoreduction by irradiation with UV light. Because of the high thermal stability of the IL, the reaction can be conducted at various temperatures, which enables the tuning of the reaction in terms of particle sizes, shapes, and connectivity. The change of the particle shapes can be assigned to the role of the cellulose as a template in conjunction with an effect provided by the ionic liquid.

In summary, the approach reported here presents a simple and partly sustainable approach in particular towards nano- and microparticles of Au, Pd, Pt or Ag with defined properties. It uses a metal salt and a reducing agent/template from renewable raw materials. The only side products of the reaction are oxidized cellulose fragments and oxidation products from the metal salt. In principle, the IL can be recycled, although purification may cause some difficulties because of the presence of small organic fragments from cellulose decomposition and the further presence of inorganic ions from the metal, respectively gold salt precursor.

As to the process of photoreduction, applicant refers to the article of Taubert et al, Gold Bulletin 2006, pages 205 - 211 wherein it was demonstrated that a complex Au (III) can be reduced by irradiation of UV light to gold platelets. It has now been found that a similar photoreduction can be achieved by irradiation of Au, Pd, Pt or Ag salts, especially Au(III) salts, which are templated by cellulose. However, a clear-cut explanatory statement on the mechanism is still a point of research. The disclosure of the above mentioned article of Taubert et al. with respect to the conditions of irradiation is herewith incorporated by reference.

Brief Description of the Drawings

The invention will now be described in more detail by means of examples and the corresponding figures. These figures illustrate: Figure 1 shows an X-ray diffraction pattern of a sample prepared via heating a solution of 20 mg of HAuCl₄*3 H₂O and 20 mg of cellulose in 1 g of [BMIM]Cl for 20 h at 110 ⁰C. Inset is a magnified view of the (111) and (200) reflections showing their narrow FWHM. Figure 2 shows SEM images of gold microcrystals prepared at different temperatures, (a) 110 ⁰C, (b) 130 ⁰C, and (c) 200 ⁰C. AU samples were grown from a solution containing 20 mg of HAuCl₄ * 3 H₂O and 20 mg of cellulose in 1 g of [BMIM]Cl. Reaction time was 20 hours. Circles in panel (b) highlight connections between individual particles. Inset in panel (c) is a magnified view of the same sample showing the rough surface and the connection between two gold plates.

Figure 3 shows SEM images of gold crystals prepared at different cellulose concentrations, (a) 40 mg of cellulose, (b) 100 mg of cellulose/g of [BMIM]Cl.

Figure 4 shows a high magnification SEM of gold particles precipitated from solutions of (a) 20 mg of cellulose and (b) 100 mg of cellulose/g [BMIM]Cl at 110 ⁰C. Figure 5 shows an TEM image (a) and electron diffraction pattern (b) of a gold particle precipitated from a solution of 100 mg of cellulose and 20 mg Of HAuCl₄ * 3 H₂O/g [BMIM]Cl at 110 ⁰C.

Figure 6 shows TGA and DTA curves of a sample grown from a solution containing 20 mg of gold salt precursor and 20 mg of cellulose/g [BMIM]Cl at 110 ⁰C. Note that the x-axis in the TGA graph is only from 96 to 101 wt%.

Figure 7 shows the Lag time until the first observation of gold particles, (a) Effect of initial cellulose concentration. Gold salt concentration is 20 mg/g [BMIM]Cl. (b) Effect of initial gold salt concentration. Cellulose concentration is 20 mg/g [BMIM]Cl. For clarity, only the error bars for one experiment are shown.

Figure 8 shows an SEM image of gold plates prepared from solutions of 40 mg of HAuCl₄ * 3 H₂O in 1 g of [BMIM]Cl via heating for 20 h at 200 ⁰C.

Figure 9 shows SEM images of Pt nanoparticles formed at 130 ⁰C (top) and 200 ⁰C (bottom) after 20 hours of heating. Figure 10 shows XRD pattern of Pd grown with the thermochemical approach at

200 ⁰C.

