Soft Matter
View Online
C
Dynamic Article Links <
Cite this: Soft Matter, 2011, 7, 2755
PAPER
www.rsc.org/softmatter
Downloaded by Instituto de Ciencia de Materiales. Biblioteca Man on 28 March 2011 Published on 27 January 2011 on http://pubs.rsc.org | doi:10.1039/C0SM01088J
Nanocomposites combining conducting and superparamagnetic components prepared via an organogel† Elena Taboada, Lise N. Feldborg, Angel Perez del Pino, Anna Roig, David B. Amabilino* and Josep Puigmartı-Luis* Received 1st October 2010, Accepted 15th December 2010 DOI: 10.1039/c0sm01088j A nanocomposite material combining an organic molecular gelator and oleate-coated iron oxide nanoparticles in proportions which range from one to fifty weight percent of the inorganic material has been prepared via the gel state. The proportion of nanoparticles and organic gelator in this mixed colloidal system gives very different characteristics to the final hybrid xerogel. Characterisation of the xerogels by transmission electron microscopy shows that at low loadings of the inorganic material a uniform distribution is observed, while above ten weight percent of nanoparticles a clear phase separation of the components (organic and inorganic) is revealed. Doping of the organic component of the xerogels by chemical oxidation results in the formation of conducting composites, whose electrical characteristics—probed by current sensing atomic force microscopy and spectroscopy—vary importantly with the amount of iron oxide colloid. The best conductors are found at low loadings of inorganic particles, at which an interesting alignment of the organic fibres is observed. The work shows that conducting materials incorporating magnetic particles can be prepared simply through the organogel route, and raises possibilities for the discovery of new properties that could come from the combination of these or related systems.
Introduction The preparation of easily processed materials which combine electrically conducting and magnetic components is a considerable challenge. The difficulty in this area resides in making a sufficiently conducting material with the magnetic component interspersed. The conducting pathways can be interrupted by the mere presence of another component. In crystalline materials this task is particularly difficult, although in layered systems it can be achieved.1 Paradoxically, it is in crystalline systems that conducting organic materials are generally at their best. A more general area is the preparation of conducting polymers incorporating magnetic nanoparticles,2 yet in these systems the nature of the electrical characteristics is distinct to the metallic type conductivity observed in crystalline molecular samples. We are interested in preparing nanostructured molecular organic conducting material3 in the form of a film that can be achieved through the gel state.4 The conducting ‘‘wires’’ are formed, thanks to the presence of supramolecular polymers, and these can be used for making conducting systems including
Institut de Ci encia de Materials de Barcelona (CSIC), Campus Universitari, 08193 Bellaterra, Catalonia, Spain. E-mail: amabilino@ icmab.es;
[email protected]; Fax: +34 93 5805729; Tel: +34 93 580 1853 † Electronic supplementary information (ESI) available: Supporting AFM images and comparitive analysis of sample conductivity and fibre alignment. See DOI: 10.1039/c0sm01088j
This journal is ª The Royal Society of Chemistry 2011
hybrid and composite materials. This approach allows the combination of dissimilar building blocks with pathways interacting and influencing upon each other in unique ways such that they can generate materials with characteristics that are distinctive to their components.5 For instance, it has been proved that when superparamagnetic ferrites or semiconductor quantum dots such as CdS are immobilised they confer magnetic and/or luminescent properties to the final hybrid gel state.6 We use the gels as a route to these materials because the fibres are ‘‘frozen’’ in the solvent matrix, which can then be evaporated, and the conducting material is then prepared by doping.7 These systems are amenable to the formation of conducting hybrid nanomaterials, as we have shown for the case of gold nanoparticles containing hydrogen bonding units which help make the two components compatible.8 Here we report a new class of self-assembled and non-covalent hybrid organogels combining conducting and superparamagnetic components formed from an amide tetrathiafulvalene (TTF) derivative (1) and organic soluble‡ magnetic iron oxide nanoparticles (g-Fe2O3 nanoparticles, abbreviated NPs here) to fulfil the objective of reaching a conducting magnetic hybrid material. The TTF unit is the one known to form a variety of interesting nanostructures and to perform ‡ We use the term soluble to infer that the whole of NP is homogeneously dissolved in the solvent. An alternative nomenclature from colloid chemistry would be to say that the material forms a homogeneous colloidal dispersion.
