1. IntroductionThe ability to produce nanostructured surfaces over macroscopic areas is likely to be of crucial importance to the integration of nanotechnology into commercial devices. The main actual techniques employed for their large scale fabrication, the light (usually ultraviolet) and electron lithography [1], start to reach their theoretical limits and even worse; they are already cost-prohibitive. About orders of magnitude cheaper than the lithographic technologies, the self-assembly [2] based approaches such as natural lithography [3] are of a growing interest in exploring their abilities. Between these two approaches seen as extreme in terms of cost, there are a multitude of other techniques [4] [5] [6] [7] [8] that try to get nanostructurted surfaces of the quality given by classical lithography and the cheapness of those obtained by self assembly. However, even if they look as being very different, electron and natural lithography have some common points. In electron lithography [9] a focused beam of electrons scans to draw custom shapes on a surface covered with an electron-sensitive film called a resist, followed by selectively chemical removal of irradiated or non-irradiated polymer. In self-assembly techniques, polystyrene or silica nanospheres self-assemble in 2D or 3D ordered close packed structures followed by infiltration of a desired material between nanosphere voids, its casting, followed by the wet chemical or oxidative thermal removal of nanospheres [10]. In both of these we have a sacrificial material (resist or nanosphere) which is removed after it ends its role in the surface structuring. However, if in electron or UV lithography the resist can be structured in “quasi-infinite” modes, in natural lithography the nanospheres can produce only few, square or hexagonal for 2D and hexagonal or cubic for 3D structures [11]. By using pre-patterned surfaces through electron or UV lithography, the number of ordering modes in nanosphere lithography can be increased, but unfortunately, also the final cost [12]. In the present work we propose an original aproach in which the sacrificial polymeric template could be replaced by crystaline materials such as ionic salts which, in some special conditions, can form arrays of single crystals. Known as dendritic crystals [13], they are common in polymer casting [14] or frequently observed when polymers such as gelatin [15] [16], organic molecules [17] or salts [18] are added even in small quantities to ionic solutions. Moreover, dendritic crystals of halite NaCl, mirabilite Na2SO4·10H2O or thenardite Na2SO4 were observed from saturated solution droplets without other additives, most of them growing at the outside of the droplet edge and being more or less influenced by the local humidity level [19]. Dendritic crystalline patterns can also appear by melt crystalizalization of some materials such as metallic alloys [20]. The cause which generates them is still under debate and few hypotesis were proposed. M.L. Neser observed the growing of dendritic crystals loaded at their ends with hemoglobin diluted and normal saline (0.9 wt%) drop was evaporated. As cause she mentions “the impurities in the solution” [21]. For the mirabilite hydrated salt case reported in [19], the authors explained that “when dehydration of mirabilite occurred, the release of water of crystallization produced small, episodic pulses of water at the edge of the drop, which generated the secondary growth of efflorescence-like crystals outside the droplet”. N. S. Bonn et al. report the growth of dendritic patterns at the end of Na2SO4 droplet drying on a hydrophilic surface due to “a mesoscopic wetting film around the contact line” [22]. G.H. Harrington et al., generate crystal patterns on smooth surfaces such as mica, glass, or calcite by using very dilute droplets of aqueous salt solution (0.1 wt% NaCl) [23]. They consider that patterns are caused by “the liquid film ruptures” at the moment at which the drops are very thin. S. Mysura made crystallization studies from thin layers of CaCl2, LiBr or LiCl aqueous solutions and presented some conclusions about the phenomena that induced the dendrite formation [24]. Thus, he observed that when the liquid film height decreases, the heat transfer coefficeint can vary several times. Also, he noted a strong dependence between evaporation rate and crystal forms on crystallization rate. Moreover, he reported abnormally high crystallization rates (30 - 50 mm/s) at extremely low supersaturations of solutions associated with the rapid growth of dendritic forms. Dendritic assemblies composed from inorganic nanoparticles can be obtained if the solutions of ionic salts are present. S. Paria et al. report the formation of self-assembled sulfur fractal patterns using colloidal sulfur particles (synthesized in situ by the reaction of sodium thiosulfate and different inorganic or organic acids) [25]. They concluded that “fractal patterns are influenced by the crystal morphology of salts” and, “particles alone form the ‘coffee-ring’ structure without forming any self-assembled fractal pattern”. If the number of reports of dendrite crystals grown from ionic solutions is rather poor, molecular solutions of organic semiconductors are often used for surface patterning through solution-shearing method [26] [27] [28]. It is worth noting that very recently a mathematical scheme was reported for an anisotropic dendritic crystal growth model by coupling diffusion and heat equations [29].