Detailed Description

Characterization. X-ray diffraction was done on a Nonius PDS 120 with CuKα radiation and position sensitive detector and on a Nonius D8 with CuKα radiation. SEM was done on a LEO 1550 Gemini operated at 20 kV. TGA and DTA were done on a Linseis L81 thermal analyzer working in perpendicular mode from 25 to 1400 ⁰C in air. Calibration was done with AI2O3. Optical microscopy was done with Zeiss Primo star at 20, 40, and 100 x. General approach of photochemical synthesis. In a typical synthesis, 10 to 200 mg of cellulose (Acros) and 10 to 100 mg of M(NO₃)_X, MCl_x, MBr_x, MI_x, M(OAc)_x, M(TfO)_x, M(acac)_x (M = Au, Pd, Pt, Ag, x = 1 - 4) were mixed with 1 g of l-butyl-3-methylimidazolium chloride (Acros). The mixture was heated to 30 ⁰C until a clear solution formed. Then the solutions were irradiated at room temperature using a 400 W UV lamp or a regular halogen office lamp. With UV lamp, a color change could be observed after a few minutes; with the office lamp the same color changes took began at about an hour. Irradiation was continued for several hours to complete the reduction. The products were recovered by repeated centrifugation and washing with water and ethanol, respectively, and drying at 60 ⁰C for 5 hours.

General approach to thermochemical synthesis. In a typical synthesis, 10 to 200 mg of cellulose (Acros) and 10 to 100 mg of M(NO₃)_X, MCl_x, MBr_x, MI_x, M(OAc)_x, M(TfO)_x, M(acac)_x (M = Au, Pd, Pt, Ag, x = 1 - 4) were mixed with 1 g of l-butyl-3-methylimidazolium chloride (Acros). The mixture was heated to the desired temperature in the range of 50 to 250 ⁰C until a clear solution formed. Heating was then continued for 20 hours. The products were recovered by repeated centrifugation and washing with water and ethanol, respectively, and drying at 60 ⁰C for 5 hours. Synthesis of Au microcrystals. In a typical synthesis, 20 mg of cellulose (Acros) and 20 mg of HAuCU * 3 H₂O (Sigma) were mixed with 1 g of l-butyl-3-methylimidazolium chloride (Acros). The mixture was heated to the desired temperature until a clear solution formed. Heating was then continued for 20 hours. The products were recovered by repeated centrifugation and washing with water and ethanol, respectively, and drying at 60 ⁰C for 5 hours.

All samples were prepared by dissolving an appropriate amount of HAuCl₄ * 3

H₂O and cellulose in [BMIM]Cl. Upon reaction of this mixture between 100 and

200 ⁰C, gold precipitates. The shiny particles deposit at the bottom of the reaction vessel and can be isolated via centrifugation. The lag time before precipitation depends on the reactant concentrations and the reaction temperature.

Figure 1 shows a typical X-ray diffraction (XRD) pattern of a sample recovered from a solution after 20 hours. All products are pure face-centered cubic (fee) gold (JCPDS 04-0784). All XRD patterns exhibit narrow reflections with full widths at half maximum (FWHM) below 0.2 degrees 2 . Estimations of the crystallite size (the coherence length) using the Scherrer equation give values well above 200 nm. Therefore, the Scherrer equation is not applicable anymore and the crystallite sizes are beyond of what can be determined from XRD. The relative intensities of the five gold reflections, however, differ from what is expected for a purely isotropic bulk gold sample. The (200)/(l 11) and (220)/(l l l) intensity ratios are 0.041 and 0.019, respectively. This is lower than the values reported for bulk, isotropic gold samples (0.52 and 0.32, JCPDS 04-0784) and suggests that the resulting gold particles are dominated by (111) facets. Usually, a high intensity of the (111) reflection and the absence of other reflections is an indication of plate-like crystals with very large (111) faces. However, as we also observe other reflections than (111), XRD indicates that the samples are not (or not entirely) plate-like, but rather have other morphologies that are dominated by (111) facets. Figure 2 and Table 1 show the effect of reaction temperature on the particle morphologies. Scanning electron microscopy (SEM) clearly shows that the samples are not uniform and contain particles with a variety of shapes and sizes in the micrometer range. This is consistent with XRD, as there, the narrow reflections and the presence of reflections besides (111) indicate that the particles are not nanoparticles and not only plate-like. Besides plates, particles with octahedral, decahedral, twinned polyhedral, and only partially developed tetrahedral shapes are observed in the samples prepared at 110 ⁰C. At higher reaction temperatures, the fraction of plates increases to ca. 100% at 200 ⁰C. Inversely, the number of polyhedral particles decreases to close to zero at a reaction temperature of 200 ⁰C. SEM further shows that, besides the number of plates, also the plate thickness increases from ca. 300 nm at 110 ⁰C to ca. 800 nm at 200 ⁰C. Furthermore, samples prepared at 130 ⁰C contain larger crystal aggregates (highlighted by circles in Figure 2b). They are not present as individual crystals anymore, but exhibit connections between several particles. These trends are increasingly found at higher reaction temperatures. Samples obtained at 160 and 200 ⁰C essentially consist of large plates with diameters of over 15 μm that form a "molten- looking" connected network.