Soft Matter, 2011, 7, 2755–2761 | 2755
View Online
Downloaded by Instituto de Ciencia de Materiales. Biblioteca Man on 28 March 2011 Published on 27 January 2011 on http://pubs.rsc.org | doi:10.1039/C0SM01088J
a variety of functions,9 while the NPs below 15 nm diameter are established superparamagnetic colloids which are attractive for a variety of applications.10 We describe how the relative proportions of the two components influence the nanostructure of the resulting hybrid, and how, in turn, the electrical properties of the doped material are modulated as a result.
Results Preparation and characterisation of the nanocomposites The nanocomposite gels were formed by dissolving organogelator 1 and a weight percent proportion of the NPs in hot hexane, and then allowing the solution to cool to room temperature unperturbed. The organogelator was synthesised using the procedure described by us previously,11 and the superparamagnetic iron oxide nanoparticles (g-Fe2O3, typically with a mean particle size of 7 nm and a polydispersity index of 10%) were synthesised adapting a literature procedure12 as reported by us previously.13 Different gel samples were prepared where the percentage in weight of the NPs contained in the final hybrid gel was varied from 1% up to 50% (Fig. 1). As can be appreciated in the photograph, all the gels are transparent and stable, even at the highest nanoparticle content studied. The change in colour arises from the absorption of iron oxide nanoparticles which are dark brown. The fact that gels can be formed with up to 50% of inorganic material is thanks to the great solubility of the NP in hexane, which is a result of the capping of the iron oxide with the oleate covering. This situation contrasts with gels formed by certain
derivatives of gold nanoparticles—which were not soluble in this organic solvent—where the maximum loading of inorganicbased matter in the hybrid of the same organogelators was a little above one percent.8 An alternative way to impregnate the gel with these nanoparticles is to allow the particles to diffuse into the gel from an isotropic solution of the NPs in hexane. When a solution of the iron colloid in hexane in a capillary tube was brought into contact (using a magnet) with the gel of 1 in hexane in the same tube a layer of the liquid is formed on the surface of the gel. However, over the period of a day the NPs do diffuse into the gel. The magnet used to move the NP solution cannot force the nanoparticles into the gel, a situation which contrasts dramatically with a suspension of crystals of 1 in hexane (formed by rapid cooling of the saturated solution) whereby the NP solution is moved back and forth in the suspension using a magnet. On the other hand, the same magnet is unable to distort the gel or remove colloidal material from it—no distortion of the meniscus is seen after contact of the magnet with the gel for one day (as is noted immediately in concentrated solutions of the NPs in hexane), probably because the concentration of nanoparticles is so low in the gels. The content of the NPs in the gel of 1 has an effect on the melting temperature of the gels (Table 1). The transition from gel to liquid was measured by observing the flow of the mixtures angled in a warmed oil bath, and repeating the process from reformed gels to attain average values. These experiments show that the presence of the NPs increases the transition temperature, and hint at an interaction between them and the fibres of the gelator 1. Even 1% of NPs is enough to increase the transition temperature by 6 C, and subsequent addition has a lesser effect. At the highest content of NPs assayed (50% by weight) the transition temperature is less than the optimum one, presumably because the areas of iron oxide colloid disrupt the gelator fibre network. The nanostructure of the hybrid gels was studied using Transmission Electron Microscopy (TEM). In order to ensure a thin enough covering on the holey carbon grid such that the sample gave good contrast, the gel samples were heated to the solution state and a drop was placed on the grid and allowed to evaporate (this procedure is necessary to avoid depositing too much material on the grid which results in a film which is too opaque to the electron beam). No staining was used to visualise the colloids. The resulting samples were observed in several places, and representative TEM images are shown in Fig. 2.
Table 1 Temperatures for the gel–liquid transition as a function of nanoparticle content in the gel of 1 in hexane Weight percent of NP in hexane gel Gel–liquid transition temperature/ C of 1
Fig. 1 A photograph of the gels formed in hexane by organogelator 1 and different weight percents of iron oxide nanoparticles (NPs).
2756 | Soft Matter, 2011, 7, 2755–2761
0 1 5 10 25 50
42 48 50 50 49 48
This journal is ª The Royal Society of Chemistry 2011
Downloaded by Instituto de Ciencia de Materiales. Biblioteca Man on 28 March 2011 Published on 27 January 2011 on http://pubs.rsc.org | doi:10.1039/C0SM01088J
View Online
Fig. 3 TEM image of 1–NP xerogel containing 5% in weight of NPs, acquired by casting a hot solution of the gel-forming solution onto a holey carbon grid.
Fig. 2 TEM images of the different 1–NP xerogels. Xerogel containing (A) 1%, (B) 5%, (C) 10%, (D) 15%, (E) 25%, (F) 50% and (G) 0% in weight of NPs. All the TEM images were acquired by casting hot solutions of the gel-forming solution onto holey carbon grids.