The present paper shows how several types of ordered polycrystalline structures can be grown onto a solid substrate and suggests some possible applications of these. Ordered structures are grown far from equilibrium conditions where geometric confinement, temperature and concentration gradients occur. These patterned surfaces can be used in electronics as large scale integrated circuits or as masks as that in conventional optical lithography.
2. Experimental2.1. Dendrite Crystal Growth1) Growth in confined volumes (thin films). A thin liquid film of 4 µm in thickness has been formed by vertical dipping and dragging out of a microscope glass slides in ionic salts aqueous solution (1 wt%). The slide was horizontally kept till it dried.
2) Growth in confined volumes (thin films) on a hot plate. A thin liquid film of 4 µm in thickness has been formed by vertical dipping and dragging out of a microscope glass slide in ionic salts aqueous solution (1 wt%). The slide was horizontally deposited (t = 5 - 10 s) onto a heater (T = 300˚C - 500˚C). A quasi-ordered array could be observed onto entire surface on optical microscope.
3) Growth on a tilted substrate. Sulfur powder was disolved in toluene near saturation. The liquid is pourred onto a tilted (10°) microscope glass slide thus it forms a thick liquid film pinned from substrate margins. After toluene natural evaporation, crystalized sulfur lines can be observed onto the slide.
2.2. Ring Crystal Growth1) Growth in confined volumes (droplets) on a hot plate. Aqueus solution of NaCl (10 wt% ) was sprayed onto a glass substrate heated at 300˚C. A large number of NaCl rings has been observed on entire surface by using optical microscope.
2) Growth in confined volumes (drops) in laser beam. A 5 mm in diameterdrop, of aqueus solution of NaCl (0.1 wt%) was deposited onto a glass substrate. A IR laser beam (λ = 10.6 µm, P = 100 W) pulse (t = 0.2 s) heats and explosive evaporates the NaCl solution drop [30]. NaCl ripples formed onto the glass substrate.
2.3. Crystals as Template1) Crystal onto crystal array. Crystalized sulfur lines were obtained onto a glass slide by technique described in (2.1c). A K2CrO4 was grown onto sulfur line by the technique described in (2.1a). Intersected suphur and A K2CrO4 net was observed on optical microscope.
2) Nanoparticle array. Iron oxide nanoparticles synthesized by laser pyrolysis [31] were dispersed (c1 = 0.1 wt%) in a mixture of ethanol/toluene (c2 = 30% vol.) and ultrasonicated. From this dispersion, a nanoparticle film was deposited onto a K2CrO4 dendritic surface obtained in (2.1.2) through dip coating technique. After film dried, it was immersed in distilled water and ultrasonicated for 5 minutes, extracted, and naturally dried again. Nanoparticle parallel lines could be seen using on optical microscope.
Structural characterization using X-ray diffraction was performed with a Panalytical X’Pert Pro MPD equipment in Brag-Bretano configuration, continuous scanning (10˚ - 70˚) deg. 2θ, with a step of 0.02˚ and a time of 0.02˚. 20 sec/step acquisition.
3. Results and DiscussionsIt is known that concentrated aqueous solutions of ionic salts form specific three dimensional single-crystals (Figure 1(a)) when water evaporates [32] [33] [34]. When crystallization takes place on a solid substrate the single-crystals appear as a collection of many isolated or partially clusterized similar crystals, only their size and numer varying. Sometime, when the solutions are highly diluted, dendritic shapes are obtained (small and flat single-crystals bonded in fractal like structures) (Figure 1(b)) [22] [35]. These dendritic self-assembled crystals are hardly reproducible (at least as shape).