Further investigation of the SEM images reveals that with increasing temperature, the surface of the precipitates becomes rougher and the surfaces exhibit steps and other surface defects, which are not observed at lower temperatures. Overall, higher reaction temperatures lead to samples where, although the individual particles have a rather ill-defined morphology, the overall morphology of the material on a micrometer to hundreds of micrometer scale is more uniform than at lower reaction temperatures. Table 1. Approximate number fraction f(shape) (%) of particles with different shapes determined from SEM and plate thickness as a function of reaction temperature. Data are given for samples grown with 20 mg of HAuCU * 3 H₂OZg [BMIM]Cl and 20 mg of cellulose/g [BMIM]Cl.^a Temperature f(plate) Plate thickness f(oct) f(dec) f(poly) f(tet) (⁰C) (nm)^b

110⁰C 6 300 ± 40 26 10 49 9

130⁰C 19 250 ± 30 29 1 44 7

160⁰C 36 750 ± 50 64

200⁰C Ca. 100 800 ± 50 a f(plate): fraction of plates; f(oct): fraction of octahedral particles; f(dec): fraction of decahedra; f(poly): fraction of complex (twinned) polyhedra; f(tet): fraction of (truncated and poorly developed) tetrahedra. b Thicknesses of plates obtained at 160 and 200 ⁰C are approximate because the plates are not smooth or often do not have a clear edge, which makes an accurate thickness determination more difficult than in the other samples.

Figure 3 and Table 2 show the effect of cellulose concentration on the particle morphology. At low cellulose concentration, the samples again consist of plates and polyhedral gold particles. As the cellulose concentration increases, the amount of gold plates decreases slightly, but all samples obtained at low temperatures still mainly contain polyhedral gold crystals. Moreover, although the number of gold plates slightly decreases with increasing concentration, the other particles still have many different shapes. There is thus no focusing effect in the sense that above a certain threshold cellulose concentration, there are only, for example, octahedral particles in the sample. Variation of the gold salt concentration reveals that also here, there is no focusing on a certain morphology or particle size with increasing gold concentration (data not shown). In summary, SEM shows that the strongest effect on mineralization is exerted by the reaction temperature. At higher temperatures, there are clear morphology changes from polyhedral particles to interconnected plates with rough surfaces and rounded edges. In contrast, changes in either the gold salt or the cellulose concentration have rather incremental effects in that the relative number of different particle morphologies seems to change only slightly. Table 2. Approximate number fraction (%) of particle morphologies as a function of cellulose concentration. Data are given for samples grown with 20 mg of HAuCU * 3 H₂O/g [BMIM]Cl and different cellulose concentrations at 110 ⁰C. mg cellulose/g [BMIM]Cl f(plate) f(oct) f(dec) f(poly) f(tet)

40 6 24 9 53 8

100 3 28 8 55 6

Figure 4 shows that the cellulose does, however, lead to a peculiar variation in the gold particles. Even though the overall morphologies and particle sizes do not depend on the cellulose concentration, particles grown at low cellulose concentrations and temperatures below 160 ⁰C have a flat and smooth surface. In contrast, particles grown at higher cellulose concentrations have a rougher surface. SEM suggests that these structures could be due to adsorbed and mineralized cellulose because some of the structures resemble fibers deposited on a surface.