In the images of the unstained xerogels containing between 1 and 10% of the nanoparticles the dominant feature is the fibres formed by 1 with the darkest contrast coming from the inorganic material (Fig. 3). As in the xerogel formed by the organic molecule on its own,7 twisted bundles of fibres are observed with widths ranging from a few tens of nanometres to approximately 100 nm (although larger clumps exist). It is not possible to determine the lengths of the fibres in a quantitative manner, as they cross and intertwine. The nanoparticles are imaged as dark dots located in or at the edges of the fibres, sometimes alone and This journal is ª The Royal Society of Chemistry 2011
sometimes in lines containing up to half a dozen particles. Quite rarely, isolated nanoparticles are observed which are apparently not associated with any nanofibre. The great propensity for the nanoparticles to adhere or be incorporated into the fibres is particularly remarkable given that they contain no hydrogen bonding unit and are presumably held to the fibres purely by van der Waals interactions. This same fact is the reason that some isolated particles are seen. If the interaction between particle and fibre were stronger, one would expect to observe no isolated particles.8 A clear difference in the structure of the nanocomposite is seen when the mixture contains greater than 10% of the nanoparticles. Above this value two important observations are worthy of note: firstly, the fibres cannot be imaged in these samples because of the high content of absorbing inorganic colloid, which means that the organic fibres can only be inferred by the presence of clear tracks in the images. Secondly, no nanoparticles are seen in the clear tracks, implying that the organic and inorganic components have phase separated. The clear non-covalent ‘‘misunderstanding’’ between the two components is evidenced when a closer look is taken at the xerogels containing 15 and 50% in weight of NPs (Fig. 4). The TEM image of the 1–NP xerogel containing 15% of particles (Fig. 4) shows obvious separation of the two components, the inorganic and the organic. The NPs start to organise (Fig. 4A (1)) and self-assemble (Fig. 4A (2)) challenging a large range organization when raising the percentage of the inorganic component in the final hybrid gel (Fig. 4B). In the later case, the assembly of the NPs is different from the self-assembly examined when Soft Matter, 2011, 7, 2755–2761 | 2757
Downloaded by Instituto de Ciencia de Materiales. Biblioteca Man on 28 March 2011 Published on 27 January 2011 on http://pubs.rsc.org | doi:10.1039/C0SM01088J
View Online
Fig. 4 TEM images of 1–NP xerogels containing 15% (A) and 50% (B) in weight of NPs relative to 1, respectively, and (C) a solution of NP dropcast from hexane. The highlighted regions in (A) show regions where particles line a fibre (1) and where they assemble to form a domain (2). All the TEM images are acquired using holey carbon grids on unstained samples.
a solution of NPs is dropcast over a TEM holey carbon grid (Fig. 4C). Unlike the solution of NPs, the hybrid xerogel gives rise to small and large domains of NPs where mono- and bilayers of NPs are observed due to the separation and the noncooperation between two components (Fig. 4B). The separation of the inorganic and organic components reminds us of phase separation in polymer based nanocomposites,14 although surface drying effects may play a role,15 the nature of the images does not imply this is dominant given the even distribution of the particles arising from their high solubility. The fibres in the clear tracks in the samples with high iron colloid content should also have NPs associated with them if the NPs are bound to the fibres by van der Waals interactions, and this observation might favour a mechanism in which the NPs align themselves with the fibres during their formation. In any case, there is clearly a strengthening interaction at the interphase between the two colloids, as demonstrated in the higher phase transition of the gel-to-liquid change (Table 1). It is also interesting to reflect that phase separated gold nanoparticles have been used recently to probe the early stages of growth of gel fibres.16 The situation contrasts with the case where clear noncovalent interactions take place between the components.8,15 It is therefore interesting to see the effects that this kind of phase separation has on the conducting properties of these materials once charge carriers have been introduced through doping. Conductivity of the doped nanocomposites Current sensing (CS) atomic force microscope (AFM) measurements of doped xerogels containing 1, 5, 15 and 25% in weight of NPs were performed for each sample separately on highly oriented pyrolytic graphite (HOPG). The doping process consisted of exposure of the xerogels to iodine vapours for two minutes in a sealed chamber.7 In the CS-AFM measurements topography and current images are recorded simultaneously with the Pt/Ir coated tip in permanent contact with the sample while applying a voltage between them. The current maps from these AFM experiments show areas with clear fibre-like morphology which is relatively uniform across the sample when the percentage of NPs is under the threshold phase separation limit (