It is known that temperature is an important parameter which dictates the density of nucleation and subsequent, the crystals size. By increasing the temperature during solvent evaporation such as: the glass slide is setteled on a heater (Figure 1(c)), droplets from a spray impact the glass slide setteled on a heater (Figure 1(d)), a IR laser beam pulse irradiates a drop setteled on a glass slide (Figure 1(e)), or an IR laser beam balleates a liquid film formed onto a glass slide (Figure 1(f)), we found that diluted ionic liquids films form different dendritic crystals on a large surface, every time. NaCl crystallizes in a beautiful fractal shape on square centimeters (Figures 1(a)-(f)). Other ionic salts such as:
BaCl2, CuCl2, Na2SiO3, Ni(CH3COO)2, and others give rise to specific crystal fractals. Different ionic salts form different shapes of dendritic crystals (Figures 1(g)-(i)), the most interesting we found to be the grid which forms through K2CrO4 crystallization (Figure 1(h)). We found that these parallel and long (few mm) of micrometers wide crystals could be used as template or as mask in a lithography process. Molecular solutions such as 9, 10-bis (phenylethynyl) anthracene (BPEA) used by Y. Li et al. in [8] can be able to form single crystal grid through solution-shearing by using a heater. Our atempt in repeating their experiments, by using sulfur dissolved in toluene, failed because sulphur crystallizes as micrometric spheroidal crystals (orthorhombic phase containing S8 molecules ) when solution fast evaporates (Figure 1(j)). However, by keeping their idea of tilting the substrate but using a free evaporation, a concentrated solution, and a tick liquid film (2 mm) we succeed to obtain long sulfur single crystals in a parallel array (Figure 1(k) and Figure 2(a), top image).
By using as template sulfur or K2CrO4 grids we may pattern surfaces in a useful way thus some devices to be fabricated. We shall forward present three of these. If onto the surface from Figure 2(a) we form a thin K2CrO4 solution film and evaporate it on a hot plate, we will not obtain a K2CrO4 polycrystalline gride as in Figure 2(a) (middle image), but a superposed sulfur and K2CrO4 net as in Figure 2(a) (bottom image), a net able to diffract the white light. Keeping the approach but changing the basic materials we could form such net from much more interesting materials. In a second example, onto a K2CrO4 grid formed on a hot plate (Figure 2(b), top image), is deposited a thin film of iron oxide nanoparticles (Figure 2(b), middle image). After sinking in water, the K2CrO4 grid
together with all the nanoparticles deposited on it is removed, remaining thus only the nanoparticles (as a grid) which were in close contact with the glass substrate (Figure 2(b), bottom image). In the third example, we show how a dilluted sulfur solution deposited onto a K2CrO4 grid (Figure 2(c), middle image) forms a sulphur dots array (Figure 2(c), bottom image) which otherway would randomly deposit onto a clean glass surface (Figure 2(c), top image).
Even if are less impresing as accuracy, these three examples can be improved and extrapolated to some interesting devices or to the study of complex phenomena in which surfaces, interfaces, and far from equilibrium crystallization lead to unexpected observations.
A such information is revealed by X ray diffraction (Figure 3) performed on the structures sown in Figure 2(a). If K2CrO4 and sulphur grids are separately grown on a flat substrate, both of them form textured surfaces in which their crystallographic orientations have some preferred directions. If the surface is not flat but structured itself (K2CrO4 grid for example), the material which will crystalyze on such surface (sulphur in our case) will gain new crystallographic orientations. We consider that it would be useful to control the crystallographic orientation of a cristalline growth film trough the structure and morphology of the substrate.
As time as dendritic structures can be grown at normal temperatures (before boiling temperature) the role of high temperatures (above boiling point), seems to resume to the creation of a large number of non-equilibrium space and time zones, in which the substrate dewetting is somehow correlate with the ionic salt concentration of crystallization. The existence of some important temperature dependent parameters such as: liquid film thickness, ionic salts concentration of saturation, or evaporation rate, which act simultaneously, make difficult the finding of such correlations. However, we observe that not the absolute value of
a high temperature is important but rather the intensity (spatial and temporal) of the temperature gradient. This temperature gradient may correlate also with the texturation phenomenon found by optical and XRD measurements, thus it must be taken in account in experiments regarding surface patterning, complex systems or, non-equilibrium processes.