TEM shows that the particles are single crystal-like (Figure 5). The diffraction patterns of individual particles can either be assigned to a single crystal or to a twinned particle. TEM often finds pentagonal particles, which are 2D projections of the decahedra found in SEM. The decahedron is a typical twin crystal, which can be considered as a junction of five tetrahedra with twin-related adjoining faces.

Figure 6 shows thermal analysis data of some samples. Thermogravimetric analysis (TGA) of the particles shows that there is only a minor weight loss of below

5% overall up to a temperature of 1400 ⁰C. TGA shows a first weight loss between ca.

50 and 200 ⁰C. This process is followed by a minor weight gain and a final slight weight loss until 1400 ⁰C. Differential thermal analysis (DTA) shows that the first weight loss is endothermic until ca. 170 ⁰C. Thereafter, an exothermic process begins and lasts until ca. 800 ⁰C. Upon further heating, the DTA curve increases again and a sharp endothermic peak is observed at 1064 to 1066 ⁰C.

The endothermic part of the weight loss below 170 ⁰C is assigned to desorption of residual solvent molecules from the purification process. TGA/DTA therefore shows that about 1 wt% of solvent is kept in the samples after drying. This indicates that some ionic liquid is still adsorbed on the surface of the final particles even after washing and drying. The exothermic contribution to the first weight loss is assigned to the thermal decomposition (the "burning") of a small fraction of cellulose present in the sample. Samples grown with higher cellulose concentrations show a very similar mass loss. This indicates that only a small fraction of the cellulose present in the reaction solution is incorporated in the final product.

The exothermic process (decomposition of the residual organic material) lasts until ca. 800 ⁰C. The slow increase of the DTA signal above ca. 800 ⁰C indicates that all organic material is decomposed above this temperature. The sharp endothermic peak at 1064 to 1066 ⁰C is due to the melting of gold. The fact that the peak is narrow and corresponds to the bulk melting temperature of gold indicates that our particles are large particles with bulk gold behavior. They are not aggregates of gold nanoparticles, which would melt at lower temperatures. TGA/DTA data are therefore in agreement with SEM and XRD, which show that the gold is not present as a nanoscale material, but rather as a bulk- like solid with larger dimensions.

Mineralization in the presence of cellulose starts from a transparent, bright yellow solution. After an initial lag time, the precipitate starts to form, which is indicated by the appearance of a red to brown color of the reaction solution. This indicates that, prior to the formation of the large particles obtained after 20 hours, small, presumably spherical, gold particles form as the precursors for the larger particles. Upon further reaction, golden particles deposit at the bottom of the reaction vessels.

Figure 7 shows the times between the start of the reaction and the first observation of a shiny gold sediment at the bottom of the reaction vessel. The lag time decreases with increasing reaction temperature. This is a clear indication that the reaction is thermally activated, similar to the formation of CuCl nanoplatelets from an ILCP. Indeed, experiments at temperatures below 110 ⁰C have shown that there, no reaction occurs within 20 hours and no product can be retrieved.

Figure 7 also shows that the lag time decreases with increasing cellulose and with increasing gold concentration. This clearly shows that the cellulose is an integral part of the reaction and that, not surprisingly, an increased amount of cellulose increases the reduction rate of the Au(III) ions and hence nucleation efficiency of the Au(O) particles. Further measurements to determine the order of the reaction, the kinetics, and the activation energy are underway. To further evaluate the role of the cellulose additive, the precipitation process without cellulose was studied. As cellulose reduces Au(III) to Au(O) and appears to interact with the growing crystals, its absence must have a strong influence on the resulting precipitates. Indeed, sedimentation experiments show that in the absence of cellulose, there is only precipitation of gold at 200 ⁰C after 8 hours, which represents a much longer lag time than in the presence of cellulose. Below 200 ⁰C, there is no macroscopic precipitation until 20 hours of reaction time.

Figure 8 shows SEM images of a sample prepared at 200 ⁰C in the absence of cellulose. SEM shows that many particles in these samples have a sheet or plate morphology. Other crystals are either spherical or small polyhedral particles. The plates are large and reach sizes of over 50 μm, similar to an earlier study, where gold sheets have been obtained by microwave heating of gold salts in [BMIM]BF₄ or [BMIM]PF₆.

Variation of the gold concentration leads to qualitative differences similar to the samples prepared with cellulose, see Tables 1 and 2. SEM suggests that the relative number of plates vs. other shapes is slightly different from sample to sample. Furthermore, occasionally gold belts of about 160 μm in length and 20 μm in width form at 10 mg HAuCl₄ * 3 H₂O. Overall however, the main morphology of the particles formed in the absence of cellulose is large plates along with smaller particles. In summary, it could be shown that cellulose/go Id(III) solutions in the IL

[BMIM]Cl can be used for the fabrication of micrometer-sized gold particles. The morphologies, particle sizes, and fractions of particles with different morphologies depend on the synthesis conditions, especially on the reaction temperature. Especially at high temperatures, plate-like crystals are obtained. Two-dimensional (gold) nanostructures such as plates or sheets, have attracted a growing interest in materials research due to their potential applications in the areas of electrochemistry, gas sensors, efficient surface-enhanced Raman scattering substrates and new nanodevices. The ease of synthesis, the tunability, the potential to scale the process up, and the availability of the reactants make our process a good candidate for the fabrication of large amounts of complex gold structures for the above and other applications. It is assumed that further the system is assignable for metals, especially noble metals.

The fact that the control sample grown without cellulose only precipitates at 200 ⁰C, shows that cellulose is needed as a reducing agent at temperatures below 200 ⁰C. In the absence of cellulose, the reduction of the Au(III) ion is most likely achieved by either oxidation of chloride anions to chlorine, according to equation (1),

2 HAuCl₄ -> 2 Au (s) + 3 Cl₂ (g) + 2 HCl (g), Eq. (1)

or by a direct reaction of the IL with the Au(III) ions. A further possibility is the reaction of gaseous Cl₂ during the reaction; as a result of this process, some fraction of the IL could react with Cl₂ and may change chemically during the reaction. The materials obtained at 160 and 200 ⁰C are different from other reports on gold plates. Usually, large plates with smooth surfaces and rather low thicknesses are found.

These results suggest that the microphase-separated structure of the ILs is partly responsible for the formation of the gold plates, possibly because the metal ion (and hence nucleus and primary particle) distribution in the IL is not homogeneous. This is supported by a recent study showing that gold plates can also form within a crystalline

IL analog, where hydrophobic and hydrophilic domains are well-separated. There, the presence of an organic template with a long-range order is of crucial importance for the shape control over the products. The formation of platelets is in this case attributed to the stiff matrix at low temperature.

Some reports suggest that the formation of gold plates results from kinetic control of the growth rates of various faces of the gold particles by selective adsorption of organic molecules on different crystal faces. The fact that in all previous reports predominantly gold plates (and not other shapes) were obtained from ILs, suggests that the IL components also act as selective growth inhibitors, mostly along the [111] axis of the growing gold particles, which leads to the commonly observed plates.

However, in the presence of cellulose, the gold plates are thicker than particles reported in these publications. This shows that the cellulose has two roles in this reaction. First, cellulose is the reducing agent for Au(III). Second, cellulose acts as a morphology and size-directing agent, which drives the crystallization towards polyhedral particles or thick plates. The differences between the control and the cellulose-containing samples can qualitatively be explained with heterogeneous nucleation on existing crystals and the relative growth rates along different crystallo graphic axes.

The shape of a face centered cubic (fee) crystal is mainly determined by the ratio R of the growth rate along the [100] vs. the [111] direction. Crystal growth rates along different directions are proportional to their surface energies. For fee gold, the relative surface energies γ of the low-index crystallo graphic planes are γ{l 10} > γ{100} > γ{ 111}. However, the surface energies and hence the relative growth rates along different crystallographic axes in fee metals can, like in many other inorganics, be modulated through the (selective) adsorption of growth modifiers on different crystal faces. Although difficult to determine from the current experiments, SEM suggests that the cellulose additive also partly enhances growth along the [111] and the [100] axes, because the samples grown in the presence of cellulose exhibit many more polyhedral particles than the control sample. Growth enhancement can either happen by selective blocking of the other crystal faces or by enhancing heterogeneous nucleation on the (111) faces by cellulose adsorption. Indeed, TGA and Figure 4b provide evidence for cellulose incorporation, although very little, and therefore agree with the mechanism suggested above.

Besides the cellulose, also the IL has an effect on the mineralization. Many ILs have ordered structures. 1,3-dialkyl imidazolium ILs such as [BMIM]Cl form two-dimensional polymeric assemblies. It is therefore reasonable to assume that the organized structure of the IL has a template effect on the formation of the gold particles. At high concentrations of HAuCl₄ * 3 H₂O, the order within the IL may be changed or destroyed due to the higher amount of foreign ions. As a result the IL is "softer" and becomes a less efficient template for the (thin) platelet formation. As a consequence, (irregular) polyhedral particles form more often, although this process does not appear to be strong. A similar, but much stronger, effect can be observed for reactions conducted at higher temperatures. Due to increased thermal motion, the templating is less efficient and the plates grow thicker with increasing temperature. Furthermore, temperature variation could also change the nucleation (location) and kinetics, which has, besides thermodynamics and steric effects, been suggested as a major influence on crystal growth by Yacaman and colleagues. In the current case, we observe many decahedra, that is, twinned particles. It is likely that the cellulose is responsible for the presence of a high number of defects in the growing particles. The formation of twin planes that promote the creation of favorable sites for further growth may also play a key role in the formation of anisotropic particles. Unlike gold, there seems to be no formation of large platelets with Pd, Pt, and Ag. Much rather, these particles precipitate as small nanoparticles, which form large, dense aggregates; Figure 9 illustrates SEM images of Pt nanoparticles formed at 130 ⁰C (top) and 200 ⁰C (bottom) after 20 hours of heating. Pt nanoparticles have been obtained that are only a few nanometers in size. The results for Pd and Ag are similar.

XRD confirms the electron microscopy data, Figure 10 shows XRD pattern of Pd grown with the thermochemical approach at 200 ⁰C. The reflections of the metal are barely visible (arrow in panel b) and the most intense signal is from some organic residue (cellulose). Overall, however, the XRD data confirm electron microscopy, because the reflections are so broad (indicative of very small metal particles) that they are barely visible.

Claims

1. Method of preparing nano- or microcrystals of Au, Pd, Pt or Ag comprising the steps of: a) preparing an ionic liquid mixture of cellulose and at least one metal salt selected form the group consisting of Au, Pd, Pt and Ag in l-butyl-3-methylimidazolium chloride; and b) (i) thermally inducing a reduction of the metal salt with cellulose by heating of the mixture to a temperature in the range of 50 to 250 ⁰C; or

(ii) photoreduction of the metal salt by irradiation of the mixture with light having a wave length in the range of 200 to 800 nm.

2. The method of claim 1, wherein the metal salt is selected from the group consisting OfM(NOs)_x, MCl_x, MBr_x, MI_x, M(OAc)_x, M(TfO)_x, M(acac)_x and HAuCl₄ * 3 H₂O, wherein M represents Au, Pd, Pt or Ag and x is an integer from 1 to 4.

3. The method of claims 1 or 2, wherein cellulose is present in the mixture in an equimolar amount to or in a molar excess to the metal salt.

4. The method of claim 3, wherein a molar ratio of cellulose to the metal salt is from 1 :1 to 20:1.

5. The method of any of the preceding claims, wherein the metal salt is an Au salt and step b) (i) of thermally inducing the reduction is performed at a temperature in the range of 180 to 220⁰C for preparing gold microcrystals having a plate-like shape and a plate thickness in the range of 700 to 1.000 nm.

6. Nano- or Microcrystals of Au, Pd, Pt or Ag prepared by the method of any of claims 1 through 5.

7. Cellulose product containing nano- or microcrystals of Au, Pd, Pt or Ag prepared by the method of any of claims 1 through 5.